Targeted Delivery of Self-Amplifying RNA Genes

As a result of millions of years of evolution, what viruses do better than anything else - better than anything else they do, or any thing else does - is to protect and deliver genes to specific (targeted) cells. This is why it is impossible not to consider their use as a gene delivery systems for medical purposes. To avoid the complications and potential dangers of using "attenuated" or "inactivated" viruses, however, we pursue a research program that involves the design and synthesis - from purified components in test tubes - of virus-like particles (VLPs) that contain genes of interest in single-stranded RNA (ssRNA) form. The ssRNA form is essential for enabling spontaneously self-assembly of the gene into a VLP when mixed with the right viral capsid protein (CP); this isn't possible, for example, with double-stranded DNA (dsDNA). For self-assembly purposes, the "right" CP is one that will efficiently package any RNA, independent of its sequence, as long as it isn't too long or too short. The CP from tobacco mosaic virus (TMV) also works for this purpose (forming cylindrical capsids), after inserting a "packaging signal" sequence into the heterologous RNA of interest. We have found that CP from cowpea chlorotic mottle virus (CCMV) or brome mosaic virus (BMV) does this job. The other advantage of RNA genes is that it is possible to work with them in messenger-sense and self-amplifying form, enabling them to be strongly translated and replicated in the cytoplasm of targeted host cells before being translated into proteins of interest.

There are big differences between how a mammalian virus undergoes its life cycle in vitro and in vivo. By “in vitro” here we mean in sub-confluent-monolayer cell-culture plates, and by “in vivo” we mean in the context of its natural host animal. In the latter case the gene regulation network of the infected cell is strongly influenced by a wide range of physical and signaling interactions between cells, and of course by the systemic immune system of the host. A huge advantage of the former case is that one can control many important “input” parameters such as the “multiplicity of infection”, and quantify directly many “output” characteristics such as “burst size”, etc. A middle ground is provided by 3D reconstructions (“organoids”) of tissue which incorporate much of the three-dimensional cell-cell organization and yet still allow for control of viral input and output and for direct imaging and quantification. We are collaborating with Dr. John Mellnik at Path BioAnalytics – a biotech company specializing in lung organoids – in an effort to compare the activity of persistent Sindbis replicons in 2D and 3D cell culture. 

Chimeric antigen receptor (CAR) therapy is among the most exciting and promising approaches to cancer immunotherapy. The idea is to provide a patient with the particular antigen-specific cytotoxic T cells needed to kill their cancer. This is done by ex vivo transformation of the patient's T cells, i.e., by extracting T cells from the body, and then expanding/proliferating them, transforming them with gene-integrating vectors like lentiviruses to induce expression of the desired CARs or TCRs, and finally infusing them back into the body, With Prof. Otto Yang in the UCLA School of Medicine, we are pursuing an alternative approach that involves a combination of in vitro and in vivo approaches to T-cell therapy which avoids these costly and risky procedures. More explicitly, we: synthesize the mRNA encoding a particular CAR of interest; package it in vitro into virus-like particles involving one copy of the RNA molecule inside a 26-nm/icosahedrally-symmetric/180-subunit protein shell; and conjugate these VLPs with an antibody against the CD3 protein subunit in T-cell receptor complexes. These particles will be taken up by T cells in vivo and transform them into ones expressing the mRNA-encoded CAR. We are presently working on demonstrating this by incubating the VLPs with purified T cells, using a CAR that is specific against the CD19 antigen over-expressed in cancerous B cells or against the glycoprotein expressed in HIV-infected cells. The CAR expression in the resulting transformed cells is quantified by flow-cytometry, and cell-killing assays with CD19-transformed or HIV infected cells are carried out by incubating the T cells with the target cells. This work is joint with Prof. Otto Yang and Dr. Chris Hofmann in the UCLA School of Medicine. (A new approach to in vivo transformation of killer T cells; J. Mol. Biol. 437, 169369 (2025))

Cytotoxic ("killer") CD8+ T cells have evolved to kill cancer cells and virus-infected cells, and even to kill cells infected by intracellular bacteria. In all these cases, the underlying mechanism involves MC-I presentation of cancer, viral, or bacterial antigenic epitopes that are recognized by the T-cell receptor. To effectively treat extra-cellular bacteria like the antibiotic-resistant Staph aureus, however, we are developing mRNA-containing in vitro reconstituted virus-like particles (VLPs) to transform killer T cells with chimeric antigen receptors (CARs) against a bacterial-membrane-specific protein based on viral bacteriophage targets. This work involves a collaboration with Prof. Irene Chen, UCLA Chemical and Biomolecular Engineering and Prof. Otto Yang, UCLA School of Medicine.

In addition, in collaboration with Prof. Markus Mapara of the Columbia University Medical Center in New York, we are using similar mRNA-containing VLPs to transform macrophages with CARs against amyloid fibrils based on an antibody specific to these aggregates.

Many mRNA-based vaccines have been investigated for their specific potential to activate dendritic cells (DCs), the highly-specialized antigen-presenting cells of the immune system that play a key role in inducing effective CD4+ and CD8+ T-cell responses. We developed a new vaccine/gene delivery platform that demonstrates the benefits of using a self-amplifying ("replicon" mRNA that is protected in a viral-protein capsid. Purified capsid protein from the plant virus Cowpea Chlorotic Mottle Virus (CCMV) is used to in vitro assemble monodisperse virus-like particles (VLPs) containing reporter proteins (e.g., Luciferase or eYFP) or the tandem-repeat model antigen SINFEKL in RNA gene form, coupled to the RNA-dependent RNA polymerase from the Nodamura insect virus. Incubation of immature DCs with these VLPs results in increased activation of maturation markers - CD80, CD86 and MHC-II - and enhanced RNA replication levels, relative to incubation with unpackaged replicon mRNA. Higher RNA uptake/replication and enhanced DC activation were detected in a dose-dependent manner when the CCMV-VLPs were pre-incubated with anti-CCMV antibodies. In all experiments the expression of maturation markers correlates with the RNA levels of the DCs. Overall, these studies demonstrate that: VLP protection enhances mRNA uptake by DCs; coupling replicons to the gene of interest increases RNA and protein levels in the cell; and the presence of anti-VLP antibodies enhances mRNA levels and activation of DCs in vitro. In collaboration with Boehringer-Ingelheim Pharmaceuticals (Ridgefield, USA, Biberach, Germany, and Vienna, Austria) preliminary in vivo experiments involving mouse vaccinations with antigen-encoding-replicon VLPs indicate a small but significant increase in antigen-specific T cells that are doubly positive for IN and TFN induction [1]. Currently, in collaboration with Prof. Otto Yang (UCLA Department of Medicine, Infectious Diseases Division), we are extending this work by conjugating to the VLPs intact antigenic proteins corresponding to those encoded by the mRNA they contain, so that we can tune the relative strengths of antibody and T-cell responses.

In collaboration with Boehringer-Ingelheim Pharmaceuticals (Ridgefield, USA, Biberach, Germany, and Vienna, Austria)

[1] Biddlecome, A., Habte, H. H., McGrath, K. M, Sambanthamoorthy, S., Wurm, M., Sykora, M. M., Knobler, C. M., Lorenz, I. C., Lasaro, M., Elbers, K., Gelbart, W. M. (2019) Delivery of self-amplifying RNA vaccines in in vitro reconstituted virus-like particles. PLoS ONE, 14 (6): e0215031. 

We are making in vitro self-assembled virus-like particles (VLPs) made from Sindbis replicons and the structural proteins of TMV [1)] or of Sindbis virus itself [(2)].

(1) TMV was the very first virus to be reconstituted from its purified components, and - as with CCMV - its capsid protein is capable of packaging heterologous RNA as long as the TMV RNA "packaging signal" is inserted into the foreign RNA to ensure the nucleation of a capsid. The capsid that forms is a long hollow/helical cylinder of protein instead of an icosahedral spherical shell. Because the curvature (reciprocal radius) of the cylinder remains constant independent of its length, there is no limit to the length of heterologous RNA than can be packaged in this way. This fact allows us - in contrast to the situation with CCMV or BMV VLPs - to accommodate replicons of any length, like those of Sindbis, instead of being restricted to the 3X-shorter RdRps from insect viruses like Nodamura. One only needs to insert the TMV packaging signal in the RNA to be packaged. Further, in addition to making RNA-specific TMV particles in vitro, we can scale up their synthesis by making them in cellulo, by cotransfecting mammalian cells with two molecules, one containing the Sindbis replicon with the TMV packaging signal and an added (reporter or therapeutic) gene of interest, and the other a "helper molecule" that is replicated by the replicon's RdRp and that is translated to give TMV capsid protein.

(2) In a similar way, one can make Sindbis virus-like particles -- RNA-containing T=4 nucleocapsids wrapped by lipid bilayer whose trans-membrane heterodimer proteins bind one-to-one the capsid proteins -- by breaking up the bicistronic single-molecule Sindbis genome into two molecules, one a replicon containing the replicase-gene-related open reading frame (ORF) and the other a "helper" containing the ORF for the structural (capsid + membrane) proteins. 

Only the replicon, to which has been added a gene of interest, includes the packaging signal, and the enveloped particles secreted by the cell are identical to the infectious virions except for containing therapeutic RNA instead of infectious genome. RNA packaging specificity and of VLP structural integrity are enhanced by making the replicon sequence and length as close as possible to those of the wildtype genome.

In both cases -- (1) TMV, and (2) Sindbis) -- the RNA "purity" of the VLPs is checked by long-read RNA sequencing.

For many reasons, RNA replication is cytotoxic, whether or not the RNA being replicated is the infectious genome or a non-infectious form of it like the replicons we work with that involve the open reading frame for the RNA-replication-related proteins but not for any of the structural proteins. In the case of both the Nodamura and the Sindbis replicons, for example, the expression of reporter genes is much stronger than for reporter-gene mRNAs alone, and it peaks and dies out after a day or two instead of just a few hours. In an effort to learn about RNA replication, and to provide less cytotoxic/longer-lived amplification of mRNAs for protein expression, we are working with a mutant of the Sindbis replicon that contains a single point mutation in the active site of the replicase and that results in a 50-fold lower level of replication and -- most dramatically -- in a persistent level of replication. Further, the expression of the replicon survives the growth of the cells into a confluent monolayer and their passaging through many hundreds of generations. We are using deep sequencing of the Sindbis replicon and mass spectrometry determination of the host cell proteome, as a function of passaging number, to unravel how the virus "stand-in" (the persistent replicon) and the host cell manage to co-survive. This work is carried out in collaboration with Prof. Charles Rice, Rockefeller University, and with Prof. Joe LOo, UCLA Chemistry & Biochemistry.

During infection by positive-sense RNA-genome viruses, RNA replication occurs in cis when the RdRp replicates the mRNA encoding it, or in trans when the it acts on the other RNA molecules making up the viral genome or on subgenomic or defective interfering RNAs that arise in its life cycle. In order to better understand these processes we are studying the effects of combining replication-competent and replication-deficient replicons on RNA replication and on the expression of proteins encoded in genes of interest (GOls) incorporated into the replicons. We quantify the extent to which the competent replicon rescues replication of the deficient replicon in trans, leading to reduced overall RNA replication but to increased total GOl translation. These findings highlight the delicate balance between RNA replication and protein synthesis in the context of cellular cytotoxity and metabolic burden, and provide insights into the rational design of saRNA-based therapeutics.

In order for our virus-like particles to remain in circulation long enough to reach their target, or to be used in follow-up booster doses, they need to themselves to be sufficiently weakly immunogenic. In the case of TMV VLPs, conjugation of PEG molecules of different molecular weight and branching structure has been shown in mouse models to correlate with different circulation times for the VLPs and different kinds of immune responses involving B cells, T cells, and macrophages. We are exploring the related question: What is the effect of these different kinds of PEGylation on the extent to which the RNA packaged in these particles is made accessible to cellular ribosomal machinery, as quantified by reporter-gene-expression in cell culture transfections studies.