Lenalidomide effectively treats a blood disorder caused by the 5q chromosomal deletion. A study shows that the drug binds to its target, CRBN, to promote the breakdown of an enzyme encoded by a gene in the 5q region. See Article p.183
Around 60 years ago, thalidomide was developed as a sedative and sold in more than 40 countries. But the drug was soon banned because of its association with serious developmental defects, such as limb deformities, in children whose mothers had taken it while pregnant. Now, thalidomide is being re-evaluated and is recognized as an effective treatment for myeloma, a cancer of plasma cells of the immune system. Moreover, derivatives of thalidomide have been developed; these compounds, which include lenalidomide and pomalidomide, make up a class of immunomodulatory drug termed IMiDs1. As well as being effective against myeloma, lenalidomide can treat2 a type of myelodysplastic syndrome (MDS) — a disorder of blood stem cells (haematopoietic cells) — that is caused by a deletion of the long arm of chromosome 5. In this issue, Krönke et al.3 (page 183) provide a model of lenalidomide action in the context of this mutation.
The protein CRBN was identified as a direct target of thalidomide through affinity-bead technology4. CRBN functions as a substrate-recognition component of an E3 ubiquitin ligase enzyme complex that catalyses the conjugation of ubiquitin molecules to specific substrate proteins, thereby marking the proteins for degradation. CRBN is also bound by lenalidomide and pomalidomide5,6 and is now regarded as a primary target of IMiDs — this binding is required for both the damaging and the therapeutic effects of the drugs. Previous research7,8,9 showed that lenalidomide and pomalidomide promote the degradation of the transcription factors Ikaros (IKZF1) and Aiolos (IKZF3) by modulating the activity of a CRBN–ubiquitin ligase complex, and that this process underlies the drugs' efficacy against myeloma. However, it has been unclear whether all IMiDs confer the same substrate specificity on CRBN.
The deletion of the long arm of chromosome 5, called del(5q), is seen in some people with MDS, and leads to hyperproliferation of haematopoietic cells and their ineffective differentiation. Lenalidomide is known to selectively induce apoptotic cell death in cells with del(5q). The deletion means that people affected have only one copy of the genes located in that chromosomal region. Krönke et al. suggest that this haploinsufficiency might explain the efficacy of lenalidomide in this disease.
By studying the effect of lenalidomide treatment on protein ubiquitination and abundance of myeloid blood cells, the authors identify the enzyme casein kinase 1α (CK1α) as a target of ubiquitin-mediated degradation in the presence of the drug. Deletion of the CRBN gene using CRISPR/Cas9 genome-editing technology abolished this degradation, suggesting that this effect is crucially dependent on CRBN. The gene that encodes CK1α, CSNK1A1, is on the long arm of chromosome 5, so it seems that the result of this degradation is to compound the already lower than normal levels of this enzyme that result from the deletion.
CK1α regulates the activity of multiple proteins. For example, it negatively regulates p53, a tumour-suppressor protein. The CK1α inhibitor D4476 has been shown to activate p53 and induce apoptosis in cells with only one copy of CSNK1A1 (ref. 10). Krönke and colleagues demonstrate that CK1α depletion sensitizes normal human haematopoietic cells to lenalidomide. They also confirm that overexpression of CK1α confers lenalidomide resistance on cells with the del(5q) mutation. By contrast, overexpression of Ikaros did not suppress lenalidomide-mediated therapeutic effects on del(5q) cells.
Rodents have been shown to be resistant to IMiDs1,6, and the authors found that lenalidomide did not decrease CK1α levels in normal mouse cells. They demonstrate that a single amino-acid difference between the human and mouse forms of CRBN is responsible for this different response to the drug, and that mouse cells expressing mouse CRBN with the substituted human amino acid were subject to lenalidomide-dependent CK1α degradation. The authors then generated 'humanized' mouse haematopoietic cells, which expressed the modified CRBN protein and had only a single copy of CSNK1A1. They show that lenalidomide treatment induced increased apoptotic death of the cells. Moreover, the increase in apoptosis in these cells was countered when p53 levels in the cells were reduced, which fits well with the previous report of p53 involvement in this pathway10. This development represents substantial progress in the field of IMiDs and thalidomide research — at last, through genetic modification of mouse CRBN, investigators can use mice to study these drugs.
Krönke et al. then compared the effect of lenalidomide on a human MDS del(5q) cell line with that of other IMiDs, including CC-122, a compound currently undergoing phase I clinical trials for blood cancers and solid tumours. All IMiDs that were examined decreased levels of Aiolos and Ikaros, and the efficacy of CC-122 was stronger than that of lenalidomide. But only lenalidomide degraded CK1α (Fig. 1). Interestingly, high concentrations of CC-122 suppressed lenalidomide-induced CK1α degradation, which suggests that lenalidomide competes with CC-122 for binding to CRBN in cells and that lenalidomide confers a distinct substrate specificity on CRBN.
Accumulating evidence3,7,8,9 indicates that the substrate recognition of CRBN is altered in response to each ligand that binds it. X-ray crystal structures of CRBN bound to IMiDs have revealed that a common glutarimide moiety in these compounds (Fig. 1) is sufficient for this binding to occur5,6. It is possible that the remaining structure of each ligand might be important for determining the enzyme's substrate specificity. Recently, uridine, a nucleotide base, was found to bind to the same ligand-binding pocket of CRBN11, and it is likely that CRBN also has cellular ligands that alter its substrate specificity.
In plants, the hormone auxin functions as a molecular glue to attach the hormone's target TIR1, a substrate-recognition component of an SCF ubiquitin ligase, to its substrate AUX/IAA proteins12. It is unclear whether ligands such as IMiDs also function to glue CRBN to its substrate or whether they act through another mechanism. Elucidation of the 3D structure of the CRBN–ligand–substrate complex will provide a deeper understanding of the substrate-recognition process and contribute to the development of potent and clinically effective drugs.
About this article
Science China Life Sciences (2015)