Lymphoblasts Produce a Lysosomal Protease That Rapidly Degrades L-Asparaginase - Implications for Therapy in Childhood Acute Lymphoblastic Leukemia.

The efficacy of the key anti-leukaemic agent E.coli L-Asparaginase (L-Asp) is limited by variable therapeutic activity and development of antibodies/hypersensitivity. Observing more frequent clinical hypersensitivity in children with high risk acute lymphoblastic leukaemia (ALL), we speculated that L-Asp was proteolytically degraded by lymphoblasts, generating immunogenic fragments. On incubation with whole cell lysate from a lymphoblastic cell line, L-Asp was cleaved into at least three fragments. This cleavage was blocked only by MV026630, a specific inhibitor of the cysteine protease Asparaginyl Endopeptidase (AEP). Incubation of L-Asp with purified recombinant AEP produced an identical cleavage pattern at the carboxy termini of specific asparagine and aspartate residues. The level of AEP expression at diagnosis in 148 childhood ALL patients was measured using U133A, Exon 1.0 ST arrays and quantitative PCR. High levels of expression were detected in 38 patients (25%), including all 6 patients with clinical hypersensitivity. Elevated AEP expression was more commonly observed in high-risk disease (42% v 21%). Molecular modelling of AEP cleavage sites predicts that the N-terminal cleavage site (C1) provides tetramer stability while the highly-conserved second cleavage site (C2) is critical for tetramer formation and enzyme activity. This model was confirmed by analyses of recombinant L-Asp variants mutated to resist AEP cleavage. Cleavage is sequential from the N terminus. A product resistant to cleavage at C1 is not degraded by AEP, forms a tetramer and retains enzymatic activity comparable to the wild-type. The C2 mutated product shows little activity and is rapidly degraded by AEP. Known antigenic epitopes of L-Asp are retained within AEP-cleaved fragments. The contribution of AEP to antigenic processing of L-Asp is being investigated by T-cell activation assays using a synthetic peptide library and T cells from patients with known hypersensitivity. Thus patients with AEP-overexpressing lymphoblasts may not benefit from L-Asp therapy either due to inactivation, the development of antibodies or both. Replacement of a single amino acid of L-Asp can prevent this process without loss of activity. This product potentially has a longer half-life and less antigenicity. Our investigations have identified a hitherto unknown pathway for L-Asp degradation and a novel mechanism of drug resistance in childhood ALL. These observations require further validation by correlation of AEP expression with asparaginase activity and antibody formation in children receiving L-Asp during treatment for ALL. Further potential improvements in therapy could include screening for AEP expression at diagnosis and the use of non E Coli asparaginase or preferably a modified recombinant L-Asp and/or AEP inhibitors for those with high levels of expression.