Our laboratory recently reported that treatment with the d-amino acid containing peptide HYD1 induces necrotic cell death in multiple myeloma (MM) cell lines. Due to the intriguing biological activity and promising in vivo activity of HYD1, we pursued strategies for increasing the therapeutic efficacy of the parent linear peptide. These efforts led to the development of a cyclized peptidomimetic, MTI-101, with increased in vitro activity and robust in vivo activity as a single agent using two myeloma models that consider the bone marrow microenvironment. MTI-101 treatment resulted in mechanistically similar hallmarks of HYD1 induced cell death, namely the generation of ROS, depletion of ATP levels, and failure to activate caspase-3. Moreover, MTI-101 was shown to be cross-resistant in the HYD1 acquired resistant H929-60 cell line that was previously developed in our laboratory. In the present study, we pursued an unbiased chemical biology approach using biotinylated peptide affinity purification and LC-MS/MS analysis to identify binding partners of MTI-101. Using this approach, CD44 was identified as a predominant binding partner. Using an ELISA based assay, we showed that biotinylated peptide bound to full length recombinant human CD44 in a concentration dependent manner. Reducing the cell surface expression of CD44 was accompanied by the activation of caspase-3 and cell death was observed in the NCI-H929 and U266 MM cell lines, indicating that MM cells require CD44 expression for survival. Ectopic expression of CD44s correlated with increased binding of the FAM-conjugated peptide in the 8226 MM cell line, and this was further corroborated using CD44 knockout mice which also showed less peptide binding compared to wild-type. However, ectopic expression of CD44s was not sufficient to increase the sensitivity to MTI-101 induced cell death. Mechanistically, we show that MTI-101 induced a pro-survival signal through the activation of Erk1/2 and that CD44 formed a complex with Pyk2. These data corroborate with that of which was previously observed with the parental peptide being a partial agonist and inducing an autophagic survival signal. With respect to cell death, we showed that CD44 forms a complex with known death inducing proteins caspase-8, caspase-10, Rip1, Rip3, Drp1, TNFAIP8, and PGAM5. Furthermore, we demonstrated that MTI-101 induced mitochondrial fission which may be modulated by a Rip1, Rip3 or Drp1 dependent and independent pathway. Finally, we show that MTI-101 has robust activity as a single agent in the SCID-Hu bone implant and 5TGM1 in vivo model of multiple myeloma. Together these data continue to support the further development of this class of compounds as well as identify CD44 as a therapeutic target for the treatment of MM.

In our previous work we demonstrated that the integrin antagonist, HYD1, induced necrotic cell death in myeloma cell lines in vitro and in vivo as a single agent. In order to further delineate biomarkers of response to HYD1 we developed an isogenic drug resistant variant named H929-60. We show that the acquisition of resistance towards HYD1 correlates with reduced expression of the cleaved α4 integrin subunit and beta 1 integrin. Moreover, we demonstrate that HYD1 interacts with α4 integrin in myeloma cells. Consistent with reduced VLA-4 expression, the resistant variant showed ablated functional binding to fibronectin, VCAM-1 and the bone marrow stroma cell line, HS-5. The reduction in binding to extracellular matrices of the resistant variant translated to sensitivity to melphalan and bortezomib induced cell death in the bone marrow stroma co-culture model of drug resistance. Moreover, CD138 positive myeloma cells were more sensitive to HYD1 induced cell death compared to the CD138 negative fraction, and potency of HYD1 induced cell death significantly correlated with α4 integrin expression. We were also able to show that reducing α4 or β1 integrin using shRNA strategies was sufficient to cause resistance in myeloma cell lines. In addition we investigated the effects of cyclized variants of HYD1 to improve potency of the agent. One such compound, named HM-27, was determined to be 30 fold more active in H929 cells when compared to HYD1. HM-27 and HYD1 were determined to have similar mechanisms of action as H929-60 cells were shown to be resistant to both compounds when compared to H929 cells. We further characterized HM-27's mechanism of action by investigating what effects HM-27 induced Ca2+ oscillations had on HM-27 induced cell death. The increases in intracellular Ca2+ seen after treatment with HM-27 were determined to occur via release from ER stores and not through influx through plasma membrane channels. Inhibiting Ca2+ release from the ER also potentiated the effects of HM-27 in MM cells. Furthermore, inhibiting Ca2+ release from the ER was also shown to block the onset of autophagy after ER treatment. Treating cells with the lysosomotropic agent, chloroquine, was shown to potentiate the activity of HM-27 in vitro and ex vivo. HM-27 was also shown to have activity in an in vivo model with combination treatment containing bortezomib and HM-27 increasing mouse survival. Collectively our data indicate that VLA-4 expression is a critical determinant of response to HYD1 induced cell death. We also showed that increases in intracellular Ca2+ seen after treatment with HM-27 had a cytoprotective effect in MM cells. Moreover, neutralizing autophagy potentiates HM-27 induced cell death in vitro and ex vivo while combining bortezomib and HM-27 increased survival in vivo. These data continue to provide rationale for further pre-clinical development of HYD1 as a novel anticancer agent.