We have developed a new transporter structure that provides effective delivery ofMorpholino antisense oligomers into a wide variety of tissues in living mice. This transporter comprises a dendritic structure assembled around a triazine core which serves to position eight guani-dinium head groups in a conformation effective to penetrate cell membranes. This transporter structure is conjugated to a Morpholino oligomer to form a delivery-enabled product referred to as a Vivo-Morpholino. Vivo-Morpholinos are shown to effectively enter and function within cultured cells in the presence of 100% serum using a rigorous positive test system based on correction of a defined splicing error in a pre-messenger RNA. In addition, Vivo-Morpholinos are demonstrated to enter into a wide variety of tissues in a similar positive test system in trans genic mice, as evidenced by correction of the targeted splicing error in all tissues assessed, including near-complete splice correction in the small intestine, colon, stomach, liver, kidney, and a number of muscles. Finally, Vivo-Morpholinos, which target the exon-skipping ofexon 23 harboring a premature termination codon in the mdx mouse model, effectively restore the reading frame of dystrophin and restore expression of a functional dystrophin protein.

Introduction

Breaking the delivery barrier

Morpholino oligomers are a proven antisense reagent used to block translation or interfere with RNA processing, including splicing and miRNA maturation, with hundreds of successful applications primarily involving microinjection into embryos (zebrafish and frog). However, the introduction of unmodified Morpholino oligomers into in vivo systems has only been successful with leaky tissues such as damaged muscle. Systemic delivery of unmodified Morpholino oligomers into normal animals has not been successful, and delivery into tissue culture test systems requires the use of additional delivery reagents and low-serum conditions (1). We set out to develop a Morpholino oligomer transporter that could be coupled directly to Morpholino oligomers and facilitate system-wide delivery when introduced into the vascular system of animals.

Transporter design

The most developed in vivo delivery strategy employs peptides based on tat, penetratin, and, subsequently, poly-arginine peptides to achieve delivery of oligomers. However, each of these has one or more serious drawbacks including poor efficacy, expensive synthetic routes, and cumbersome or inefficient coupling strategies.

The early touted natural transporter peptides tat and penetratin have been found inefficient at delivering oligomers (2,3). The inefficiency of tat delivery across the cell membrane was thought to be due to a requirement for folding and subsequent renaturation of the protein during membrane trans-location (4,5). Penetratin coupled to steric-block peptide nucleic acid (PNA) oligomers (which, like Morpholinos, are uncharged and capable of interfering with RNA processing) has been shown to require 5–10 µM concentrations to achieve delivery into the cytosol but fails to deliver significantly to the nucleus (6,7). The inefficiency of tat and penetratin delivery to the nucleus is thought to be due to a majority of the materials entering cells via endocy-tosis and a subsequent inability to permeabilize the endocytic vesicles, as well as an inability to circumvent the electrostatic interactions with cellular heparin sulfates (8). In order to interfere with most RNA processing, and pre-messenger RNA (pre-mRNA) splicing in particular, an oligomer must be delivered to the nucleus.

A more promising delivery moiety evolved from these natural peptides, as the active components were defined as 6–9 arginine residues in a bio-available 6-aminohexanoic-spaced structure (9), with an optimal length defined as eight arginines (10). These arginine-based peptides with the 6-aminohexanoic–spaced oligoarginine (R-Ahx-R)×4 have successfully delivered Morpholinos to the nucleus and cytosol in vitro (11) and in vivo (12), including sustained induction of dystrophin expression in mdx mice (13). They have also been shown to actively or prophy-lactically knock down viral titers in various tissues of mice infected with corona virus (14), picornavirus (15), respiratory syncytial virus (16), and influenza A (17). However, it should be noted that arginine-based peptides are not generally available to the research community and that their greatest successes have been in delivering Morpholinos to the cytosol of what would be considered easily deliverable tissues like liver (18) or leaky muscle (19). The question of how well they deliver into a wide spectrum of tissues remains unanswered.

The active components of arginine peptides have been defined as the guanidinium head groups of the arginine residues; it has been further characterized that placing eight guanidinium head groups on a synthetic scaffold and coupling this scaffold to fluorescein leads to enhanced delivery as compared to arginine-based peptides (20,21). The guanidinium head groups have been predicted to interact with phosphates of phospholipids both by electrostatic attraction and multiple hydrogen bonds (22) and, as such, each guanidinium head group is capable of a charge-charge interaction plus hydrogen bonding.

We sought to exploit a nonlinear, non-peptidic and non-natural architecture for a transporter that provides cost-effective synthesis, convenient conjugation, and exhibits superior performance in transporting Morpholino oligomers across biological barriers. The design of our transporter is based on molecular transporters reported in the literature and the following factors. (i) Since guanidinium head groups of arginine-rich peptides are principally responsible for uptake into cells, we developed a synthesis scheme that makes use of a tri-functional triazine as a core scaffold to assemble and present guanidinium head groups in a nonlinear and economically feasible manner. (ii) Reports in the literature indicate that 7–15 guanidine head groups is optimal for efficient uptake (22) with 8 guanidine head groups exhibiting the most efficient internalization (10). We chose to install a total of eight guanidine head groups on two of the side chains of a triazine core, leaving the third for efficient on-column conjugation to the Morpholino oligomer with uncomplicated post-synthesis processing (Figure 1 and Reference 23). We call these novel, delivery-enabled Morpholinos Vivo-Morpholinos.