Background
The NOTCH signaling pathway is pivotal for various physiological processes including immune responses, and has been implicated in the pathogenesis in many diseases including T-cell acute lymphoblastic leukemia (T-ALL). Over 70% of T-ALL patient samples contain mutations in NOTCH1 and/or FBXW7 that result in the activation of the NOTCH pathway. Various targeted drugs are available that inhibit NOTCH signaling, but their effectiveness varies due to variable NOTCH pathway activities among individual patients. Moreover, patients' leukemic cells that lack these mutations may still require NOTCH signaling. A quick and robust quantification of NOTCH pathway activity in primary patient samples would identify patients who could benefit from NOTCH targeted treatment.
Aims
In primary human T-ALL samples, we aimed to determine the NOTCH pathway activity in relation to active, intracellular NOTCH1 (ICN1) levels and in relation to NOTCH1 and/or FBXW7 mutations. Additionally, we investigated whether the NOTCH pathway activity score is more accurate than a mutation-based activity prediction.
Methods
Our test to assess functional NOTCH pathway activity in various cell types was applied to primary human T-ALL samples. The NOTCH test infers a quantitative NOTCH pathway activity score from mRNA levels of conserved direct NOTCH target genes based on a Bayesian network model. This model describes the causal relation between up- or downregulation of NOTCH target genes and the presence of an active or inactive NOTCH transcription complex. The Bayesian model was calibrated on publically available Affymetrix U133 Plus2.0 microarray datasets of samples with an active or inactive NOTCH pathway. Following validation on multiple cell types and malignancies, we scored NOTCH pathway activation in our well-characterized cohort of 117 T-ALL patient samples and related it to clinical and biological parameters including outcome.
Results
The NOTCH pathway model was calibrated using a microarray dataset containing high-grade serous ovarian cancer-which has high NOTCH activity-and normal ovarian tissue samples that lack NOTCH activity. Validation of the test using datasets from primary cells and cell lines of various origins revealed that it measures the NOTCH activity status in different cellular contexts. In primary diagnostic T-ALL samples, we observed a significant relationship between NOTCH pathway activity scores and active, intracellular cleaved NOTCH1 (ICN1) protein levels and the presence of NOTCH1-activating mutations. We next scored NOTCH pathway activity over the four T-ALL subgroups ETP-ALL, TLX, Proliferative and TALLMO. The TLX subgroup had the highest NOTCH activity levels compared to the other subgroups, consistent with the high percentage of TLX cases with NOTCH1/FBXW7 mutations. Strikingly, the significance of the correlation between ICN1 levels and NOTCH pathway activity was mainly attributed to the strong NOTCH1-activating mutations that include NOTCH1 juxtamembrane domain mutations, or hetero-dimerization mutations combined with PEST domain or FBXW7 mutations. When assessing the event-free survival and relapse-free survival curves, we observed that patients with the lowest (lower than the 25th-percentile) NOTCH pathway activity scores had the shortest event-free survival compared to the others (p<0.05, log-rank test).
Summary/Conclusion
High NOTCH pathway activation was mostly present in-but not limited to-T-ALL samples harboring strong NOTCH1 mutations, indicating that additional mechanisms can activate NOTCH signaling. Combined with the observation that the measured NOTCH pathway activity relates to ICN1 protein levels, this indicates that the pathway activity score more accurately reflects NOTCH pathway activity than the predicted activity based on NOTCH1 mutations alone.
Disclosures
Holtzer: Philips Research: Employment. Verhaegh:Philips Research: Employment. van de Stolpe:The Netherlands: Employment; Eindhoven: Employment; Philips Research: Employment.
Border of Notch activity establishes a boundary between the two dorsal appendage tube cell types
