Rational design of chemical genetic probes of RNA function and lead therapeutics targeting repeating transcripts

Abstract

RNA is an important yet vastly underexploited target for small molecule chemical probes or lead therapeutics. Small molecules have been used successfully to modulate the function of the bacterial ribosome, viral RNAs and riboswitches. These RNAs are either highly expressed or can be targeted using substrate mimicry, a mainstay in the design of enzyme inhibitors. However, most cellular RNAs are neither highly expressed nor have a lead small molecule inhibitor, a significant challenge for drug discovery efforts. Herein, I describe the design of small molecules targeting expanded repeating transcripts that cause myotonic muscular dystrophy (DM). These test cases illustrate the challenges of designing small molecules that target RNA and the advantages of targeting repeating transcripts. Lastly, I discuss how small molecules might be more advantageous than oligonucleotides for targeting RNA.

FIGURE 1

Designing small molecules targeting RNA is challenging because there are few lead compounds. A large data set of leads is likely required to provide sufficient information to reveal general routes to target RNA because of its structural diversity. (a) Targeting approaches used for DNA were enabled, in part, by the structures of natural products, such as distamycin, bound to the minor groove. (b) Examples of the structural diversity of RNA in validated drug targets. (c) RNA forms many different structures that can be utilized as binding sties for small molecules, as is illustrated in the structures of RNA base pairs and examples of 1 × 1 nt RNA internal loops. Abbreviation: TAR, trans-activation response.

FIGURE 2

Development of small molecules that target the RNAs that cause myotonic dystrophy (DM). (a) Myotonic dystrophy type 1 (DM1) is caused by a r(CUG) repeat expansion [r(CUG)^exp] that inactivates muscleblind-like 1 protein (MBNL1). Inactivation of MBNL1 causes dysregulation of pre-mRNA splicing. A modularly assembled compound that displays a bis-benzimidazole (H, Fig. 3) improves DM1-associated defects in cell culture models of disease [25,63]. (b) DM2 is caused by an expanded r(CCUG) repeat [r(CCUG)^exp], which also binds and inactivates MBNL1. A modularly assembled compound that displays K inhibits the r(CCUG)^exp–MBNL1 complex (Fig. 3) [77]. (c) Identification of a lead small molecule to target the r(CCUG)^exp. In two-dimensional combinatorial screening (2DCS) selections, the randomized regions of the RNA libraries are kept intentionally small such that they are found as components of cellular RNAs. The RNA motif libraries are then screened for binding small molecules. Collectively, these studies identified a set of RNA motif–small molecule interactions that can be used to design compounds against RNAs that cause or contribute to disease, such as DM2. In a resin-based selection, kanamycin A was used, which led to the development of the derivative K to target r(CCUG) repeats by using a modular assembly approach [24,77].

FIGURE 3

Rationally designed, modularly assembled small molecules targeting the causative agents myotonic dystrophy type 1 (DM1) and type 2 (DM2), r(CUG)^exp and r(CCUG)^exp, respectively [24,63,77]. RNA-binding modules (H or K) were modularly assembled onto a peptoid backbone and are potent inhibitors of the r(CUG)^exp–MBNL1 complex (nH-4 compounds) or the r(CCUG)^exp–MBNL1 complex (nK-4 compounds). nH-4 compounds improved DM1-associated defects in cell culture models [25].

FIGURE 4

Small molecules that improve myotonic dystrophy type 1 (DM1)-associated defects in cellular and/or mouse models of disease. H1 was designed using chemical similarity searching [70], whereas Tri-Ac was designed using structure-aided design [55]. DCC-1 [72], pentamidine [71], 1 [75] and 2 [75] were identified via in vitro screening methods. D-Hexapeptide was identified via a screen in a Drosophila model of DM1 [76].