Synonymous codon usage regulates gene expression such that transcripts rich in optimal codons produce significantly more protein than their nonoptimal counterparts. A major unresolved issue has been understanding the mechanisms by which synonymous codons regulate gene expression. We and others have previously shown that nonoptimal codons slow translation elongation speeds and thereby trigger mRNA degradation. However, differences in transcript abundance are not always sufficient to explain differences in protein levels, suggesting there are additional mechanisms by which codon usage regulates gene expression. Using reporter assays in human and Drosophila cells, we found that transcript levels account for less than half of the variation in protein abundance. We demonstrate that the differences at the protein level are not attributable to either protein folding or stability. Instead, we find that mRNAs with nonoptimal codons are bound by fewer ribosomes and that nonoptimal codon usage represses translation initiation. Nonoptimal transcripts are also less bound by the key translation initiation factors eIF4E and eIF4G, providing a mechanistic explanation for their reduced initiation rates. Our results reveal a new mechanism of regulation by codon usage, where nonoptimal codons repress further rounds of translation.

Since the initial reported discovery of SARS-CoV-2 in late 2019, genomic surveillance has been an important tool to understand its transmission and evolution. Here, we sought to describe the underlying regional phylodynamics before and during a rapid spreading event that was documented by surveillance protocols of the United States Air Force Academy (USAFA) in late October-November of 2020. We used replicate long-read sequencing on Colorado SARS-CoV-2 genomes collected July through November 2020 at the University of Colorado Anschutz Medical campus in Aurora and the United States Air Force Academy in Colorado Springs. Replicate sequencing allowed rigorous validation of variation and placement in a phylogenetic relatedness network. We focus on describing the phylodynamics of a lineage that likely originated in the local Colorado Springs community and expanded rapidly over the course of two months in an outbreak within the well-controlled environment of the United States Air Force Academy. Divergence estimates from sampling dates indicate that the SARS-CoV-2 lineage associated with this rapid expansion event originated in late October 2020. These results are in agreement with transmission pathways inferred by the United States Air Force Academy, and provide a window into the evolutionary process and transmission dynamics of a potentially dangerous but ultimately contained variant.

Since the initial reported discovery of SARS-CoV-2 in late 2019, genomic surveillance has been an important tool to understand its transmission and evolution. Here, we describe a case study of genomic sequencing of Colorado SARS-CoV-2 samples collected August through November 2020 at the University of Colorado Anschutz Medical campus in Aurora and the United States Air Force Academy in Colorado Springs. We obtained nearly complete sequences for 44 genomes, inferred ancestral sequences shared among these local samples, and used NextStrain variant and clade frequency monitoring in North America to place the Colorado sequences into their continental context. Furthermore, we describe genomic monitoring of a lineage that likely originated in the local Colorado Springs community and expanded rapidly over the course of two months in an outbreak within the well-controlled environment of the United States Air Force Academy. This variant contained a number of amino acid-altering mutations that may have contributed to its spread, but it appears to have been controlled using extensive contact tracing and strict quarantine protocols. The genome sequencing allowed validation of the transmission pathways inferred by the United States Air Force Academy and provides a window into the evolutionary process and transmission dynamics of a potentially dangerous but ultimately contained variant.