PSMA is a well-characterized transmembrane protein, which is strongly expressed in human prostate cancer and the vascular endothelium 95 96 . Importantly, PSMA is continually recycled from the plasma membrane and is constitutively endocytosed in PSMA-positive LNCaP cells, making it an attractive portal to deliver molecules intracellularly 66 . An anti-PSMA monoclonal antibody was demonstrated to promote the internalization rate. Using a purified fusion target protein containing a modified extracellular form of PSMA, Lupold et al. previously selected from an RNA library two 2'-fluoro (2'-F)-modified RNase-resistant RNA aptamers (A-9 and A-10) with low nanomolar affinity binding constants 78 . They also quantified the affinity of each aptamer for PSMA by measuring the inhibition of N-acetylated α-linked acidic dipeptidase (NAALADase) activity. Aptamer A-9 inhibited PSMA noncompetitively with an average K_i of 2.1 nM, whereas aptamer A-10 inhibited competitively with an average K_i of 11.9 nM. Because these anti-PSMA aptamers can be internalized, they have recently been engineered for cell-type specific delivery of various cargoes 53 54 61 , such as chemotherapeutic agents, drug-encapsulated nanoparticles, toxins, enzymes and siRNAs.

Three independent groups have successfully employed the anti-PSMA RNA aptamers to specifically deliver siRNAs to target cells (Figure 1). In a proof of concept study 49 , Chu et al. recently reported successful non-covalent conjugation of biotinylated anti-PSMA aptamer (A-9) with biotinylated 27-mer lamin A/C or GAPDH siRNAs via a modular streptavidin connector (Figure 1a). To enhance siRNA release in the cytoplasm, a reducible disulfide linker was designed between the sense strand of siRNA and the biotin group. By using such a streptavidin connector, two aptamers and two siRNAs were elegantly assembled into a multivalent construct, displaying effective PSMA receptor-mediated internalization of aptamer-siRNAs and specific silencing of the targeted transcripts in tumor cells.

A somewhat different approach was developed by Giangrande and colleagues 18 , in which a 2'-F-modified anti-PSMA aptamer (A-10) was covalently appended to the sense strand of a 21-mer siRNA portion, which in turn was hybridized to the 21-mer antisense strand (Figure 1b). The resulting aptamer-siRNA chimeric RNA was shown to be selectively internalized into cells expressing PSMA, and to effectively knock down expression of the targeted the tumor survival genes (PLK1 and BCL2) both in cell culture and in vivo after intratumoral delivery. Because this delivery system consists only of RNA components, it offers several potential advantages as a therapeutic agent, including lack of immunogenicity, the possibility for chemical synthesis, and stabilizing modifications for in vivo application.

Although aptamer-siRNA chimeras can be directly administered to localized targets (for example, intratumoral delivery), systemic administration will be required for many diseases. In general, greater therapeutic doses are required for systemic administration, leading to higher costs and side effects. Most recently, Giangrande and colleagues addressed this issue by optimizing their previous PSMA-siRNA chimeric design to achieve enhanced inhibition of prostate cancer xenograft growth via systemic administration (Figure 1c) 17 . First, the aptamer portion of the PSMA A10-Plk1 chimera was truncated from 71 to 39 nucleotides, while still maintaining high binding affinity. Most importantly, the truncated version (containing the aptamer and sense strand of the siRNA) comprising a total of 64 nucleotides makes this amenable to chemical synthesis. Second, the silencing potency was enhanced through structural modifications of the siRNA portion, enabling more efficient incorporation of the siRNA by the cellular RNAi machinery. The group then added a 2-nucleotide (UU) overhang at the 3' end of the siRNA duplex, but also swapped the positions of the passenger and guide strands of the siRNA. These modifications favor Dicer recognition and loading of the guide strand (containing the two base 3' overhang) into an RNA-induced silencing complex (RISC), hence increasing the silencing activity and specificity. By appending a polyethylene glycol moiety with molecular weight 20 kDa onto the siRNA passenger, the circulating half-life of the chimeric molecule was substantially increased and the bioavailability was markedly improved, leading to prolonged silencing in vivo. As a result of these efforts, the optimized second-generation aptamer-siRNA chimeras (Figure 1c) resulted in pronounced regression of PSMA-expressing tumors after systemic administration in athymic mice. Additionally, the therapeutic dose of the new chimera was dramatically reduced from 1 nmol on each of 10 successive days (10 × 1 nmol) to 0.25 nmol in every other day for a total of 10 days (5 × 0.25 nmol), minimizing both the cost of treatment and the risk of harmful side effects.

Other efforts to further refine aptamer-mediated siRNA delivery and targeting efficiency are being attempted through multimerization of the aptamer portion. Previous studies with aptamers have revealed that multivalent versions of aptamers can increase the potency and antitumor response, and promote receptor activation 97 98 99 100 . The multivalent aptamer-siRNA construct has also been recently exploited for facilitating receptor internalization, further improving the therapeutic potential. Wullner et al. generated two different bivalent anti-PSMA aptamer-siRNA chimeras in which the siRNAs targeted eukaryotic elongation factor 2 77 . Their modifications included using the siRNA itself as a linker to join the two aptamers or appending the siRNAs onto the 3' ends of each aptamer. Compared with the monovalent aptamer-siRNA chimeras (55% target knockdown), these bivalent aptamer-siRNA constructs resulted in an almost complete loss of PSMA-positive cell viability, suggesting that bivalent aptamers definitely promote internalization of chimeras. These efforts have encouraged new thinking in the design of multiple aptamer-siRNA conjugates.

The CD4 receptor, a glycoprotein expressed on the surface of certain subsets of T lymphocytes 101 102 103 , is a primary receptor used by HIV-1 to gain entry into host T cells. It was previously reported that overexpressed CD4 protein in T helper cells can be endocytosed 104 . Aptamers targeting CD4 were produced by immobilizing soluble, recombinant CD4 antigen onto Sepharose beads, allowing elution of unbound oligonucleotides and retention of bound species, which were further amplified for the next selection rounds 89 . Using this approach, 2'-F-modified RNA aptamers with high CD4 affinity were identified. The ability of CD4 aptamers to block functional T cell responses was tested using an allogeneic mixed lymphocyte reaction (MLR), a complex in vitro assay of T-cell recognition and responsiveness, in which the comparative standard is the W3/25 CD4 monoclonal antibody, which binds to the same site as the tested aptamer clones. Thus, the ability to block MLR correlates with CD4 binding activity. These aptamers showed inhibitory effects in a CD4-specific manner. This CD4-specific aptamer has been assembled into a multifunctional nano-device for targeted delivery of siRNAs in a T-cell line engineered to overexpress CD4 22 50 .

Recently, anti-CD4 RNA aptamers have been exploited for targeted delivery of siRNAs 22 105 . The self-assembling bacteriophage phi29 RNA (pRNA) was joined with the anti-CD4 aptamer and allowed to form a nano-complex with a pRNA-siRNA chimera. It was previously demonstrated that pRNAs can be accurately assembled through interlocking right- and left-hand loops into various oligomers (dimer, trimer, hexamer) ranging in size from nanometers to micrometers 51 106 . The pRNA itself can be fused with various agents (folate, aptamer, siRNA, dye, antitumor drugs), while still allowing oligomerization of the pRNAs 107 . For example, as shown in Figure 2, two pRNA molecules were respectively fused with siRNAs (against survivin, green fluorescent protein (GFP), Bcl2 antagonist of cell death (BAD) or luciferase) and the anti-CD4 aptamer. Through the interaction of right and left interlocking loops, the two chimeric pRNAs could be precisely dimerized into a stable nanovector of approximately 25 nm in diameter. The nano-scale RNA dimer was also shown to be effectively internalized into a CD4-overexpressing T cell line, and the siRNAs consequently knocked down the expression levels of the targeted surviving or enhanced GFP mRNAs. The direct correlation between CD4 expression level and the internalization/silencing activity of the siRNAs also provided proof of anti-CD4 aptamer-mediated cell-specific siRNA delivery. Similarly, a trimeric conjugate was engineered in the same way. Three chimeric pRNA building blocks (one fused with the CD4 aptamer, another with an siRNA and a third with a fluorescent molecule) were assembled into a multifunctional nano-device, which elicited siRNA-mediated target knockdown and also provided molecular imaging via the fluorescent dye. This self-assembling nano device may improve the in vivo kinetics and enhance the therapeutic efficacies of the delivered siRNAs.

The HIV-1-encoded gp120 protein, a glycoprotein envelope on the surface of HIV-1, plays an important role in viral entry into CD4 cells 101 103 . The interaction of gp120 and CD4 triggers HIV-1 entry and initiates cell fusion 108 109 110 111 . Recently, a chimeric Fab gp120 antibody fragment-protamine fusion was demonstrated to facilitate receptor-specific siRNA uptake into cells expressing the HIV-1 gp120 envelope protein, suggesting this protein as a new molecular target for receptor-mediated siRNA delivery 40 . Several 2'-F modified anti-HIV gp120 RNA aptamers have been isolated with the use of a BIAcore biosensor system (Stevenage, United Kingdom) 90 91 112 113 or conventional nitrocellulose filter binding of aptamers to recombinant proteins 20 . The selected aptamers can specifically bind to and be rapidly internalized into cells expressing the HIV-1 envelope protein. In addition, the aptamers alone can neutralize HIV-1 infectivity.

Recently, we used gp120 aptamer-siRNA chimeras for cell type-specific delivery of siRNAs in cultured cells and in a HIV-1 infected Rag-Hu mouse model. The gp-120 aptamer (Figure 3a), was covalently linked to siRNAs that target the HIV-1 tat/rev common exon 19 . Because both the aptamer and the siRNA can inhibit HIV-1 replication by respectively blocking the gp120-CD4 receptor interaction and silencing HIV-1 tat/rev expression, this novel anti-gp120 aptamer-siRNA chimera possesses a dual inhibitory function. Treatment of HIV-1-infected cells with these chimeras resulted in the selective gp120-mediated internalization of the aptamer-siRNA by endocytosis and the specific silencing of the targeted mRNA transcript. Interestingly, a small change in the length of the siRNA portion of the chimera from 21bp to 27bp resulted in enhanced silencing potency. This was the result of Dicer processing of the 27-mer from the aptamer and perhaps a more efficient handoff of the processed siRNA to RISC. These results demonstrated that HIV-1 gp120 expressed on the surface of HIV-1-infected cells represents a unique target for aptamer-mediated siRNA delivery.

In a study by Zhou et al. 20 , a 'sticky bridge' strategy was developed to non-covalently conjugate the aptamer with various siRNAs (Figure 3b). In this design format, one pair of complementary GC-rich sticky bridge sequences was chemically attached to the 3' end of the aptamer. The complement to this sequence was attached to one of the two siRNA, strands and the aptamer and siRNA were joined by Watson-Crick base pairing. A flexible three-carbon atom hinge (C3) was added as a spacer between the adhesive (sticky) sequence and the aptamer to allow spatial and structural flexibility. Importantly, this sticky bridge-based strategy can be used to facilitate the effective interchange of different siRNAs with a single aptamer, which is required to avert viral resistance to the siRNA component. We combined three different siRNAs with the gp120 aptamer: one against the HIV-1 tat/rev gene, and two siRNAs targeting the HIV host dependency factors CD4 and transportin 3, respectively. The specific binding and internalization of the aptamer-siRNA conjugates into gp120-expressing cells was demonstrated by confocal microscopy, and the aptamer-'sticky bridge'-siRNA combinations downregulated targeted gene expression and suppressed HIV replication in cell culture. Additionally, the aptamer-siRNA combinations also served as dual-function inhibitors, providing additive efficacy. These results demonstrated the potential use of aptamer-siRNA conjugates as a systemic, cell type-specific, siRNA cocktail delivery system for anti-HIV-1 therapy.

Most recently, we tested the anti-HIV efficacy of these aptamer-siRNA dual inhibitors in a humanized mouse model (P. Neff et al., manuscript submitted to Science Translational Medicine). In this model system, the humanized Rag2^-/-γc^-/- mice (RAG-hu) were treated with human CD34 hematopoietic progenitor cells, which engraft and differentiate in a variety of human hematopoietic lineages. The mature T cells and monocytes were infected with the HIV-1 NL4.3 virus. After 3 weeks of viral replication, the animals were injected intravenously once weekly with the aptamer-siRNA conjugates. We observed a dramatic decrease in viral load in all the treated animals, in most cases to undetectable levels within a week after the intravenous administration of the chimera. The suppression of viral load averaged three logs of reduction relative to controls, and persisted throughout and beyond the treatment period in several of the animals. Most importantly, the aptamer-siRNA treatment completely prevented T-cell depletion mediated by viral infection. Therefore, the capacity to achieve marked viral suppression in vivo together with restoration of CD4 T cell levels using aptamer-siRNA constructs should pave the way for implementing novel therapeutic strategies for treating HIV disease. In particular, these dual-action constructs will be useful for treatment of patients who do not respond to highly active anti-retroviral therapy, the standard multi-drug treatment that has proved so effective in battling AIDS.