Lenalidomide (LEN) and its analogue, pomalidomide, promote erythroid lineage competence and in vitro colony-forming capacity. In patients with non-del(5q) myelodysplastic syndrome (MDS), LEN restores erythropoiesis in a subset of patients (List et al. N Eng J Med 2005;352:549). Investigations by Ebert et al. showed that such responders to LEN treatment display repression of erythroid specific genes and that LEN restored transcriptional response to erythropoietin (Epo) (Ebert BL et al. PLoS Medicine 2008;5(2):e35), suggesting that LEN enhances Epo receptor (R) signal fidelity. We previously reported that LEN induces cellular expression of JAK2 associated EpoR in a concentration-dependent manner, however, the mechanism of regulation is unclear (Basiorka et al. Blood 2011,118: 2382a). Recent investigations implicated inhibition of the cereblon RING (really interesting new gene) finger domain containing E3-ubiquitin ligase complex as a key target of the immunomodulatory drugs (IMiDs) responsible for the teratogenic effects of thalidomide and the cytotoxic effects of LEN in multiple myeloma (Ito T et al. Science 2010; 327:1345–50; Zhu YW et al. Blood 2011;118:4771–9). We recently showed that LEN also interacts with the RING finger E3 ubiquitin ligase, murine double minute 2 (MDM2) to inhibit ligase ubiquitination and stabilize the protein (Wei et al. Oncogene, MS#ONC-2011-01840R, 2012). Because EpoR turnover is regulated by ubiquitination and proteasomal degradation, we evaluated the effects of LEN on the E3-ubiquitin ligase, RNF41, which regulates steady state or ligand independent, Janus kinase (JAK2) associated Type I receptor internalization (Wauman et al. J Cell Science. 2011;124:921–932). We hypothesized that LEN upregulates JAK2/EpoR expression through inhibition of RNF41 function, thereby increasing EpoR expression and enhancing JAK2 competent receptor signaling.
Treatment of the UT-7 erythroid progenitor cell line with cycloheximide ±1μM LEN showed that LEN stabilized cellular EpoR (T1/2, LEN >72h vs. 56h). To determine if the effects of LEN on receptor turnover are restricted to Type 1 cytokine receptors, we examined the effects of LEN on cellular expression of IL3-R (Type 1) and c-Kit (Type 2). LEN up-regulated IL3-R expression in a concentration-dependent fashion, whereas c-Kit expression was unchanged, confirming Type 1 receptor specificity. To determine if LEN alters EpoR/RNF41 interaction, we assessed protein association after LEN treatment. Immunoprecipitation (IP) of either EpoR or RNF41 followed by immunoblot (IB) for the binding partner showed that LEN promoted EpoR/RNF41 association in a concentration dependent manner. To investigate the effects of LEN on RNF41 function, we assessed protein specific ubiquitination after proteasomal inhibition with bortezomib followed by LEN treatment. IP of RNF41 and EpoR followed by ubiquitin IB showed that LEN inhibited RNF41 auto-ubiquitination in a concentration-dependent fashion accompanied by a corresponding decrease in EpoR ubiquitination, suggesting that LEN inhibits RNF41 ubiquitination to increase EpoR accumulation. To confirm that RNF41 is the principal target of LEN responsible for EpoR stabilization, we transfected HEK293T cells with EpoR and/or RNF41 expression vectors using the calcium phosphate method. Steady state EpoR expression was lower in EpoR/RNF41 cells compared with cells transfected with EpoR alone. Moreover, EpoR upregulation by LEN was abrogated in EpoR/RNF41 cells indicating that cellular RNF41 is a critical determinant of EpoR upregulation by LEN. Immunohistochemical staining of 16 bone marrow biopsies from non-del(5q) LEN-treated MDS patients are in progress to determine the relationship between cellular RNF41 level in erythroid precursors and clinical response.
Our findings suggest that LEN acts as a broad RING finger E3-ubiquitin ligase inhibitor, whose targets extend to the Type 1 cytokine receptor specific, RNF41. RNF41 inhibition by LEN promotes accumulation of signaling competent JAK2/EpoR complexes that may augment Epo responsiveness. Further investigation is warranted to determine if erythroid expression level of RNF41 may serve as a biomarker for response to LEN in patients with non-del(5q) MDS.
Asterisk with author names denotes non-ASH members.