PDF file - 1112K, Supplementary Figure 1: Surface Expression Of ROR1 Associates With Breast Cancer Cell-lines That Have High-metastatic Potential. Supplementary Figure 2: High-level Expression of ROR1 In Breast Cancer Is Associated With Shorter Lung, Bone and Brain Metastasis-free Survival. Supplementary Figure 3: High-Level Expression Of ROR1 In Breast Cancer Is Associated With Shorter Metastasis-Free Survival, and Independent from their ER, PR and HER2 status. Supplementary Figure 4. Expression of ROR1 By Breast Cancer Cell Lines Is Associated With Features Of EMT. Supplementary Figure 5: Silencing ROR1 Reduces Expression Of CXCR4. Supplementary Figure 6. Silencing ROR1 Regulates EMT Genes expression. Supplementary Figure 7: Silencing ROR1 Effects Modest Late-Growth Inhibition Of Orthotopic Xenografts At The Site Of Injection But Strong Inhibition Of Experimental Pulmonary Metastases. Supplementary Figure 8: Immunohistochemistry of Experimental Metastatic Foci. Supplementary Figure 9. Silencing ROR1 Reduces Pulmonary Metastasis And Bone Metastasis Of MDA-MB-231 Derived Cell lines LM2-4175 And BoM-1833 In Vivo. Supplementary Figure 10: Silencing ROR1 Inhibits Migration Of HS-578T and BT549 Migration In Vitro. Supplementary Table 1: Summary Of Human Breast Cancer Cell Line Characteristics With Respect To EMT Status, Subgrouping, Invasiveness And Matrigel Morphology. Supplementary Table 2: Correlation Of ROR1 Status With Various Clinical Features Of Breast Cancer. Supplementary Table 3: Expression Of ROR1 Serves As An Independent Risk Factor For Short Metastasis-free Survival.

Key Points Nurse-like cells express Wnt5a to induce ROR1-dependent stimulation of NF-κB, which leads to autocrine IL-6-induced STAT3 activation in CLL cells. Cirmtuzumab inhibits Wnt5a-induced, ROR1-dependent stimulation of NF-κB, and thereby represses autocrine IL-6-dependent STAT3 activation in CLL.

ROR1 is an oncoembryonic receptor tyrosine kinase expressed on chronic lymphocytic leukemia (CLL) B cells, but not on most normal post-partum tissues. It functions as a receptor for Wnt5a, which is present at high levels in the plasma of patients with CLL relative to that of age-matched controls. Wnt5a/ROR1 is known to activate pro-survival signals in CLL cells, but detailed mechanisms are not fully understood. We found that monocyte-derived Nurse-Like Cells (NLC) could induce CLL cells to activate STAT3 (pY705) and that this effect could be blocked by neutralizing antibodies to Wnt5a, which we found expressed at high levels by NLC. We observed that this effect also could be blocked by cirmtuzumab, a humanized mAb that can inhibit ROR1-signaling, indicating this effect of NLC on CLL cells was Wnt5a and ROR1 dependent. We performed a time-course study examining for pSTAT3 in isolated serum-starved CLL cells treated with exogenous Wnt5a. This revealed that Wnt5a induced delayed activation of STAT3, appearing first at 3 hours. As such, we hypothesized that the noted activation of STAT3 in CLL cells was indirect, being caused by a factor(s) made by isolated CLL cells in response to Wnt5a. We examined harvested supernatants of CLL cells cultured with or without Wnt5a using a human cytokine array assay. Multiple proinflammatory factors including IL-6, IL-8, CCL2, CCL3, CCL4, and CXCL1 were detected in the medium of CLL cells cultured for 24 hours with Wnt5a that were either not detected or found at lower levels in the medium of CLL cells cultured without Wnt5a. Moreover, the medium harvested from CLL cells cultured with Wnt5a, but not the medium of CLL cells cultured without, could induce pSTAT3 within 30 minutes in serum-starved CLL cells; this effect could be inhibited by the anti-IL-6-receptor mAb, tocilizumab, but not by cirmtuzumab, even when used at concentrations that could block the capacity of Wnt5a to induce latent activation of STAT3 and production of IL-6. Since the genes encoding these proinflammatory factors are targets of nuclear factor kappa B (NF-κB), we hypothesized that they were induced through Wnt5a/ROR1-dependent activation of NF-κB. Consistent with this notion, we found that BAY 11-7082 or BMS-345541, which can each inhibit activation of NF-κB, also could block Wnt5a-induced CLL-cell activation of STAT3 and production of IL-6. Furthermore, Wnt5a could induce phosphorylation of NF-κB p65 in CLL cells within 30 minutes, and this effect could be blocked by cirmtuzumab. Using real-time PCR array to evaluate for expression of NF-κB target genes, we found Wnt5s could induce up-regulation of NF-κB target genes in CLL cells, and that this effect could be blocked by cirmtuzumab. Moreover, these studies revealed that Wnt5a/ROR1/NF-κB signaling could induce expression of genes encoding the noted proinflammatory factors in CLL cells. To examine the in vivo significance of these findings, we collected plasma and CLL cells from patients treated with cirmtuzumab in a phase I clinical trial (Choi, MY, et al, Cell Stem Cell 22:951, 2018). RNAseq and ELISA respectively were used to examine the transcriptomes of negatively-selected CLL cells and the concentrations of IL-6 in plasma collected before and after treatment. Consistent with our in vitro findings, treatment with cirmtuzumab downregulated CLL-cell expression of NF-κB target genes in vivo by gene set enrichment analysis (n = 3, NES = 2.10, FDR q = 0.01). The levels of IL-6 in plasma also were significantly decreased in patients after therapy (p = 0.02, n = 5, Paired Student t test). Collectively, these studies indicate that Wnt5a/ROR1-dependent signaling induced by NLC may play a major role in the noted activation of NF-κB in CLL, leading to the production of factors, such as IL-6, which are posted to contribute to pathogenesis. Moreover, these data suggest that some of the noted clinical effects of therapy with cirmtuzumab may be due to suppression of Wnt5a-induced, ROR1-dependent activation of NF-κB in patients with CLL.

ROR1 is a receptor tyrosine kinase-like orphan receptor for Wnt5a that is expressed by cells during embryogenesis and by the neoplastic cells of a variety of cancers, including chronic lymphocytic leukemia (CLL). ROR1 can induce activation of β-catenin-independent non-canonical Wnt-signaling. Studies reveal a cross-talk between the non-canonical Wnt-signaling pathway and the β-catenin-dependent canonical Wnt-signaling pathway, which we previously found was also activated in CLL (Lu D, et al, PNAS 101:31118-3123, 2004). Prior studies indicated that silencing a related Wnt5a receptor, ROR2, could augment canonical Wnt-signaling induced by Wnt3a. In this study, we examined whether genetic silencing of ROR1 or inhibition of ROR1-signaling also could influence canonical Wnt signaling. To inhibit ROR1 signaling we used the humanized anti-ROR1 mAb cirmtuzumab, which is being evaluated in patients with CLL (Choi MY, et al, Cell Stem Cell, 22:951, 2018). Surprisingly, we found that CRISPR/Cas9 deletion of ROR1 in 293T cells also could enhance the capacity of Wnt3a to activate canonical Wnt-signaling, albeit to a lesser extent than CRISPR/Cas9 deletion of ROR2; conversely, re-introduction of ROR1 into ROR1-deleted 293T cells suppressed Wnt3a-induced activation of canonical Wnt-signaling. In contrast, treatment of wildtype 293T cells with cirmtuzumab did not enhance Wnt3a-induced activation of canonical Wnt-signaling, but nonetheless suppressed ROR1-dependent non-canonical signaling induced by Wnt5a. We examined the influence of ROR1 on canonical Wnt-signaling in CLL. First, we examined whether the relative expression of ROR1 influenced the relative levels of genes induced by activation of the canonical Wnt-signaling pathway. Gene set enrichment analysis (GSEA) of gene-expression data on CLL cells of different patients (n=448, GSE13204) revealed that CLL cells with low-level expression of ROR1 (ROR1Low) did not have increased levels of genes induced by activation of canonical Wnt-signaling relative to those noted in CLL cells with high-level expression of ROR1 (ROR1Hi). Nonetheless, ROR1Hi CLL cells did have increased levels of genes induced by activation of non-canonical Wnt signaling compared to ROR1Low CLL cells. As in 293T cells, siRNA-mediated knockdown of ROR1 in ROR1Hi CLL cells could enhance the capacity of Wnt3a to increase the levels of genes induced by canonical Wnt-signaling (e.g. MYC, CCND1). However, treatment of the same CLL cells with cirmtuzumab did not enhance the levels of such genes in response to Wnt3a, even at concentrations that exceeded those required to inhibit Wnt5a-induced ROR1-dependent non-canonical Wnt-signaling. We examined whether these findings also applied to CLL cells treated with cirmtuzumab in vivo. For this, we performed gene-set enrichment analyses on the transcriptomes of CLL cells collected from patients before and after treatment with cirmtuzumab in a recently completed phase I clinical trial (Choi MY, et al, Cell Stem Cell, 22:951, 2018). Although treatment with cirmtuzumab repressed expression of genes induced by activation of non-canonical Wnt signaling, we did not observe changes in the levels of genes induced by activation of the canonical Wnt signaling pathway. Collectively, this study demonstrates that cirmtuzumab can inhibit non-canonical Wnt signaling without enhancing canonical Wnt signaling in CLL, in contrast to what we observed in CLL cells silenced for ROR1. As such, treatment with cirmtuzumab may represent a more selective approach to suppressing non-canonical Wnt5a signaling than strategies aimed at genetic down-modulation or silencing of ROR1.

Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is an oncoembryonic protein expressed on chronic lymphocytic leukemia (CLL) that can serve as a receptor for Wnt5a, which can promote leukemia cell migration, proliferation, and survival. We found Wnt5a could induce ROR1 to complex with DOCK2 (dedicator of cytokinesis 2) and induce activation of Rac1/2; these effects could be blocked by cirmtuzumab, a humanized anti-ROR1 monoclonal antibody. We find that silencing DOCK2 specifically impaired the capacity of Wnt5a to induce activation of Rac1/2 or enhance CLL cell proliferation. We generated truncated forms of ROR1 and found the cytoplasmic proline-rich domain (PRD) of ROR1 was required for Wnt5a to induce ROR1 to complex with DOCK2 and activate Rac1/2 in the CLL cell-line MEC1. We introduced single amino acid substitutions of proline (P) to alanine (A) in the ROR1-PRD at potential binding sites for the Src-homology 3 domain of DOCK2. In contrast to wild-type ROR1, or other ROR1 P→A variants, ROR1P808A was unable to recruit DOCK2 in response to Wnt5a. Moreover, unlike MEC1 cells transfected with wild-type ROR1 or ROR1 with P→A substitutions at positions 784, 826, or 841, MEC1 cells transfected to express ROR1P808A did not have a growth advantage over MEC1 cells that do not express ROR1. This study reveals that the recruitment of DOCK2 may be critical for the capacity of Wnt5a to enhance CLL proliferation, which may contribute to the observed increased tendency for disease progression in patients who have CLL cells that express high levels of ROR1.

Cirmtuzumab may enhance the therapeutic activity of ibrutinib by inhibiting ROR1-dependent signaling pathway in patients with chronic lymphocytic leukemia (CLL). Mantle cell lymphoma (MCL) is B-cell malignancy that also expresses ROR1. In this study, we found that the plasma of patients with MCL had high levels of Wnt5a, a ROR1 ligand, that were comparable to those found in patients with CLL; in contrast Wnt5a was virtually undetectable in the plasma of age-matched healthy adults. We also found that Wnt5a induced Rac1 activation in the primary MCL cells. Cirmtuzumab, but not ibrutinib, could inhibit the capacity of Wnt5a to induce primary MCL cells to activate Rac1. Addition of exogenous Wnt5a in vitro significantly enhanced the numbers of MCL cell divisions and the proportion of dividing MCL cells entering S/G2 in MCL cells over time in the presence of CD154 and IL-4/10. Treatment of the MCL cells with cirmtuzumab, but not ibrutinib, blocked Wnt5a-enhanced proliferation of MCL cells. This study indicates that cirmtuzumab and ibrutinib may have complementary activity in the treatment of patients with MCL.

Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is an oncoembryonic antigen that is expressed on CLL cells, but not on normal postpartum tissues. We found that ROR1 was a receptor for Wnt5a, which could enhance CLL-cell proliferation (Yu J, et al, JCI 126:585, 2016; Yu J, et al, Leukemia 31:2608, 2017; Hasan MK, et al, Blood 132:170, 2018). We performed mass spectrometry-based proteomics to interrogate immune-precipitates of Wnt5a-activated ROR1 and identified Ca2+/calmodulin-dependent protein kinase II (CaMKII), a serine/threonine-specific protein kinase. Recent studies demonstrated that high levels of different isoforms of CaMK, especially CaMKII, were expressed in several cancers. CaMKII phosphorylates nearly 40 different proteins, including enzymes, ion channels, kinases, and transcription factors and can play a critical role in the regulation of proliferation and survival of various cancer cells (Wang YY, et al, Oncotarget 20:11725, 2015; Chi M, et al, Sci Rep 6:33132, 2016; Hoffman A, et al, Cell Signal 26:748, 2014). We validated our mass spectrometry finings in co-immunoprecipitation studies and immunoblot analyses. These studies showed that Wnt5a could induce the specific phosphorylation and recruitment of CaMKII to ROR1. Moreover, these studies demonstrated that Wnt5a-induced recruitment of CaMKII to ROR1 could be blocked by cirmtuzumab, a first-in-class humanized anti-ROR1 mAb undergoing clinical testing in patients with CLL (Choi MY, et al, Cell Stem Cell 22:951, 2018). CaMKs serve as a ubiquitous intracellular receptor for Ca2+ and their activity are regulated through Ca2+ signaling. Therefore, we examined calcium influx in CLL cells by flow cytometry. Treatment of CLL cells cultured overnight in serum-free medium induced calcium influx within seconds. Moreover, the relative capacity of Wnt5a to induce calcium influx was associated with the relative expression level of ROR1 on CLL cells of different patients. Furthermore, Wnt5a could enhance calcium influx induced by treatment with anti-µ in CLL cells that expressed ROR1, but not in CLL cells that lacked expression of ROR1. Knockdown of ROR1 using ROR1 siRNA or treatment with cirmtuzumab inhibited the capacity of Wnt5a to enhance calcium influx induced by treatment with anti-µ. Collectively, these data demonstrate that Wnt5a-induced ROR1-signaling in CLL can influence BCR-signaling and activation of the Ca2+/calmodulin-dependent protein kinase II signaling pathway.

Cirmtuzumab is a humanized monoclonal antibody (mAb) that targets ROR1, an oncoembryonic orphan receptor for Wnt5a found on cancer stem cells (CSCs). Aberrant expression of ROR1 is seen in many malignancies and has been linked to Rho-GTPase activation and cancer stem cell self-renewal. For patients with chronic lymphocytic leukemia (CLL), self-renewing, neoplastic B cells express ROR1 in 95% of cases. High-level leukemia cell expression of ROR1 is associated with an unfavorable prognosis. We conducted a phase 1 study involving 26 patients with progressive, relapsed, or refractory CLL. Patients received four biweekly infusions, with doses ranging from 0.015 to 20 mg/kg. Cirmtuzumab had a long plasma half-life and did not have dose-limiting toxicity. Inhibition of ROR1 signaling was observed, including decreased activation of RhoA and HS1. Transcriptome analyses showed that therapy inhibited CLL stemness gene expression signatures in vivo. Cirmtuzumab is safe and effective at inhibiting tumor cell ROR1 signaling in patients with CLL.

Small-molecule drugs that inhibit survival signals in leukemia cells have demonstrated high efficacy in the treatment of patients (pts) with chronic lymphocytic leukemia (CLL). Although not directly cytotoxic for leukemia cells, these drugs block key signaling pathways that govern leukemia-cell trafficking, proliferation, and survival. However, non-specific inhibition of enzymes involved in other signaling pathways in non-neoplastic cells may result in toxicity, which can limit the therapeutic index of such drugs, at least in some settings.

Wnt5a can activate Rho GTPases in chronic lymphocytic leukemia (CLL) cells by inducing the recruitment of ARHGEF2 to ROR1. Mass spectrometry on immune precipitates of Wnt5a-activated ROR1 identified 14-3-3ζ, which was confirmed by co-immunoprecipitation. The capacity of Wnt5a to induce ROR1 to complex with 14-3-3ζ could be blocked in CLL cells by treatment with cirmtuzumab, a humanized mAb targeting ROR1. Silencing 14-3-3ζ via small interfering RNA impaired the capacity of Wnt5a to: (1) induce recruitment of ARHGEF2 to ROR1, (2) enhance in vitro exchange activity of ARHGEF2 and (3) induce activation of RhoA and Rac1 in CLL cells. Furthermore, CRISPR/Cas9 deletion of 14-3-3ζ in ROR1-negative CLL cell-line MEC1, and in MEC1 cells transfected to express ROR1 (MEC1-ROR1), demonstrated that 14-3-3ζ was necessary for the growth/engraftment advantage of MEC1-ROR1 over MEC1 cells. We identified a binding motif (RSPS 857 SAS) in ROR1 for 14-3-3ζ. Site-directed mutagenesis of ROR1 demonstrated that serine-857 was required for the recruitment of 14-3-3ζ and ARHGEF2 to ROR1, and activation of RhoA and Rac1. Collectively, this study reveals that 14-3-3ζ plays a critical role in Wnt5a/ROR1 signaling, leading to enhanced CLL migration and proliferation.

Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is an oncoembryonic antigen expressed on chronic lymphocytic leukemia (CLL) cells, but not on normal postpartum tissues. The ligand that binds to ROR1 is Wnt5a, which we find present at high levels in the plasma of patients with CLL relative to that of age-matched healthy adults. Wnt5a can induce CLL-cell activation of Rho-GTPases in a ROR1-dependent manner, thereby enhancing leukemia-cell proliferation, migration, and survival. Currently, the cell-source(s) of Wnt5a in CLL is not known. Good candidates include Nurse-Like cells (NLCs), which are monocyte-derived accessory cells that develop in the context of CLL cells and that are found within the leukemia-cell microenvironment in lymphoid tissues. Already NLCs have been shown to elaborate factors that can promote leukemia-cell activation/survival, including chemokines (e.g. CXCL12, CXCL13) and members of the tumor-necrosis-factor family (e.g. BAFF and APRIL). Because activated monocytes can produce Wnt5a, we hypothesize that NLC also may produce this non-canonical Wnt factor. For this study, we generated NLCs from blood mononuclear cells of CLL patients through co-culture with leukemia B cells using established methods. NLCs, which stained positive for CD163 and CD68, were isolated free of leukemia cells via adherence. We isolated RNA from purified CLL cells or purified NLC of the same patient for real-time PCR to measure the levels of WNT5A transcripts. WNT5A transcripts were abundant in the RNA isolated from NLC, but were negligible in RNA isolated from CLL cells each patient tested (N = 5). We evaluated for Wnt5a via ELISA in the culture-medium of isolated NLCs of five different patients. The cultured medium of NLCs from each patient had detectable Wnt5a that increased in concentration over time. Furthermore, ultra-high resolution confocal fluorescent microscopy of NLC-CLL co-cultures showed Wnt5a was expressed at high-levels in NLCs, but was undetectable in CLL cells of each patient tested (N = 3). We performed co-culture studies and trans-well assays to assess the survival and migration of CLL cells when cultured with or without NLCs, and with or without a neutralizing mAb specific for Wnt5a or cirmtuzumab, which is a humanized anti-ROR1 mAb that blocks Wnt5a-induced, ROR1-dependent signaling. We found that anti-Wnt5a or cirmtuzumab, but not a mAb of irrelevant specificity, each could comparably and significantly inhibit the protective effects of NLCs for CLL cells in vitro (p= 0.008, N = 3 and p= 0.0032, N = 3, respectively by paired student's t-test). Moreover, anti-Wnt5a or cirmtuzumab each could comparably and significantly inhibit the migration of CLL cells enhanced by co-culture with NLCs in vitro (p < 0.0001, N = 3, by paired student's t-test). We also examined for Rho-GTPase activation in CLL cells co-cultured with NLCs. CLL cells co-cultured with NLCs, but not CLL cells cultured alone, had high-level activation of Rac1 and RhoA. The activation of RhoA and Rac1 in co-cultured CLL cells could be blocked by neutralizing anti-Wnt5a or cirmtuzumab, but not by control mAb of irrelevant specificity. We conclude that NLCs express Wnt5a, which can enhance CLL-cell migration and survival via a ROR1-dependent pathway. We speculate that the high-level Wnt5a noted in the plasma of patients with CLL is produced primarily by NLCs, which reside in lymphoid tissues. If so, then the high-level Wnt5a present in the plasma of CLL patients may extend the protective effects of NLCs beyond the CLL microenvironment in which NLCs reside. Treatment with cirmtuzumab to block ROR1-dependent Wnt5a-signaling may mitigate such effects, potentially enhancing the clearance of leukemia cells when used alone or in combination with drugs that target other survival-signaling pathways in CLL.

Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is an oncoembryonic protein expressed on chronic lymphocytic leukemia (CLL) B cells, but not on virtually all normal adult tissues. Studies have found that high-level leukemia-cell expression of ROR1 associated with enhanced leukemia-cell proliferation, aggressive disease, and shorter overall survival compared to that of patients with CLL cells that had low-to-negligible expression of ROR1 (Cui B. et al., Blood, 128:2931, 2016). Analysis of ROR1-immune precipitates of CLL-cell lysates via mass spectrometry and immunoblot analyses revealed that ROR1 was associated with DOCK2 (Dedicator of Cytokinesis 2) in freshly isolated CLL cells. DOCK2 is a member of the DOCK-A subfamily of guanine exchange factors (GEFs) specific for Rac1 and Rac2 (Rac1/2) that is expressed primarily in leukocytes, including CLL B cells. DOCK proteins differ from other GEFs in that they do not possess the canonical tandem Dbl-homology (DH) and Pleckstrin-homology (PH) domains that elicit nucleotide exchange. Instead DOCK2 possess a DOCK homology region 2 (DHR2) domain, which mediates Rac activation by stabilizing Rac in its nucleotide-free state. DOCK2 also contains a N-terminal SH3 domain. We found that ROR1 dissociated from DOCK2 in CLL cells cultured in serum-free medium, unless the cells were stimulated with exogenous Wnt5a, which we found present at high-levels in the plasma of CLL patients relative to that of age-matched healthy adults. Wnt5a could induce ROR1 to complex with DOCK2 and cause activation of Rac1/2; this effect could be inhibited by silencing DOCK2 in primary CLL cells. Furthermore, these effects of Wnt5a on CLL cells also could be completely blocked by treatment of the CLL cells with cirmtuzumab, a humanized anti-ROR1 mAb, which is undergoing clinical evaluation in patients with CLL. We find that silencing DOCK2 specifically impaired the capacity of Wnt5a to enhance CXCL12-directed migration or CLL-cell proliferation in vitro. We corroborated these findings using the CLL-cell-line MEC1, which also expresses DOCK2, but does not express ROR1. MEC1 cells can be transduced to express ROR1 or various mutant forms of ROR1, allowing us to examine the structure-function relationships required for ROR1-DOCK2 interactions. We confirmed that DOCK2 complexes with ROR1 in response to Wnt5a in MEC1-ROR1 cells. We generated truncated forms of ROR1 and found the cytoplastmic proline-rich domain (PRD) of ROR1 was required for ROR1 to complex with DOCK2 and activate Rac1/2 upon stimulation with Wnt5a. We next introduced single amino-acid substitutions of proline (P) to alanine (A) in the ROR1-PRD at positions 784, 808, 826, or 841 in potential SH3-binding sites. In contrast to wild-type ROR1, or other ROR1P==>A mutants, ROR1P(808)A had impaired capacity to recruit DOCK2 to ROR1 in response to Wnt5a. Moreover, unlike MEC1 cells transfected with wild-type ROR1 or ROR1 with P->A substitutions at positions 784, 826, or 841, MEC1 cells transfected to express ROR1P(808)A did not have a growth advantage over that of MEC1 cells that do not express ROR1. This study reveals that the recruitment of DOCK2 may be critical for the capacity of Wnt5a to enhance CLL proliferation, which may contribute to the observed increased tendency for disease progression in CLL patients who have leukemia cells that express high-levels of ROR1.

Signaling via the B cell receptor (BCR) plays an important role in the pathogenesis and progression of chronic lymphocytic leukemia (CLL) or mantle cell lymphoma (MCL). This is underscored by the clinical effectiveness of an inhibitor of Bruton's tyrosine kinase (BTK), ibrutinib, which can block BCR-signaling. However, ibrutinib cannot induce complete responses (CR) or durable remissions without continued therapy, particularly in patients with MCL. This suggests that ancillary pathways contribute to the growth/survival of MCL/CLL that are independent of BCR-signaling and/or inhibition of BTK. Studies have found that Wnt5a can promote activation of Rho GTPase Rac1 to enhance CLL-cell proliferation and survival via a ROR1-dependent,but BTK-independent pathway (Yu J, et al., Leukemia 31:1333, 2017). Furthermore, Wnt5a-induced Rac1 activation could be blocked by cirmtuzumab, a humanized anti-ROR1 mAb, which significantly could enhance the capacity of ibrutinib to clear CLL cells in vivo . MCL cells also express ROR1. Moreover, we find that the median level of ROR1 on primary MCL cells is higher than that on CLL cells; the median absolute mean fluorescence intensity of ROR1 (ROR1 DMFI) on MCL (87.5 ± 18.5 (S.D.), N = 8) was significantly higher than the median ROR1 DMFI on CLL cells (38.5 ± 18.5 (S.D.), N = 1567; (Cui B, et al., Blood 128:2931, 2016)). We also found that the plasma of patients with MCL had high levels of Wnt5a that were comparable to those found in patients with CLL; in contrast Wnt5a was virtually undetectable in the plasma of age-matched healthy adults. We cultured primary MCL cells of different patients with ibrutinib, cirmtuzumab, or both ibrutinib and cirmtuzumab for 2 h, and then stimulated the cells with exogenous Wnt5a for 30 min. In parallel, the same MCL cells were cultured without exogenous Wnt5a. We found that Wnt5a induced Rac1 activation in the primary MCL cells. Cirmtuzumab, but not ibrutinib, could inhibit the capacity of Wnt5a to induce primary MCL cells to activate Rac1. We induced proliferation of MCL cells by co-culturing MCL cells with HeLa cells expressing CD154 (HeLaCD154) and recombinant interleukin (IL)-4 and IL-10. Addition of exogenous Wnt5a to co-cultures of MCL cells with HeLaCD154 cells and IL-4/10 significantly enhanced the proportion of dividing MCL entering S/G2 and the numbers of cell divisions over time. Treatment of the MCL cells with cirmtuzumab, but not ibrutinib, blocked Wnt5a-enhanced survival and proliferation of MCL cells from each of 3 different patients. This study demonstrates that Wnt5a can induce Rac1 activation, which leads to enhanced proliferation and survival of primary MCL cells via a ROR1-dependent signaling pathway that can be blocked by cirmtuzumab. On the other hand, this signaling pathway cannot be blocked by ibrutinib, even at concentrations that completely inhibit BTK or BCR-signaling. This implies that cirmtuzumab and ibrutinib may have additive, if not synergistic, activity in the treatment of patients with MCL.

Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is an oncoembryonic cell surface protein, which is expressed on the neoplastic B cells of patients with chronic lymphocytic leukemia (CLL), but not on virtually all normal post-partem tissues. ROR1 functions as a receptor for Wnt5a, which induces activation of Rho-GTPases, AKT, and HS-1 and promotes migration, proliferation, and survival of CLL B cells. We observed that freshly isolated CLL cells also had high levels of pSTAT3 (Y705), which attenuated over time in culture in serum-free media unless the CLL cells were stimulated with exogenous Wnt5a. This effect of Wnt5a could be blocked by cirmtuzumab, a humanized anti-ROR1 mAb that currently is undergoing clinical evaluation in patients with CLL. We examined the capacity of recombinant Wnt5a (rWnt5a) to induce activation of STAT3 in isolated CLL cells by performing a time-course study, evaluating for pSTAT3 in serum-starved CLL cells at time 0, 5', 15', 30' and 1, 2, 3, 4, and 24 hours after stimulation with exogenous Wnt5a. Surprisingly, activation of STAT3 in CLL cells was delayed, appearing first at 3 hours after addition of rWnt5a, and peaking at 24 hours, the last time point examined. In contrast, activation of CLL cells via treatment of CLL cells with interleukin (IL)-6, induced activation of STAT3 within 15' of the addition of this cytokine to the cultures. We hypothesized that the activation of STAT3 in CLL cells by Wnt5a was indirect, being caused by a factor(s) made by isolated CLL cells in response to stimulation with Wnt5a. To test this hypothesis, we examined CLL cells, and harvested supernatants of cultured CLL cells, at various times after stimulation with rWnt5a. RNA isolated from CLL cells was examined via real-time PCR for cytokine transcripts. This revealed a prominent, time-dependent increase in transcripts encoding IL-6 following stimulation of CLL cells with rWnt5a. Consistent with this observation we found Wnt5a induced CLL cells to secrete IL-6, which was detected via ELISA at increasing concentrations over time in culture-supernatants of CLL cells treated with rWnt5a, but not in the culture supernatants of unstimulated CLL cells. Culture supernatants collected from CLL cells stimulated for 24 hours with rWnt5a could induce activation of STAT3 within 15' in serum-starved CLL cells; this effect could be inhibited significantly by the anti-IL-6-receptor mAb, tocilizumab, but not by cirmtuzumab, even when used at concentrations that could block the capacity of concomitantly-added Wnt5a to induce latent STAT3 activation at 24 hours, suggesting that cirmtuzumab could inhibit Wnt5a-induced CLL-cell production of cytokines, such as IL-6. We examined for IL-6 in cultures of CLL cells treated for 24 hours with rWnt5a and cirmtuzumab, or rWnt5a and an antibody of irrelevant specificity. We found the concentration of IL-6 in the supernatants of CLL cells treated with rWnt5a and cirmtuzumab (110.4 ± 18.9 pg/ml) was significantly lower than that found in the supernatants of CLL cells treated with rWnt5a and control antibody (416.2 ± 53.7 pg/ml, p = 0.0007, Students t test). This study reveals that Wnt5a can induce CLL to secrete cytokines, such as IL-6, via a ROR1-dependent pathway, leading to latent, autocrine activation of STAT3.

ROR1 (Receptor tyrosine kinase-like orphan receptor 1) is an evolutionarily conserved, oncoembryonic antigen that is expressed on CLL cells, but not on normal postpartum tissues. In prior studies we found that ROR1 was a receptor for Wnt5a, which could activate RhoGTPases and enhance chemokine-directed leukemia-cell migration. Moreover, Wnt5a-enhanced migration could be inhibited by cirmtuzumab, a first-in-class humanized mAb specific for a functional epitope of the ROR1 extracellular domain; this mAb is undergoing phase I clinical testing in patients with CLL. However, it is not known whether ROR1 contributes to cell migration only through its recruitment/activation of guanine exchange factors (GEFs) or whether it also can complex with proteins that directly interact with the actin cytoskeleton. We performed mass spectrometry-based proteomics to interrogate immune-precipitates of Wnt5a-activated ROR1 from CLL cells and identified HS1 (or HCLS1, hematopoietic cell-specific Lyn substrate 1), a conserved cytoplasmic protein that can interact with the actin cytoskeleton. Moreover, tyrosine phosphorylation of HS1 has been found in prior studies to be associated with enhanced leukemia-cell migration and adverse prognosis in CLL, although the stimuli/mechanism(s) contributing to HS1 activation were unknown. We validated that Wnt5a induced the association of ROR1 with HS1 in primary CLL cells via co-immunoprecipitation studies and immunoblot analyses. Moreover, we found that Wnt5a could induce tyrosine phosphorylation of HS1 in CLL cells previously deprived of Wnt5a. On the other hand, treatment of CLL cells with the anti-ROR1 mAb cirmtuzumab inhibited the capacity of Wnt5a to induce the association of ROR1 with HS1 or tyrosine phosphorylation of HS1. We speculated that the Src-homology 3 (SH3) domain of HS1 allowed it to dock onto the cytoplasmic proline-rich domain (PRD) of ROR1. Consistent with this notion we found that a mutant-form of ROR1 lacking the PRD was unable to complex with HS1 or facilitate F-actin polymerization in response to Wnt5a. Accordingly, we used site-directed mutagenesis to introduce single amino acid substitutions of proline (P) to alanine (A) in the P-X-X-P- motifs of the ROR1 PRD at positions 784, 808, 826, 841, or 850. Each mutant form of ROR1, except for P841A-ROR1, could complex with HS1, permit tyrosine phosphorylation of HS1, and allow for enhanced leukemia-cell motility. Collectively, these data indicate that Wnt5a can induce HS1 to dock at P841 in the PRD of ROR1 and undergo tyrosine phosphorylation, thereby enhancing F-actin polymerization and planar-cell-polarity leukemia-cell migration. Moreover, we found that cirmtuzumab can inhibit activation of HS1 and F-actin polymerization in response to Wnt5a, potentially accounting in part for its capacity to inhibit Wnt5a-enhanced leukemia-cell migration.

Signaling via the B cell receptor (BCR) plays an important role in the pathogenesis and progression of chronic lymphocytic leukemia (CLL). This is underscored by the clinical effectiveness of an inhibitor of Bruton's tyrosine kinase (BTK), ibrutinib, which can block BCR-signaling. However, ibrutinib cannot induce complete responses (CR) or durable remissions without continued therapy, suggesting that ancillary pathways contribute to CLL growth/survival that are independent of BCR-signaling. ROR1 is a receptor for Wnt5a, which can promote activation of Rac1 to enhance CLL-cell proliferation and survival. We hypothesized that the effects of ibrutinib on blocking BCR-signaling might be offset by non-canonical Wnt-signaling via ROR1. If so, then inhibition of both ROR1- and BCR-signaling might have an enhanced anti-tumor effect.

Evolutionarily conserved receptor tyrosine kinase–like orphan receptor-1 and -2 (ROR1/2) are considered distinct receptors for Wnt5a and are implicated in noncanonical Wnt signaling in organogenesis and cancer metastasis. We found that Wnt5a enhanced proliferation and migration of chronic lymphocytic leukemia (CLL) cells and that these effects were blocked by the humanized anti-ROR1 mAb cirmtuzumab (UC-961). Treatment of CLL cells with Wnt5a induced ROR1 to oligomerize with ROR2 and recruit guanine exchange factors (GEFs), which activated Rac1 and RhoA; siRNA-mediated silencing of either ROR1 or ROR2 or treatment with UC-961 inhibited these effects. Using the ROR1-deficient CLL cell line MEC1, we demonstrated that ectopic ROR1 expression induced ROR1/ROR2 heterooligomers, which recruited GEFs, and enhanced proliferation, cytokine-directed migration, and engraftment potential of MEC1 cells in immune-deficient mice. Notably, treatment with UC-961 inhibited engraftment of ROR1+ leukemia cells in immune-competent ROR1-transgenic mice. Molecular analysis revealed that the extracellular Kringle domain is required for ROR1/ROR2 heterooligomerization and the cysteine-rich domain or intracellular proline-rich domain is required for Wnt5a-induced recruitment of GEFs to ROR1/ROR2. This study identifies an interaction between ROR1 and ROR2 that is required for Wnt5a signaling that promotes leukemia chemotaxis and proliferation.

ROR1 is an oncoembryonic orphan receptor found on chronic lymphocytic leukemia (CLL) B cells, but not on normal postpartum tissues. ROR1 is a receptor for Wnt5a that may complex with TCL1, a coactivator of AKT that is able to promote development of CLL. We found the CLL cells of a few patients expressed negligible ROR1 (ROR1Neg), but expressed TCL1A at levels comparable to those of samples that expressed ROR1 (ROR1Pos). Transcriptome analyses revealed that ROR1Neg cases generally could be distinguished from those that were ROR1Pos in unsupervised gene-expression clustering analysis. Gene-set enrichment analyses demonstrated that ROR1Neg CLL had lower expression and activation of AKT signaling pathways relative to ROR1Pos CLL, similar to what was noted for leukemia that respectively developed in TCL1 vs ROR1xTCL1 transgenic mice. In contrast to its effect on ROR1Pos CLL, Wnt5a did not enhance the proliferation, chemotaxis, or survival of ROR1Neg CLL. We examined the CLL cells from 1568 patients, which we randomly assigned to a training or validation set of 797 or 771 cases, respectively. Using recursive partitioning, we defined a threshold for ROR1 surface expression that could segregate samples of the training set into ROR1-Hi vs ROR1-Lo subgroups that differed significantly in their median treatment-free survival (TFS). Using this threshold, we found that ROR1-Hi cases had a significantly shorter median TFS and overall survival than ROR1-Lo cases in the validation set. These data demonstrate that expression of ROR1 may promote leukemia-cell activation and survival and enhance disease progression in patients with CLL.

ROR1 is a type-1 tyrosine kinase-like orphan-receptor that ordinarily is expressed during embryogenesis, but that also is found on leukemia cells of patients (pts) with chronic lymphocytic leukemia (CLL). We identified patients with CLL cells that had negligible expression of ROR1, despite otherwise satisfying all standard criteria for diagnosis of CLL by iwCLL criteria.

Evolutionarily-conserved ROR1 and ROR2 are each considered a distinct receptor for Wnt5a, implicated in non-canonical Wnt-signaling involved in organogenesis or cancer metastasis. However, in a companion study we found that Wnt5a could induce ROR1 to complex with ROR2 on leukemia cells of patients with chronic lymphocytic leukemia (CLL). Wnt5a-induced ROR1:ROR2 recruited and activated guanine exchange factors (GEF), which in turn activated Rac1 and RhoA, thereby enhancing leukemia-cell proliferation and migration. To map the domain(s) of ROR1 required for formation of ROR1:ROR2 complexes we studied MEC1, a cell line generated from a patients with CLL that we found expressed Wnt5a and ROR2, but not ROR1. Transfection of MEC1 cells to express ROR1 significantly enhanced MEC1 proliferation and chemokine-induced migration, effects that could be blocked by treatment of MEC1-ROR1 cells with a humanized anti-ROR1 monoclonal antibody (mAb), cirmtuzumab (UC-961). We generated MEC1 cells that expressed vectors encoding full-length ROR1, or each of various truncated-forms of ROR1 lacking distinct structural domains. Fluorescence confocal microscopy and co-immune-precipitation studies found that each of the truncated forms of ROR1, except the truncated ROR1 lacking the extracellular kringle (KNG) domain, could form complexes with ROR2 on transfected MEC1 cells; formation of such complexes could be reduced by treatment of the cells with neutralizing antibodies to Wnt5a. Nevertheless, despite forming complexes with ROR2, only the full-length ROR1 formed complexes with ROR2 that could recruit and activate GEF and thereby activate RhoA and Rac1 in MEC1-ROR1 cells. Moreover, only MEC1-ROR1 cells transfected with full-length ROR1 had enhanced proliferation and chemokine-induced migration relative to that of parental MEC1 cells, which only expressed ROR2; again such effects could be blocked by treatment of the ROR1-transfected MEC1 cells with cirmtuzumab. We also examined the activity of cirmtuzumab against MEC1-ROR1 cells engrafted into immune-deficient Rag2−/− γc−/− mice. First we observed that Rag2−/− γc−/− engrafted with MEC1-ROR1 cells had significantly greater splenic and marrow involvement with CD19+ human leukemia than did littermates infused with equal numbers of MEC1 cells that did not express ROR1. However, treatment of mice engrafted with MEC1-ROR1 with cirmtuzumab significantly inhibited the growth of MEC1-ROR1, completely abrogating the growth advantage that MEC1-ROR1 cells had over that of MEC1 cells in vivo. MEC1-ROR1 cells harvested from mice treated with UC-961 had lost or attenuated expression of ROR1, suggesting that cirmtuzumab selected against MEC1 cells that expressed ROR1. Collectively, these studies reveal that the KRG domain of ROR1 is necessary for it to complex with ROR2 in MEC1 cells, but that full-length ROR1 is required for activation of ROR1:ROR2 complexes leading to the activation of RhoA and Rac1. Furthermore, these studies demonstrate that the anti-ROR1 mAb cirmtuzumab can block the formation of such complexes and impair the capacity of ROR1 to enhance leukemia cells migration and proliferation in vitro and in vivo, providing rationale for ongoing clinical evaluation of this antibody in patients with CLL or other cancers that are complemented by ROR1-dependent, non-canonical Wnt5a signaling.