Thymidine Kinase 1 (TK1) is primarily known as a cancer biomarker with good prognostic capabilities for both hematological and solid malignancies. However, recent studies targeting TK1 at protein and mRNA levels have shown that TK1 may be useful as a therapeutic target. In order to examine the use of TK1 as a therapeutic target, it is necessary to develop therapeutics specific for it. Single domain antibodies (sdAbs), represent an exciting approach for the development of immunotherapeutics due to their cost-effective production and higher tumor penetration than conventional antibodies. In this study, we isolated sdAb fragments specific to human TK1 from a human sdAb library. A total of 400 sdAbs were screened through 5 rounds of selection by monoclonal phage ELISA. The most sensitive sdAb fragments were selected as candidates for preclinical testing. The sdAb fragments showed specificity for human TK1 in phage ELISA, Western blot analysis and had an estimated limit of detection of 3.9 ng/ml for the antibody fragments 4-H-TK1_A1 and 4-H-TK1_D1. The antibody fragments were successfully expressed and used for detection of membrane associated TK1 (mTK1) through flow cytometry on cancer cells [lung (~95%), colon (~87%), breast (~53%)] and healthy human mononuclear cells (MNC). The most sensitive antibody fragments, 4-H-TK1_A1 and 4-H-TK1_D1 were fused to an engineered IgG1 Fc fragment. When added to cancer cells expressing mTK1 co-cultured with human MNCs, the anti-TK1-sdAb-IgG1_A1 and D1 were able to elicit a significant antibody-dependent cell-mediated cytotoxicity (ADCC) response against lung cancer cells compared to isotype controls (P<0.0267 and P<0.0265, respectively). To our knowledge this is the first time that the isolation and evaluation of human anti-TK1 single domain antibodies using phage display technology has been reported. The antibody fragments isolated here may represent a valuable resource for the detection and the targeting of TK1 on tumor cells.

Since the clinical relevance of hypoxanthine guanine phosphoribosyltransferase (HPRT) as both a cancer biomarker and a potential tumor target is increasing, we isolated human nanobodies (Nbs) against HPRT and tested their potential to detect and target cytosolic and membrane associated HPRT in lung, breast and colon cancer cells. HPRT is a purine salvage pathway enzyme that catalyzes the conversion of hypoxanthine to inosine monophosphate and guanine to guanosine monophosphate. Until recently, HPRT was considered as a housekeeping gene and its clinical relevance was mainly limited to congenital central nervous disorders such as Lesch-Nyhan disease. However, recent gene expression analysis across all cancers in The Cancer Genome Atlas (TCGA) database and analysis of HPRT expression within cancer patients' tissues has revealed that the expression of HPRT is elevated in the majority of malignant tissues in comparison to normal tissues. Moreover, multiple studies have recently shown that HPRT may be associated with the cell membrane in lung, colon cancer and lymphoma cells, and not in normal cells. Thus, HPRT is a potential immunotherapeutic target for cancer. Using phage display technology we isolated 384 nanobody clones from the Christ human single domain antibody (dAb) library against human HPRT. Ten of the most sensitive Nbs were sequenced and evaluated to determine their sensitivity using dose response curves. Four parameter logistic regression analysis of the top 10 HPRT nanobodies showed R squared values ranging from 0.9388-0.9989 and a sigmoidal shape denoting a concentration-dependent antigen-ligand relationship. The anti-HPRT Nbs were able to detect minimum HPRT amounts ranging from 10-20 ng/ml. Validation of the anti-HPRT Nbs was conducted with a siRNA HPRT knockdown in A549 cells. A549 siRNA HPRT knockdown cell lysate showed at least a 5-fold reduction in the overall signal in phage ELISA compared to A549 wild type cell lysate (p&lt;0.001) indicating the specificity of the Nbs to HPRT. In addition, flow cytometry analysis showed that the anti-HPRT Nbs were able to detect membrane associated HPRT on A549 (20%), NCI-H460 (23%), HT-29 (41%) , SW620 (46%) and MDA-MB-231 (32%) cancer cell lines. Anti-HPRT Nbs did not show significant binding to normal mononuclear cells compared to cancer cells (p&lt;0.002). Anti-HPRT Nbs were successfully expressed and purified using the pET-scFv expression system. Human anti-HPRT Nbs may be used for the detection of both cytosolic and membrane associate HPRT. Future directions of this research will explore the uses of anti-HPRT-Nb-based therapies and their use in other applications such as chimeric antigen receptors (CARs) for cell adoptive therapies. Citation Format: Edwin J. Velazquez, Tyler B. Humpherys, Toni O. Mortimer, David M. Bellini, Kathryn R. Smith, Rachel M. Morris, Abigail Goodman, Kim L. O'Neill. Isolation and evaluation of human nanobodies against hypoxanthine guanine phosphoribosyltransferase (HPRT) and their potential use for the detection and targeting of lung, breast and colon cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1840.

The purpose of this study was to analyze the regulatory gene cancer network of the tumor biomarker Thymidine Kinase 1 (TK1) across all cancers and identify the genes with meaningful correlations that interact and drive tumorigenesis along with TK1. In addition, we sought to find a feasible mechanism for TK1 membrane expression in cancer cells. Using data from The Cancer Genome Atlas (TCGA), the pathway commons protein interaction database, the Biogrid database and the R graphite library, we analyzed the correlation of mRNA expression levels of all proteins that were involved in TK1 metabolic pathways or physically interacted with TK1. By searching the pathway commons protein interaction database and data drawn from the graphite library in R, we found that 1495 proteins were involved with TK1 metabolism. Using Dijkstra's algorithm we found that 294 proteins had 1-2 degrees of interaction with TK1. Analysis of the Biogrid database showed that 191 proteins physically interact with TK1. The genes with higher average Pearson's correlation values across all 25 cancers had the following functions: cell cycle regulation, DNA synthesis, replication and repair, transcription factors (E2F family), transport of mono and dicarboxylate for energy production pathways, transport of nucleosides, pyrimidine metabolism, DNA methylation of tumor suppressor genes, mRNA stabilization, aging, regulation of cell proliferation, control of cyclin kinases, apoptosis, and formation of channels on cell membranes. Among the genes that had significant correlation values with TK1 (P&lt;0.001) were: RRM2(2), CDK1, CDK2, CDK4(0), CDC20(0), CDC25A, MKI67(0), MCM3, E2Fs, CD6, PXMP2, DTYMK(1), BRCA1, EZH2(0), NUDT1(3), SLC29A2(2), SLC25A10(2), SLC16A3(3), SLC25A19(3), PRIM1, CARHSP1(0), NKIRAS2(0), METTL23(0) CENPB(0), FAF1(0), COPS6 (0), TRIM28 (0), MSH2(0), AAMP (0). Further analysis revealed that 476 of the proteins from the common pathways and Biogrid had transmembrane (tm) domains. Although it had a moderate correlation value, SEZ6L protein, which is overexpressed on the cell membrane of lung, colon and gastric cancers was a physical interactor of TK1. Additional analysis showed that the nucleoside transporter SLC29A2, that transports both adenosine and thymidine, had a tm domain and 2 degrees of interaction with TK1. Like the purinosome, we hypothesize that TK1 in malignant cells may be complexing with other pyrimidine synthesis proteins. Moreover, the physical interaction with other proteins containing tm domains could provide a path for TK1 membrane expression. Overexpression, secretion and membrane association of TK1 may be an evolutionary advantage that supply enough enzymes to modify the nucleotide pool balance necessary for DNA replication in tumor proliferating cells. Future directions will focus on the TK1 protein-protein interactions found in this analysis. Citation Format: Edwin J. Velazquez, David M. Bellini, Rachel A. Skabelund, Tyler B. Humpherys, Jonathan R. Skidmore, Brett E. Pickett, Stephen R. Piccolo, Kim L. O'Neill. Bioinformatic analysis of the tumor biomarker thymidine kinase 1: Elucidating its cancer gene network and membrane expression across all cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2134.