No description available.

Single-particle cryogenic electron microscopy (cryo-EM) has become a powerful mainstay tool in high resolution structural biology thanks to advances in hardware, software and sample preparation technology. In my thesis, I utilized this technique to unravel the function of various challenging biological macromolecules. My first focus was bacterial ribosomal biogenesis: understanding how bacteria assemble their ribosomes. Ribosomes are the factories of the cell, responsible for manufacturing all proteins. Ribosomes themselves are huge, with the bacterial version made of 52 proteins and 4566 RNA nucleotides. How these components assemble has long been a mystery. Early groundbreaking work sketched out a biogenesis pathway using purified components in vitro – but under non-physiological conditions. We sought to understand how the bacterial ribosome – specifically the large subunit 50S – is built inside the cell. To achieve this, we engineered a conditional knock-out bacterial strain that lacked one specific ribosomal protein (L17). This caused the cells to accumulate incomplete intermediates along the 50S biogenesis pathway. These intermediates were purified and examined with mass spectrometry and single-particle cryo-EM. Two major hurdles arose in this project: firstly, the biogenesis intermediates exhibited a preferred orientation when vitrified for cryo-EM analysis. This means that instead of showing many different views required for reconstruction of the 3D structure, the intermediates only adopted one view on the cryo-EM grid. To overcome this problem, we engineered a method to induce additional views on the microscope by tilting the stage. Using another test protein that also exhibited preferred orientation (hemagglutinin), we optimized and characterized this new tilt methodology and showed it was generally applicable to overcoming preferred orientation, regardless of type of specimen. We also created a software tool, called 3DFSC (3dfsc.salk.edu), for other microscopists to calculate the degree of directional anisotropy in their structures due to preferred orientation. Using this tilt strategy finally enabled the structural elucidation of our 50S intermediates. The second challenge in the project was the large amount of heterogeneity present in the sample. Through hierarchical 3D classification schemes using the latest software tools, we obtained 14 different 50S intermediate structures, all from imaging a single cryo-EM grid. By analyzing the missing components of each intermediate, and corroborating these observations with mass spectrometry data, we outlined the first in vivo 50S assembly pathway, and showed that ribosome assembly occurs step-wise and in parallel pathways. My second focus was on pushing the resolution limits of single-particle cryo-EM using adeno-associated virus (AAV) serotype 2 homogeneous virus-like particles (VLPs) that lack DNA. Exploiting several technical advances to improve resolution, including use of gold grids, per-particle CTF refinement, and correction for Ewald sphere curvature, we managed to obtain a 1.86 Å resolution reconstruction of the AAV2L336C variant VLP, the highest resolution icosahedral virus reconstruction solved by single-particle cryo-EM to date. Using our structure, we were able to show improvements using Ewald sphere curvature correction and shed light on the mechanistic basis as to why the L336C mutation resulted in defects in genome packaging and infectivity compared to the WT viral particles. My third focus was the understanding of small membrane proteins involved in infectious diseases. Membrane proteins are a challenge to work with due to the need for them to be extracted from the lipid bilayer for studies as compared to soluble proteins. Infectious diseases have a huge burden on society, with the top three infectious agents accounting for 2.7 million deaths in 2016. The third most deadly infectious disease is malaria, a mosquito-borne parasite which kills 450,000 people annually. One drug used early on for treating malaria was chloroquine but its usefulness waned due to development of resistance. Chloroquine resistance is mediated by the chloroquine resistance transporter (PfCRT). Although small (49 kDa) for single-particle cryo-EM, we solved its structure by using fragment antibody technology to add mass and help with image alignment and 3D reconstruction. The 3.2 Å structure resembles other drug metabolite transporters, and the chloroquine resistance mutations map to a ring around the central cavity, suggesting this central pore as the drug binding site. Tuberculosis (TB) is the top killer, above malaria and HIV/AIDS, being responsible for 1.3 million deaths. In TB, a common antibiotic target is the bacterium’s cell wall synthesis machinery. One family of such enzymes is the arabinosyltransferases, which synthesize the critical arabinose sugars. Using single-particle cryo-EM, we solved two high resolution structures of one such essential enzyme, AftD. Due to the low yield of the protein, a picoliter automated sample dispensing robot was crucial to allow for initial cryo-EM analysis. We then performed mutagenesis studies in M. smegmatis, a TB model organism, which uncovered the critical amino acid residues in the active site and determined that a bound acyl-carrier-protein was likely involved in allosteric inhibition of AftD’s active site. Another member of the family, EmbB, is the target of a widely used frontline TB drug called ethambutol. We have solved the high resolution structures of the apo and putative drug-bound states of EmbB, allowing us to map out, for the first time, both the active site and drug-resistance mutations of this crucial enzyme. The atomic structures of the functional pockets of Mycobacterial AftD and malarial PfCRT will hopefully enable structure-based drug design to improve existing drugs or potentially even develop new treatments against these infectious maladies. In conclusion, the continual and breathtaking improvements in single-particle cryo-EM methodology has been instrumental in allowing the elucidation of the aforementioned biological macromolecules from ribosome biogenesis intermediates, to AAV2 vehicle, Plasmodium drug resistance transporter to mycobacterial glycosyltransferases – structures of which help explain biological function.

No description available.

Kinases play a crucial role in regulating cellular signaling cascades, making them therapeutic targets for several human diseases. In human cancers, mis-regulation and mutations of kinases such as EGFR (epidermal growth factor receptor) have been found to drive malignant transformation. Due to the conserved structural elements of protein kinases, the majority of kinase inhibitors available have a tendency to inhibit multiple targets. The biological impact of this promiscuity is insufficiently defined and the prevalence of cellular compensatory mechanisms additionally varies the clinical responses to drug treatment. In order to understand the relationship between selectivity and efficacy, prior to clinical trials, it is essential to characterize how inhibitors interact with the kinome within a cellular context. Monitoring inhibitor-target interactions generally involves in vitro assaying with purified proteins or protein domains, which compromises the native integrity of the kinases. Cellbased assays either gain outcomes from bulk populations that average out cell variance or phenotypic assays that lack molecular resolution. To obtain information on drug interactions on a single cell level, we have developed a method to measure the direct binding of kinase inhibitors to their targets in situ and in vivo. Kinase inhibitors are chemically tagged with fluorophores that serve as acceptors to genetically tagged donor fluorophores on the enzyme and the interaction is measured using FRET-FLIM. With epidermal growth factor receptor (EGFR) and irreversible EGFR inhibitors as the model system, this approach has been applied to image inhibitor-kinase interactions in live and fixed cells. Using this method, a small panel of tyrosine kinase targets, and labeled inhibitors, we were able to investigate the cross-specificity within the panel. Additionally it was found that the specificity of inhibitors for specific kinase conformations enables the distinction between EGFR in the active and inactive conformation by the inhibitor-probes. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2011.

No description available.

Thesis (MScMed (Dept. of Biomedical Sciences. Anatomy and Histology) --University of Stellenbosch, 2007. / Thyroid nodules are a common clinical problem encountered by physicians, surgeons and radiologists who deal with the head and neck region. However, most follicular lesions of the thyroid are benign, so that the indications for surgery should be as accurate as possible. The morphological difficulty on fine-needle aspiration biopsy (FNAB) of reliably distinguishing preoperatively between benign and malignant lesions has led to a search for ancillary methods that can assist with the diagnosis. The aim of the first study was to develop a cytological scoring system to improve diagnostic accuracy of fine-needle aspiration biopsy of papillary carcinomas with special reference to the follicular variant of papillary carcinoma. The objective of the second study was the application of immunodiagnostic markers Galectin-3 and HBME-1 to histology tissue sections and their corresponding fine-needle aspiration cytology smears to assess their value in distinguishing benign from malignant thyroid lesions. In the first study 16 different cytological features such as background, architecture and cellular morphology were quantatively assessed and scored. Only 14 of the 16 variables were statistically significant. The statistical analysis demonstrated that a score ≥ 4 was indicative of a papillary carcinoma with a sensitivity of 96%. A score < 4 suggested a benign multinodular goiter with a specificity of 97%. In the second study Galectin-3 and HBME-1 were applied to histology tissue sections and their corresponding fine-needle aspiration cytology smears. Statistical analyses showed that the sensitivity of immunohistochemistry for diagnosing malignancy was better than the immunocytochemistry, but the specificity of immunocytochemistry was superior. Furthermore the diagnostic accuracy of immunohistochemistry (86%) and immunocytochemistry (88%) using co-expression of these two antibodies was excellent. In this study on immunocytochemistry, papillary carcinomas were clearly identified with a 100% co-expression in the classic and 71% in the follicular variant of papillary carcinoma. For the surgeon the identification of papillary carcinoma is critical, as this determines the extent of surgery. Similary, the confirmation of a non-neoplastic lesion may prevent surgery. In most cases follicular neoplasms, benign or malignant, will usually be excised for histopathology, prior to definite therapy. These studies show that the implementation of ancillary methods such as a scoring system and immunodiagnostic markers can improve the diagnostic accuracy of thyroid fine-needle aspiration biopsies in our laboratory. This may lead to better management of thyroid nodules. However, it is advisable that cytopathologists always take all the clinical features and image analyses into consideration before making a diagnosis.