Date of Award

Spring 1-1-2018

Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Marcelo Sousa

Second Advisor

Thomas Perkins

Third Advisor

Michael Stowell

Fourth Advisor

Joseph Falke

Fifth Advisor

Amy Palmer


Pathogenic bacteria have developed a wide range of tools for circumventing or overcoming the host’s defenses. Over time, these tools have become increasingly complex, allowing bacteria to live and thrive within a wide variety of host environments. One such tool is the Type III Secretion System (T3SS), a needle-like complex that allows bacteria to directly inject proteins, known as effectors, from their cytoplasm into host cells. Once inside host cells, effector proteins have a wide range of effects, from shutting down the host immune response to rearranging the host cytoskeleton to accommodate invading bacteria.

Because the T3SS needle presents a narrow channel (< 2 nm), effector proteins must be mechanically unfolded before passing through. Proteins are unfolded by a molecular motor that associates with the base of needle and pulls protein into the channel. While this motor can unfold and secrete many proteins, it is unable to unfold proteins that have high mechanical stability. This indicates a need for effectors to be mechanically labile no matter their function. This may be one of the reasons effectors have very low sequence and structural similarity to other members of their protein super-families. This spurred our investigation into how effectors respond to mechanical force. To investigate effector protein stability, I used atomic force microscopy (AFM) to mechanically unfold the proteins.

Here I show that effector proteins of the T3SS unfold at very low force, despite containing a wide variety folds and functions. This supports our hypothesis that to facilitate efficient secretion, effectors evolved to be mechanically labile. Because effector proteins unfold at such low force, it was critical for me to utilize site-specific attachment to both the AFM tip and surface, increasing both the amount of data I could collect and the quality of collected data. Site-specific attachment resulted in a 70-fold improvement in the yield of high quality data, allowing rapid characterization of mechanically labile α-helical proteins.

Combining site-specific attachment with modified cantilevers allowed the collection of unfolding data for 5 effector proteins, finding they all unfold at low force (<20 pN), making them some of the most mechanically labile proteins studied to date by AFM-SMFS. Comparing the mechanical stability of effector proteins to their in vivo secretion rates, showed that unfolding force does not always correlate with in vivo secretion rate. However, the distance to the transition state does correlate with in vivo secretion rate.

To elucidate how effector proteins have evolved to be efficiently secreted, the mechanical stability of an effector protein, NleC, was compared with a non-secreted homologue, protealysin. While the initial unfolding event of NleC occurs below the detection limit of our AFM platform, the unfolding of an intermediate along the unfolding pathway was measured. When compared to the unfolding of protealysin, the unfolding intermediate of NleC was shown to be significantly less stable, supporting our hypothesis that effector proteins have evolved to unfold at low force to facilitate efficient secretion. We now have an array of tools that allow for the efficient mechanical characterization of diverse proteins with high-precision.