Silencing disease genes in the laboratory and the clinic

Abstract

Synthetic nucleic acids are commonly used laboratory tools for modulating gene expression and have the potential to be widely used in the clinic. Progress towards nucleic acid drugs, however, has been slow and many challenges remain to be overcome before their full impact on patient care can be understood. Antisense oligonucleotides (ASOs) and small interfering RNAs (siRNAs) are the two most widely used strategies for silencing gene expression. We first describe these two approaches and contrast their relative strengths and weaknesses for laboratory applications. We then review the choices faced during development of clinical candidates and the current state of clinical trials. Attitudes towards clinical development of nucleic acid silencing strategies have repeatedly swung from optimism to depression during the past 20 years. Our goal is to provide the information needed to design robust studies with oligonucleotides, making use of the strengths of each oligonucleotide technology.

Figure 1

Comparison of the ASO and siRNA mechanisms. (A) ASOs must be stable as single-stranded oligonucleotides and find their target alone. (B) siRNAs are delivered as duplexes, then taken up by Argonaute (AGO), part of the RNA-induced silencing complex. Thus the “antisense oligonucleotide” that guides AGO to complementary mRNA is always present in the cell as a duplex or as part of a protein complex. Note that RISC also engages in complex biology with endogenous small RNAs [71].

Figure 2

Structures of some common chemically modified nucleotides. (A) Replacement of a nonbridging phosphate oxygen with sulfur gives the phosphorothioate (PS) linkage, which dramatically increases nuclease stability. (B) Various sugar modifications are compatible with unmodified DNA/RNA synthesis and increase binding affinity and nuclease stability. (C) PNA and PMO modifications have a neutral backbone structure that is dramatically different from the sugar-phosphate backbone of natural oligonucleotides.

Figure 3

Typical mismatched and scrambled controls. (A) A duplex with several pairs of bases exchanged (from one strand to the other) should maintain very similar properties to the parent duplex but lose affinity for its target mRNA. (B) A scrambled control can be made by moving blocks of bases with respect to the parent oligomer.

Figure 4

Mechanism of action of Mipomersen.

Figure 5

Mode of action of drug candidates PRO051 and AVI-4568. (A) This DMD patient is missing dystrophin exon 50. Splicing of the pre-mRNA gives mature mRNA that is out-of-frame, and so no functional dystrophin can be produced. (B) In the presence of a splice-switching ASO that favors exclusion of exon 51, the cell splices exon 49 to exon 52, which restores the reading frame and causes translation of a shorter but partially functional dystrophin protein.

Figure 6

miR-122 is a liver-specific miRNA that regulates multiple pathways. Therapeutic inhibition of miR-122 by Miravirsen blocks HCV replication and lowers plasma cholesterol.