Conformational dynamics of the Sec secretion system

Protein biogenesis is an intricate process for proteins that function at extracytoplasmic cellular compartments, or beyond, and these account for more than one third of any proteome. Their folding process is temporally and topologically uncoupled from protein synthesis. It is made even more challenging by the need for targeting to export machineries and translocation through them. Most exported proteins cross the cytoplasmic membrane through the ubiquitous and essential Sec secretion system in poorly-characterized unfolded states. While membrane proteins will be exported co-translationally in bacteria, secretory proteins follow a mostly post-translational route, which is the focus of this study. Once the appropriate cellular compartment is reached, exported proteins acquire their native states. In the cytoplasm, Sec-dependent export requires non-native protein states, as well as specific and productive protein recognition to prevent illicit protein export.

To date, the key solution to these challenges is attributed to signal peptides that exported proteins carry amino-terminally of their mature domains during their ribosome-membrane trafficking, and to chaperones: Signal peptides provide the targeting signal, delay protein folding and activate the export channel; Chaperones act as holdases, preventing protein folding and aggregation. However, signal peptides and chaperones are dispensable for targeting and subsequent secretion of at least some secretory mature domains in vitro and in vivo, under physiological conditions. Apparently, inherent properties of secretory mature domains may have an important contribution. The identification of these features is impeded by technical challenges in the biophysical analysis of non-native protein states that are essential for secretion. The aim of this study was to characterize the conformational properties of these transient protein states, in order to gain insights into protein folding of this unique protein class and into the molecular basis of their recognition by their receptors during targeting.

To this end, during my PhD research I developed the method of Hydrogen/Deuterium eXchange-Mass Spectrometry (HDX-MS) at the LMB (Rega Institute), currently unique in Belgium, for the characterization of targeting-competent, non-folded, secretory proteins. The tool is uniquely suited for monitoring protein folding, capturing folding intermediates and quantifying conformational differences between different states (e.g. a preprotein, that is carrying a signal peptide, vs a mature domain). Under appropriate conditions, it enables the conformational analysis of disordered protein states and even the localization of residual structure. Therefore, it was a cornerstone in understanding dynamic processes within the non-native secretory states, very difficult to monitor with most other approaches.

For a thorough characterization of the properties that are relevant to ribosome-membrane trafficking, conformational properties derived from HDX-MS, and complementary biophysics tools, were correlated with targeting and secretion, mainly using in vitro biochemical assays and mutagenesis analysis. Our pipeline was applied to more than 20 E.coli secretory proteins, rendering this study the first secretome-representative-wide analysis of translocation-competence.

Our collective findings demonstrate that secretory proteins represent a unique structural class, the primary sequences of which were optimized by evolution at multiple levels for secretion. Firstly, in addition to signal peptides, secretory mature domains contain independent, linear or three-dimensional, targeting signals (MTSs) that are essential for secretion. These dock onto their receptor, SecA, at a distinct but adjacent groove to that of signal peptides. Secretory proteins are therefore bivalent, dynamic ligands.

Secondly, most secretory mature domain sequences have evolved for delayed folding, in a signal peptide-independent manner. They adopt flexible/disordered states that mediate the exposure of the autonomous MTSs. This evolutionary selection could be achieved even with fine divergences from their cytoplasmic structural homologues.

Thirdly, the early mature domain of secretory proteins has a novel dual role: it is a conformational rheostat of mature domain disorder and a physical linker for signal peptides. Via this tether signal peptides finely tune the disordered ensemble of the mature domain that follows and each signal peptide – mature domain pair was selectively matched on the basis of these dynamics. Exaggerated mature domain manipulation by signal peptides is observed in a few mature domains that fold into rigid states in the absence of their signal peptides.

These novel mechanisms are essential for efficient protein secretion and have wider ramifications in protein folding and proteostasis.