Angiogenesis, the formation of new blood vessels, is an essential physiological process for wound healing, reproduction, and development. In cancer, angiogenesis is crucial to sustain solid tumor growth and correlates with metastases. Studies of bone marrow biopsies from children with leukemia indicate that leukemic cells induce angiogenesis in the bone marrow.1 The central role of angiogenesis in cancer promoted the hypothesis that cancer may be treated with anti-angiogenic drugs. Moreover, other studies2 proposed that endothelial cells of tumor blood vessels within specific tumors express unique receptors, or zip codes, such that targeted delivery of anti-cancer drugs might be feasible and, simultaneously, could minimize collateral damage of healthy cells.
To map the molecular heterogeneity within the human vasculature, a combinatorial peptide bacteriophage library was screened by in vivo phage display of normal and tumor blood vessels in cancer patients undergoing terminal wean.3,4 The phage are engineered to display a short peptide sequence as a fusion protein on the pIII coat protein, and phage libraries typically have a peptide diversity of 109. Importantly, intravenously injected phage extravasate from leaky tumor blood vessels in vivo to bind to receptor proteins present on the surface of tumor cells, as well as the extracellular matrix and perivascular cells. In addition to in vivo phage display, an in vitro phage display method, Biopanning and Rapid Analysis of Selective Interactive Ligands (BRASIL), was developed and used to profile the expression of cell surface receptors of cultured human cancer cells in the NCI-60 panel (a National Cancer Institute panel of 60 human cancer cell lines from different histologic origins and grades).5,6 Recovered phage display peptide sequences that act as ligands to bind to accessible receptors expressed on the luminal surface of endothelial cells or on the surface of tumor cells. The experimental design enriches for phage that bind to accessible receptors and are internalized upon receptor binding.
In vivo phage display studies facilitated construction of a human vascular map from phage that specifically bind to receptors expressed on the vascular beds of numerous organs or tumor cells. As expected, various receptors that are uniquely expressed on the vascular beds in specific tumors were identified, as were receptors that are present on the extracellular matrix, perivascular cells, and tumor cells. Surprisingly, several common receptors expressed in different tumor cell types were also identified. For instance, the National Center for Biotechnology Information Basic Local Alignment Search Tool (BLAST) recognized one such peptide, GRRAGGS, to be identical to a short peptide sequence within interleukin 11 (IL-11).3,7 Subsequent in vitro and in vivo analyses identified the IL-11 receptor alpha (IL-11Rα) as the cognate binding partner. In fact, high levels of IL-11Rα expression in tissue sections from prostate cancer patients correlated with disease progression.8 IL-11Rα overexpression was also confirmed in breast cancer,9 osteosarcoma,10 and prostate cancer metastases.11
High levels of IL-11Rα in prostate cancer metastases indicated IL-11Rα might also be expressed in the bone marrow in leukemia and lymphoma patients.12 Indeed, this was confirmed by immunohistochemistry and flow cytometry in all of the leukemia, myeloma, and lymphoma cell lines tested — MOLT-4, OCI-AML3, K562, KMB7, THP-1, HL-60, CCRF-CEM, TF-1, SR-786, TUR, RPMI-8226, and U937. Additionally, IL-11Rα was detected in patient-derived bone marrow samples, including acute myeloid leukemia (AML; n=33), myelodysplastic syndrome (n=4), myeloproliferative syndrome (n=2), and B-cell malignancies (n=4). Myelodysplastic syndrome bone marrow specimens showed focal disease involvement, whereas other cases showed significant disease involvement ranging from 40 to 60 percent.
A ligand-directed drug candidate, bone metastasis targeting peptidomimetic-11 (BMTP-11), was developed; it consists of the CGRRAGGSC peptide fused to the antibacterial apoptotic peptide, D(KLAKKLAK)2 through a glycinylglycine linker. D(KLAKKLAK)2 disrupts mitochondrial membranes and is toxic to eukaryotic cells upon cell internalization. We hypothesized that the targeting peptide would guide the apoptotic peptide to the target site and, upon cellular internalization, lead to cell death in tumor cells that express IL-11Rα. Following preclinical studies, a first-in-man phase 0 clinical trial of BMTP-11 in castrate-resistant prostate cancer patients with osteoblastic bone metastasis (NCT00872157) confirmed selective BMTP-11 localization and apoptosis induction in tumors in the bone marrow.11 In vitro studies showed that BMTP-11 preferentially induces cell death in MOLT-4 leukemia cells compared with normal white blood cells.12 Interestingly, the myristoylated BMTP-11 analog increased cell death by ninefold in cultured OCI-AML3, K562, and MOLT-4 cells compared to the efficacy of the parental BMTP-11 in MOLT-4 cells in the same time frame.
Schematic of Drug Development Pipeline. Phage display in humans or with cultured tumor cells identified common and/or unique receptors that can be exploited for targeted drug delivery. In the case of BMTP-11 or myristoylated BMTP-11, the targeting peptide ligand, CGRRAGGSC, was synthetically fused to the apoptotic peptide, D(KLAKKLAK)2, and validated in vivo and in vitro, respectively. BMTP-11 was advanced to preclinical studies, and following FDA approval, BMTP-11 was tested as a new investigational drug in a first-in-man clinical trial.
Schematic of Drug Development Pipeline. Phage display in humans or with cultured tumor cells identified common and/or unique receptors that can be exploited for targeted drug delivery. In the case of BMTP-11 or myristoylated BMTP-11, the targeting peptide ligand, CGRRAGGSC, was synthetically fused to the apoptotic peptide, D(KLAKKLAK)2, and validated in vivo and in vitro, respectively. BMTP-11 was advanced to preclinical studies, and following FDA approval, BMTP-11 was tested as a new investigational drug in a first-in-man clinical trial.
Other cancer-specific signatures have also been discovered and characterized. For example, screening human leukemia and lymphoma cell lines and patient-derived AML and acute lymphoblastic leukemia (ALL) bone marrow using a subtractive cell-targeting technology identified a cell-internalizing peptide motif, Phe-Phe/Tyr-X-Leu-Arg-Ser (FF/YXLRS), where X is any amino acid residue.13 Further analyses revealed that FF/YXLRS binds to the neuropilin-1 (NRP-1) receptor. Treatment of cultured MOLT-4, CCRF-CEM, OCI-AML3, HL-60, K562, SR-786, U937, or RPMI-8226 cells with the NRP-1 targeting peptide, CGFYWLRSC, fused to D(KLAKKLAK)2 decreased cell viability in the 5- to 30-µm range. Analyses of bone marrow samples from AML and ALL patients confirmed NRP-1 expression compared to normal bone marrow. These data, together with the promising results of BMTP-11 in targeting prostate cancer metastases and the improved efficacy of the myristoylated BMTP-11 analog against leukemia and lymphoma cells indicate that tumor-specific molecular zip codes exist and can be effectively exploited to design drugs to systemically treat cancers (see Figure for our drug development pipeline).
