Aptamers and aptamer targeted delivery

Abstract

When aptamers first emerged almost two decades ago, most were RNA species that bound and tagged or inhibited simple target ligands. Very soon after, the ‘selectionologists’ developing aptamer technology quickly realized more potential for the aptamer. In recent years, advances in aptamer techniques have enabled the use of aptamers as small molecule inhibitors, diagnostic tools and even therapeutics. Aptamers are now being employed in novel applications. We review, herein, some of the recent and exciting applications of aptamers in cell-specific recognition and delivery.

Introduction

Aptamers are nucleic acid binding species capable of tightly binding to and discreetly distinguishing target ligands. In fact, they have often been described as nucleic acid versions of antibodies. However, unlike antibodies, aptamers have yet to elicit immunogenicity in vivo.^1–5 Moreover, these molecules are readily amenable to chemical synthesis (thereby decreasing production costs) and can be easily modified during the synthesis process, making them more adaptable for different applications.

Early aptamer selections focused on RNA or DNA that bound small molecule ligands, such as dyes,⁶ ATP,⁷ or soluble proteins such as thrombin⁸ or polymerases^{9, 10}. Since those initial selections, hundreds of targets have been successfully selected against, with the targets for selection ranging from simple organic molecules to proteins, complexes, and, more recently, whole cells and organisms (see the Aptamer Database at http://aptamer.icmb.utexas.edu for a comprehensive compilation¹¹). By varying key parameters of a selection, aptamers with finely tuned physical and functional properties have been selected, and the applications appear almost limitless. For example, RNA aptamers can and have been encoded in expression cassettes and expressed in vivo for use in gene therapy applications (reviewed in ^12–14), whereas DNA aptamers, due to their ease of production, are well suited for diagnostic applications.^{15, 16} Additionally, the inclusion of modified RNA containing 2′deoxy, 2′F, 2′NH₃ or 2′OMe into selections (reviewed in ¹⁷), have allowed for stabilized aptamers that can now be used in complex biological solutions, such as blood and serum, paving the way for in vivo use of aptamers.

In the current review, we will summarize some advances in aptamer selections and some of the newer applications of aptamers in the most recent few years. In particular, we will focus on developments in targeting cells and cell surface receptors and delivery to cells using aptamers which target the cell surface. For more details about aptamers for gene therapy, aptazymes, riboswitch-type aptamers or other earlier generation aptamers, there are several reviews already available.^{12, 14, 16, 18}

The process of in vitro selection (SELEX)

Aptamers are nucleic acid binding species generated by iterative rounds of in vitro selection, or SELEX (Figure 1A).^{6, 9} Briefly, pools of random-sequence RNA or ssDNA are incubated with target molecules under carefully chosen selection conditions. Binding species are partitioned away from non-binders, amplified to generate a new pool, and the process is repeated until a desired ‘phenotype’ is achieved or until sequence diversity is significantly diminished.

Figure 1.

The process of in vitro selection (SELEX). A). In a traditional selection, random sequence pools (squiggly lines) are incubated with a target (pink orbs). Non-binding species are washed away, and the bound species are eluted and amplified to regenerate the pool for the next round. B). In more complex target selections, the pool is first incubated with a non-target (green orbs) in a negative selection step. Non-bound species from the negative step are added to the target, and the selection proceeds as usual. Over many cycles of selection, high-affinity sequences (aptamers) can be isolated.

Multiple cycles of selection and amplification winnow an initial pool containing upwards of 10¹⁵ molecules to only those few species that have the highest affinities and specificities for a target. Previously selected aptamers have typically bound their targets with Kd values in the nanomolar to picomolar range and can discriminate against very closely related targets (e.g., proteins that differ by only a few amino acids^{19, 20} or targets that differ by slight modifications.²¹ By including a negative selection step into the selection cycle in which pool sequences are first exposed to non-desired targets to remove non-specific binders, aptamers with very discrete specificities can be isolated (Figure 1B). This very important negative selection step has been crucial in identifying aptamers which can discriminate between closely related protein targets, and more recently, in the development of selections against complex targets such as whole cells.

Whole cell SELEX

Instead of the traditional positive selection for a known target and a negative selection step against a non-target such as a support resin or the filter used for immobilization of the target, selections can be used to parse out a target by subtracting the ‘background’ in the negative step. This approach has been successfully applied to known targets on the cell surface. For example, by combining positive rounds of selection against a cell line engineered to express the RET receptor tyrosine kinase with negative selections against the non-RET expressing parental cell line, Cerchia et al. were able to elicit aptamers which bound that specific receptor.^{22, 23}

This type of ‘toggle’ selection can also be taken to the extreme in which the experiment is performed not by knowing what the target is, but by knowing what one doesn’t want the target to be. In one of the earliest examples of using this subtraction selection technique to parse out aptamers that recognize differences between cells, Ulrich et al. raised aptamers against infective stages (trypomastigotes) of the parasite Trypanosoma cruzi. By including negative selections against the cells of the non-infective stage (epimastigotes), the authors aimed to find aptamers against epitopes specific to the infectious form. The aptamers, which proved to inhibit cell invasion by the parasite, were further used to identify host factors that T. cruzi utilizes for infection.²⁴

Using this or similar approaches, aptamers have been selected which can distinguish rat glioblastoma from microgial cells²⁵ and recognize populations of T cells,^{26, 27} liver cells,²⁸ lung cells,²⁹ and dendritic cells.³⁰ In one example of whole cell selection, the Tan group performed selections to distinguish between B-lymphocytes and T-lymphocytes. By performing positive and negative selections, respectively, against CEM and Ramos cells, aptamers were selected that could mostly distinguish the target cell type in mixed cultures of cells.²⁷ In a follow-up study, the targets of the aptamers were identified. One aptamer:target pair, sdc8 and its target, protein tyrosine kinase 7, was shown to be internalized, a trait that could prove useful for a number of applications, as will be further described below.³¹

To illustrate one application of whole cell selections, the Krylov group demonstrated how these selections could funnel directly into biomarker discovery. The group selected ssDNA aptamers against immature and mature dendritic cells and produced two sets of aptamers: one with preference for immature dendritic cells and the other for mature cells. The aptamers were tethered to beads and used to bind cells. These cells were lysed, washed, and the aptamer-bound targets were eluted and analyzed by mass spectrometry. The analysis revealed markers known to be expressed on these cells as well as several markers not previously associated with these cells.³⁰

These works and others like them illustrate how these target-naïve selections could be used for biomarker discovery. Importantly, these and other selections have also shown that aptamers can be selected against different states of a cell (age, differentiation, health, progression of disease, etc.). For example, aptamers have been selected that can distinguish differentiated from non-differentiated neuronal cells³² and cells infected by the vaccinia virus from non-infected cells.³³

Targeting cells with aptamers

A toolbox of aptamers that can recognize different cells and different disease or health states of a cell could be readily adapted for in vivo diagnostic as well as therapeutic purposes. While these may ultimately be adapted from aptamers identified ‘blindly’ from whole cell SELEX, the current advances have been made in adapting aptamers that target specific proteins and/or cell surface receptors. A number of these aptamers have been adapted for diagnostic and imaging purposes. For example, Borbas et al. and Hicke et al. both used radiolabeled forms of anti-MUC-1 and anti-tenascin C aptamers, respectively, to image tumors in mice.^{34, 35}

Although aptamers identified by whole cell SELEX have not yet been applied to in vivo imaging and diagnostic applications, they have been adapted for in vitro diagnostic assays. Shangguan et al. screened aptamers generated against T-cell acute lymphoblastic leukemia cancer cells for their ability to type cancer patient samples.²⁷ Although there was some cross-binding between the different types of samples, for the most part, the aptamers bound to their intended cell types. This suggests that a larger subset of aptamers could more definitively type a given cancer sample, as each type of cancer would produce a specific binding profile.³⁶ However, care must be taken in interpreting these types of results, and the Ellington group has recently published a summary of factors that might be considered in using aptamers for cell typing.³⁷

Aptamer-mediated delivery to cells

In addition to biomarker discovery and diagnostics, aptamers can also be selected that not only recognize a cell, but also be endocytosed, potentiating the escorted delivery of appended cargoes into cells.³⁸ How to specifically deliver drugs to certain sites and not to others is a conundrum that is tantamount to the success of a therapeutic. By slightly augmenting the cell-specific aptamer selections described above, aptamers can be selected that can specifically enter and deliver agents into cells.

For the most part, targeted delivery using aptamers have been performed using aptamers selected against receptors known to be internalized, and a variety of different cargoes have now been successfully delivery both in vitro as well as in vivo (Figure 2). For example, the Farokhzad and Langer labs have used an aptamer selected against the prostate specific membrane antigen (PSMA) to deliver nanoparticles to PSMA-expressing LnCap cells for imaging³⁹ and drug delivery.^{40, 41} Both the Ellington and Sullenger labs also made use of anti-PSMA aptamers, but for delivery of siRNA to LnCap cells. In a simple and elegant design, McNamara et al. delivered anti-plk and anti-bcl2 siRNAs to LnCap cells by appending the siRNA sense strand directly to the anti-PSMA aptamer. This chimera was hybridized to the antisense strand, incubated on cells and effectively knocked down gene expression. The group went on to demonstrate that this delivery system was effective in vivo.⁴²

Figure 2.

Aptamer-mediated delivery strategies. A). One strand of siRNA (red) is directly linked to the aptamer (black), and the complement siRNA is annealed via hybridization.^{42, 47} B). siRNA or other cargoes are biotinylated and loaded onto streptavidin. A biotinylated aptamer mediates delivery of the entire particle.⁴³ C). Toxins or other drugs are directly conjugated to the aptamer for delivery.^{48, 50} Finally, D). Aptamers are directly bound to nanoparticles.^{58, 59}

In a similar manner, Chu et al. delivered siRNA by loading the four binding pockets of streptavidin with biotinylated siRNA and biotinylated aptamer.⁴³ These constructs also showed knockdown on par with controls in which siRNA was directly transfected into cells. By using a streptavidin scaffold, the group created a more modular complex, in which different siRNA or different aptamers could eventually be co-delivered, or varying ratios of each component could be used. The multivalent construct also has the added potential benefit of garnering avidity effects, a property that was recently exploited by another group who demonstrated siRNA delivery using bivalent aptamer chimeras.⁴⁴

The Rossi Lab also used aptamer-mediated delivery to inhibit HIV-1. Zhou et al. recognized that gp120 on the surface of an infected cell might allow for targeting siRNA. Indeed, this viral antigen had previously been targeted for siRNA delivery utilizing an anti-GP120 single chain antibody protamine conjugate.⁴⁵ They created chimeras of anti-gp120 aptamers⁴⁶ with anti-Tat siRNA. The aptamers, in of themselves, were capable of inhibiting viral replication to some extent, as they could block HIV-1:cell interactions. ^46,47 However, the combination of aptamer with aptamer-delivered siRNA essentially obliterated infection in culture.⁴⁷

Proteins, too, have been targeted to cells using aptamers. For example, the Ellington lab demonstrated almost a 1000-fold increase in the toxicity of the ribosomal toxin gelonin when conjugated to an anti-PSMA aptamer. The toxin normally has difficulty entering the cell, and is, therefore, relatively non-toxic. Enhanced toxicity was only observed for PSMA expressing LNCaP cells. A non-expressing PSMA line, PC3 actually showed a decreased response to the toxin-conjugate when compared to the toxin alone.⁴⁸

More recently, the Neufeld lab, was able to deliver the enzyme α-L-iduronidase to the lysosomes of cells deficient in this enzyme using an anti-mouse transferrin receptor aptamer. The aptamer conjugates were able to restore enzyme activity in the cells suggesting a potential therapeutic approach for the treatment of lysosomal storage diseases.⁴⁹

This ability to deliver siRNA or other gene modulators to specific cells via receptor-targeted aptamers should be a great tool for researchers and clinicians alike, perhaps allowing for site-specific control of disease or development genes. Additionally, aptamers identified from selections against cells may also be capable of internalizing and similarly adapted for delivery. For example, Xiao et al. demonstrated that an aptamer previously selected naively against T-cells bound to protein kinase 7 and could be internalized.³¹ They subsequently adapted this aptamer for targeted delivery of the drug doxorubicin using an acid labile drug conjugate.⁵⁰

The next 20 years…

Aptamer selections have progressed from targeting and inhibiting the function of isolated proteins or molecules and simple in vitro proof-of-principle concepts to targeting epitopes on whole cell surfaces. Because aptamers can now be selected without prior knowledge of specific cell surface markers or cell surface architecture, the process of in vitro selection can be used to distinguish between different cell or disease types. Indeed a problem with treatment and diagnosis of many diseases is precisely that the specific disease marker may not be known. With cancerous cells, for example, the type of cancer, degree of progression or severity and location may all affect marker types and expression levels. Into the future, selections against whole cells, or diseased tissues should provide us with panels of aptamers that can specifically target phenotypes. These in turn can be adapted to serve as valuable diagnostics or further adapted for the targeted delivery of siRNA or other therapeutics. Further still, when one considers how peptides that home to tumors, tumor vasculature and tissues can be identified from phage display libraries in whole animals,^{51, 52} it seems that similar approaches for aptamers are likely on the horizon.

In addition to advances in aptamer selections and aptamer-mediated delivery, new mechanisms for function are also beginning to emerge. For example, while aptamers have historically been used as antagonists to their targets, the Sullenger lab has recently demonstrated the use of an aptamer as an agonist. Using an aptamer targeting the cell surface receptor OX40, a TNF family receptor that functions in maintaining T-cell proliferation, the group showed that while one aptamer was inadequate for stimulating the receptor and maintaining T cell activation, aptamer dimers were capable of enhancing T cell activation both in tissue culture and when injected into mice.⁵³ The aptamer dimers were generated via molecular assembly using DNA-based scaffolds that would hybridize to portions of the anti-OX40 aptamer. In this manner, these affinity agents were readily transformed into activators. Similar effects have previously been seen when comparing the function of antibody fragments which bind a single receptor with whole antibodies which can crosslink receptors. The ease with which aptamers can readily and modularly be assembled presents a new paradigm for combinatorial drug design and aptamer engineering.

The progress in aptamer development since its inception in the Gold and Szostak labs has been rapid and impressive. With one aptamer drug already FDA approved (Macugen) and many more in clinical trials, aptamers have already demonstrated therapeutic use. ^1–5 Additionally, new methods of generating aptamers as well as refinements in the traditional selections are currently ongoing. Some of these include robotically performed selections,⁵⁴ FACS-based selections,²⁶ CE-based selections,⁵⁵ and single bead selections.^{56, 57} The newest applications of aptamers, only some of which have been reviewed here, and leading edge technologies that are being created will play significant roles in RNA-mediated biotechnology.