The Influence Of Guest Sequence On Heterologous RNA Encapsulation By An Engineered Non-viral Protein Capsid

RNA plays important roles in many biological processes,such as cell proliferation,differentiation,cell cycle,invasion,and apoptosis.Therefore,a novel scenario for treating diseases at the RNA level,especially for cancers that are hard to cure using conventional drugs,has been put forward.RNA-based therapeutics,including small interfering RNAs and antisense oligonucleotides,enable targeted regulation of gene expression to achieve the aim of treating diseases.The attractive features of RNA-based therapeutics are high specificity,high efficiency,high flexibility,and low immune responses.However,many challenges accompany their use and limit their usefulness,such as low stability,poor tissue distribution,and inefficient cell uptake.Simple RNA viruses,which are composed of an RNA genome and a protein capsid,provide inspiration for the development of systems for the delivery of RNA drugs.The viral capsid acts as a nanocontainer that protects the genome from degradation and delivers it into a new host cell.The assembly of new virus particles requires encapsulation of the viral RNA by the virus capsid,which is a challenging molecular recognition problem given the high abundance of cellular RNAs in the local environment.Nevertheless,virus capsids preferentially package their cognate RNAs over the host cell RNAs.An understanding of how virus capsids recognize their own genomes could lead to improved methods for loading drug delivery vehicles with therapeutic RNAs.Viral protein capsids have been widely used as vessels for RNA storage and delivery.However,the use of viral capsids for drug delivery entails ineluctable risks,including the possibility of infection(for human viruses)and difficulty in controlling and predicting their behaviors.A previous study has shown that an engineered non-viral protein capsid called Aquifex aeolicus lumazine synthase(Aa LS)can be engineered to act as a container for RNA.The Aa LS capsid has a symmetric dodecahedral structure that is assembled from 60 identical subunits.This scaffold presents an appealing alternative to viral capsids,due to its simpler structure,extremely high stability,and efficient production in Escherichia coli.A variant of Aa LS,called Aa LS-pos,was engineered to possess a positively supercharged interior surface,which endowed this capsid with the ability to encapsulate RNA via charge complementarity.Production of Aa LS-pos in E.coli leads to the formation of encapsulation complexes with a variety of endogenous cellular RNAs.However,guest selection by Aa LS-pos is not completely random.A distinct guest size limit exists which is determined by the finite space available in the hollow interior of the Aa LS-pos capsid.RNAs larger than ～400 nt are excluded from the capsid.Below the guest size cut-off,a different pattern is observed for the most abundant RNAs found in the capsid compared to the whole cell.Given that all RNAs have the same density of negative charge,there must be some additional factors that account for the non-random distribution of cellular RNA guests below the size limit.We hypothesize that base sequence represents an important additional factor for determining the RNA guest preferences of Aa LS-pos.In order to determine whether guest sequence influences RNA encapsulation by Aa LS-pos,eight reporter RNAs,which have constant length but contain large regions of highly diverse sequence,were constructed.For RNA 1-1,2-1,3-1 and 4-1,each encoding DNA consisted of a T7 promoter,followed by an arbitrary sequence region(ASR 1-4),the Broccoli-F30 aptamer,then a linker and ending with the T7 terminator.For RNA 2-2,3-2 and 4-2,the Broccoli-F30 aptamer was replaced by an identical ASR 6.For RNA 5,a different and longer arbitrary sequence(ASR 5)was inserted that spanned both the original ASR site and the Broccoli-F30 site.Encapsulation complexes were generated via co-production of reporter RNA and Aa LS-pos in E.coli cells.q-RT-PCR analysis showed significant differences in the yields of full-length reporter RNA encapsulated by Aa LS-pos.The encapsulation yield of the best reporter RNA guest is higher than the worst by up to ～200-fold.In their partitioning between the capsid and the bulk cellular environment,the reporter RNAs with an alternative sequence(ASR 6)in place of Broccoli-F30 aptamer exhibited higher enrichments in Aa LS-pos than those with Broccoli-F30 aptamer,and the observed enrichment factors differed by up to ～28,000-fold,suggesting that the stable structure of the Broccoli-F30 aptamer may hamper RNA-protein complex formation resulting in a lower encapsulation yield.Four of the reporter RNAs had the Broccoli-F30 aptamer,which enabled fluorescence measurements as an additional method to estimate the encapsulation yields.The fluorescence assay indicated higher levels than those found by q-RT-PCR,suggesting that reporter RNAs are degraded and partially degraded forms are more frequently encapsulated by Aa LS-pos than their full-length counterparts.Further,the encapsulation yields of the full-length reporter RNAs show inverse correlations with both expression level and the secondary structural stabilities of the ASRs.These findings confirm that guest sequence has a substantial influence on RNA encapsulation.I propose that this influence is manifested through conformational flexibility of the RNA and three-dimensional shape complementarity with the capsid.These sequence-dependent properties likely determine the extent to which an RNA guest can maximize favorable charge-pairing interactions with the interior surface of Aa LS-pos.Knowledge of sequence-dependent properties may be helpful for furthering our understanding of how viruses encapsulate their own genomes in cells and be conducive to the development of novel RNA delivery systems.