Programmed death-ligand 1 (PD-L1) is a type 1 transmembrane protein that has been reported to play a vital role in mediating suppressed immunity. The interaction between PD-L1 and PD-1 delivers a negative signal that reduces the proliferation of these T cells and induces apoptosis at the same time. Antibodies that can block the Programmed death-ligand 1 (PD-L1) on tumor cells have been shown to alleviate cancer-induced immunosuppression. While antibodies have a great potential in various therapeutic uses, many drawbacks such as the high cost of production, huge molecular size, and poor permeability impose restrictions on the extensive use of full-length antibodies. These limitations have necessitated research for finding alternatives to antibodies, such as peptides, that have lower molecular weight and similar properties as antibodies but do not have the lengthy and complicated approach of producing antibodies. In this study, a novel approach based on molecular interactions of the PD1-PD-L1 complex was developed to design peptides against PD-L1 using Knob-Socket model as basis. Three generations of peptides, α-helix, over-packed and salt bridge function peptides, were designed. All designed peptides were docked in the Molecular Operating Environment (MOE) and the AutoDock Vina software for the docking energy and the detail interaction information. Synthesis and characterization of selected peptides were performed after simulation studies. Surface Plasmon Resonance (SPR) studies showed that α-helix and over-packed peptides can’t bind to the PD-L1 protein with no response on sensorgrams, while peptides with salt bridge function had a higher binding response than those two generations of peptides. In confocal microscopic studies, PD-L1 positive breast cancer cell line MDA-MB-231 was used to determine the binding specificity of the salt bridge function peptides to PD-L1 in vitro, while another breast cancer cell line (MCF-7, without PD-L1) was used as a control. After incubation with peptides, significant fluorescence intensities were detected on the MDA-MB-231 cells, while only background fluorescence was observed on MCF-7 cells. In conclusion, this study demonstrated that peptides against PD-L1 designed using the Knob-Socket model and molecular interaction between PD-L1-PD1 complex showed feasibility to bind specifically with PD-L1 receptors.

To investigate whether higher order packing interactions confer protein-DNA specificity, a modified Knob-Socket (KS) model was used to analyze the interface of bZIP-DNA crystal structures. The KS analysis identified a nine-residue quadripartite recognition core consisting of four contiguous KS pockets P1, P2, N3, and N4 that each pack one of the four DNA half-site bases in the target sequence. Only one base per base pair packs, and these interactions are split across the DNA strands: the first two positive strand positions 1p and 2p pack into P1 and P2 while the last two negative strand positions 3n and 4n pack into N3 and N4. Amino acid sequence analysis of the four KS pocket regions indicates that the primary mechanism recognition is packing or non-packing of the 5-methyl group of dT as well as 5-methylcytosine. P1 shows little packing of dT; P2 packs dT but including two Asn residues in this pocket seems to block packing in this region; N3 also packs dT, but including a Phe also blocks packing; N4 consistently packs dT. This analysis demonstrates that there is an amino acid code to DNA recognition, allowing for multi-residue recognition and packing of the 5-methyl group.

Cancer has been a pervasive and deadly problem for many years. No treatments have been developed that effectively destroy cancer cells while also keeping healthy cells safe. In this work, the knob-socket construct is used to analyze two systems involved in cancer pathways, the PDZ domain and the Bcl-BH3 complex. Application of the knob-socket model in mapping the packing surface topology (PST) allows a direct analysis of the residue groups important for peptide specificity and affinity in both of these systems. PDZ domains are regulatory proteins that bind the C-terminus of peptides involved in the signaling pathway of cancer progression. The domain includes five -strands, two -helices, and six coils/turns. In this study, the PST of all eight solved crystal structures of T-cell lymphoma invasion and metastasis 1 (Tiam1) PDZ domains are mapped to reveal details of ligand-domain binding pockets and packing interactions. Four main interactions were identified in the comparison of the PST maps and a consensus sequence was calculated using knob-socket interaction data. In the case of the Bcl-BH3 complex, binding of these two proteins prevents an unhealthy cell from undergoing apoptosis. In the knob-socket mapped protein-ligand interactions, the helical ligand consists of 8 to 10 residues that specifically interact with four helices on the binding protein: the N-terminus of Helix2, the main bodies of Helix3 and Helix4 and the C-terminus of Helix5. Among all of the interactions that were analyzed, there were three amino acids from the ligand, glycine, leucine, and isoleucine, that always packed into the hydrophobic groove that is key for ligand recognition. By using knob-socket analysis to map quaternary packing structure, it was possible to identify the quaternary-level protein interactions that define ligand specificity and binding strength. From this analysis, possible protein mimetics can be developed that could be used as cancer treatments.

Human Epidermal Growth Factor Receptor 2 (HER2) is a cell surface receptor tyrosine kinase and plays a role in the signal pathways leading to cell proliferation and differentiation. Overexpression of HER2 is found in various cancers including breast, ovarian, gastric, colon, and non-small-cell lung cancers, which makes it an attractive target for cancer therapy. Specific antibodies, peptides and small molecules are developed by scientists to bind with HER2 as therapeutical agents, dimerization inhibitors and biological makers. Among these molecules, antibodies showed excellent binding affinity and specificity toward HER2. However, uses of antibodies are limited by their high cost of production, long development time, limited ability to penetrate tumor tissue and immunogenicity. Many of these limitations are due to the high molecular weight of antibodies. Compared to antibodies, peptides and small molecule that selectively recognize HER2 have advantages in solubility, permeability and immunogenicity. So far, the design of all peptides and small molecules for binding with HER2 either utilize phage display technique or rely on computational screen of large library of millions of small molecules. These approaches all suffer from the drawbacks of tedious, labor intensive, and time consuming as well as uncertainty of outcome. In this study, it was hypothesized that a novel approach based on molecular interactions of HER2-Pertuzumab complex and Knob-Socket model can be developed to design antibody mimics for targeting HER2. All designed antibody mimics were simulated and docked with HER2 using Molecular Operating Environment (MOE) software to estimate binding energy and analyze the detail interaction map. A series of mimics were then synthesized and characterized. HER2 positive breast cancer cells MDA-MB-361 and ZR-75-1 were used in confocal microscopic and flow cytometric studies to evaluate the binding specificity of all antibody mimics to HER2 in vitro, while human embryonic kidney cell (HEK293) was used as control. After incubation with antibody mimics, high fluorescence intensities were observed on MDA-MB-361 and ZR-75-1 cells, while only background fluorescence were observed on HEK293 cells. Surface plasma resonance (SPR) studies showed that all antibody mimics bind to HER2 protein with KD value in range of 55.4 nM- 525.5 nM. Western blot technique was used to evaluate inhibition capability of antibody mimics on phosphorylation of HER2 downstream signaling Akt and MAPK pathways that were crucial for cell differentiation and survival. When treated with antibody mimics at 10µM for 24 h, more than 85% phosphorylation of Akt pathway was inhibited while phosphorylation of MAPK pathway was not affected. This finding proved that antibody mimics could bind to HER2 extracellular domain and selectively inhibit the dimerization between HER2 and HER3 to block phosphorylation of Akt pathway in a similar way as Pertuzumab. In addition, in vivo studies on tumor bearing nude mice were carried out to investigate the distribution and binding specificity of antibody mimics towards HER2 positive tumor after injecting through vein tail. Signal intensity ratio (SIR) of tumor to muscle revealed about 10-fold increase in tumor retention of HER2-PEP11 compared to the Cy7.5 carboxylic acid and Cy7.5-HER2-PEP22, which confirmed excellent in vivo binding specificity of antibody mimic HER2-PEP11 to HER2 positive tumor. In conclusion, this study demonstrated that a rational design of antibody mimics with both binding specificity and affinity towards HER2 based on the molecular interaction between Pertuzumab and HER2 and Knob-Socket model is feasible.

Protein sequences contain the information in order for a protein to fold to a unique compact, three-dimensional native structure. The forces that drive protein structures to form compact folds are largely dominated by burial of hydrophobic amino acids, which results in non-specific packing of amino acid side-chains. The knob-socket model attempts to organize side-chain packing into tetrahedral packing motifs. This tetrahedral motif is characterized with a three residues on the same secondary structure forming the base of the tetrahedron packing with a side-chain from a separate secondary structure. The base of the motif is termed the socket, and the other side-chain is called the knob. Here, we focus on extending the knob-socket model to understand tertiary and quaternary structure. First, single knobs sometimes pack into more than one socket in real structures. We focus on understanding the topology and amino acid preferences of these tertiary packing surfaces. The main results from the study of tertiary packing surfaces is that they have a preferred handedness, some interactions are ancillary to the packing interaction, there are specific amino preferences for specific positions in packing surfaces, and there is no relationship between side-chain rotamer of the knob packing into the tertiary packing surface. Next, we examine the application of the knob-socket to irregular and mixed packing in protein structure. The main conclusions from these efforts show canonical packing modes between secondary structures and highlight the important of coil secondary structure in providing many of the knobs for packing. Third, we investigate protein quaternary structure with a clique analysis of side-chain interactions. We identify a possible pseudo knob-socket interaction, and compare knob-socket interactions between tertiary and quaternary structure. Lastly, we discuss the workflow used in CASP12 to predict side-chain contacts and atomic coordinates of proteins.

Proteins are molecular machines of life in the truest sense. Being the expressors of genotype, proteins have been a focus in structural biology. Since the first characterization and structure determination of protein molecule more than half a century ago1, our understanding of protein structure is improving only incrementally. While computational analysis and experimental techniques have helped scientist view the structural features of proteins, our concepts about protein folding remain at the level of simple hydrophobic interactions packing side-chain at the core of the protein. Furthermore, because the rate of genome sequencing is far more rapid than protein structure characterization, much more needs to be achieved in the field of structural biology. As a step in this direction, my dissertation research uses computational analysis and experimental techniques to elucidate the fine structural features of the tertiary packing in proteins. With these set of studies, the knowledge of the field of structural biology extends to the fine details of higher order protein structure.