A Bioinformatics Tool for Identifying Intratumoral Microbes from the ORIEN Dataset

Evidence supports significant interactions among microbes, immune cells, and tumor cells in at least 10%-20% of human cancers, emphasizing the importance of further investigating these complex relationships. However, the implications and significance of tumor-related microbes remain largely unknown. Studies have demonstrated the critical roles of host microbes in cancer prevention and treatment responses. Understanding interactions between host microbes and cancer can drive cancer diagnosis and microbial therapeutics (bugs as drugs). Computational identification of cancer-specific microbes and their associations is still challenging due to the high dimensionality and high sparsity of intratumoral microbiome data, which requires large datasets containing sufficient event observations to identify relationships, and the interactions within microbial communities, the heterogeneity in microbial composition, and other confounding effects that can lead to spurious associations. To solve these issues, we present a bioinformatics tool, microbial graph attention (MEGA), to identify the microbes most strongly associated with 12 cancer types. We demonstrate its utility on a dataset from a consortium of nine cancer centers in the Oncology Research Information Exchange Network. This package has three unique features: species-sample relations are represented in a heterogeneous graph and learned by a graph attention network; it incorporates metabolic and phylogenetic information to reflect intricate relationships within microbial communities; and it provides multiple functionalities for association interpretations and visualizations. We analyzed 2,704 tumor RNA sequencing samples and MEGA interpreted the tissue-resident microbial signatures of each of 12 cancer types. MEGA can effectively identify cancer-associated microbial signatures and refine their interactions with tumors.

SIGNIFICANCE

Studying the tumor microbiome in high-throughput sequencing data is challenging because of the extremely sparse data matrices, heterogeneity, and high likelihood of contamination. We present a new deep learning tool, MEGA, to refine the organisms that interact with tumors.

Profiling the molecular and clinical landscape of glioblastoma utilizing the Oncology Research Information Exchange Network brain cancer database

Background

Glioblastoma exhibits aggressive growth and poor outcomes despite treatment, and its marked variability renders therapeutic design and prognostication challenging. The Oncology Research Information Exchange Network (ORIEN) database contains complementary clinical, genomic, and transcriptomic profiling of 206 glioblastoma patients, providing opportunities to identify novel associations between molecular features and clinical outcomes.

Methods

Survival analyses were performed using the Logrank test, and clinical features were evaluated using Wilcoxon and chi-squared tests with q-values derived via Benjamini-Hochberg correction. Mutational analyses utilized sample-level enrichments from whole exome sequencing data, and statistical tests were performed using the one-sided Fisher Exact test with Benjamini-Hochberg correction. Transcriptomic analyses utilized a student's t-test with Benjamini-Hochberg correction. Expression fold changes were processed with Ingenuity Pathway Analysis to determine pathway-level alterations between groups.

Results

Key findings include an association of MUC17, SYNE1, and TENM1 mutations with prolonged overall survival (OS); decreased OS associated with higher epithelial growth factor receptor (EGFR) mRNA expression, but not with EGFR amplification or mutation; a 14-transcript signature associated with OS > 2 years; and 2 transcripts associated with OS < 1 year.

Conclusions

Herein, we report the first clinical, genomic, and transcriptomic analysis of ORIEN glioblastoma cases, incorporating sample reclassification under updated 2021 diagnostic criteria. These findings create multiple avenues for further investigation and reinforce the value of multi-institutional consortia such as ORIEN in deepening our knowledge of intractable diseases such as glioblastoma.

The tumor microbiome as a predictor of outcomes in patients with metastatic melanoma treated with immune checkpoint inhibitors

Emerging evidence supports the important role of the tumor microbiome in oncogenesis, cancer immune phenotype, cancer progression, and treatment outcomes in many malignancies. In this study, we investigated the metastatic melanoma tumor microbiome and potential roles in association with clinical outcomes, such as survival, in patients with metastatic disease treated with immune checkpoint inhibitors (ICIs). Baseline tumor samples were collected from 71 patients with metastatic melanoma before treatment with ICIs. Bulk RNA-seq was conducted on the formalin-fixed paraffin-embedded (FFPE) tumor samples. Durable clinical benefit (primary clinical endpoint) following ICIs was defined as overall survival ≥24 months and no change to the primary drug regimen (responders). We processed RNA-seq reads to carefully identify exogenous sequences using the {exotic} tool. The 71 patients with metastatic melanoma ranged in age from 24 to 83 years, 59% were male, and 55% survived >24 months following the initiation of ICI treatment. Exogenous taxa were identified in the tumor RNA-seq, including bacteria, fungi, and viruses. We found differences in gene expression and microbe abundances in immunotherapy responsive versus non-responsive tumors. Responders showed significant enrichment of several microbes including Fusobacterium nucleatum, and non-responders showed enrichment of fungi, as well as several bacteria. These microbes correlated with immune-related gene expression signatures. Finally, we found that models for predicting prolonged survival with immunotherapy using both microbe abundances and gene expression outperformed models using either dataset alone. Our findings warrant further investigation and potentially support therapeutic strategies to modify the tumor microbiome in order to improve treatment outcomes with ICIs.

A bioinformatics tool for identifying intratumoral microbes from the ORIEN dataset

Evidence supports significant interactions among microbes, immune cells, and tumor cells in at least 10-20% of human cancers, emphasizing the importance of further investigating these complex relationships. However, the implications and significance of tumor-related microbes remain largely unknown. Studies have demonstrated the critical roles of host microbes in cancer prevention and treatment responses. Understanding interactions between host microbes and cancer can drive cancer diagnosis and microbial therapeutics (bugs as drugs). Computational identification of cancer-specific microbes and their associations is still challenging due to the high dimensionality and high sparsity of intratumoral microbiome data, which requires large datasets containing sufficient event observations to identify relationships, and the interactions within microbial communities, the heterogeneity in microbial composition, and other confounding effects that can lead to spurious associations. To solve these issues, we present a bioinformatics tool, MEGA, to identify the microbes most strongly associated with 12 cancer types. We demonstrate its utility on a dataset from a consortium of 9 cancer centers in the Oncology Research Information Exchange Network (ORIEN). This package has 3 unique features: species-sample relations are represented in a heterogeneous graph and learned by a graph attention network; it incorporates metabolic and phylogenetic information to reflect intricate relationships within microbial communities; and it provides multiple functionalities for association interpretations and visualizations. We analyzed 2704 tumor RNA-seq samples and MEGA interpreted the tissue-resident microbial signatures of each of 12 cancer types. MEGA can effectively identify cancer-associated microbial signatures and refine their interactions with tumors.

Abstract 6074: Germline and somatic genomic profiling of urothelial carcinoma

Urothelial carcinoma (UC) presents most frequently as bladder cancer and is the most common cancer of the urinary system in the United States. UC relapse and progression are common and impose a significant negative impact on the lives of patients and healthcare resources. To elucidate the biological mechanisms of UC and find novel biomarkers, we analyzed the clinical and genomic data in the Oncology Research Information Exchange Network (ORIEN). We conducted gene-based and gene-set based association tests on rare germline variants comparing the exome sequencing of 336 UC patients and genome sequencing of 366 healthy controls (42 unrelated individuals from the Centre d'Etudes du Polymorphisme Humain [CEPH] families and 324 from the University of Utah Heritage 1000 [H1K] Projects). The analysis of loss-of-function (LoF) variants revealed that the forkhead box (FOX) J2 gene set was significantly associated with UC at the genome-wide level (Bonferroni-corrected p-value=0.01). Genes in this gene set contain a motif that matches the FOXJ2 transcription factor binding site. Firth-penalized Cox proportional hazard regression on overall survival identified LoF variants in genes down-regulated in naive CD8 T cells to be associated with worse prognosis (Bonferroni-corrected p-value=0.032, hazard ratio=28.2, and 95% confidence interval 6.66 to 119.0). In exome sequencing of tumor tissues, we searched for driver genes and pathways by testing for higher variant allelic fractions than the genome average. The tests yielded eleven genes with genome-wide significance (Table 1). By gene set analysis using the MSigDB Hallmark database, significant pathways included the P53 pathway (p&lt;2 × 10−16), Wnt beta-catenin signaling (p&lt;2 × 10−16), E2F targets (p=8 × 10−13), PI3K/AKT/mTOR signaling (4 × 10−7), and apoptosis (p=9 × 10−7). These results reveal the germline predisposition variants and somatic oncogenic drivers in UC and suggest immune evasion as a contributing factor for poor clinical outcomes in UC patients.

Table 1. Genes with high allelic fractions Gene P-value Number of Variants TP53 &lt;2 × 10−16 203 RB1 2.42 × 10−16 67 ELF3 2.06 × 10−7 38 TSC1 4.05 × 10−7 32 KMT2D 7.59 × 10−7 91 ZFP36L1 8.80 × 10−7 33 CDKN1A 2.47 × 10−6 40 FGFR3 1.11 × 10−5 37 ARID1A 1.28 × 10−5 62 SMARCA4 1.53 × 10−5 17 PIK3CA 2.00 × 10−5 53

Citation Format: Bing-Jian Feng, Wendy Kohlmann, David A. Nix, Aaron Atkinson, Kenneth M. Boucher, Courtney Carroll, Jill Kolesar, Eric A. Singer, Stephen B. Edge, Kamal Sahu, Alejandro Sanchez, Mikaela Larson, Michelle L. Churchman, Laura Graham, John D. Carpten, Yousef Zakharia, Lindsey Byrne, Rohit K. Jain, Kenneth G. Nepple, Ahmad Shabsigh, Jad Chahoud, Sumati Gupta. Germline and somatic genomic profiling of urothelial carcinoma. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6074.

Replicative Instability Drives Cancer Progression

In the past decade, defective DNA repair has been increasingly linked with cancer progression. Human tumors with markers of defective DNA repair and increased replication stress exhibit genomic instability and poor survival rates across tumor types. Seminal studies have demonstrated that genomic instability develops following inactivation of BRCA1, BRCA2, or BRCA-related genes. However, it is recognized that many tumors exhibit genomic instability but lack BRCA inactivation. We sought to identify a pan-cancer mechanism that underpins genomic instability and cancer progression in BRCA-wildtype tumors. Methods: Using multi-omics data from two independent consortia, we analyzed data from dozens of tumor types to identify patient cohorts characterized by poor outcomes, genomic instability, and wildtype BRCA genes. We developed several novel metrics to identify the genetic underpinnings of genomic instability in tumors with wildtype BRCA. Associated clinical data was mined to analyze patient responses to standard of care therapies and potential differences in metastatic dissemination. Results: Systematic analysis of the DNA repair landscape revealed that defective single-strand break repair, translesion synthesis, and non-homologous end-joining effectors drive genomic instability in tumors with wildtype BRCA and BRCA-related genes. Importantly, we find that loss of these effectors promotes replication stress, therapy resistance, and increased primary carcinoma to brain metastasis. Conclusions: Our results have defined a new pan-cancer class of tumors characterized by replicative instability (RIN). RIN is defined by the accumulation of intra-chromosomal, gene-level gain and loss events at replication stress sensitive (RSS) genome sites. We find that RIN accelerates cancer progression by driving copy number alterations and transcriptional program rewiring that promote tumor evolution. Clinically, we find that RIN drives therapy resistance and distant metastases across multiple tumor types.

The impact of HRD in patients with pancreatic adenocarcinoma undergoing surgical resection: An updated analysis

4132

Background: Limited data is available regarding which mutations in the homologous recombination repair (HRR) pathway beyond BRCA can be targeted with platinum-based chemotherapy in the perioperative setting in patients with pancreatic ductal adenocarcinoma (PDAC). In this updated analysis, we assess the outcome of patients with homologous recombination deficiency (HRD) in response to platinum vs. non-platinum based perioperative chemotherapy in resected PDAC and have included additional variants linked to HRD. Methods: Patients with resectable PDAC, diagnosed between 1999-2020 from the participating members of the Oncology Research Information Exchange Network (ORIEN) were included in the study. Patient’s germline and somatic whole exome sequencing (WES) data were analyzed for known pathogenic and likely pathogenic variants according to ClinVar for the following HRR pathway genes: BRCA1, BRCA2, PALB2, BRIP1, BRAD1, BLM, BAP1, ATM, RAD51C, RAD51, RAD50, RAD54B, CHECK2, NBN, FANCA/B/C/D2/E/F/G, ARID1A, MRE11 and XRCC2. The Kaplan Meier method was used to compare median overall survival (mOS) between patients with and without HRR pathway mutations in response to perioperative platinum vs non-platinum-based chemotherapy. Multivariate cox proportional hazard model was used to calculate HR and 95% CIs adjusting for age, sex and pathologic stage. Results: The ORIEN cohort included 311 patients with resectable PDAC and available WES data. A total of 22 patients (7%) had an HRR pathway mutation. Of these, 8 (36%) received perioperative platinum-based chemotherapy and 9 (41%) a non-platinum based regimen, 4 patients (23%) received no perioperative systemic treatment. Frequency of HRR variants detected: BRCA2 n=8 (2.6%), BRCA1 n=3 (1%), ATM n=2 (0.6%), ARID1A n=1 (0.3%), BRIP1 n=1 (0.3%), CHECK2 n=1 (0.3%), FANCG n=1 (0.3%), PALPB2 n=1 (0.3%), RAD50 n=4 (1.3%), RAD51C n=1 (0.3%). The mOS for patients with HRR mutations exposed to perioperative platinum-based chemotherapy was 3.5 years (95% CI 3.4-NA), patients with HRR mutation but no platinum exposure had a mOS of 1.2 years (CI 0.9-NA). In patients with no HRR mutation exposed to platinum-based chemotherapy mOS was 2.7 years (CI 2.3-3.9) and in those without exposure mOS was 2.9 years, p=0.43. Comparison of risk of death between the 4 groups is demonstrated in the table. Conclusions: There was a trend towards improved survival in patients with PDAC who harbored a HRR pathway mutation and were treated with perioperative platinum-based chemotherapy compared to those with no platinum exposure. Our results highlight the importance of identifying patients with HRD beyond BRCA and the need for large prospective studies in the perioperative setting to further assess their predictive role. [Table: see text]

Predictors of immunotherapeutic benefits in patients with advanced melanoma and other malignancies treated with immune checkpoint inhibitors utilizing ORIEN “real-world” data

2618

Background: Despite the significant improvements in treating cancer with immune checkpoint inhibitors (ICIs), many patients (pts) do not achieve disease control. Using Oncology Research Information Exchange Network (ORIEN) Avatar real-world data conducted under the Total Cancer Care protocol we investigate predictive biomarkers of ICI benefits in pts with advanced malignancies. Methods: Clinical data were normalized as part of ORIEN Avatar. RNA-seq was performed on tumor samples following the RSEM pipeline and gene expressions were quantified as Transcript Per Million (TPM). Gene expressions (GE) were log2(TPM+1) transformed. Mann-Whitney U-test was used to compute differences between groups, and Kaplan-Meier survival analysis was performed. Results: Pts (n=1214) with 27 cancer types treated with ICIs were retrieved from the database, where 1143 and 875 patients were profiled by WES and RNA-seq, respectively. 804 pts had both WES and RNA-seq data. The top six cancer types were renal cell carcinoma (n=206), non-small cell lung cancer (n=173), head and neck cancer (n=157), melanoma (n=154), sarcomas (n=99) and bladder cancer (n=87). The ICI regimens included therapy with atezolizumab (n=87), avelumab (n=12), cemiplimab (n=6), ipilimumab (n=47), nivolumab (n=424), pembrolizumab (n=525) and ipilimumab+nivolumab (n=113). Median overall survival (OS) for the entire cohort was 21.9 months. Patients had significant improvement in OS if ICI was given in the first line ( P<0.0001). Previously published GE signatures were tested in the melanoma cohort. Signatures related to IFNg, effector T cells, chemokines, MHC II and Tertiary Lymphoid Structures were significantly prognostic and predictive of ICI benefits in advanced melanoma (Table). Analyses in the other cancer types are ongoing. Conclusions: GE data analyses validate the predictive value of immune related gene signatures following ICI immunotherapy in melanoma. Ongoing analyses are investigating these signatures in other malignancies and integrating the GE data with data related to TMB, somatic mutations and germline genetic variations.[Table: see text]

DACH1 mutation frequency in endometrial cancer is associated with high tumor mutation burden in a nationwide cohort

e17634

Background: DACH1 is a novel transcriptional repressor and tumor suppressor gene. DACH1 mutations are associated with poor prognostic features and reduced overall survival in endometrial cancer, with an increased prevalence in the Appalachian region of Kentucky. Preliminary studies have suggested an association with an increase in tumor mutation burden. In this follow up study, we utilized the nationwide Oncology Research Information Exchange Network (ORIEN) to determine the frequency of DACH1 mutations in patients with endometrial cancer through a multi-institution analysis and evaluate its impact on RNA expression, clinical correlates, and outcomes. Methods: We obtained clinical and genomic data for 691 patients with endometrial cancer from nine U.S. institutions within the ORIEN collaborative. We examined the clinical attributes of the cancers with DACH1 status by comparing whole-exome sequencing (WES), RNA Sequencing (RNASeq), microsatellite instability (MSI), and tumor mutational burden (TMB). Results: Appalachian women with endometrial cancer had an increased frequency of DACH1 mutations (6/41 patients, 15%) compared to the non-Appalachian endometrial cancer population (24/581 patients, 4.1%) with p-value = 0.010, with the non-Appalachian DACH1 mutation frequency mirroring the rate of DACH1 gene mutation seen in TCGA at 3.8%. DACH1 mutated patients have a higher tumor mutation burden compared to DACH1 wild-type (32.2 vs. 4.62, p-value = 0.001) though no differences in microsatellite instability between DACH1 mutated and wild-type were present (p-value = 0.350). DACH1 mutations showed significant gene co-occurrence patterns with POLE, MLH1, MSH2, MSH6 and PMS2. Conclusions: DACH1 mutations are prevalent in Kentucky patients with endometrial cancer, particularly those from the Appalachian region. These results were again reflected in the TCGA PanCancer Atlas as well as the ORIEN multi-institution cohort. These mutations are associated with high tumor mutational burden and co-occur with genome destabilizing gene mutations. These findings suggest DACH1 as a candidate biomarker for future trials with immunotherapy, particularly in endometrial cancers.

A deep learning approach utilizing clinical and molecular data for identifying prognostic biomarkers in patients treated with immune checkpoint inhibitors: An ORIEN pan-cancer study

2619

Background: Immune checkpoint inhibitors (ICIs) have made significant improvements in the treatment of cancer patients (pts), but many continue to experience primary or secondary resistance. Here, we leveraged clinical and genomic data to identify prognostic biomarkers in pts treated with ICIs utilizing a pan-cancer approach. Methods: Pts were enrolled to the Total Cancer Care protocol across 18 cancer centers within the Oncology Research Information Exchange Network (ORIEN). RNA-seq was performed on tumors following the RSEM pipeline and gene expressions were quantified as Transcript Per Million (TPM) and were logarithmically normalized. An Auto-Encoder Survival Deep Network (AE-SDN) architecture was developed that combined the reconstruction loss of AE with Cox regression for modeling time to event. For comparison, immunoscore for each pt was calculated based on the estimated densities of tumor CD3+ and CD8+ T cells (Galon, 2020) utilizing CIBERSORTx. The quality of overall survival (OS) predictions was assessed using Harrell’s concordance index (C-index). Log-rank test was used to assess stratified group differences (by ICI or cancer histology) along with Kaplan-Meier (KM) survival analysis of AE-SDN and immunoscore. Results: Pts (n=522) with 4 cancer types including melanoma (n=125), renal cell carcinoma (n=149), non-small cell lung cancer (n=128) and head and neck cancer (n=120) treated with 6 ICI regimens were included in this analysis. ICI regimens were nivolumab (n=219), pembrolizumab (n=202), ipilimumab+nivolumab (n=69), ipilimumab (n=30), avelumab (n=1) and cemiplimab (n=1). The Table summarizes the overall C-index and associated 95% CIs and log-rank P values for the entire cohort (regardless of histology) resulting from our proposed AE-SDN model and the separate estimated immunoscore categorization. AE-SDN top selected genes were mostly related to immunity, carcinogenesis and tumor suppression. The corresponding KM plots showed significantly wider separations of the survival curves in favor of our proposed AE-SDN model relative to the immunoscore with more than 20% improvement in prediction power. Conclusions: Deep network machine learning analysis is a promising approach to identifying relevant prognostic biomarkers in cancer pts treated with ICI. This may lead to novel therapeutic predictive signatures and identification of mechanisms of ICI resistance. Our AE-SDN gene expression signature was significantly prognostic and outperformed the estimated CD3+, CD8+ T Cell immunoscore. Further refinements to our prediction power are ongoing along with more advanced neural network architectures to elucidate related functional pathways. [Table: see text]

Inherited germline variants in urothelial cancer: A multicenter whole-exome sequencing analysis

451

Background: Outside of Lynch syndrome, few genetic factors have been associated with urothelial cancer (UC). This project seeks to describe the frequency of germline variants in a multi-center UC cohort. Methods: Patients diagnosed with UC from 1980 to 2019 and consented to the Total Cancer Care protocol by members of the Oncology Research Information Exchange Network (ORIEN) were included in the study. The ORIEN program includes a convenience sample of cases with both germline and tumor samples available, though this analysis was germline only. Whole exome sequencing data were analyzed using a GATK best practices for germline variant analysis. A series of quality control (allele frequency > = 0.3, read depth > = 20, genotype quality > = 20), annotation, functional impact (loss of function/splicing), and region (cancer predisposition genes/pathways) filters were applied to identify variants of significant consequence. Results: 348 exomes were evaluated. This identified 60 variants classified as pathogenic/likely pathogenic (P/LP) and 17 novel, high impact variants in genes associated with high, moderate, or recessively inherited cancer risk in 77 individuals (22.1%). 15 (4.5%) had P/LP variants in genes associated with a high or moderate cancer risk, and 10 (3%) of these variants were considered to be clinically actionable with associated with cancer screening, risk reduction or treatment recommendations. The high/moderate risk genes with P/LP variants included CHEK2 (n = 5), FANCM (n = 4), MSH6 (n = 2) and single cases of MSH2, BRCA2, ERCC3, and BRIP1. The average age of diagnosis in those with and without a high/moderate risk P/LP variant was 67.6 (57-80 yrs) and 68.9 (45-91 yrs). Family history of cancer was reported for 73% of those with a germline LP/PV and 60% of those without, but this trend was not significant (p =.44). Conclusions: Individuals with UC have a high frequency of germline variants that warrant further study for association with cancer risk. A smaller, but significant portion also carry germline variants that may be clinically relevant to them and/or family members. Currently UC is not included in genetic testing criteria. This study adds to the growing body of research indicating that diverse types of cancer patients harbor germline variants. Strategies for expanding testing should be considered.

The impact of germline and somatic mutations in the homologous recombination repair pathway in pancreatic cancer patients who undergo perioperative chemotherapy

602

Background: Limited data is available regarding the effects of germline and somatic mutations in the homologous recombination repair (HRR) pathway in patients with resectable pancreatic cancer and exactly which mutations can be targeted with platinum-based chemotherapy. We aimed to assess the impact of HRR pathway mutations in a large cohort of pancreatic patients who underwent curative intent surgical resection. Methods: Patients with resectable pancreatic cancer who underwent perioperative chemotherapy, diagnosed from 1999-2020 from the participating members of the Oncology Research Information Exchange Network (ORIEN) were included in the study. Patients with germline and somatic whole exome sequencing data were analyzed for known pathogenic and likely pathogenic variants according to ClinVar in the following HRR pathway genes: BRCA1, BRCA2, PALB2, BRIP1, BRAD1, ATM, RAD51C, RAD51, RAD50, CHECK2, FANCC, FANCA, MRE11 and XRCC2. The Kaplan Meier method was used to compare median overall survival (OS) between patients with adenocarcinoma, with and without HRR pathway mutations. Multivariate cox proportional hazard model was used to calculate HR and 95% CI adjusting for age at diagnosis, sex and pathologic stage. Results: During the study period, the ORIEN cohort included 417 patients with resectable pancreatic cancer and whole exome sequencing. Of these 313 (75%) patients had adenocarcinoma and 104 (25%) neuroendocrine tumor. A total of 19 patients (5%) had an HRR pathway mutation - 15 (5%) in the adenocarcinoma group and 4 (4%) in the neuroendocrine group. In the adenocarcinoma group, 97 (31%) patients underwent platinum-based perioperative chemotherapy. Median OS was 2.8 years (IQR 2.5-3.3) in the adenocarcinoma group without HRR pathway mutation and 3.8 years (IQR 3.4-NA) in the group with HRR pathway mutation (HR 0.6: 95% CI 0.3-1.4, p = 0.76). Conclusions: There was a trend towards improved survival in patients with adenocarcinoma receiving perioperative platinum-based chemotherapy with HRR pathway mutations compared to those without a mutation. This finding supports previous data in the literature regarding the prognostic role of HRR pathway alterations in pancreatic cancer. Larger prospective studies are needed to assess the predictive role of these mutations in the perioperative setting in response to platinum-based chemotherapy.

Germline deletion of ETV6 in familial acute lymphoblastic leukemia

Recent studies have identified germline mutations in TP53, PAX5, ETV6, and IKZF1 in kindreds with familial acute lymphoblastic leukemia (ALL), but the genetic basis of ALL in many kindreds is unknown despite mutational analysis of the exome. Here, we report a germline deletion of ETV6 identified by linkage and structural variant analysis of whole-genome sequencing data segregating in a kindred with thrombocytopenia, B-progenitor acute lymphoblastic leukemia, and diffuse large B-cell lymphoma. The 75-nt deletion removed the ETV6 exon 7 splice acceptor, resulting in exon skipping and protein truncation. The ETV6 deletion was also identified by optimal structural variant analysis of exome sequencing data. These findings identify a new mechanism of germline predisposition in ALL and implicate ETV6 germline variation in predisposition to lymphoma. Importantly, these data highlight the importance of germline structural variant analysis in the search for germline variants predisposing to familial leukemia.

PAX5-driven subtypes of B-progenitor acute lymphoblastic leukemia

Recent genomic studies have identified chromosomal rearrangements defining new subtypes of B-progenitor acute lymphoblastic leukemia (B-ALL), however many cases lack a known initiating genetic alteration. Using integrated genomic analysis of 1,988 childhood and adult cases, we describe a revised taxonomy of B-ALL incorporating 23 subtypes defined by chromosomal rearrangements, sequence mutations or heterogeneous genomic alterations, many of which show marked variation in prevalence according to age. Two subtypes have frequent alterations of the B lymphoid transcription-factor gene PAX5. One, PAX5alt (7.4%), has diverse PAX5 alterations (rearrangements, intragenic amplifications or mutations); a second subtype is defined by PAX5 p.Pro80Arg and biallelic PAX5 alterations. We show that p.Pro80Arg impairs B lymphoid development and promotes the development of B-ALL with biallelic Pax5 alteration in vivo. These results demonstrate the utility of transcriptome sequencing to classify B-ALL and reinforce the central role of PAX5 as a checkpoint in B lymphoid maturation and leukemogenesis.

Abstract IA12: Modeling and targeting CREBBP mutations in relapsed acute lymphoblastic leukemia

Genes encoding epigenetic regulators are commonly mutated at relapse in acute lymphoblastic leukemia, but the way in which such mutations promote drug resistance and their potential for therapeutic targeting is largely unknown. We previously identified mutations in CREBBP, encoding CREBBP binding protein, a histone and non-histone acetylase and transcriptional coregulator in relapsed ALL and non-Hodgkin’s lymphoma. In view of the known role of CREBBP mediation of the transcriptional response to glucocorticoids, we hypothesized that CREBBP mutations promote glucocorticoid resistance, and sought to comprehensively define the prevalence and spectrum of mutations in relapsed ALL, and elucidate their role in resistance.

In studies of childhood ALL from St. Jude and the Children’s Oncology Group, CREBBP mutations were observed in 37 (21.2%) of patients, representing the most common target of sequence mutation at relapse. The mutations were either preserved from diagnosis to relapse, or evolved as new mutations or from minor clones to become clonal mutations at relapse. There were hotspots of mutation at R1446 and between P1488-P1494 in the histone acetyltransferase domain, residues critical for recognition of acetylation substrates, and are known or predicted to impair histone acetylation in biochemical assays and in Crebbp-/-;Ep300-/- mouse embryonic fibroblasts reconstituted with mutant CREBBP.

To examine the role of CREBBP mutations in drug resistance, we generated isogenic CREBBP wild-type and mutant human ALL cell lines, and conventional and conditional Crebbp Q1501P (=human Q1500) and R1447H (=human R1446) mouse strains. Mice bearing germline Q1501P or R1447H alleles exhibited dysmorphology, reduced fertility, lymphoproliferation, and reduced lifespan. This phenotype was most pronounced in Crebbp R1447 mice, preventing establishment of a stable line; thus, we generated a conditional Crebbp-R1447H knockin strain (CREBBP-R1447H-CKI) that expresses CREBBP R1447H following Cre-mediated recombination.

We generated BCR-ABL1 pre-B ALL models by transducing E13.5 fetal liver cells with BCR-ABL1 p185 retrovirus. Consistently, leukemias derived from heterozygous or homozygous CREBBP-Q1501P or heterozygous CREBBP-R1447H-CKI mice showed profound and selective resistance to prednisolone and dexamethasone. Similarly, NALM6 (DUX4/ERG) human ALL cells harboring CREBBP mutations induced by genome editing (NALM6-C001) also showed complete resistance to glucocorticoids, without change in expression of the glucocorticoid receptor (GR). Chromatin immunoprecipitation and sequencing showed that this was accompanied by marked attenuation of recruitment of GR to chromatin, attenuation of histone 3 lysine 18 (H3K18) and H3K27 acetylation, and reduced expression of H3K18Ac-marked genes following dexamethasone treatment. Global acetylome profiling showed that CREBBP mutations resulted in perturbed acetylation of multiple histone and non-histone targets, including CREBBP itself, IKZF1, and MYC, with reciprocal increase in EP300 acetylation.

To identify agents that kill CREBBP mutant leukemic cells and that may reverse steroid resistance, we screened NALM6-C001 with a collection of 11,293 chemicals (5,832 unique compounds enriched for FDA-approved drugs) dosed at 10 mM in the presence or absence of dexamethasone for 48 or 96 hrs. 223 active compounds and 220 compounds targeting epigenetic modifications were tested in 10-point dose response screen. Seventeen compounds with IC50 &lt;10 nM were identified, including a selective CREBBP/EP300 bromodomain inhibitor. This compound restored dexamethasone responsiveness of CREBBP-R1447H-CKI leukemic cells, which were insensitive to steroids or drug alone.

These results demonstrate the high prevalence of CREBBP in relapsed ALL, and show that these mutations have a central role in driving glucocorticoid resistance. These data also suggest that these mutations create a therapeutic vulnerability through EP300 acetylation that may be exploited by bromodomain inhibition.

Citation Format: Yunchao Chang, David W. Woessner, Wenwei Lin, Taosheng Chen, Beisi Xu, Yiping Fan, Haiyan Tan, Junmin Peng, Lawryn Kasper, Michelle L. Churchman, Daniela S. Gerhard, Mignon L. Loh, Stephen P. Hunger, Aman Seth, Karen Gascoigne, Charles G. Mullighan. Modeling and targeting CREBBP mutations in relapsed acute lymphoblastic leukemia [abstract]. In: Proceedings of the AACR Special Conference: Pediatric Cancer Research: From Basic Science to the Clinic; 2017 Dec 3-6; Atlanta, Georgia. Philadelphia (PA): AACR; Cancer Res 2018;78(19 Suppl):Abstract nr IA12.

Germline Genetic IKZF1 Variation and Predisposition to Childhood Acute Lymphoblastic Leukemia

Somatic genetic alterations of IKZF1, which encodes the lymphoid transcription factor IKAROS, are common in high-risk B-progenitor acute lymphoblastic leukemia (ALL) and are associated with poor prognosis. Such alterations result in the acquisition of stem cell-like features, overexpression of adhesion molecules causing aberrant cell-cell and cell-stroma interaction, and decreased sensitivity to tyrosine kinase inhibitors. Here we report coding germline IKZF1 variation in familial childhood ALL and 0.9% of presumed sporadic B-ALL, identifying 28 unique variants in 45 children. The majority of variants adversely affected IKZF1 function and drug responsiveness of leukemic cells. These results identify IKZF1 as a leukemia predisposition gene, and emphasize the importance of germline genetic variation in the development of both familial and sporadic ALL.

Modulation of Navitoclax Sensitivity by Dihydroartemisinin-Mediated MCL-1 Repression in BCR-ABL+ B-Lineage Acute Lymphoblastic Leukemia

Purpose: BCR-ABL⁺ B-ALL leukemic cells are highly dependent on the expression of endogenous antiapoptotic MCL-1 to promote viability and are resistant to BH3-mimetic agents such as navitoclax (ABT-263) that target BCL-2, BCL-X_L, and BCL-W. However, the survival of most normal blood cells and other cell types is also dependent on Mcl-1 Despite the requirement for MCL-1 in these cell types, initial reports of MCL-1-specific BH3-mimetics have not described any overt toxicities associated with single-agent use, but these agents are still early in clinical development. Therefore, we sought to identify approved drugs that could sensitize leukemic cells to ABT-263.Experimental Design: A screen identified dihydroartemisinin (DHA), a water-soluble metabolite of the antimalarial artemisinin. Using mouse and human leukemic cell lines, and primary patient-derived xenografts, the effect of DHA on survival was tested, and mechanistic studies were carried out to discover how DHA functions. We further tested in vitro and in vivo whether combining DHA with ABT-263 could enhance the response of leukemic cells to combination therapy.Results: DHA causes the downmodulation of MCL-1 expression by triggering a cellular stress response that represses translation. The repression of MCL-1 renders leukemic cells highly sensitive to synergistic cell death induced by ABT-263 in a mouse model of BCR-ABL⁺ B-ALL both in vitro and in vivo Furthermore, DHA synergizes with ABT-263 in human Ph⁺ ALL cell lines, and primary patient-derived xenografts of Ph⁺ ALL in culture.Conclusions: Our findings suggest that combining DHA with ABT-263 can improve therapeutic response in BCR-ABL⁺ B-ALL. Clin Cancer Res; 23(24); 7558-68. ©2017 AACR.

Ikaros: Exploiting and targeting the hematopoietic stem cell niche in B-progenitor acute lymphoblastic leukemia

Genetic alterations of IKZF1 encoding the lymphoid transcription factor IKAROS are a hallmark of high-risk B-progenitor acute lymphoblastic leukemia (ALL), such as BCR-ABL1-positive (Ph+) and Ph-like ALL, and are associated with poor outcome even in the era of contemporary chemotherapy incorporating tyrosine kinase inhibitors. Recent experimental mouse modeling of B-progenitor ALL has shown that IKZF1 alterations have multiple effects, including arresting differentiation, skewing lineage of leukemia from myeloid to lymphoid, and, in Ph+ leukemia, conferring resistance to tyrosine kinase inhibitor (TKI) therapy without abrogating ABL1 inhibition. These effects are in part mediated by acquisition of an aberrant hematopoietic stem cell-like program accompanied by induction of cell surface expression of stem cell and adhesion molecules that mediate extravascular invasion and residence in the niche and activation of integrin signaling pathways. These effects can be exploited therapeutically using several approaches. IKZF1 alterations also result in upregulation of RXRA that encodes part of the heterodimeric retinoic acid X receptor. Rexinoids, a synthetic class of retinoids that bind specifically to retinoid "X" receptors such as bexarotene potently reverse aberrant adhesion and niche mislocalization in vivo and induce differentiation and cell cycle arrest. Focal adhesion kinase inhibitors block the downstream integrin-mediated signaling, reverse adhesion, and niche mislocalization. Both agents act synergistically with TKIs to prolong survival of Ph+ ALL in mouse and human xenograft model, with long-term remission induced by focal adhesion kinase inhibitors. Therefore, these findings provide important new conceptual insights into the mechanisms by which IKZF1 alterations result in drug resistance and indicate that therapeutic strategies directed against the pathways deregulated by mutation, rather than attempting to restore IKZF1 expression directly, represent promising therapeutic approaches in this disease.

Deregulation of DUX4 and ERG in acute lymphoblastic leukemia

Chromosomal rearrangements deregulating hematopoietic transcription factors are common in acute lymphoblastic leukemia (ALL). Here we show that deregulation of the homeobox transcription factor gene DUX4 and the ETS transcription factor gene ERG is a hallmark of a subtype of B-progenitor ALL that comprises up to 7% of B-ALL. DUX4 rearrangement and overexpression was present in all cases and was accompanied by transcriptional deregulation of ERG, expression of a novel ERG isoform, ERGalt, and frequent ERG deletion. ERGalt uses a non-canonical first exon whose transcription was initiated by DUX4 binding. ERGalt retains the DNA-binding and transactivation domains of ERG, but it inhibits wild-type ERG transcriptional activity and is transforming. These results illustrate a unique paradigm of transcription factor deregulation in leukemia in which DUX4 deregulation results in loss of function of ERG, either by deletion or induced expression of an isoform that is a dominant-negative inhibitor of wild-type ERG function.

Synergism of FAK and tyrosine kinase inhibition in Ph+ B-ALL

BCR-ABL1⁺ B progenitor acute lymphoblastic leukemia (Ph⁺ B-ALL) is an aggressive disease that frequently responds poorly to currently available therapies. Alterations in IKZF1, which encodes the lymphoid transcription factor Ikaros, are present in over 80% of Ph⁺ ALL and are associated with a stem cell-like phenotype, aberrant adhesion molecule expression and signaling, leukemic cell adhesion to the bone marrow stem cell niche, and poor outcome. Here, we show that FAK1 is upregulated in Ph⁺ B-ALL with further overexpression in IKZF1-altered cells and that the FAK inhibitor VS-4718 potently inhibits aberrant FAK signaling and leukemic cell adhesion, potentiating responsiveness to tyrosine kinase inhibitors, inducing cure in vivo. Thus, targeting FAK with VS-4718 is an attractive approach to overcome the deleterious effects of FAK overexpression in Ph⁺ B-ALL, particularly in abrogating the adhesive phenotype induced by Ikaros alterations, and warrants evaluation in clinical trials for Ph⁺ B-ALL, regardless of IKZF1 status.

Functional Conservation in the SIAMESE-RELATED Family of Cyclin-Dependent Kinase Inhibitors in Land Plants

The best-characterized members of the plant-specific SIAMESE-RELATED (SMR) family of cyclin-dependent kinase inhibitors regulate the transition from the mitotic cell cycle to endoreplication, also known as endoreduplication, an altered version of the cell cycle in which DNA is replicated without cell division. Some other family members are implicated in cell cycle responses to biotic and abiotic stresses. However, the functions of most SMRs remain unknown, and the specific cyclin-dependent kinase complexes inhibited by SMRs are unclear. Here, we demonstrate that a diverse group of SMRs, including an SMR from the bryophyte Physcomitrella patens, can complement an Arabidopsis thaliana siamese (sim) mutant and that both Arabidopsis SIM and P. patens SMR can inhibit CDK activity in vitro. Furthermore, we show that Arabidopsis SIM can bind to and inhibit both CDKA;1 and CDKB1;1. Finally, we show that SMR2 acts to restrict cell proliferation during leaf growth in Arabidopsis and that SIM, SMR1/LGO, and SMR2 play overlapping roles in controlling the transition from cell division to endoreplication during leaf development. These results indicate that differences in SMR function in plant growth and development are primarily due to differences in transcriptional and posttranscriptional regulation, rather than to differences in fundamental biochemical function.

Efficacy of Retinoids in IKZF1-Mutated BCR-ABL1 Acute Lymphoblastic Leukemia

Alterations of IKZF1, encoding the lymphoid transcription factor IKAROS, are a hallmark of high-risk acute lymphoblastic leukemia (ALL), however the role of IKZF1 alterations in ALL pathogenesis is poorly understood. Here, we show that in mouse models of BCR-ABL1 leukemia, Ikzf1 and Arf alterations synergistically promote the development of an aggressive lymphoid leukemia. Ikzf1 alterations result in acquisition of stem cell-like features, including self-renewal and increased bone marrow stromal adhesion. Retinoid receptor agonists reversed this phenotype, partly by inducing expression of IKZF1, resulting in abrogation of adhesion and self-renewal, cell cycle arrest, and attenuation of proliferation without direct cytotoxicity. Retinoids potentiated the activity of dasatinib in mouse and human BCR-ABL1 ALL, providing an additional therapeutic option in IKZF1-mutated ALL.

Interleukin-7 receptor mutants initiate early T cell precursor leukemia in murine thymocyte progenitors with multipotent potential

Early T cell precursor acute lymphoblastic leukemia (ETP-ALL) exhibits lymphoid, myeloid, and stem cell features and is associated with a poor prognosis. Whole genome sequencing of human ETP-ALL cases has identified recurrent mutations in signaling, histone modification, and hematopoietic development genes but it remains to be determined which of these abnormalities are sufficient to initiate leukemia. We show that activating mutations in the interleukin-7 receptor identified in human pediatric ETP-ALL cases are sufficient to generate ETP-ALL in mice transplanted with primitive transduced thymocytes from p19(Arf-/-) mice. The cellular mechanism by which these mutant receptors induce ETP-ALL is the block of thymocyte differentiation at the double negative 2 stage at which myeloid lineage and T lymphocyte developmental potential coexist. Analyses of samples from pediatric ETP-ALL cases and our murine ETP-ALL model show uniformly high levels of LMO2 expression, very low to undetectable levels of BCL11B expression, and a relative lack of activating NOTCH1 mutations. We report that pharmacological blockade of Jak-Stat signaling with ruxolitinib has significant antileukemic activity in this ETP-ALL model. This new murine model recapitulates several important cellular and molecular features of ETP-ALL and should be useful to further define novel therapeutic approaches for this aggressive leukemia.

Expression of arf tumor suppressor in spermatogonia facilitates meiotic progression in male germ cells

The mammalian Cdkn2a (Ink4a-Arf) locus encodes two tumor suppressor proteins (p16^Ink4a and p19^Arf) that respectively enforce the anti-proliferative functions of the retinoblastoma protein (Rb) and the p53 transcription factor in response to oncogenic stress. Although p19^Arf is not normally detected in tissues of young adult mice, a notable exception occurs in the male germ line, where Arf is expressed in spermatogonia, but not in meiotic spermatocytes arising from them. Unlike other contexts in which the induction of Arf potently inhibits cell proliferation, expression of p19^Arf in spermatogonia does not interfere with mitotic cell division. Instead, inactivation of Arf triggers germ cell–autonomous, p53-dependent apoptosis of primary spermatocytes in late meiotic prophase, resulting in reduced sperm production. Arf deficiency also causes premature, elevated, and persistent accumulation of the phosphorylated histone variant H2AX, reduces numbers of chromosome-associated complexes of Rad51 and Dmc1 recombinases during meiotic prophase, and yields incompletely synapsed autosomes during pachynema. Inactivation of Ink4a increases the fraction of spermatogonia in S-phase and restores sperm numbers in Ink4a-Arf doubly deficient mice but does not abrogate γ-H2AX accumulation in spermatocytes or p53-dependent apoptosis resulting from Arf inactivation. Thus, as opposed to its canonical role as a tumor suppressor in inducing p53-dependent senescence or apoptosis, Arf expression in spermatogonia instead initiates a salutary feed-forward program that prevents p53-dependent apoptosis, contributing to the survival of meiotic male germ cells.

The intimately linked Arf and Ink4a genes, encoded in part by overlapping reading frames within the Cdkn2a locus, are induced by oncogenic stress, activating the p53 and Rb tumor suppressors, respectively, to inhibit proliferation of incipient cancer cells. As such, expression of the p19^Arf and p16^Ink4a proteins is undetected in most normal mouse tissues. However, p19^Arf is physiologically expressed in mitotically dividing spermatogonia, the progenitor cells that differentiate to form meiotic spermatocytes in which Arf expression is extinguished. We show that, instead of provoking cell cycle arrest or death, Arf expression in spermatogonia facilitates survival of their meiotic progeny, ensuring production of normal numbers of mature sperm. When Arf is ablated, meiotic defects ensue, along with p53-dependent cell death of spermatocytes, indicating an unexpected role of p53 in monitoring meiotic progression. Surprisingly, it is the absence of p19^Arf rather than its induction that enforces p53 expression in this setting. Co-inactivation of Ink4a compensates for Arf loss by fueling proliferation of spermatogonial progenitors, but does not correct meiotic defects triggered by Arf loss. Although the Arf and Ink4a tumor suppressors are expected to restrain cellular self-renewal, Arf plays an unexpected role in male germ cells by facilitating their proper meiotic progression.

Abstract SY29-03: Regulation of cellular self-renewal by the ARF

The Ink4-Arf locus encodes three tumor suppressor proteins, two of which (p16Ink4a and p15Ink4b) inhibit the activities of cyclin D-dependent kinases (Cdk4 and Cdk6) and a third (p19Arf) that inhibits the Mdm2 E3 ligase to activate p53. Arf induction in response to aberrantly increased and persistent signaling thresholds mediated by activated oncogenes triggers a p53-dependent transcriptional program that can lead to cell cycle arrest or apoptosis, thereby eliminating incipient tumor cells. Arf inactivation compromises these p53-dependent protective effects and synergizes with oncogene activation to drive tumor progression. The Ink4-Arf locus is epigenetically silenced in self-renewing stem cells by Polycomb Group (PcG) complexes but is remodeled as stem cells differentiate to yield progeny that have lost their unlimited self-renewal potential. Thus, although it is generally not engaged in most normal tissues, the Ink4-Arf locus is “poised” to respond to oncogenic stress signals in non-stem cells and can thereby act as a differentiation stage-specific tumor suppressor. Deletion of Ink4-Arf is a frequent event in many forms of cancer, helping to reconfer the ability of tumor cells to self-renew indefinitely. The fact that INK4A-ARF deletions and p53 mutations occur in a mutually exclusive manner in many human tumors provides the strongest argument that disruption of the ARF-HDM2-p53 “pathway” is a hallmark of cancer in humans as well as in mice.

We have produced several strains of mice in which functional Arf coding sequences in the mouse genome have been replaced by reporter genes. These include cassettes encoding green fluorescent protein (GFP) or Cre recombinase that are driven by the cellular Arf promoter. These experiments revealed that, apart from monitoring latent oncogenic signals in living animals, the Arf promoter is transiently engaged in several normal cell types. For example, Arf is expressed in the male germ cell lineage in mitotically amplifying spermatogonia but not in the primary spermatocytes that derive from them. Notably, in this setting, Arf expression is compatible with cell division and does not trigger apoptosis or cellular senescence. However, inactivation of Arf alters the homeostatic relationship between spermatogonia and spermatocytes and leads to a progressive depletion of male germ stem cells and to precocious infertility as animals age. Surprisingly, although the Ink4a gene is co-expressed with Arf in spermatogonia, its inactivation results in increased sperm production, thereby offsetting the effects of Arf loss. In short, despite their intimate linkage and ability to be epigenetically silenced en bloc, Ink4a and Arf inactivation lead to opposing phenotypes in this setting. We suspect that the ability of transient Arf and Ink4a expression to “gate” stem cell self-renewal may reflect a primordial function of this locus distinct from its role in tumor suppression. This may potentially explain why this gene cluster is evolutionarily conserved in mammals, despite its strong propensity to undergo deletion in cancer cells.

Citation Format: Charles J. Sherr, Adam Gromley, Michelle L. Churchman, Emmanuel J. Volanakis. Regulation of cellular self-renewal by the ARF [abstract]. In: Proceedings of the AACR 101st Annual Meeting 2010; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr SY29-03

Novel plant-specific cyclin-dependent kinase inhibitors induced by biotic and abiotic stresses

The EL2 gene of rice (Oryza sativa), previously classified as early response gene against the potent biotic elicitor N-acetylchitoheptaose and encoding a short polypeptide with unknown function, was identified as a novel cell cycle regulatory gene related to the recently reported SIAMESE (SIM) gene of Arabidopsis thaliana. Iterative two-hybrid screens, in vitro pull-down assays, and fluorescence resonance energy transfer analyses showed that Orysa; EL2 binds the cyclin-dependent kinase (CDK) CDKA1;1 and D-type cyclins. No interaction was observed with the plant-specific B-type CDKs. The amino acid motif ELERFL was identified to be essential for cyclin, but not for CDK binding. Orysa;EL2 impaired the ability of Orysa; CYCD5;3 to complement a budding yeast (Saccharomyces cerevisiae) triple CLN mutant, whereas recombinant protein inhibited CDK activity in vitro. Moreover, Orysa;EL2 was able to rescue the multicellular trichome phenotype of sim mutants of Arabidopsis, unequivocally demonstrating that Orysa;EL2 operates as a cell cycle inhibitor. Orysa;EL2 mRNA levels were induced by cold, drought, and propionic acid. Our data suggest that Orysa;EL2 encodes a new type of plant CDK inhibitor that links cell cycle progression with biotic and abiotic stress responses.

SIAMESE, a plant-specific cell cycle regulator, controls endoreplication onset in Arabidopsis thaliana

Recessive mutations in the SIAMESE (SIM) gene of Arabidopsis thaliana result in multicellular trichomes harboring individual nuclei with a low ploidy level, a phenotype strikingly different from that of wild-type trichomes, which are single cells with a nuclear DNA content of approximately 16C to 32C. These observations suggested that SIM is required to suppress mitosis as part of the switch to endoreplication in trichomes. Here, we demonstrate that SIM encodes a nuclear-localized 14-kD protein containing a cyclin binding motif and a motif found in ICK/KRP (for Interactors of Cdc2 kinase/Kip-related protein) cell cycle inhibitor proteins. Accordingly, SIM was found to associate with D-type cyclins and CDKA;1. Homologs of SIM were detected in other dicots and in monocots but not in mammals or fungi. SIM proteins are expressed throughout the shoot apical meristem, in leaf primordia, and in the elongation zone of the root and are localized to the nucleus. Plants overexpressing SIM are slow-growing and have narrow leaves and enlarged epidermal cells with an increased DNA content resulting from additional endocycles. We hypothesize that SIM encodes a plant-specific CDK inhibitor with a key function in the mitosis-to-endoreplication transition.