Negative Hyperselection of Resistance Mutations for Panitumumab Maintenance in RAS Wild-Type Metastatic Colorectal Cancer (PanaMa Phase II Trial, AIO KRK 0212).

PURPOSE: We evaluated additional mutations in RAS wild-type (WT) metastatic colorectal cancer (mCRC) as prognostic and predictive biomarkers for the efficacy of added panitumumab to a 5-fluorouracil plus folinic acid (FU/FA) maintenance as pre-specified analysis of the randomized PanaMa trial. PATIENTS AND METHODS: Mutations (MUT) were identified using targeted next-generation sequencing (NGS; Illumina Cancer Hotspot Panel v2) and IHC. RAS/BRAF V600E/PIK3CA/AKT1/ALK1/ERBB2/PTEN MUT and HER2/neu overexpressions were negatively hyperselected and correlated with median progression-free survival (PFS) and overall survival (OS) since start of maintenance treatment, and objective response rates (ORR). Univariate/multivariate Cox regression estimated hazard ratios (HR) and 95% confidence intervals (CI). RESULTS: 202 of 248 patients (81.5%) of the full analysis set (FAS) had available NGS data: hyperselection WT, 162 (80.2%); MUT, 40 (19.8%). From start of maintenance therapy, hyperselection WT tumors were associated with longer median PFS as compared with hyperselection MUT mCRC (7.5 vs. 5.4 months; HR, 0.75; 95% CI, 0.52-1.07; P = 0.11), OS (28.7 vs. 22.2 months; HR, 0.53; 95% CI, 0.36-0.77; P = 0.001), and higher ORR (35.8% vs. 25.0%, P = 0.26). The addition of panitumumab to maintenance was associated with significant benefit in hyperselection WT tumors for PFS (9.2 vs. 6.0 months; HR, 0.66; 95% CI, 0.47-0.93; P = 0.02) and numerically also for OS (36.9 vs. 24.9 months; HR, 0.91; 95% CI, 0.61-1.36; P = 0.50), but not in hyperselection MUT tumors. Hyperselection status interacted with maintenance treatment arms in terms of PFS (P = 0.06) and OS (P = 0.009). CONCLUSIONS: Extended molecular profiling beyond RAS may have the potential to improve the patient selection for anti-EGFR containing maintenance regimens.

Serial Analysis of Gene Mutations and Gene Expression during First-Line Chemotherapy against Metastatic Colorectal Cancer: Identification of Potentially Actionable Targets within the Multicenter Prospective Biomarker Study REVEAL.

Most metastatic colorectal cancer (mCRC) patients succumb to refractory disease due to secondary chemotherapy resistance. To elucidate the molecular changes associated with secondary resistance, we recruited 64 patients with mCRC and hepatic metastases before standard first-line chemotherapy between 2014 and 2018. We subjected DNA from primary tumor specimens (P), hepatic metastasis specimens after treatment (M), and liquid biopsies (L) taken prior to (pre), during (intra), and after (post) treatment to next generation sequencing. We performed Nanostring expression analysis in P and M specimens. Comparative bioinformatics and statistical analysis revealed typical mutational patterns with frequent alterations in TP53, APC, and KRAS in P specimens (n = 48). P and pre-L (n = 42), as well as matched P and M (n = 30), displayed a similar mutation spectrum. In contrast, gene expression profiles classified P (n = 31) and M (n = 23), distinguishable by up-regulation of immune/cytokine receptor and autophagy programs. Switching of consensus molecular subtypes from P to M occurred in 58.3% of cases. M signature genes SFRP2 and SPP1 associated with inferior survival, as validated in an independent cohort. Molecular changes during first-line treatment were detectable by expression profiling rather than by mutational tumor and liquid biopsy analyses. SFRP2 and SPP1 may serve as biomarkers and/or actionable targets.

A RAS-Independent Biomarker Panel to Reliably Predict Response to MEK Inhibition in Colorectal Cancer.

Background: In colorectal cancer (CRC), mutations of genes associated with the TGF-ß/BMP signaling pathway, particularly affecting SMAD4, are known to correlate with decreased overall survival and it is assumed that this signaling axis plays a key role in chemoresistance. Methods: Using CRISPR technology on syngeneic patient-derived organoids (PDOs), we investigated the role of a loss-of-function of SMAD4 in sensitivity to MEK-inhibitors. CRISPR-engineered SMAD4R361H PDOs were subjected to drug screening, RNA-Sequencing, and multiplex protein profiling (DigiWest®). Initial observations were validated on an additional set of 62 PDOs with known mutational status. Results: We show that loss-of-function of SMAD4 renders PDOs sensitive to MEK-inhibitors. Multiomics analyses indicate that disruption of the BMP branch within the TGF-ß/BMP pathway is the pivotal mechanism of increased drug sensitivity. Further investigation led to the identification of the SFAB-signature (SMAD4, FBXW7, ARID1A, or BMPR2), coherently predicting sensitivity towards MEK-inhibitors, independent of both RAS and BRAF status. Conclusion: We identified a novel mutational signature that reliably predicts sensitivity towards MEK-inhibitors, regardless of the RAS and BRAF status. This finding poses a significant step towards better-tailored cancer therapies guided by the use of molecular biomarkers.

Clonality and timing of relapsing colorectal cancer metastasis revealed through whole-genome single-cell sequencing.

Recurrence of tumor cells following local and systemic therapy is a significant hurdle in cancer. Most patients with metastatic colorectal cancer (mCRC) will relapse, despite resection of the metastatic lesions. A better understanding of the evolutionary history of recurrent lesions is required to identify the spatial and temporal patterns of metastatic progression and expose the genetic and evolutionary determinants of therapeutic resistance. With this goal in mind, here we leveraged a unique single-cell whole-genome sequencing dataset from recurrent hepatic lesions of an mCRC patient. Our phylogenetic analysis confirms that the treatment induced a severe demographic bottleneck in the liver metastasis but also that a previously diverged lineage survived this surgery, possibly after migration to a different site in the liver. This lineage evolved very slowly for two years under adjuvant drug therapy and diversified again in a very short period. We identified several non-silent mutations specific to this lineage and inferred a substantial contribution of chemotherapy to the overall, genome-wide mutational burden. All in all, our study suggests that mCRC subclones can migrate locally and evade resection, keep evolving despite rounds of chemotherapy, and re-expand explosively.

DNA methylation reveals distinct cells of origin for pancreatic neuroendocrine carcinomas and pancreatic neuroendocrine tumors.

BACKGROUND: Pancreatic neuroendocrine neoplasms (PanNENs) fall into two subclasses: the well-differentiated, low- to high-grade pancreatic neuroendocrine tumors (PanNETs), and the poorly-differentiated, high-grade pancreatic neuroendocrine carcinomas (PanNECs). While recent studies suggest an endocrine descent of PanNETs, the origin of PanNECs remains unknown. METHODS: We performed DNA methylation analysis for 57 PanNEN samples and found that distinct methylation profiles separated PanNENs into two major groups, clearly distinguishing high-grade PanNECs from other PanNETs including high-grade NETG3. DNA alterations and immunohistochemistry of cell-type markers PDX1, ARX, and SOX9 were utilized to further characterize PanNECs and their cell of origin in the pancreas. RESULTS: Phylo-epigenetic and cell-type signature features derived from alpha, beta, acinar, and ductal adult cells suggest an exocrine cell of origin for PanNECs, thus separating them in cell lineage from other PanNENs of endocrine origin. CONCLUSIONS: Our study provides a robust and clinically applicable method to clearly distinguish PanNECs from G3 PanNETs, improving patient stratification.

Mitogen-activated protein kinase activity drives cell trajectories in colorectal cancer.

In colorectal cancer, oncogenic mutations transform a hierarchically organized and homeostatic epithelium into invasive cancer tissue lacking visible organization. We sought to define transcriptional states of colorectal cancer cells and signals controlling their development by performing single-cell transcriptome analysis of tumors and matched non-cancerous tissues of twelve colorectal cancer patients. We defined patient-overarching colorectal cancer cell clusters characterized by differential activities of oncogenic signaling pathways such as mitogen-activated protein kinase and oncogenic traits such as replication stress. RNA metabolic labeling and assessment of RNA velocity in patient-derived organoids revealed developmental trajectories of colorectal cancer cells organized along a mitogen-activated protein kinase activity gradient. This was in contrast to normal colon organoid cells developing along graded Wnt activity. Experimental targeting of EGFR-BRAF-MEK in cancer organoids affected signaling and gene expression contingent on predictive KRAS/BRAF mutations and induced cell plasticity overriding default developmental trajectories. Our results highlight directional cancer cell development as a driver of non-genetic cancer cell heterogeneity and re-routing of trajectories as a response to targeted therapy.

Correction to: The majority of ß-catenin mutations in colorectal cancer is homozygous.

An amendment to this paper has been published and can be accessed via the original article.

The majority of ß-catenin mutations in colorectal cancer is homozygous.

BACKGROUND: ß-catenin activation plays a crucial role for tumourigenesis in the large intestine but except for Lynch syndrome (LS) associated cancers stabilizing mutations of ß-catenin gene (CTNNB1) are rare in colorectal cancer (CRC). Previous animal studies provide an explanation for this observation. They showed that CTNNB1 mutations induced transformation in the colon only when CTNNB1 was homozygously mutated or when membranous ß-catenin binding was hampered by E-cadherin haploinsufficiency. We were interested, if these mechanisms are also found in human CTNNB1 mutated CRCs. RESULTS: Among 869 CRCs stabilizing CTNNB1 mutations were found in 27 cases. Homo- or hemizygous CTNNB1 mutations were detected in 74% of CTNNB1 mutated CRCs (13 microsatellite instabile (MSI-H), 7 microsatellite stabile (MSS)) but only in 3% (1/33) of extracolonic CTNNB1 mutated cancers. In contrast to MSS CRC, CTNNB1 mutations at codon 41 or 45 were highly selected in MSI-H CRC. Of the examined three CRC cell lines, ß-catenin and E-cadherin expression was similar in cell lines without or with hetereozygous CTNNB1 mutations (DLD1 and HCT116), while a reduced E-cadherin expression combined with cytoplasmic accumulation of ß-catenin was found in a cell line with homozygous CTNNB1 mutation (LS180). Reduced expression of E-cadherin in human MSI-H CRC tissue was identified in 60% of investigated cancers, but no association with the CTNNB1 mutational status was found. CONCLUSIONS: In conclusion, this study shows that in contrast to extracolonic cancers stabilizing CTNNB1 mutations in CRC are commonly homo- or hemizygous indicating a higher threshold of ß-catenin stabilization to be required for transformation in the colon as compared to extracolonic sites. Moreover, we found different mutational hotspots in CTNNB1 for MSI-H and MSS CRCs suggesting a selection of different effects on ß-catenin stabilization according to the molecular pathway of tumourigenesis. Reduced E-cadherin expression in CRC may further contribute to higher levels of transcriptionally active ß-catenin, but it is not directly linked to the CTNNB1 mutational status.

Malignant transformation and genetic alterations are uncoupled in early colorectal cancer progression.

BACKGROUND: Colorectal cancer (CRC) development is generally accepted as a sequential process, with genetic mutations determining phenotypic tumor progression. However, matching genetic profiles with histological transition requires the analyses of temporal samples from the same patient at key stages of progression. RESULTS: Here, we compared the genetic profiles of 34 early carcinomas with their respective adenomatous precursors to assess timing and heterogeneity of driver alterations accompanying the switch from benign adenoma to malignant carcinoma. In almost half of the cases, driver mutations specific to the carcinoma stage were not observed. In samples where carcinoma-specific alterations were present, TP53 mutations and chromosome 20 copy gains commonly accompanied the switch from adenomatous tissue to carcinoma. Remarkably, 40% and 50% of high-grade adenomas shared TP53 mutations and chromosome 20 gains, respectively, with their matched carcinomas. In addition, multi-regional analyses revealed greater heterogeneity of driver mutations in adenomas compared to their matched carcinomas. CONCLUSION: Genetic alterations in TP53 and chromosome 20 occur at the earliest histological stage in colorectal carcinomas (pTis and pT1). However, high-grade adenomas can share these alterations despite their histological distinction. Based on the well-defined sequence of CRC development, we suggest that the timing of genetic changes during neoplastic progression is frequently uncoupled from histological progression.

Reduced replication origin licensing selectively kills KRAS-mutant colorectal cancer cells via mitotic catastrophe.

To unravel vulnerabilities of KRAS-mutant CRC cells, a shRNA-based screen specifically inhibiting MAPK pathway components and targets was performed in CaCo2 cells harboring conditional oncogenic KRASG12V. The custom-designed shRNA library comprised 121 selected genes, which were previously identified to be strongly regulated in response to MEK inhibition. The screen showed that CaCo2 cells expressing KRASG12V were sensitive to the suppression of the DNA replication licensing factor minichromosome maintenance complex component 7 (MCM7), whereas KRASwt CaCo2 cells were largely resistant to MCM7 suppression. Similar results were obtained in an isogenic DLD-1 cell culture model. Knockdown of MCM7 in a KRAS-mutant background led to replication stress as indicated by increased nuclear RPA focalization. Further investigation showed a significant increase in mitotic cells after simultaneous MCM7 knockdown and KRASG12V expression. The increased percentage of mitotic cells coincided with strongly increased DNA damage in mitosis. Taken together, the accumulation of DNA damage in mitotic cells is due to replication stress that remained unresolved, which results in mitotic catastrophe and cell death. In summary, the data show a vulnerability of KRAS-mutant cells towards suppression of MCM7 and suggest that inhibiting DNA replication licensing might be a viable strategy to target KRAS-mutant cancers.

DNA methylation profiling reliably distinguishes pulmonary enteric adenocarcinoma from metastatic colorectal cancer.

Pulmonary enteric adenocarcinoma is a rare non-small cell lung cancer subtype. It is poorly characterized and cannot be distinguished from metastatic colorectal or upper gastrointestinal adenocarcinomas by means of routine pathological methods. As DNA methylation patterns are known to be highly tissue specific, we aimed to develop a methylation-based algorithm to differentiate these entities. To this end, genome-wide methylation profiles of 600 primary pulmonary, colorectal, and upper gastrointestinal adenocarcinomas obtained from The Cancer Genome Atlas and the Gene Expression Omnibus database were used as a reference cohort to train a machine learning algorithm. The resulting classifier correctly classified all samples from a validation cohort consisting of 680 primary pulmonary, colorectal and upper gastrointestinal adenocarcinomas, demonstrating the ability of the algorithm to reliably distinguish these three entities. We then analyzed methylation data of 15 pulmonary enteric adenocarcinomas as well as four pulmonary metastases and four primary colorectal adenocarcinomas with the algorithm. All 15 pulmonary enteric adenocarcinomas were reliably classified as primary pulmonary tumors and all four metastases as well as all four primary colorectal cancer samples were identified as colorectal adenocarcinomas. In a t-distributed stochastic neighbor embedding analysis, the pulmonary enteric adenocarcinoma samples did not form a separate methylation subclass but rather diffusely intermixed with other pulmonary cancers. Additional characterization of the pulmonary enteric adenocarcinoma series using fluorescence in situ hybridization, next-generation sequencing and copy number analysis revealed KRAS mutations in nine of 15 samples (60%) and a high number of structural chromosomal changes. Except for an unusually high rate of chromosome 20 gain (67%), the molecular data was mostly reminiscent of standard pulmonary adenocarcinomas. In conclusion, we provide sound evidence of the pulmonary origin of pulmonary enteric adenocarcinomas and in addition provide a publicly available machine learning-based algorithm to reliably distinguish these tumors from metastatic colorectal cancer.

Robust in-silico identification of Cancer Cell Lines based on RNA and targeted DNA sequencing data.

Cancer cell lines (CCL) are an integral part of modern cancer research but are susceptible to misidentification. The increasing popularity of sequencing technologies motivates the in-silico identification of CCLs based on their mutational fingerprint, but care must be taken when identifying heterogeneous data. We recently developed the proof-of-concept Uniquorn 1 method which could reliably identify heterogeneous sequencing data from selected sequencing technologies. Here we present Uniquorn 2, a generic and robust in-silico identification method for CCLs with DNA/RNA-seq and panel-seq information. We benchmarked Uniquorn 2 by cross-identifying 1612 RNA and 3596 panel-sized NGS profiles derived from 1516 CCLs, five repositories, four technologies and three major cancer panel-designs. Our method achieves an accuracy of 96% for RNA-seq and 95% for mixed DNA-seq and RNA-seq identification. Even for a panel of only 94 cancer-related genes, accuracy remains at 82% but decreases when using smaller panels. Uniquorn 2 is freely available as R-Bioconductor-package 'Uniquorn'.

DNA copy number changes define spatial patterns of heterogeneity in colorectal cancer.

Genetic heterogeneity between and within tumours is a major factor determining cancer progression and therapy response. Here we examined DNA sequence and DNA copy-number heterogeneity in colorectal cancer (CRC) by targeted high-depth sequencing of 100 most frequently altered genes. In 97 samples, with primary tumours and matched metastases from 27 patients, we observe inter-tumour concordance for coding mutations; in contrast, gene copy numbers are highly discordant between primary tumours and metastases as validated by fluorescent in situ hybridization. To further investigate intra-tumour heterogeneity, we dissected a single tumour into 68 spatially defined samples and sequenced them separately. We identify evenly distributed coding mutations in APC and TP53 in all tumour areas, yet highly variable gene copy numbers in numerous genes. 3D morpho-molecular reconstruction reveals two clusters with divergent copy number aberrations along the proximal-distal axis indicating that DNA copy number variations are a major source of tumour heterogeneity in CRC.

SoFIA: a data integration framework for annotating high-throughput datasets.

MOTIVATION: Integrating heterogeneous datasets from several sources is a common bioinformatics task that often requires implementing a complex workflow intermixing database access, data filtering, format conversions, identifier mapping, among further diverse operations. Data integration is especially important when annotating next generation sequencing data, where a multitude of diverse tools and heterogeneous databases can be used to provide a large variety of annotation for genomic locations, such a single nucleotide variants or genes. Each tool and data source is potentially useful for a given project and often more than one are used in parallel for the same purpose. However, software that always produces all available data is difficult to maintain and quickly leads to an excess of data, creating an information overload rather than the desired goal-oriented and integrated result. RESULTS: We present SoFIA, a framework for workflow-driven data integration with a focus on genomic annotation. SoFIA conceptualizes workflow templates as comprehensive workflows that cover as many data integration operations as possible in a given domain. However, these templates are not intended to be executed as a whole; instead, when given an integration task consisting of a set of input data and a set of desired output data, SoFIA derives a minimal workflow that completes the task. These workflows are typically fast and create exactly the information a user wants without requiring them to do any implementation work. Using a comprehensive genome annotation template, we highlight the flexibility, extensibility and power of the framework using real-life case studies. AVAILABILITY AND IMPLEMENTATION: https://github.com/childsish/sofia/releases/latest under the GNU General Public License CONTACT: liam.childs@hu-berlin.de SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Loss of prolyl hydroxylase-2 in myeloid cells and T-lymphocytes impairs tumor development.

The tumor microenvironment plays a pivotal role during cancer development and progression. The balance between suppressive and cytotoxic responses of the tumor immune microenvironment has been shown to have a direct effect on the final outcome in various human and experimental tumors. Recently, we demonstrated that the oxygen sensor HIF-prolyl hydroxylase-2 (PHD2) plays a detrimental role in tumor cells, stimulating systemic growth and metastasis in mice. In our current study, we show that the conditional ablation of PHD2 in the hematopoietic system also leads to reduced tumor volume, intriguingly generated by an imbalance between enhanced cell death and improved proliferation of tumor cells. This effect seems to rely on the overall downregulation of protumoral as well as antitumoral cytokines. Using different genetic approaches, we were able to confine this complex phenotype to the crosstalk of PHD2-deficient myeloid cells and T-lymphocytes. Taken together, our findings reveal a multifaceted role for PHD2 in several hematopoietic lineages during tumor development and might have important implications for the development of tumor therapies in the future.

Loss of epithelial hypoxia-inducible factor prolyl hydroxylase 2 accelerates skin wound healing in mice.

Skin wound healing in mammals is a complex, multicellular process that depends on the precise supply of oxygen. Hypoxia-inducible factor (HIF) prolyl hydroxylase 2 (PHD2) serves as a crucial oxygen sensor and may therefore play an important role during reepithelialization. Hence, this study was aimed at understanding the role of PHD2 in cutaneous wound healing using different lines of conditionally deficient mice specifically lacking PHD2 in inflammatory, vascular, or epidermal cells. Interestingly, PHD2 deficiency only in keratinocytes and not in myeloid or endothelial cells was found to lead to faster wound closure, which involved enhanced migration of the hyperproliferating epithelium. We demonstrate that this effect relies on the unique expression of ß3-integrin in the keratinocytes around the tip of the migrating tongue in an HIF1α-dependent manner. Furthermore, we show enhanced proliferation of these cells in the stratum basale, which is directly related to their attenuated transforming growth factor ß signaling. Thus, loss of the central oxygen sensor PHD2 in keratinocytes stimulates wound closure by prompting skin epithelial cells to migrate and proliferate. Inhibition of PHD2 could therefore offer novel therapeutic opportunities for the local treatment of cutaneous wounds.

HIF prolyl hydroxylase 2 (PHD2) is a critical regulator of hematopoietic stem cell maintenance during steady-state and stress.

Hypoxia is a prominent feature in the maintenance of hematopoietic stem cell (HSC) quiescence and multipotency. Hypoxia-inducible factor (HIF) prolyl hydroxylase domain proteins (PHDs) serve as oxygen sensors and may therefore regulate this system. Here, we describe a mouse line with conditional loss of HIF prolyl hydroxylase 2 (PHD2) in very early hematopoietic precursors that results in self-renewal of multipotent progenitors under steady-state conditions in a HIF1α- and SMAD7-dependent manner. Competitive bone marrow (BM) transplantations show decreased peripheral and central chimerism of PHD2-deficient cells but not of the most primitive progenitors. Conversely, in whole BM transfer, PHD2-deficient HSCs replenish the entire hematopoietic system and display an enhanced self-renewal capacity reliant on HIF1α. Taken together, our results demonstrate that loss of PHD2 controls the maintenance of the HSC compartment under physiological conditions and causes the outcompetition of PHD2-deficient hematopoietic cells by their wild-type counterparts during stress while promoting the self-renewal of very early hematopoietic progenitors.

Hypoxia-inducible factors as key regulators of tumor inflammation.

Low levels of oxygen or hypoxia is often an obstacle in health, particularly in pathological disorders like cancer. The main family of transcription factors responsible for cell survival and adaptation under strenuous conditions of hypoxia are the "hypoxia-inducible factors" (HIFs). Together with prolyl hydroxylase domain enzymes (PHDs), HIFs regulates tumor angiogenesis, proliferation, invasion, metastasis, in addition to resistance to radiation and chemotherapy. Additionally, the entire HIF transcription cascade is involved in the "seventh" hallmark of cancer; inflammation. Studies have shown that hypoxia can influence tumor associated immune cells toward assisting in tumor proliferation, differentiation, vessel growth, distant metastasis and suppression of the immune response via cytokine expression alterations. These changes are not necessarily analogous to HIF's role in non-cancer immune responses, where hypoxia often encourages a strong inflammatory response.

HIF-1α is a protective factor in conditional PHD2-deficient mice suffering from severe HIF-2α-induced excessive erythropoiesis.

Erythropoiesis must be tightly balanced to guarantee adequate oxygen delivery to all tissues in the body. This process relies predominantly on the hormone erythropoietin (EPO) and its transcription factor hypoxia inducible factor (HIF). Accumulating evidence suggests that oxygen-sensitive prolyl hydroxylases (PHDs) are important regulators of this entire system. Here, we describe a novel mouse line with conditional PHD2 inactivation (cKO P2) in renal EPO producing cells, neurons, and astrocytes that displayed excessive erythrocytosis because of severe overproduction of EPO, exclusively driven by HIF-2α. In contrast, HIF-1α served as a protective factor, ensuring survival of cKO P2 mice with HCT values up to 86%. Using different genetic approaches, we show that simultaneous inactivation of PHD2 and HIF-1α resulted in a drastic PHD3 reduction with consequent overexpression of HIF-2α-related genes, neurodegeneration, and lethality. Taken together, our results demonstrate for the first time that conditional loss of PHD2 in mice leads to HIF-2α-dependent erythrocytosis, whereas HIF-1α protects these mice, providing a platform for developing new treatments of EPO-related disorders, such as anemia.

Inhibition of HIF prolyl hydroxylase-2 blocks tumor growth in mice through the antiproliferative activity of TGFß.

Virtually all solid tumors are dependent on a vascular network to provide them with the right amount of nutrients and oxygen. In that sense, low oxygen tension or hypoxia leads to an adaptive response that is transcriptionally regulated by the hypoxia-inducible factors (HIF), which are tightly controlled by the HIF prolyl hydroxylases (PHD). In this study, we show that inhibition of the oxygen sensor PHD2 in tumor cells stimulates vessel formation but paradoxically results in a profound reduction of tumor growth. This effect relies on the antiproliferative nature of the TGFß signaling pathway, in a largely HIF-independent manner. Moreover, our findings reveal that PHD2 has an essential function in controlling the dual nature of TGFß during tumorigenesis and may offer an alternative opportunity for anticancer therapy.