Upregulation of Alternatively Glycosylated Smoothened (SMO) Is Involved in Doxorubicin (DXR) Resistance in Diffuse Large B-Cell Lymphoma

Diffuse large B-cell lymphoma (DLBCL) is the most common lymphoid malignancy in adults. Chemoresistance remains the most urgent clinical challenge in the management of DLBCL patients. The salvage therapy options for patients with relapsed disease are poor (responses rates less than 20%) and survival is usually short. Therefore, the development of less toxic yet effective therapeutic approaches for DLBCL patients will address a significant immediate medical need. Smoothened (SMO), a Frizzled-class G-protein-coupled-receptor (GPCR), is activated by cholesterol and regulated indirectly by Hedgehog (Hh) ligands through the release of PTCH1-mediated suppression. SMO contains an N-terminal extracellular segment with the cysteine-rich domain (CRD), a heptahelical transmembrane domain (TMD), and a C-terminal intracellular segment. Although SMO localizes in cilia to function as Hh signal transducer, our data support that SMO is also a constituent of lipid rafts, and plays a fundamental role in lipid raft organization in DLBCL. SMO in doxorubicin (DXR) resistant cell lines shows increased glycosylation of its extracellular segments and drug resistance is sensitive to modulation of cellular cholesterol levels, both features associated with protein localization to lipid raft and lipid raft composition, a poorly understood aspect in drug resistance. To improve our current therapeutic approaches for lymphoma patients we need to increase our understanding of the cellular mechanisms that contribute to tumor progression and drug resistance.

We established a DXR-resistant cell line (SUDHL4) through long-term exposure to escalating doses of DXR (DXR-resistant cell line: virtually no cell death up to a DXR concentration of 120 nM; parental cell line DXR-IC₅₀: 40nM). We observed that DXR-resistant SUDHL4 cells were also resistant to escalating doses of vincristine in comparison with the parental (sensitive) cells. Studies on the response of DXR resistant lymphoma cells to DXR confirmed high levels of DNA damage in resistant cells (measured by phosphorylation of histone H2AX) confirming that DXR resistance was not due to increase drug efflux. In addition, in this experimental model, both parental and DXR-resistant lymphoma cells have mutated TP53 (missense mutation p.V157A) showing that TP53 deficiency was not a factor in inducing DXR resistance.

Several lines of evidence support a direct role of SMO in the DXR-resistance observed in DLBCL cells. First, DXR-resistant lymphoma cells in comparison with the parental cells showed a strong increase in SMO mRNA, but not its canonical downstream target of signaling, the transcription factor GLI1. Concordantly, SMO protein levels were substantially elevated in DXR-resistant cells (as detected by immunofluorescence and western blot). Moreover, upregulation of the expression of SMO was detected by immunohistochemistry in 4 of 5 relapsed DLBCL tumors in comparison with the matched tumors at the time of the initial diagnosis and before chemotherapy treatment. Second, we found that SMO is localized in the lipid rafts and mediates the assembly of a TRAF6 E3 ligase-dependent machinery that enhances AKT signaling. Additionally, we observed that SMO is highly glycosylated in DXR-resistant SUDHL4 cells but not in the parental cell line; previously, it has been found that SMO has 4 potential N-glycosilation sites that are dispensable for canonical Hh signaling but needed to induce a non-canonical SMO signal through Gα1. Finally, depletion of SMO using shRNA and a lentiviral-tetracycline inducible system and pharmacologically inhibition of SMO using the classic SMO inhibitor (cyclopamine-KAAD) or a cholesterol synthesis inhibitor (lovastatin) re-established sensitivity to DXR in DXR-resistant cells. Interestingly, lovastatin did not modify DXR sensitivity in the parental cell line. In summary, this study suggests a novel function of SMO in the development of DXR-resistance in DLBCL and opens new opportunities to regain or enhance DXR sensitivity through targeting this novel mechanism.