The partitioning of translation to the outer membrane of the endoplasmic reticulum is a problem that has been the subject of inquiry since the discovery of the ribosome. The large degree to which ribosomes were found to be tethered to the membrane led to intense investigation of a series of related questions regarding the identity of those mRNAs that are translated on the endoplasmic reticulum, and the functions of that localization in cell stress. In this dissertation, I approach each of these questions in turn and work to reconcile my observations with those models that have been previously proposed. A theme of this work is the application of modern methods, particularly deep sequencing technology, to address problems that had largely been considered solved. The most prominently featured method is ribosome profiling, which is paired with classical biochemical and cell biological techniques. I arrive at several conclusions: 1) a significant fraction of all mRNAs is well represented on the endoplasmic reticulum membrane, 2) the properties of translation diverge substantially between membrane-associated and free ribosomes, and 3) the compartmentalization of translation can serve as an important variable in cell stress.
