Astrocyte-specific neurofibromatosis 1 gene deletion is insufficient for astrocytoma formation in mice

To gain insight into the role of the NF1 (Neurofibromatosis type 1) gene during neural development and in tumorigenesis, we have utilized the bacteriophage P1, Cre/loxP system to generate a conditional allele at the NF1 locus (NF1 flox) that permits temporal and spatial ablation of function through Cre-mediated recombination. We have been using these mice to assess the scope of NF1 requirement in distinct cell types. At the center of this approach is to identify the cells that give origin to the tumors most frequently found in NF1 patients: neurofibromas, neurofibrosarcomas, and astrocytomas. We have hypothesized that specific stem cells must lose NF1 by LOH to begin this process. I will discuss the consequences of NF1 loss in neurons, Schwann cells, and neural precursors. Distinct tumor phenotypes appear in each case. In malignant tumors, our mouse models indicate that the p53 pathway must also become mutated to cooperate with loss of NF1. Additionally, we have genetic evidence that the haploin-sufficient state is essential for tumor appearance. These data suggest that profilactic therapies preceding tumor appearance should be considered for NF1. Acknowledgements:  Funded by NINDS, NNFF, and DOD.     

A potential role for NF1 mRNA editing in the pathogenesis of NF1 tumors.

Neurofibromatosis type I (NF1) is a common disorder that predisposes to neoplasia in tissues derived from the embryonic neural crest. The NF1 gene encodes a tumor suppressor that most likely acts through the interaction of its GTPase-activating protein (GAP)-related domain (GRD) with the product of the ras protooncogene. We have previously identified a site in the NF1 mRNA, within the first half of the NF1 GRD, which undergoes base-modification editing. Editing at that site changes a C to a U, thereby introducing an in-frame stop codon. NF1 RNA editing has been detected in all cell types studied, to date. In order to investigate the role played by editing in NF1 tumorigenesis, we analyzed RNA from 19 NF1 and 4 non-NF1 tumors. We observed varying levels of NF1 mRNA editing in different tumors, with a higher range of editing levels in more malignant tumors (e.g., neurofibrosarcomas) compared to benign tumors (cutaneous neurofibromas). Plexiform neurofibromas have an intermediate range of levels of NF1 mRNA editing. We also compared tumor and nontumor tissues from several NF1 individuals, to determine the extent of variability present in the constitutional levels of NF1 mRNA editing and to determine whether higher levels are present in tumors. The constitutional levels of NF1 mRNA editing varied slightly but were consistent with the levels observed in non-NF1 individuals. In every case, there was a greater level of NF1 mRNA editing in the tumor than in the nontumor tissue from the same patient. These results suggest that inappropriately high levels of NF1 mRNA editing does play a role in NF1 tumorigenesis and that editing may result in the functional equivalent of biallelic inactivation of the NF1 tumor suppressor.     

The haploinsufficient hematopoietic microenvironment is critical to the pathological fracture repair in murine models of neurofibromatosis type 1

Germline mutations in the NF1 tumor suppressor gene cause neurofibromatosis type 1 (NF1), a complex genetic disorder with a high predisposition of numerous skeletal dysplasias including short stature, osteoporosis, kyphoscoliosis, and fracture non-union (pseudoarthrosis). We have developed murine models that phenocopy many of the skeletal dysplasias observed in NF1 patients, including reduced bone mass and fracture non-union. We also show that the development of these skeletal manifestations requires an Nf1 haploinsufficient background in addition to nullizygous loss of Nf1 in mesenchymal stem/progenitor cells (MSCs) and/or their progenies. This is replicated in two animal models of NF1, PeriCre(+);Nf1(flox/-) and Col2.3Cre(+);Nf1(flox/-) mice. Adoptive transfer experiments demonstrate a critical role of the Nf1+/- marrow microenvironment in the impaired fracture healing in both models and adoptive transfer of WT bone marrow cells improves fracture healing in these mice. To our knowledge, this is the first demonstration of a non-cell autonomous mechanism in non-malignant NF1 manifestations. Collectively, these data provide evidence of a combinatory effect between nullizygous loss of Nf1 in osteoblast progenitors and haploinsufficiency in hematopoietic cells in the development of non-malignant NF1 manifestations.     

Astrocyte-specific inactivation of the neurofibromatosis 1 gene (NF1) is insufficient for astrocytoma formation

Individuals with the neurofibromatosis 1 (NF1) inherited tumor syndrome develop low-grade gliomas (astrocytomas) at an increased frequency, suggesting that the NF1 gene is a critical growth regulator for astrocytes. In an effort to determine the contribution of the NF1 gene product, neurofibromin, to astrocyte growth regulation and NF1-associated astrocytoma formation, we generated astrocyte-specific Nf1 conditional knockout mice (Nf1(GFAP)CKO) by using Cre/LoxP technology. Transgenic mice were developed in which Cre recombinase was specifically expressed in astrocytes by embryonic day 14.5. Successive intercrossing with mice bearing a conditional Nf1 allele (Nf1flox) resulted in GFAP-Cre Nf1flox/flox (Nf1(GFAP)CKO) animals. No astrocytoma formation or neurological impairment was observed in Nf1(GFAP)CKO mice after 20 months, but increased numbers of proliferating astrocytes were observed in several brain regions. To determine the consequence of Nf1 inactivation at different developmental times, the growth properties of embryonic day 12.5 and postnatal day 2 Nf1 null astrocytes were analyzed. Nf1 null astrocytes exhibited increased proliferation but lacked tumorigenic properties in vitro and did not form tumors when injected into immunocompromised mouse brains in vivo. Collectively, our results suggest that loss of neurofibromin is not sufficient for astrocytoma formation in mice and that other genetic or environmental factors might influence NF1-associated glioma tumorigenesis.     

Poster Sessions BP02: Oligodendrocytes and Schwann Cells. RAS activation regulates angiogenic and anti‐angiogenic factor expression in NF1‐derived Schwann cells

Neurofibromatosis Type 1 tumors are highly vascularized and contain Schwann cells with hyperactivated Ras. In vitro, the NF1‐derived neurofibromin deficient Schwann cells have an angiogenic profile, which favors angiogenesis and sustains the growth of the NF1‐derived tumors. This study examined the relationship of the activation state of Ras as it related to the expression of angiogenic and antiangiogenic factors in both cultured NF1‐derived Schwann cells and normal human Schwann cells. Western blot analysis of normal human Schwann cells revealed low expression of angiogenic vascular endothelial growth factor (VEGF) as well as low expression of the antiangiogenic pigment epithelium derived factor (PEDF). Relative to normal human Schwann cells, NF1‐derived Schwann cells have increased RAS activity and a three‐fold increase in VEGF expression. Surprisingly, PEDF was also expressed in the NF1‐derived Schwann cells at approximately the same level as VEGF expression. Using a retroviral construct, we introduced the GAP‐related domain of neurofibromin into the NF1‐derived Schwann cells to reduce the level of activated Ras. Relative to the untreated NF1‐derived Schwann cells the Schwann cells expressing the GAP‐related domain expressed about one‐half the VEGF but twice the PEDF. We conclude that decreasing the Ras activity in NF1‐drived Schwann cells will not only decrease proliferation, but also slow tumor angiogenesis due to the decreased expression of angiogenic and increased expression of antiangiogenic factors.     

Somatic NF1 mutational spectrum in benign neurofibromas: mRNA splice defects are common among point mutations

Neurofibromas, benign tumors that originate from the peripheral nerve sheath, are a hallmark of neurofibromatosis type 1 (NF1). Although loss of heterozygosity (LOH) is a common phenomenon in this neoplasia, it only accounts for part of the somatic NF1 mutations found. Somatic point mutations or the presence of "two hits" in the NF1 gene have only been reported for a few neurofibromas. The large size of the NF1 gene together with the multicellular composition of these tumors has greatly hampered their molecular characterization. Here, we present the somatic NF1 mutational analysis of the whole set of neurofibromas studied by our group and consisting in 126 tumors derived from 32 NF1 patients. We report the identification of 45 independent somatic NF1 mutations, 20 of which are reported for the first time. Different types of point mutations together with LOH affecting the NF1 gene and its surrounding region or extending along the 17q arm have been found. Among point mutations, those affecting the correct splicing of the NF1 gene are common, coinciding with results reported on germline NF1 mutations. In most cases, we have been able to confirm that both copies of the NF1 gene are inactivated. We have also found that both somatic and germline mutations can be expressed at the RNA level in the neoplastic cells. Furthermore, we have observed that the study of more than one tumor derived from the same patient is useful for the identification of the germline mutation. Finally, we have noticed that the culture of neurofibromas and their fibroblast clearance facilitates LOH detection in cases in which it is difficult to determine.     

Pigment cell‐related manifestations in neurofibromatosis type 1: an overview

Neurofibromatosis type 1 (NF1) is an autosomal dominant neurocutaneous disorder, affecting approximately 1 in 3500 individuals. The most commonly seen tumors in NF1 patients are the (sub)cutaneous neurofibromas. However, individuals with NF1 typically present in childhood with well‐defined pigmentary defects, including café‐au‐lait macules (CALMs), intertriginous freckling and iris Lisch nodules. NF1 is considered a neurocristopathy, primarily affecting tissues derived from the neural crest. Since the pigment producing melanocyte originates in the neural crest, the presence of (hyper)pigmentary lesions in the NF1 phenotype because of changes in melanocyte cell growth and differentiation is to be expected. We want to discuss the pigmentary cutaneous manifestations of NF1 represented by CALMs and intertriginous freckles and the pigmentary non‐cutaneous manifestations represented by iris Lisch nodules. Several hypotheses have been suggested in explaining the poorly understood etiopathogenesis of CALMs. Whether other pigmentary manifestations might share similar etiopathogenic mechanisms remains obscure. Additional attention will be drawn to a readily seen phenomenon in NF1: hyperpigmentation overlying (plexiform) neurofibromas, which could suggest common etiopathogenetic‐environmental cues or mechanisms underlying CALMs and neurofibromas. Finally, we want to address the relationship between malignant melanoma and NF1.     

Characterization of a novel transgenic mouse line expressing Cre recombinase under the control of the Cdx2 neural specific enhancer

Several genetically modified mouse models have been generated in order to drive expression of the Cre recombinase in the neuroectoderm. However, none of them specifically targets the posterior neural plate during neurulation. To fill this gap, we have generated a new transgenic mouse line in which Cre expression is controlled by a neural specific enhancer (NSE) from the Caudal‐related homeobox 2 (Cdx2) locus. Analyses of Cre activity via breeding with R26R‐YFP reporter mice have indicated that the Cdx2NSE‐Cre mouse line allows for recombination of LoxP sites in most cells of the posterior neural plate as soon as from the head fold stage. Detailed examination of double‐transgenic embryos has revealed that this novel Cre‐driver line allows targeting the entire posterior neural tube with an anterior limit in the caudal hindbrain. Of note, the Cdx2NSE regulatory sequences direct Cre expression along the whole dorso‐ventral axis (including pre‐migratory neural crest cells) and, accordingly, YFP fluorescence has been also observed in multiple non‐cranial neural crest derivatives of double‐transgenic embryos. Therefore, we believe that the Cdx2NSE‐Cre mouse line represents an important novel genetic tool for the study of early events occurring in the caudal neuroectoderm during the formation of both the central and the peripheral nervous systems. genesis 51:777–784. © 2013 Wiley Periodicals, Inc.     

KiSS1 suppresses TNFα‐induced breast cancer cell invasion via an inhibition of RhoA‐Mediated NF‐κB activation

Tumor necrosis factor‐alpha (TNFα) induces cancer development and metastasis, which is prominently achieved by nuclear factor‐kappa B (NF‐κB) activation. TNFα‐induced NF‐κB activation enhances cellular mechanisms including proliferation, migration, and invasion. KiSS1, a key regulator of puberty, was initially discovered as a tumor metastasis suppressor. The expression of KiSS1 was lost or down‐regulated in different metastatic tumors. However, it is unclear whether KiSS1 regulates TNFα‐induced NF‐κB activation and further tumor cell migration. In this study, we demonstrate that KiSS1 suppresses the migration of breast cancer cells by inhibiting TNFα‐induced NF‐κB pathway and RhoA activation. Both KiSS1 overexpression and KP10 (kisspeptin‐10) stimulation inhibited TNFα‐induced NF‐κB activity, suppressed TNFα‐induced cell migration and cell attachment to fibronectin in breast cancer cells while KP10 has little effect on cancer cell proliferation. Furthermore, KP10 inhibited TNFα‐induced cell migration and RhoA GTPase activation. Therefore, our data demonstrate that KiSS1 inhibits TNFα‐induced NF‐κB activation via downregulation of RhoA activation and suppression of breast cancer cell migration and invasion. J. Cell. Biochem. 107: 1139–1149, 2009. © 2009 Wiley‐Liss, Inc.     

Progesterone and calcitriol attenuate inflammatory cytokines CXCL1 and CXCL2 in ovarian and endometrial cancer cells

Cytokines/chemokines are key players in cancer‐related inflammation. Increasing evidence suggests that chemokines produced by tumor cells are the mediators of metastasis. Thus, agents that can downregulate chemokines expression have potential against cancer metastasis. We have previously shown inhibition of ovarian and endometrial cancer cell growth with progesterone and calcitriol. In the present study, we evaluated the effect of these two agents on the expression of inflammatory genes. Using a RT‐PCR array of inflammatory cytokines/chemokines and their receptors, we found a marked attenuation of CXCL1 and CXCL2 (GRO‐α and ‐β) in cancer cells by both treatments. Knockdown of NFκB resulted in a reduced expression of CXCL1 and CXCL2 and the inhibitory effect of progesterone and calcitriol on the expression of chemokines was abrogated in NFκB‐silenced cancer cells. Silencing of IκBα increased the expression of CXCL1 and CXCL2 in cancer cells, which can be attributed to the increased activation of NFκB‐p65, caused by the lack of its inhibitor. Progesterone and calcitriol‐induced inhibition was abolished in IκBα‐knockdown cells. Our results demonstrate that suppression of IκBα phosphorylation by progesterone and calcitriol contributes to the reduced expression of CXCL1 and CXCL2. Downregulation of CXCL1 and CXCL2 was associated with a marked inhibition of metastasis‐promoting genes. Overall, our results indicate that progesterone and calcitriol inhibit IκBα phosphorylation, NFκB activation, and the expression of NFκB regulated metastasis promoting genes. These results provide attractive data for the possible use of progesterone and calcitriol in the management of endometrial and ovarian tumors. J. Cell. Biochem. 113: 3143–3152, 2012. © 2012 Wiley Periodicals, Inc.     

Inactivation of NF1 in CNS causes increased glial progenitor proliferation and optic glioma formation

The gene responsible for neurofibromatosis type 1 (NF1) encodes a tumor suppressor that functions as a negative regulator of the Ras proto-oncogene. Individuals with germline mutations in NF1 are predisposed to the development of benign and malignant tumors of the peripheral and central nervous system (CNS). Children with this disease suffer a high incidence of optic gliomas, a benign but potentially debilitating tumor of the optic nerve; and an increased incidence of malignant astrocytoma, reactive astrogliosis and intellectual deficits. In the present study, we have sought insight into the molecular and cellular basis of NF1-associated CNS pathologies. We show that mice genetically engineered to lack NF1 in CNS exhibit a variety of defects in glial cells. Primary among these is a developmental defect resulting in global reactive astrogliosis in the adult brain and increased proliferation of glial progenitor cells leading to enlarged optic nerves. As a consequence, all of the mutant optic nerves develop hyperplastic lesions, some of which progress to optic pathway gliomas. These data point to hyperproliferative glial progenitors as the source of the optic tumors and provide a genetic model for NF1-associated astrogliosis and optic glioma.     

Nf1-dependent tumors require a microenvironment containing Nf1+/-- and c-kit-dependent bone marrow

Interactions between tumorigenic cells and their surrounding microenvironment are critical for tumor progression yet remain incompletely understood. Germline mutations in the NF1 tumor suppressor gene cause neurofibromatosis type 1 (NF1), a common genetic disorder characterized by complex tumors called neurofibromas. Genetic studies indicate that biallelic loss of Nf1 is required in the tumorigenic cell of origin in the embryonic Schwann cell lineage. However, in the physiologic state, Schwann cell loss of heterozygosity is not sufficient for neurofibroma formation and Nf1 haploinsufficiency in at least one additional nonneoplastic lineage is required for tumor progression. Here, we establish that Nf1 heterozygosity of bone marrow-derived cells in the tumor microenvironment is sufficient to allow neurofibroma progression in the context of Schwann cell Nf1 deficiency. Further, genetic or pharmacologic attenuation of c-kit signaling in Nf1+/- hematopoietic cells diminishes neurofibroma initiation and progression. Finally, these studies implicate mast cells as critical mediators of tumor initiation.     

Regulation of mTOR complex 2 signaling in neurofibromatosis 2-deficient target cell types

Inactivating mutations in the neurofibromatosis 2 (NF2) tumor suppressor gene results in the development of schwannomas and meningiomas. Using NF2-deficient meningioma cells and tumors, together with the normal cellular counterparts that meningiomas derive, arachnoid cells, we identified merlin as a novel negative regulator of mTOR complex 1 (mTORC1). We now show that merlin positively regulates the kinase activity of mTORC2, a second functionally distinct mTOR complex, and that downstream phosphorylation of mTORC2 substrates, including Akt, is reduced upon acute merlin deficiency in cells. In response to general growth factor stimulation, Akt signaling is attenuated in merlin RNA interference-suppressed human arachnoid and Schwann cells by mechanisms mediated by hyperactive mTORC1 and impaired mTORC2. Moreover, Akt signaling is impaired differentially in a cell type-dependent manner in response to distinct growth factor stimuli. However, contrary to activation of mTORC1, the attenuated mTORC2 signaling profiles exhibited by normal arachnoid and Schwann cells in response to acute merlin loss were not consistently reflected in NF2-deficient meningiomas and schwannomas, suggesting additional genetic events may have been acquired in tumors after initial merlin loss. This finding contrasts with another benign tumor disorder, tuberous sclerosis complex, which exhibits attenuated mTORC2 signaling profiles in both cells and tumors. Finally, we examined rapamycin, as well as the mTOR kinase inhibitor, Torin1, targeting both mTOR complexes to identify the most efficacious class of compounds for blocking mTOR-mediated signaling and proliferation in merlin-deficient meningioma cells. These studies may ultimately aid in the development of suitable therapeutics for NF2-associated tumors.     

Comparative Oncogenomics Implicates the Neurofibromin 1 Gene (NF1) as a Breast Cancer Driver

Identifying genomic alterations driving breast cancer is complicated by tumor diversity and genetic heterogeneity. Relevant mouse models are powerful for untangling this problem because such heterogeneity can be controlled. Inbred Chaos3 mice exhibit high levels of genomic instability leading to mammary tumors that have tumor gene expression profiles closely resembling mature human mammary luminal cell signatures. We genomically characterized mammary adenocarcinomas from these mice to identify cancer-causing genomic events that overlap common alterations in human breast cancer. Chaos3 tumors underwent recurrent copy number alterations (CNAs), particularly deletion of the RAS inhibitor Neurofibromin 1 (Nf1) in nearly all cases. These overlap with human CNAs including NF1, which is deleted or mutated in 27.7% of all breast carcinomas. Chaos3 mammary tumor cells exhibit RAS hyperactivation and increased sensitivity to RAS pathway inhibitors. These results indicate that spontaneous NF1 loss can drive breast cancer. This should be informative for treatment of the significant fraction of patients whose tumors bear NF1 mutations.     

Specific and spatial labeling of P0‐Cre versus Wnt1‐Cre in cranial neural crest in early mouse embryos

P0‐Cre and Wnt1‐Cre mouse lines have been widely used in combination with loxP‐flanked mice to label and genetically modify neural crest (NC) cells and their derivatives. Wnt1‐Cre has been regarded as the gold standard and there have been concerns about the specificity of P0‐Cre because it is not clear about the timing and spatial distribution of the P0‐Cre transgene in labeling NC cells at early embryonic stages. We re‐visited P0‐Cre and Wnt1‐Cre models in the labeling of NC cells in early mouse embryos with a focus on cranial NC. We found that R26‐lacZ Cre reporter responded to Cre activity more reliably than CAAG‐lacZ Cre reporter during early embryogenesis. Cre immunosignals in P0‐Cre and reporter (lacZ and RFP ) activity in P0‐Cre/R26‐lacZ and P0‐Cre/R26‐RFP embryos was detected in the cranial NC and notochord regions in E8.0–9.5 (4–19 somites) embryos. P0‐Cre transgene expression was observed in migrating NC cells and was more extensive in the forebrain and hindbrain but not apparent in the midbrain. Differences in the Cre distribution patterns of P0‐Cre and Wnt1‐Cre were profound in the midbrain and hindbrain regions, that is, extensive in the midbrain of Wnt1‐Cre and in the hindbrain of P0‐Cre embryos. The difference between P0‐Cre and Wnt1‐Cre in labeling cranial NC may provide a better explanation of the differential distributions of their NC derivatives and of the phenotypes caused by Cre‐driven genetic modifications.     

The tumor suppressor neurofibromin confers sensitivity to apoptosis by Ras-dependent and Ras-independent pathways

Neurofibromatosis type 1 (NF1) is characterized by a high incidence of benign and malignant tumors attributed to loss of function of Nf1, which encodes neurofibromin, a tumor suppressor with Ras-GAP activity. Neurofibromin deficiency typically causes chronic activation of Ras, considered the major contributor to manifestation of NF1. Resistance to radio- and chemotherapy are typical of NF1-associated tumors, but the underlying mechanism is unknown. Here, we investigated interrelationships between neurofibromin expression, Ras activity, and sensitivity to apoptosis. Neurofibromin-deficient mouse embryonic fibroblasts (MEFs) and human NF1 tumor cells were more resistant than neurofibromin-expressing cells to apoptosis. Moreover, Nf1(-/-), Nf1(+/-), and Nf1(+/+) MEFs exhibited gene-dosage-related resistance to apoptosis. Resistance of the Nf1-deficient cells was mediated by two survival pathways: a Ras-dependent pathway, and a Ras-independent pathway promoted by the lack of an NF1-GRD-independent proapoptotic action of neurofibromin. Therefore, besides its Ras-dependent growth inhibition, neurofibromin can exert tumor suppression via a proapoptotic effect.     

Hypomorphic phenotype in mice with pituitary‐specific knockout of steroidogenic factor 1

Summary: The bacteriophage Cre recombinase provides a powerful approach for tissue‐specific gene inactivation. Using a Cre transgene driven by the common alpha subunit of glycoprotein hormones (αGSU‐Cre), we have previously inactivated steroidogenic factor 1 (SF‐1) in the anterior pituitary, causing hypogonadotropic hypogonadism with sexual infantilism, sterility, and severe gonadal hypoplasia. We now explore the molecular mechanisms underlying a hypomorphic gonadal phenotype in mice carrying two floxed SF‐1 alleles (F/F) relative to mice carrying one recombined and one floxed allele (F/R). Because their Cre‐mediated disruption of the locus encoding SF‐1 was less efficient, αGSU‐Cre, F/F mice retained some gonadotropin‐expressing cells in the anterior pituitary, thereby stimulating some gonadal function. This novel in vivo model for exploring the effects of differing levels of gonadotropins on gonadal development highlights the need for careful genotype‐phenotype comparisons in studies using Cre recombinase to produce tissue‐specific knockouts. genesis 30:65–69, 2001. © 2001 Wiley‐Liss, Inc.     

Multiple roles of NF1 in the melanocyte lineage

NF1 is a tumour suppressor gene, germline mutations of which lead to neurofibromatosis type 1 syndrome. Patients develop benign tumours from several types of cells including neural crest‐derived cells. NF1 somatic mutations also occur in 15% of sporadic melanoma, a cancer originating from melanocytes. Evidence now suggests the involvement of NF1 mutations in melanoma resistance to targeted therapies. Although NF1 is ubiquitously expressed, genetic links between NF1 and genes involved in melanocyte biology have been described, implying the lineage‐specific mechanisms. In this review, we summarize and discuss the latest advances related to the roles of NF1 in melanocyte biology and in cutaneous melanoma.     

Cell-nonautonomous induction of ovarian and uterine serous cystadenomas in mice lacking a functional Brca1 in ovarian granulosa cells

Women with germline mutations in BRCA1 have a 40% risk of developing ovarian cancer by age 70 and are also predisposed to cancers of the fallopian tubes. Given that ovulatory activity is a strong risk factor for sporadic ovarian cancer, we hypothesized that reduced BRCA1 expression might predispose to gynecological cancers indirectly, by influencing ovarian granulosa cells. These cells secrete sex steroids that control the ovulatory cycle and influence the growth of ovarian epithelial tumors. Granulosa cells also secrete mullerian inhibiting substance (MIS), a hormone that inhibits both the formation of female reproductive organs in male embryos and the proliferation of ovarian epithelial tumor cells. We tested this hypothesis by using the Cre-lox system to inactivate the Brca1 gene in mouse ovarian granulosa cells. A truncated form of the Fsh receptor promoter served as the Cre driver. Here, we show that indeed, inactivation of the Brca1 gene in granulosa cells led to the development of cystic tumors in the ovaries and uterine horns. These tumors carried normal Brca1 alleles, supporting the view that Brca1 may influence tumor development indirectly, possibly through an effector secreted by granulosa cells.