Affibody molecules are small (7 kDa) affinity proteins of non-immunoglobulin origin that have been generated to specifically interact with a large number of clinically important molecular targets. In this thesis, Affibody molecules have been employed as tracers for radionuclide molecular imaging of HER2- and IGF-1R-expressing tumors, paper I-IV, and for surface knock-down of EGFR, paper V. In paper I, a tag with the amino acid sequence HEHEHE was fused to the N-terminus of a HER2-specific Affibody molecule, (ZHER2), and was shown to enable facile IMAC purification and efficient tri-carbonyl 99mTc-labeling. In vivo evaluation of radioactivity uptake in different organs showed an improved biodistribution, including a 10-fold lower radioactivity uptake in liver, compared to the same construct with a H6-tag. In paper II, it was further shown that an N-terminally placed HEHEHE-tag on ZHER2 provided lower unspecific uptake of radioactivity in liver compared to its H6-tagged counterpart even when radiolabeling was at the C-terminus using alternative chemistries to attach 99mTc, 111In or 125I. In paper III, the H6-tag’s composition and position was varied with regards to charge, hydrophobicity and its C- or N-terminal placement on ZHER2. Among the ten variants investigated, it was found that an N-terminal HEHEHE-tag provided the most favorable overall biodistribution profile and that introduction of hydrophobic and positively charged amino acids provoked liver uptake of radioactivity. In paper IV, the HEHEHE-tag was shown to enable IMAC purification and tri-carbonyl 99mTc-labeling of an IGF-1R-specific Affibody molecule and improved its overall biodistribution when compared to the same construct with a H6-tag. In paper V, the aim was to develop an intracellular receptor-entrapment system to reduce the surface levels of EGFR. An EGFR-specific Affibody molecule was expressed as a fusion to different mutants of an intracellular transport protein in SKOV-3 cells, resulting in a collection of cell lines with 50%, 60%, 80% and 96% reduced surface level of EGFR. Analysis of the proliferation rate of these cell lines showed that a modest reduction (15%) in proliferation occurs between 60% and 80% reduction of the surface level of EGFR. / <p>QC 20130129</p>

This thesis is about protein profiling in blood-derived samples using suspension bead ar- rays built with protein affinity reagents, and the evaluation of binding characteristics and potential disease relation of such profiles. A central aim of the presented work was to discover and verify disease associated protein profiles in blood-derived samples such as serum or plasma. This was based on immobiliz- ing antigens or antibodies on color-coded beads for a multiplexed analysis. This concept generally allow for a dual multiplexing because hundreds of samples can be screened for hundreds of proteins in a miniaturized and parallelized fashion. At first, protein antigens were used to study humoral immune responses in cattle suffering from a mycoplasma infec- tion (Paper I). Here, the most immunogenic of the applied antigens were identified based on reactivity profiles from the infected cattle, and were combined into an antigen cocktail to serve as a diagnostic assay in a standard ELISA set-up. Next, antibodies and their em- ployment in assays with directly labeled human samples was initiated. This procedure was applied in a study of kidney disorders where screening of plasma resulted in the discovery of a biomarker candidate, fibulin-1 (Paper II). In parallel to the disease related applica- tions, systematic evaluations of the protein profiles were conducted. Protein profiles from 2,300 antibodies were classified on the bases of binding properties in relation to sample heating and stringent washing (Paper III). With a particular focus on heat dependent de- tectability, a method was developed to visualize those proteins that were captured to the beads in an immunoassay by using Western blotting (Paper IV). In conclusion, this thesis presents examples of the possibilities of comparative plasma profiling enabled by protein bead arrays. / <p>QC 20130208</p>

Ryanodine receptor 1 (RyR1) is a homotetrameric calcium channel located in the sarcoplasmic reticulum (SR) of skeletal muscle. We employed metal affinity chromatography followed by liquid chromatography mass spectrometry from HEK-293 cells to purify affinity tagged cytosolic RyR1, with interacting proteins. In total, we identified 703 proteins with high confidence (>99%). Of the putative RyR1 interacting proteins, five candidates [calcium homeostasis endoplasmic reticulum protein (CHERP), ER-golgi intermediate compartment 53kDa protein (LMAN1), T-complex protein (TCP), phosphorylase b kinase (PHBK) and four and half LIM domains protein 1 (FHL1)], were selected for interaction studies. Immunofluorescence analysis showed that CHERP co-localizes with RyR1 in the SR of rat soleus muscle. Calcium transient assays in HEK293 cells over-expressing RyR1 with siRNA suppressed CHERP or FHL1, showed reduced calcium release via RyR1. In conclusion, we have identified RyR1 interacting proteins in CHERP and FHL1 which may represent novel regulatory mechanisms involved in excitation-contraction coupling.

As the demand of indoor Location-Based Services (LBSs) increases, there is a growing interest in developing an accurate indoor positioning and tracking system on mobile devices. The core location determination problem can be reformulated as a sparse natured problem and thus can be solved by applying the Compressive Sensing (CS) theory. This thesis proposes a compact received signal strength (RSS) based real-time indoor positioning and tracking systems using CS theory that can be implemented on personal digital assistants (PDAs) and smartphones, which are both limited in processing power and memory compared to laptops. The proposed tracking system, together with a simple navigation module is implemented on Windows Mobile-operated smart devices and their performance in different experimental sites are evaluated. Experimental results show that the proposed system is a lightweight real-time algorithm that performs better than other traditional fingerprinting methods in terms of accuracy under constraints of limited processing and memory resources.

Gesture Recognition provides an efficient human-computer interaction for interactive and intelligent computing. In this work, we address the problem of gesture recognition using the theory of random projection and by formulating the recognition problem as an $\ell_1$-minimization problem. The gesture recognition uses a single 3-axis accelerometer for data acquisition and comprises two main stages: a training stage and a testing stage. For training, the system employs dynamic time warping as well as affinity propagation to create exemplars for each gesture while for testing, the system projects all candidate traces and also the unknown trace onto the same lower dimensional subspace for recognition. A dictionary of 18 gestures is defined and a database of over 3,700 traces is created from 7 subjects on which the system is tested and evaluated. Simulation results reveal a superior performance, in terms of accuracy and computational complexity, compared to other systems in the literature.

Ryanodine receptor 1 (RyR1) is a homotetrameric calcium channel located in the sarcoplasmic reticulum (SR) of skeletal muscle. We employed metal affinity chromatography followed by liquid chromatography mass spectrometry from HEK-293 cells to purify affinity tagged cytosolic RyR1, with interacting proteins. In total, we identified 703 proteins with high confidence (>99%). Of the putative RyR1 interacting proteins, five candidates [calcium homeostasis endoplasmic reticulum protein (CHERP), ER-golgi intermediate compartment 53kDa protein (LMAN1), T-complex protein (TCP), phosphorylase b kinase (PHBK) and four and half LIM domains protein 1 (FHL1)], were selected for interaction studies. Immunofluorescence analysis showed that CHERP co-localizes with RyR1 in the SR of rat soleus muscle. Calcium transient assays in HEK293 cells over-expressing RyR1 with siRNA suppressed CHERP or FHL1, showed reduced calcium release via RyR1. In conclusion, we have identified RyR1 interacting proteins in CHERP and FHL1 which may represent novel regulatory mechanisms involved in excitation-contraction coupling.

Gesture Recognition provides an efficient human-computer interaction for interactive and intelligent computing. In this work, we address the problem of gesture recognition using the theory of random projection and by formulating the recognition problem as an $\ell_1$-minimization problem. The gesture recognition uses a single 3-axis accelerometer for data acquisition and comprises two main stages: a training stage and a testing stage. For training, the system employs dynamic time warping as well as affinity propagation to create exemplars for each gesture while for testing, the system projects all candidate traces and also the unknown trace onto the same lower dimensional subspace for recognition. A dictionary of 18 gestures is defined and a database of over 3,700 traces is created from 7 subjects on which the system is tested and evaluated. Simulation results reveal a superior performance, in terms of accuracy and computational complexity, compared to other systems in the literature.

As the demand of indoor Location-Based Services (LBSs) increases, there is a growing interest in developing an accurate indoor positioning and tracking system on mobile devices. The core location determination problem can be reformulated as a sparse natured problem and thus can be solved by applying the Compressive Sensing (CS) theory. This thesis proposes a compact received signal strength (RSS) based real-time indoor positioning and tracking systems using CS theory that can be implemented on personal digital assistants (PDAs) and smartphones, which are both limited in processing power and memory compared to laptops. The proposed tracking system, together with a simple navigation module is implemented on Windows Mobile-operated smart devices and their performance in different experimental sites are evaluated. Experimental results show that the proposed system is a lightweight real-time algorithm that performs better than other traditional fingerprinting methods in terms of accuracy under constraints of limited processing and memory resources.

Previous studies in Dr. Yang’s laboratory have established a grafting, design, and subdomain approach in order to investigate the properties behind Ca2+-binding sites located in Ca2+-binding proteins by employing engineered proteins. These approaches have not only enabled us to isolate Ca2+-binding sites and obtain their Ca2+-binding affinities, but also to investigate conformational changes and cooperativity effects upon Ca2+ binding. The focus of my thesis pertains to optimizing the expression and purification of engineered proteins with tailored functions. Proteins were expressed in E. coli using different cell strains, vectors, temperatures, and inducer concentrations. After rigorous expression optimization procedures, proteins were further purified using chromatographic and/or refolding techniques. Expression and purification optimization of proteins is essential for further analyses, since the techniques used for these studies require high protein concentrations and purity. Evaluated proteins had yields between 5-70 mg/L and purities of 80-90% as confirmed by SDS-PAGE electrophoresis.

Calcium regulates many biological processes through interaction with proteins with different conformational, dynamic, and metal binding properties. Previous studies have shown that the electrostatic environment plays a key role in calcium binding affinity. In this research, we aim to dissect the contribution of the electrostatic environment to calcium binding affinity using protein design. Many natural calcium binding proteins undergo large conformational changes upon calcium binding which hampers the study of these proteins. In addition, cooperativity between multiple calcium binding sites makes it difficult to study site-specific binding affinity. The design of a single calcium binding site into a host system eliminates the difficulties that occur in the study of calcium binding affinity. Using a computer algorithm we have rationally designed several calcium binding sites with a pentagonal bipyramidal geometry in the non-calcium dependent cell adhesion protein CD2 (CD2-D1) to better investigate the key factors that affect calcium binding affinity. The first generation proteins are all in varying electrostatic environments. The conformational and metal binding properties of each of these designed proteins were analyzed. The second generation designed protein, CD2.6D79, was designed based on criteria learned from the first generation proteins. This protein contains a novel calcium binding site with ligands all from the â-strands of the non-calcium dependent cell adhesion protein CD2. The resulting protein maintains native secondary and tertiary packing and folding properties. In addition to its selectivity for calcium over other mono and divalent metal ions, it displays strong metal binding affinities for calcium and its analogues terbium and lanthanum. Furthermore, our designed protein binds CD48, the ligand binding partner of CD2, with an affinity three-fold stronger than CD2. The electrostatic potential of the calcium binding site was modified through mutation to facilitate the study of the effect of electrostatic interactions on calcium binding affinity. Several charge distribution mutants display varying metal binding affinities based on their charge, distance to the calcium binding site, and protein stability. This study will provide insight into the key site factors that control calcium binding affinity and calcium dependent biological function.