ORP4L: Can Targeting an MCS Component Provide Tools for Eradication of Leukemia?

Abstract

The study commented here reports that the OSBP homologue ORP4L is aberrantly induced in CD34⁺CD38⁻ leukemia stem cells (LSCs) of patients with acute myeloid leukemia and acts as an accessory factor for phospholipase C β3 (PLCβ3), a PLC isoform with a dominant role in these cells. Our mechanistic data suggest that ORP4L extracts and presents PI4,5P₂ for catalysis by PLCβ3, thus controlling Ca²⁺ oscillations, bioenergetics, and survival of the cells. A small molecular compound LYZ-81 is described as a specific inhibitor of ORP4L, which can be employed to eradicate LSCs in vitro and in vivo in NOD/SCID mice. Our observations identify ORP4L as a potential target for new leukemia therapies.

Commentary to: Zhong, W., Xu, M., Li, C., Zhu, B., Cao, X., Li, D., … Yan, D. (2019). ORP4L Extracts and presents PIP₂ from plasma membrane for PLCβ3 catalysis: Targeting it eradicates leukemia stem cells. Cell Reports, 26(8), 2166–2177.e9. doi:10.1016/j.celrep.2019.01.082.

Leukemia stem cells (LSCs) are a subpopulation of abnormal hematopoietic stem cells (HSCs) that propagate leukemias and are responsible for the high frequency of relapse characteristic of current leukemia therapies (Elert, 2013; Lapidot et al., 1994; Schepers, Campbell, & Passegue, 2015). LCSs were initially defined as having the CD34⁺CD38⁻ phenotype (Bonnet & Dick, 1997), similar to normal HSCs, and despite the identification of also other immunophenotypically defined leukemia-initiating cell populations, CD34⁺CD38⁻ cells remain the best-characterized and most potent LSC phenotype in xenograft mouse models and retransplantation experiments. In a recent study performed in the laboratory of D. Yan at Jinan University, Guangzhou, China (Zhong et al., 2019), we provide evidence suggesting that targeting a membrane contact site (MCS) component, OSBP-related protein 4 (ORP4, also designated OSBP2), could help us develop new tools to combat leukemias by exterminating LSCs.

Unlike its ubiquitously expressed essential homologue oxysterol-binding protein (OSBP/OSBP1), ORP4 is expressed only in a limited number of tissues and cell types, and its knock-out in mice has relatively benign phenotypic effects featured by male sterility, consistent with its abundant expression in testis (Udagawa et al., 2014). Burgett et al. (2011) identified ORP4, together with OSBP, as a target of antiproliferative natural compounds termed ORPphilins, which inhibit the growth of a number of cancer cell lines. Consistently, work by the group of N. Ridgway showed that deep knock-down of ORP4 in cancer-derived or immortal cell lines resulted in quiescence, while in untransformed epithelial cells this caused apoptotic cell death, suggesting that ORP4 promotes the survival of rapidly proliferating cells (Charman et al., 2014). We thereafter demonstrated that ORP4L in aberrantly induced in acute T-cell lymphocytic leukemia (T-ALL) cells but absent in healthy human T-cells and plays an important role in G-protein coupled signaling, endoplasmic reticulum (ER) Ca²⁺ release and mitochondrial oxidative phosphorylation in the malignant T-ALL cells, crucial for the survival of these cells (Zhong et al., 2016).

In the latest study (Zhong et al., 2019), we further pursued the mechanism of function of ORP4L, this time in CD34⁺CD38⁻ leukemia stem cells isolated from patients suffering from acute myeloid leukemia. Like for T-ALL cells, ORP4L was shown to be expressed in the LSCs but not in normal HSCs. Also in this cell type, ORP4L executes a key role in the Ca²⁺ control of mitochondrial energy metabolism, its key interaction partner being phospholipase C β3 (PLCβ3), an isoform with a dominant role in the malignant cells. Our analyses by employing a cell-free reconstituted system provided evidence that ORP4L in fact extracts phosphatidylinositol-4,5-bisphosphate (PI4,5P₂) from regions of the plasma membrane (PM), introducing it to PLCβ3 for hydrolysis into diacylglycerol (DAG) and inositol trisphosphate (IP₃), which mediates cytosolic and mitochondrial Ca²⁺ oscillations crucial for active bioenergetics (Figure 1). This type of function had not previously been suggested or documented for any of the ORP family members, nor was it appreciated that PLCs would require such an accessory component for activity. However, such lipid “liftase” activity has been reported for phosphatidylinositol transfer proteins (PITPs) stimulating PI kinase activity (Cockcroft, 2012) and for saposins (SAPs, sphingolipid activator proteins) that present glycosphingolipids for lysosomal hydrolases (Sandhoff, 2013). Moreover, molecular dynamics simulations suggested quite significant affinity of the PLC reaction product IP₃ for the ORP4L inositol phosphate binding cleft (suggested cost of pulling it out was 17.85 ± 2.19 kcal/mol), raising the possibility that distinct mechanisms may be required for its release and routing to the receptors in the ER.

Figure 1.

A schematic of ORP4L function as a cofactor of PLCβ3. According to the data in Zhong et al. (2019), ORP4L pulls out PIP₂ from the PM, introducing it to PLCβ3 for catalysis to produce the second messengers DAG and IP₃. Whether ORP4L could, via interaction with VAPs mediated by its two phenylalanines in an acidic tract (FFAT) motif, directly couple IP₃ generation at the PM and its channeling to IPTR in the ER is an interesting subject of future study. PM = plasma membrane; PIP₂ = PI4,5P₂; ER = endoplasmic reticulum; IPTR = IP₃ receptors; DAG = diacylglycerol; IP₃ = inositol trisphosphate; VAP = VAMP-associated protein; PLCβ3 = phospholipase C β3.

ORP4L was further demonstrated to be necessary for the survival of LSCs, its knock-down reducing the colony-forming capacity of the LSCs in vitro and the in vivo engraftment of LSCs in NOD/SCID mice. We next employed the ORPphilin OSW-1, reducing its general cytotoxicity by generating a derivative, LYZ-81, selective for ORP4L and avoiding the ubiquitous essential protein OSBP. The IC₅₀ values of this compound for LSCs and HSCs were 3.27 nM and 1.29 μM, respectively, and it bound ORP4L at a K_d of 1.05 ± 0.24 nM, while its K_d for OSBP was 5.9 ± 1.9 μM. Of note, OSW-1 displays a similar nM K_d for both proteins. LYZ-81 competed for PI4,5P₂ binding to ORP4L, inhibited the extraction of PI4,5P₂ from membranes and production of the second messengers DAG and IP₃. By employing this small molecular compound, we demonstrated that one can also by pharmacologic targeting of ORP4L eradicate human LSCs both in vitro and in vivo. Although LYZ-81 may not be an optimal lead compound for the development of pharmaceuticals, this study identifies ORP4L as a potential target of future leukemia therapy development.

ORP4L interacts with the endoplasmic reticulum (ER) VAMP-associated proteins (VAPs) and binds cholesterol, oxysterols, and phosphoinositides (Charman et al., 2014; Wang, JeBailey, & Ridgway, 2002; Wyles, Perry, & Ridgway, 2007). It has been localized at the trans-Golgi network (TGN), ER-Golgi contact sites, and the PM as well as on vimentin intermediate filaments (Charman et al., 2014; Pietrangelo & Ridgway, 2018; Zhong et al., 2016). The data by N. Ridgway’s group suggest that the OSBP-, lipid-, and VAP-regulated interactions of ORP4L with ER-Golgi contact sites are involved in the maintenance of Golgi and TGN structure (Pietrangelo & Ridgway, 2018). It is thus far unclear whether the PM localization of ORP4L, which is consistent with its interaction with the PM signaling enzyme PLCβ3, could in fact reflect its presence at ER-PM contact sites. This is not necessarily the case, as the PM targeting of ORP4L did not depend on its interaction with VAPs nor did ORP4L colocalize with the ER protein mCherry-VAPA at the PM (Pietrangelo & Ridgway, 2018). Furthermore, our earlier data indicated that VAP binding by ORP4L may not be necessary for its PLCβ3 scaffolding activity (Zhong et al., 2016). Also, whether the function of ORP4L at ER-Golgi MCSs could be involved in the discovered control of PLC activity is unknown. These issues definitely require further study: Considering that ORP4L does bind VAPs, one is tempted to speculate that ORP4L, which scaffolds a PLCβ3 signaling complex and introduces PI4,5P₂ to PLCβ3 (Zhong et al., 2016, 2019), could also couple the generation of IP₃ by the PLC to its direct routing to the ER IP₃ receptors at an ER-PM contact site (Figure 1).

To conclude, the study commented here (Zhong et al., 2019) reveals a new mechanism of ORP function in promoting PLC signaling. Importantly, it also suggests that targeting ORP4L, which is aberrantly induced in LSCs, may have significant therapeutic potential.