Cassidy Dobson’s research is focused on examining structural interactions of an Ixodes scapularis (the common deer tick) salivary protein, Salp15. Salp15 in conjunction with the Borrellia burgdorferi protein Outer Surface Protein C (OspC) are both able to facilitate host immune cell evasion. It has been observed that Salp15 may assist in this evasive mechanism by interacting with the CD4 molecule on T helper cells while shielding OspC from detection. It can therefore be hypothesized that the interaction of these two proteins would limit an immune response due to the unresponsiveness of CD4 to orchestrate its normal immune response. I am primarily focusing on elucidating the structural interaction between CD4, Salp15, and OspC in hopes of understanding the underlying mechanisms of Lyme disease.

Transition metal regulation in the cell is a tightly controlled process.  In E. coli, trace amounts of nickel are required for cellular viability.  If the cellular nickel concentration becomes too high, the regulatory protein NikR binds to the promoter of the nikABCDE, and this effectively shuts down import of nickel into the cell.  The precise mechanism of NikR repression is not understood.  NikR contains four ‘high-affinity’ (Kd=pm) nickel binding sites and two ‘low-affinity’ (Kd=mM-nM) nickel binding sites.  In vitro NikR binds many third row transition metals with high-affinity and binds Cu2+ 1000x tighter than Ni2+, however in vivo, the protein is only responsive to Ni2+.  One possibility that we are exploring is that NikR binds each transition metal with a different coordination geometry and/or uses different ligand selection for each metal.  Another area of interest is the low-affinity site.  This site is not conserved among all NikR homologues.  We are working toward understanding if this site is necessary for a structural change in the overall protein confirmation or serves as a functional sensor of Ni concentration.

In both eukaryotic and prokaryotic organisms, newly synthesized proteins destined for secretion or integration into a membrane traverse the secretory (Sec) pathway.  In bacteria most proteins containing an N-terminal signal sequence are post-translationally targeted to the preprotein translocase SecA.  A key step in the Sec pathway is the recognition of the signal sequence by SecA.  After signal sequence binding, the SecA-preprotein complex associates with the SecYEG translocon and through ATP hydrolysis by SecA and the proton motive force, the preprotein is translocated across the inner membrane.  During the translocation process SecA undergoes conformational changes from a closed, inactive state to one or more open, active states.  The closed and inactive form of SecA has been described by the available crystal structures but little structural information is known about the open and active state of SecA.  Following up initial experiments performed by Song and Kim (J Biochem, 1997), we have determined that low concentrations of urea generates a SecA intermediate with ATPase activity similar to the membrane activated form of the protein.  Further characterization of this intermediate by CD and tryptophan fluorescence demonstrates that in low concentrations of urea SecA undergoes a conformational change with a small loss of secondary structure. We are in the process of defining the signal sequence binding site of the active and open form of SecA through crosslinking analysis. A comparison of the protease digestion patterns of signal sequence crosslinked SecA in both active and inactive forms suggests that the active conformation has a different signal sequence binding site which may represent the functional mode of signal sequence recognition prior to translocation.  The future direction of this work is to determine the region of SecA responsible for signal sequence recognition in the active form of the protein. 

I will be developing protein surface mapping methods using residue-specific covalent labeling combined with mass spectrometry to elucidate the structural features of Cu-induced β-2-microglobulin (β2m) amyloid formation.  Method development and optimization will be performed using model proteins that have well-known structures.  Appropriate reaction controls will be developed to ensure that the covalent labels, which are used as the structural probes, do not themselves disrupt the protein structure. 

The optimized methods will be used to study human β2m, a monomeric protein that has been shown to aggregate into amyloid fibrils in dialysis patients leading to a disorder known as dialysis-related amyloidosis.  Under conditions that lead to β2m amyloid formation, reactions of β2m with modification labels will be used to identify local protein substructures and specific amino acids that are important in the oligomeric assembly of β2m.  Specifically, the surface mapping methods will be used to test a proposed model for the Cu2+-induced amyloid formation of β2m wherein oligomeric assembly is initiated by interactions between the N- and C-terminal regions of two different monomers.  Lastly, the selective covalent modification approach will be used for residue-level characterization of protein stability as a function of chemical denaturant for particular regions of β2m. The goal of these latter experiments will be to identify the protein region destabilized by Cu2+ binding.

Nonenveloped viruses exploit a combination of established cellular pathways and unique viral mechanisms to complete their life cycle.  Simian Virus 40 (SV40) is a nonenveloped virus, which is internalized via endocytosis and eventually trafficked to the endoplasmic reticulum (ER). From the ER, the viral genome must be delivered to the nucleus via an unidentified mechanism for replication to take place.  Our lab has proposed that the minor structural coat proteins VP2 and VP3 are involved in the translocation of either a sub-viral particle containing the viral genome or the genome itself from the ER to the cytosl.  Once in the cytosol, the viral genome is delivered to the nucleus for replication, transcription of viral genes and virus assembly.  Afterwards, the viral progeny must be liberated from the host cell in a timely manner through a cytolytic process.  Our lab has proposed that the combination of VP2 and/or VP3 with the late viral protein VP4 is responsible for viral release from the host cell.  Taken together, the regulated localization of the viral proteins VP2/3/4 to host membranes or viral particles is essential to virus viability.

A variety of approaches have been used to determine the localization of VP2/3/4 in the presence of all host cell membranes in order to elucidate the mechanism of viral egression from the ER during infection or the host cell during release.   Each protein has been expressed in mammalian cells with a C-terminal GFP tag.  We found that when VP2, VP3, or VP4 are expressed individually they localize predominantly within the nucleus but also in other regions of the cell.  VP2/3/4 all contain a C-terminal nuclear localization signal (NLS) so future work will be done to determine if mutation of the NLS perturbs nuclear localization and allows secondary membrane/organelle(s) localization to be revealed.  In an alternative approach, digitonin-permeabilized cells were used to determine membrane insertion and localization of VP2 and VP3.  Sedimentation and detergent treatment experiments revealed that both VP2 and VP3 co-sediment with the nuclear fraction and a significant percentage is detergent soluble or membrane inserted.  Future work will include cell fractionation experiments to determine the specific organelle localization of each protein individually or in combination.  Once the membrane specificity for each or all of the proteins is determined in vitro fluorescence studies will be used to characterize the lipid/protein interactions.  Liposomes will be used which mimic the lipid composition of the target membranes determined by the aforementioned techniques.  Thus the interaction between viral proteins and representative host cell membranes will be characterized in order to elucidate the mechanism of both viral translocation out of the ER during infection and viral release.

The ER glucosyltransferase post-translationally reglucosylates non-native and slow folding domains during glycoprotein maturation

Misfolding of secretory proteins can irreparably impair their efficient transport to the cell surface and extracellular space. This can result in numerous debilitating human diseases such as cystic fibrosis, diabetes mellitus, hereditary emphysema, familial hypercholesterolemia, and others. In the endoplasmic reticulum (ER), there is a highly regulated quality control system to ensure high fidelity protein folding and proper segregation and disposal of terminally misfolded proteins coordinated by a network of chaperones and foldases. A distinct subset of chaperones, such as calnexin (CNX) and calreticulin (CRT), promote proper folding of glycosylated secretory cargo. CNX and CRT rely on the transfer of a covalently linked highly hydrophilic 14- member carbohydrate to secretory glycoproteins to facilitate their interaction with actively folding proteins. A central component in the quality control of glycoprotein folding in the ER is UDP-glucose: glycoprotein glucosyltranferase (GT), which is believed to act as a folding sensor that monitors glycoprotein maturation. Through the transfer of a single glucose to the glucosidase-trimmed glycans of non-native secretory proteins that have been released from CNX and CRT, it is thought that GT can reinitiate lectin chaperone binding to favor native state conformations or to sequester terminally misfolded proteins from the active folding environment. We seek to understand the critical role of GT in monitoring the maturation of secretory cargo and how this process could go awry in human disease.

Little is known of the folding sensor capabilities of GT during active folding events in the intact ER. One problem that has hampered previous GT studies in the intact ER is the differentiation of glycans modified by GT from those that are newly transferred to secretory proteins. To overcome this, we used a mutant mammalian cell line that transfers glycans that bear no glucoses, allowing for the isolation of glycoproteins modified by GT reglucosylation. Using the well-characterized substrate influenza hemagglutinin (HA), we show that GT reglucosylates N-linked glycans that support lectin chaperone binding in the slow folding stem domain, once the nascent chain is released from the ribosome. In addition, these glycans are continually reglucosylated during post-translational folding, indicating inherent targeting to this region. In contrast, the fast folding globular domain is not readily recognized by GT. During delayed disulfide bond formation, GT reglucosylation can be initiated on the normally overlooked globular domain. Additionally, we determined that mutation of HA to either an oligomerization-deficient form or one lacking the stabilizing large-loop disulfide bond resulted in late term reglucosylation by GT when compared to wild-type protein, highlighting the ability of GT to respond to deficiencies in secretory protein maturation. Therefore, GT post-translationally reglucosylates glycans on slow folding or non-native domains for chaperone recruitment to critical regions and can act as a folding sensor by driving lectin chaperone association with “off-pathway” protein species. Having established the capabilities of GT with a well-characterized substrate, we are now pursuing the involvement of GT in the maturation of the cystic fibrosis transmembrane conductance regulator and α1-antitrypsin, the respective causative agents in cystic fibrosis and hereditary emphysema.

Hsp70 proteins are found in all organisms and take part in a variety of fundamental cellular processes, such as nascent polypeptide folding, disaggregation and refolding, protein complex assembly and disassembly, and polypeptide translocation across membranes, among other roles. These functions are all accomplished through the binding and release of polypeptide chains at one Hsp70 domain, which is dynamically regulated by the other Hsp70 domain in a poorly understood mechanism with limited structural information. My research makes use of Hsp70s extensive evolutionary record in finding residues that share conservational dependencies with each other, and in one case, across the two domains. To test the computational results, I'm mutating residues in an E. coli Hsp70 to determine the functional importance of implicated residues and to chemically crosslink the domains. The characterization of mutations and crosslinking could allow for a better understanding of intradomain allosteric pathways and the interdomain interface, and could stabilize the two domains docked onto each other for further structural studies.