Newly synthesized proteins must form their native structures in the crowded environment of the cell, while avoiding non-native conformations that can lead to aggregation. Yet remarkably little is known about the progressive folding of polypeptide chains during chain synthesis by the ribosome, or of the influence of this folding environment on productive folding in vivo. P22 tailspike is a homotrimeric protein that is prone to aggregation via misfolding of its central β-helix domain in vitro. We have produced stalled ribosome:tailspike nascent chain complexes of four fixed lengths in vivo, in order to assess co-translational folding of newly synthesized tailspike chains as a function of chain length. Partially synthesized, ribosome-bound nascent tailspike chains populate stable conformations with some native-state structural features even prior to the appearance of the entire β-helix domain, regardless of the presence of the chaperone trigger factor, yet these conformations are distinct from the conformations of released, refolded tailspike truncations. These results suggest that organization of the aggregation-prone β-helix domain occurs co-translationally, prior to chain release, to a conformation that is distinct from the accessible energy minimum conformation for the truncated free chain in solution.As a protein is synthesized by the ribosome, it begins to fold into a three-dimensional shape, and must simultaneously avoid aggregation with other proteins in the crowded cell. In this seemingly hostile environment, where total protein concentrations can exceed 200-300 mg/ mL, de novo protein folding during and after translation is, for many proteins, more efficient than in vitro refolding1. In other words, many proteins that can fold productively in the cell will aggregate severely under in vitro refolding reactions, presumably due to differences between the dominant folding pathway used. For example, experiments with both bacterial and firefly luciferase have demonstrated that these nascent chains adopt conformations cotranslationally that are not populated during refolding from denaturant2,3, and these cotranslational conformations fold to the native state much more efficiently than the conformations populated during refolding from denaturant. However, there continues to be debate as to what extent large, multi-domain proteins can fold co-translationally in prokaryotes4-7.While molecular chaperones certainly play a role in efficient protein folding in the cell, less than 20% of E. coli cytoplasmic proteins require an interaction with one of the three major chaperone systems (trigger factor (TF), DnaK/DnaJ, or GroEL/ES) in order to fold correctly under normal growth conditions8. And remarkably, both chaperone systems responsible for Correspondence should be addressed to P.L.C. (pclark1@nd.edu). Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript wil...