Search

Goal: To evaluate Notch inhibitor monotherapy and combination (combo) therapy in Adenoid Cystic Carcinoma (ACC) Patient Derived Xenograft (PDX) models with and without activating Notch mutations (mt). ACC is a rare salivary gland malignancy with no standard of care. Chemotherapy resistance limits treatment to surgery/radiation and ~60% of patients (pts) will recur. ~22% of ACC pts have Notch activating mutations that are associated with a distinct phenotype, aggressive disease and poor prognosis. In a P1 solid tumor study, 2 ACC pts were treated with AL101, an investigational gamma secretase inhibitor (J Clin Oncol 36, 2018 abstract 2515). One pt with an NRR (Negative Regulatory Region) activating mutation had a prolonged partial response, the 2nd pt with a PEST mutation had stable disease. Both had an extended PK profile with a sustained pharmacodynamic response (>50% inhibition of HES1 in peripheral blood). These results prompted us to evaluate the effect of AL101 in tumors that harbor/lack Notch activating mutations. Design: 4 ACC PDX models were evaluated: ACCx9 (Notch1 mt: NRR activating mutation), ACCx11 (Notch1 mt, tandem duplication), ACCx6 (Notch1 wt), ACCx5M1 (Notch1 VUS, not predicted to be activating). Activated Notch1 (nuclear IHC stain) was confined to the Notch mt models. Tumors were implanted into 6-12 week old athymic nude female mice. Upon reaching 150-300 mm3 tumor volume, mice were randomized to treatment (N=5), or vehicle (N=10) arms. AL101 was dosed at 7.5 mg/kg po (days 1-4 of Q7D) as single agent or in combo with cisplatin (3 mg/kg ip; qw) or everolimus (10 mg/kg; po; qd). Results: Significant tumor growth inhibition (TGI) was seen with AL101 monotherapy compared to vehicle treatment in both Notch mt models (ACCx9; 110% TGI P0.99, ACCx5M1: 60% TGI P=0.26). Cisplatin and everolimus monotherapy had no significant effect on Notch mt models (cisplatin: ACCx9; 43% TGI P=0.98, ACCx11; 50% TGI P=0.99, everolimus: ACCx9; 51% TGI P=0.6, ACCx11; 54% TGI P=0.99). AL101 with cisplatin or everolimus had no further benefit in the ACCx9 model (both 106% TGI P=0.98 or P>0.99 respectively). The ACCx11 model had a non-significant benefit (104% TGI P=0.17 for cisplatin combo and 107% TGI P=0.18 for everolimus combo). Interestingly, the ACCx6 Notch wt model was sensitive to everolimus (75% TGI P=0.027) with no significant benefit of adding AL101. Conclusion: AL101 monotherapy had a significant anti-tumor effect in ACC PDX tumors with Notch activating mutations and lacked effectiveness in tumors lacking such mutations. Neither cisplatin nor everolimus monotherapy treatment had a significant effect in Notch mt models, alone or in combination with AL101. These data support the clinical development of AL101 as a targeted monotherapy for ACCs with Notch activating mutations. Citation Format: Renata ferrarotto*, Genia Alpert*, Udi Gluschnaider, Rami Rauch, Adi Mondshine, Oz Solomon, Bill Kramer, Evgeny Izumchenko, John Heymach, Andrea Vergara-Silva, Jon Aster, Matti Davis. AL101 mediated tumor inhibition in Notch mutated ACC PDX models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4885.

1064 Background: The Notch pathway is activated during mammary gland development and has been implicated as a key driver in breast cancer. There is an urgent need to identify new therapeutic strategies for triple-negative breast cancer (TNBC), a sub-type associated with poor prognosis and no available targeted therapies. Notch gain of function (GOF) genetic alterations are potential tumor drivers found in ~10% of TNBC. This motivated the development of Notch inhibitors, including AL101 a pan-Notch, gamma secretase inhibitor (J Clin Oncol 36, 2018 abstract 2515). AL101 is currently being evaluated in Adenoid Cystic Carcinoma patients with activating Notch mutations (NCT03691207, ACCURACY trial). Here, we aim to test the activity of AL101 in TNBC patient derived xenograft (PDX) models with Notch activating genetic alterations. Methods: Gene expression cluster analysis was performed for 38 TNBC PDX tumors using a list of 21 Notch target genes. Seven tumors, bearing a “Notch-on” signature, were enriched with mutated/fusion (M/F) Notch genes and clustered separately from all other tumors. Of 9 models selected for study, 4 had a Notch-on signature and were expected to respond to AL101. Tumors were implanted into female athymic nude mice. Once tumors reached an average size of 150-300 mm3, mice (n = 5/group) were randomized to Vehicle or AL101 treatment arms (3 mg/kg, PO, 4on/3off) until tumors reached 1500 mm3 or day 60. Results: As measured by tumor growth inhibition (TGI), AL101 was more potent in tumors with a putative Notch-on signature. Within these 4 models, M/F genes were present in Notch1-NRR GOF (103% TGI p = 0.0004); Notch2-fusion (62%TGI p = 0.036); Notch3-fusion (75% TGI p = 0.032); or Notch4-fusion (147% TGI p < 0.00001). Tumors lacking the Notch signature did not respond significantly to AL101: WT Notch (43% TGI p = 0.0104; 64% TGI p = 0.13); Notch1 with a predicted loss of function mutation (12% TGI p = 0.53), Notch1 Variant of Unknown Significance (VUS) (30% TGI p = 0.44), Notch2 VUS (41% TGI p = 0.44). Conclusions: We demonstrate that in TNBC PDX models, the presence of a Notch-on signature and Notch GOF mutations/fusions correlates with potent response to AL101. These data support the clinical development of AL101 as a targeted therapy for TNBC with Notch GOF alterations.

2515Background: Notch signaling can be deregulated in human cancer. AL101 (formerly BMS-906024), a gamma secretase inhibitor that potently inhibits all Notch receptors, was evaluated. The primary o...

NF-kappaB, a ubiquitous, inducible transcription factor involved in immune, inflammatory, stress and developmental processes, is retained in a latent form in the cytoplasm of non-stimulated cells by inhibitory molecules, IkappaBs. Its activation is a paradigm for a signal-transduction cascade that integrates an inducible kinase and the ubiquitin-proteasome system to eliminate inhibitory regulators. Here we isolate the pIkappaBalpha-ubiquitin ligase (pIkappaBalpha-E3) that attaches ubiquitin, a small protein which marks other proteins for degradation by the proteasome system, to the phosphorylated NF-kappaB inhibitor pIkappaBalpha. Taking advantage of its high affinity to pIkappaBalpha, we isolate this ligase from HeLa cells by single-step immunoaffinity purification. Using nanoelectrospray mass spectrometry, we identify the specific component of the ligase that recognizes the pIkappaBalpha degradation motif as an F-box/WD-domain protein belonging to a recently distinguished family of beta-TrCP/Slimb proteins. This component, which we denote E3RSIkappaB (pIkappaBalpha-E3 receptor subunit), binds specifically to pIkappaBalpha and promotes its in vitro ubiquitination in the presence of two other ubiquitin-system enzymes, E1 and UBC5C, one of many known E2 enzymes. An F-box-deletion mutant of E3RS(IkappaB), which tightly binds pIkappaBalpha but does not support its ubiquitination, acts in vivo as a dominant-negative molecule, inhibiting the degradation of pIkappaBalpha and consequently NF-kappaB activation. E3RS(IkappaB) represents a family of receptor proteins that are core components of a class of ubiquitin ligases. When these receptor components recognize their specific ligand, which is a conserved, phosphorylation-based sequence motif, they target regulatory proteins containing this motif for proteasomal degradation.

Anergy induction initiates a transcription program involving the upregulation of several ubiquitin ligases (E3s). New data show how these E3s contribute to the establishment of anergy.

β-TrCP, the substrate recognition subunit of a Skp1–Cul1–F-box (SCF) ubiquitin ligase, is ubiquitously expressed from two distinct paralogs, targeting many regulatory proteins for proteasomal degradation. We generated inducible β-TrCP hypomorphic mice and found that they are surprisingly healthy, yet have a severe testicular defect. We show that the two β-TrCP paralogs have a nonredundant role in spermatogenesis. The testicular defect is tightly associated with cell adhesion failure within the seminiferous tubules and is fully reversible upon β-TrCP restoration. Remarkably, testicular depletion of a single β-TrCP substrate, Snail1, rescued the adhesion defect and restored spermatogenesis. Our studies highlight an unexpected functional reserve of this central E3, as well as a bottleneck in a specific tissue: a single substrate whose stabilization is incompatible with testicular differentiation.

SummaryThe Hedgehog (Hh) signaling pathway regulates development in animals ranging from flies to humans. Although its framework is conserved, differences in pathway components have been reported [1–4]. A kinesin-like protein, Costal2 (Cos2), plays a central role in the Hh pathway in flies [3, 5]. Knockdown of a zebrafish homolog of Cos2, Kif7, results in ectopic Hh signaling, suggesting that Kif7 acts primarily as a negative regulator of Hh signal transduction [6]. However, in vitro analysis of the function of mammalian Kif7 and the closely related Kif27 has led to the conclusion that neither protein has a role in Hh signaling [4]. Using Kif7 knockout mice, we demonstrate that mouse Kif7, like its zebrafish and Drosophila homologs, plays a role in transducing the Hh signal. We show that Kif7 accumulates at the distal tip of the primary cilia in a Hh-dependent manner. We also demonstrate a requirement for Kif7 in the efficient localization of Gli3 to cilia in response to Hh and for the processing of Gli3 to its repressor form. These results suggest a role for Kif7 in coordinating Hh signal transduction at the tip of cilia and preventing Gli3 cleavage into a repressor form in the presence of Hh.

β-TrCP/E3RS (E3RS) is the F-box protein that functions as the receptor subunit of the SCFβ-TrCP ubiquitin ligase (E3). Surprisingly, although its two recognized substrates, IκBα and β-catenin, are present in the cytoplasm, we have found that E3RS is located predominantly in the nucleus. Here we report the isolation of the major E3RS-associated protein, hnRNP-U, an abundant nuclear phosphoprotein. This protein occupies E3RS in a specific and stoichiometric manner, stabilizes the E3 component, and is likely responsible for its nuclear localization. hnRNP-U binding was abolished by competition with a pIκBα peptide, or by a specific point mutation in the E3RS WD region, indicating an E3–substrate-type interaction. However, unlike pIκBα, which is targeted by SCFβ-TrCP for degradation, the E3-bound hnRNP-U is stable and is, therefore, a pseudosubstrate. Consequently, hnRNP-U engages a highly neddylated active SCFβ-TrCP, which dissociates in the presence of a high-affinity substrate, resulting in ubiquitination of the latter. Our study points to a novel regulatory mechanism, which secures the localization, stability, substrate binding threshold, and efficacy of a specific protein-ubiquitin ligase.