This is a PDF file of a peer-reviewed paper that has been accepted for publication. Although unedited, the content has been subjected to preliminary formatting. Nature is providing this early version of the typeset paper as a service to our authors and readers. The text and figures will undergo copyediting and a proof review before the paper is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers apply.

The increasing importance of in vitro -transcribed (IVT) mRNA for synthesizing the encoded therapeutic protein in vivo demands the manufacturing of pure mRNA products. The major contaminant in the IVT mRNA is double-stranded RNA (dsRNA), a transcriptional by-product that can be removed only by burdensome procedure requiring special instrumentation and generating hazardous waste. Here we present an alternative simple, fast, and cost-effective method involving only standard laboratory techniques. The purification of IVT mRNA is based on the selective binding of dsRNA to cellulose in an ethanol-containing buffer. We demonstrate that at least 90% of the dsRNA contaminants can be removed with a good, >65% recovery rate, regardless of the length, coding sequence, and nucleoside composition of the IVT mRNA. The procedure is scalable; purification of microgram or milligram amounts of IVT mRNA is achievable. Evaluating the impact of the mRNA purification in vivo in mice, increased translation could be measured for the administered transcripts, including the 1-methylpseudouridine-containing IVT mRNA, which no longer induced interferon (IFN)-α. The cellulose-based removal of dsRNA contaminants is an effective, reliable, and safe method to obtain highly pure IVT mRNA suitable for in vivo applications.

No abstract

BNT162b2, a lipid nanoparticle (LNP) formulated nucleoside-modified messenger RNA (mRNA) encoding the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein (S) stabilized in the prefusion conformation, has demonstrated 95% efficacy to prevent coronavirus disease 2019 (COVID-19). Recently, we reported preliminary BNT162b2 safety and antibody response data from an ongoing placebo-controlled, observer-blinded phase 1/2 vaccine trial1. We present here antibody and T cell responses from a second, non-randomized open-label phase 1/2 trial in healthy adults, 19-55 years of age, after BNT162b2 prime/boost vaccination at 1 to 30 µg dose levels. BNT162b2 elicited strong antibody responses, with S-binding IgG concentrations above those in a COVID-19 human convalescent sample (HCS) panel. Day 29 (7 days post-boost) SARS-CoV-2 serum 50% neutralising geometric mean titers were 0.3-fold (1 µg) to 3.3-fold (30 µg) those of the HCS panel. The BNT162b2-elicited sera neutralised pseudoviruses with diverse SARS-CoV-2 S variants. Concurrently, in most participants, S-specific CD8+ and T helper type 1 (TH1) CD4+ T cells had expanded, with a high fraction producing interferon-γ (IFNγ). Using peptide MHC multimers, the epitopes recognised by several BNT162b2-induced CD8+ T cells when presented on frequent MHC alleles were identified. CD8+ T cells were shown to be of the early-differentiated effector-memory phenotype, with single specificities reaching 0.01-3% of circulating CD8+ T cells. In summary, vaccination with BNT162b2 at well tolerated doses elicits a combined adaptive humoral and cellular immune response, which together may contribute to protection against COVID-19.

In vivo genome editing using nuclease-encoding mRNA corrects SP-B deficiency (2015) Nature Biotechnology, 33 (6), pp. 584-586.In vivo genome editing using nuclease-encoding mRNA corrects SP-B deficiencyTo the Editor:Nuclease-mediated genome editing holds great potential to knock out or repair diseasecausing genes. An ideal nuclease delivery vehicle is short-lived, does not integrate into the genome, and can enter target cells efficiently. These requirements have not yet been achieved simultaneously by any nuclease delivery vector. We and others have used modified mRNA, which is non-integrating and provides a transient pulse of protein expression, as an alternative to traditional viral vectors [1][2][3][4][5] . This approach allowed us to deliver therapeutic proteins in mouse models of Surfactant Protein B (SP-B) deficiency 3 and experimental asthma 4 . Here we apply it to deliver site-specific nucleases, demonstrating the value of nuclease-encoding chemically modified (nec) mRNA as a tool for in vivo genome editing. We chose a well-established transgenic mouse model of SP-B deficiency 6 in which SP-B cDNA is under the control of a tetracycline-inducible promoter 7 . Administration of doxycycline drives SP-B expression levels similar to those in wild-type mice (Supplementary Fig. 1), whereas cessation of doxycycline leads to phenotypic changes similar to those of the human disease, including thickened alveolar walls, heavy cellular infiltration, increased macrophages and neutrophils, interstitial edema, augmented cytokines in the lavage, a decline in lung function, and fatal respiratory distress leading to death within days 8,9 . We inserted a constitutive CAG promoter immediately upstream of the SP-B cDNA to allow doxycycline-independent expression and prolonged life in treated mice.First, we customized a panel of ZFNs and TALENs targeting the transgenic SP-B cassette ( Fig. 1a and Supplementary Fig. 2). We chose TALEN #1 (T1) and ZFN #3 (Z3) owing to their high activity and proximity to the desired site of promoter integration (Figs. 1a,b; amino acid sequences in Supplementary Fig. 4) and compared delivery by plasmid 1 DNA and mRNA. mRNA delivery resulted in higher levels of double-strand break (DSB)-induction ( Fig. 1c and Supplementary Fig. 3; P < 0.05) and homology-directed repair (HDR) ( Fig. 1d, P < 0.05). As Z3 mRNA was more efficient than T1 mRNA in both cases, Z3 was chosen for further experimentation. Comparison with a Z3-encoding AAV serotype 6 vector (AAV6) ("Z3 AAV") shows the relatively transient expression of Z3 mRNA (Fig. 1e), limiting the time during which off-target cleavage activity could occur.To optimize Z3 expression in the mouse lung, we administered a panel of 3xFLAG-tagged Z3 mRNAs with various modification schemes 2,5,10 , with or without complexation to biocompatible, biodegradable nanoparticles (NPs) made of chitosan-coated poly (lactic-coglycolic) acid (chit-PLGA) 11,12 . Following intratracheal (i.t.) delivery, NP-complexation significantly increased mRNA expression levels ( Supplem...