Oncogene (2009) 28, 1594–1604 & 2009 Macmillan Publishers Limited All rights reserved 0950-9232/09 $32.00 www.nature.com/onc ORIGINAL ARTICLE The role of the Birt–Hogg–Dube´ in mTOR activation and renal tumorigenesis

TR Hartman1, E Nicolas2, A Klein-Szanto3, T Al-Saleem3, TP Cash4, MC Simon4,5,6 and EP Henske1,7

1Department of Medical Oncology, Fox Chase Cancer Center, Philadelphia, PA, USA; 2Genomics Core Facility, Fox Chase Cancer Center, Philadelphia, PA, USA; 3Department of Pathology, Fox Chase Cancer Center, Philadelphia, PA, USA; 4Abramson Family Cancer Research Institute, University of Pennsylvania, Philadelphia, PA, USA; 5Howard Hughes Medical Institute, University of Pennsylvania, Philadelphia, PA, USA; 6Department of Cell and Developmental Biology, University of Pennsylvania, Philadelphia, PA, USA and 7Division of Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA

Birt–Hogg–Dube´ (BHD) syndrome is a tumor-suppressor Introduction disorder characterized by tumors, cystic lung disease and . Very little is known Birt–Hogg–Dube´ (BHD) syndrome is an autosomal about the molecular pathogenesis of BHD. Clinical dominant disorder characterized by hamartomas of skin similarities between BHD and complex follicles, spontaneous and renal cell (TSC) suggest that the BHD and TSC may carcinoma (RCC) (Birt et al., 1977; Toro et al., 1999; function within a common pathway. The TSC proteins Zbar et al., 2002). Unlike most other genetic disorders inhibit the activity of the mammalian target of rapamycin associated with renal tumors, BHD patients develop complex 1 (TORC1), and in Schizosaccharomyces pombe, multiple histological tumor types, including oncocyto- Bhd and Tsc1/Tsc2 have opposing roles in the regulation mas (which are considered benign) and chromophobe, of amino-acid homeostasis. We report here that in clear cell and papillary carcinomas (Pavlovich et al., mammalian cells, downregulation of BHD reduces the 2002, 2005). The BHD gene was cloned in 2002 of ribosomal protein S6, an indicator of (Nickerson et al., 2002). Nearly all reported human TORC1 activity. To determine whether folliculin, the germ-line BHD are predicted to result in product of the BHD gene, regulates mammalian target of premature protein truncation (Khoo et al., 2002; rapamycin activity in vivo, we generated a mouse with Nickerson et al., 2002; Painter et al., 2005; Schmidt targeted inactivation of the Bhd gene. The mice developed et al., 2005; Vocke et al., 2005). Inactivating mutations spontaneous oncocytic cysts and tumors composed of cells of the remaining allele have been identified in renal that resemble the renal cell carcinomas in BHD patients. carcinomas from BHD patients, indicating that BHD is The cysts and tumors had low levels of phospho-S6. Taken a tumor-suppressor gene (Vocke et al., 2005). together, these data indicate that folliculin regulates the BHD encodes a 64-kDa protein, folliculin (FLCN), activity of TORC1, and suggest a new paradigm in which which has no significant homology to other human both inappropriately high and inappropriately low levels of proteins. The function of folliculin is not completely TORC1 activity can be associated with renal tumori- understood. Two folliculin interacting proteins have genesis. been reported, FLCN-interacting protein 1 (FNIP1) and Oncogene (2009) 28, 1594–1604; doi:10.1038/onc.2009.14; its homolog FLCN-interacting protein 2 (FNIP2/ published online 23 February 2009 FNIPL) (Baba et al., 2006; Hasumi et al., 2008; Takagi et al., 2008). Both FNIP1 and FNIP2 interact with Keywords: BHD; mTOR; folliculin; renal carcinoma; AMP-activated protein kinase (AMPK), suggesting that oncocytoma folliculin, FNIP1 and FNIP2 may be involved in energy or nutrient sensing through the AMPK and mammalian target of rapamycin (mTOR) signaling pathways (Baba et al., 2006; Hasumi et al., 2008; Takagi et al., 2008). The clinical hallmarks of BHD, including facial hamartomas (folliculomas), lung cysts, pneumothorax and renal tumors, are similar to certain manifestations of tuberous sclerosis complex (TSC). TSC is a tumor- Correspondence: Dr EP Henske, Division of Pulmonary and Critical suppressor gene syndrome caused by mutations in the Care Medicine, Brigham and Women’s Hospital, Harvard Medical TSC1 or TSC2 gene (European 16 School, One Blackfan Circle, Boston, MA 02115, USA. E-mail: [email protected] Tuberous Sclerosis Consortium, 1993; van Slegtenhorst Received 9 October 2008; revised 10 December 2008; accepted 26 et al., 1997; O’Callaghan et al., 1998). TSC patients can December 2008; published online 23 February 2009 develop facial hamartomas (angiofibromas), renal BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1595 angiomyolipomas and less frequently oncocytomas and expression for further experiments. The two cell lines renal carcinomas, including clear cell and chromophobe with the highest levels of BHD expression were both subtypes (Barbour and Casali, 1978; Henske et al., 1995, glioblastoma derived: SF-268 and U251 (Figure 1b), 1996; Crino et al., 2006). Women with TSC can develop with levels 4- to 6-fold higher than a kidney-derived cell lymphangioleiomyomatosis, which is associated with line, CAKI-1. cystic lung disease and pneumothorax (Smolarek et al., 1998; Urban et al., 1999; Carsillo et al., 2000). BHD downregulation inhibits the phosphorylation of The clinical similarities between TSC and BHD ribosomal protein S6 suggest that folliculin and the TSC proteins may As noted earlier, in S. pombe the BHD homolog function within a common cellular pathway. The functions in a pathway that opposes the evolutionarily TSC1 and TSC2 proteins, hamartin and tuberin, conserved Tsc1/Tsc2/Rheb/Tor2 pathway. To determine respectively, heterodimerize and inhibit mTOR by the whether folliculin functions in a pathway opposing that small GTPase Rheb, which is the target of tuberin’s of TSC in mammalian cells, we treated U251 cells with highly conserved GTPase-activating domain (Castro TSC2, BHD or control small-interfering RNA (siRNA), et al., 2003; Garami et al., 2003; Inoki et al., 2003; and used levels of phospho-ribosomal protein S6(Ser Saucedo et al., 2003; Stocker et al., 2003; Zhang et al., 235/236) to monitor the levels of mTOR activity. 2003; Tee et al., 2003b). Expression profiling of Downregulation of TSC2 enhanced phospho-S6levels Schizosaccharomyces pombe yeast containing deletions relative to control siRNA, as expected based on the of either the Tsc or Bhd revealed that multiple TSC/Rheb/TORC1 signaling pathway, whereas down- amino-acid permeases and transporters that are down- regulation of BHD decreased phospho-S6levels in regulated in Dtsc1 and Dtsc2 are upregulated in Dbhd, serum starvation conditions (Figure 2a). To confirm and that Dbhd S. pombe have elevated intracellular levels this in a different cell type, we downregulated BHD in of specific amino acids that are low in Dtsc1 and Dtsc2 human embryonic kidney (HEK) 293 cells. Down- (van Slegtenhorst et al., 2007). These data indicate that regulation of BHD again resulted in a decrease in in yeast Bhd and Tsc1/Tsc2 function in common phospho-S6(Figure 2b). These data suggest that in pathways, but surprisingly with opposing roles. mammalian cells, folliculin and the TSC proteins have We report here that downregulation of BHD in opposing roles in target of rapamycin complex 1 mammalian cells leads to mTOR inhibition. Moreover, (TORC1) regulation, similar to S. pombe. Finally, to mice carrying a Bhd develop oncocytic renal determine whether folliculin regulates mTOR activation tumors resembling those found in BHD patients, and in a human kidney epithelial cell line, we downregulated the tumors have low levels of phosphorylation of BHD in HK-2 cells. As in U251 and HEK293 cells, ribosomal protein S6, an indicator of low mTOR folliculin downregulation reduced phospho-S6levels in activity. These data suggest that both inappro- serum starvation conditions (Figure 2c). Folliculin priate mTOR inhibition in BHD and inappropriate downregulation also decreased phospho-S6in HEK293 mTOR activation in TSC are associated with renal and U251 cells grown in full serum. Interestingly, tumorigenesis. levels of endogenous folliculin appeared to decrease after 15 min of serum stimulation in HK-2 cells (Figure 2c compare lanes 3 and 5). To determine if this was cell type specific, we stimulated U251 cells Results with serum and again observed lower endogenous folliculin levels after serum stimulation (Figure 2d BHD mRNA is highly expressed in testis, ovary, brain compare lanes 4 and 6). and lung, and in glioblastoma-derived cell lines To define the expression pattern of BHD in human tissues, we used quantitative real-time RT– Development of mice with targeted disruption of Bhd PCR (qRT–PCR) to compare BHD mRNA levels in These results in cell culture systems suggested that cells 20 different normal human tissues, using RNA speci- with folliculin downregulation have decreased levels of mens pooled from at least 3 donors (Ambion, Austin, TORC1 activity. To determine whether this is also true TX, USA). The highest levels of BHD transcript were in vivo, we generated mice with targeted inactivation of found in testis, ovary, brain and lung (Figure 1a). The Bhd using an embryonic stem (ES) cell line from level in the kidney was approximately 45% of that in the BayGenomics. This ES cell line was created using a testis, and the liver (which had the lowest level of BHD gene trap vector technique and contains a bgeo (b- transcript) was approximately 25% of the testis. Our galactosidase/neomycin) cassette integrated between qRT–PCR results of the tissue-specific expression of 8 and 9 in the Bhd gene, resulting in a truncated BHD mRNA vary somewhat from a recently published folliculin protein (Figure 3a). Disruption of Bhd mRNA study by Hasumi et al. (2008) that found the highest by the gene trap vector in the ES cells was confirmed by levels of expression in the pituitary gland, cervix and reverse (RT)–PCR using an upstream pancreas. These differences are possibly due to different primer in Bhd 7 and a downstream primer in the sources of tissue mRNA. We next tested 19 human b-gal sequence of the gene trap. The Bhd mutant cancer cell lines derived from ovary, brain and lung, to RT–PCR band (366 bp) can be distinguished from identify those with high levels of endogenous BHD wild-type Bhd using a downstream primer in exon 9,

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1596

Figure 1 Endogenous BHD mRNA levels in tissues and cell lines. (a) Quantitative real-time RT–PCR was used to identify tissues with high levels of BHD mRNA. A panel of normal human tissue RNA (Ambion) was utilized, with each sample being a pool from at least three donors. (b) RNA was prepared from 20 human cancer cell lines derived from kidney, ovary, brain and lung. Quantitative real- time RT–PCR was used to detect BHD mRNA levels. Relative mRNA levels are normalized to a renal cell line, CAKI-1.

which results in a 250 bp RT–PCR product (Figures 3b (Figure 4c) were composed of oncocytic cells with and c). When Bhd heterozygotes are mated, litters foamy-appearing , perinuclear clearing and consist of a ratio of B33% wild-type and B67% of Bhd uniform-appearing nuclei, resembling oncocytoma cells. heterozygous pups. No Bhd homozygous mutant pups Twenty-four Bhd þ /À mice were killed between 9 and were identified. To confirm reduced levels of Bhd 17 months of age. No macroscopic abnormalities were mRNA in the heterozygous mice compared to wild-type observed in any organ except the kidneys, which had mice, we isolated RNA from the kidneys of age-matched small but visible cysts (Figure 4a). Microscopically, both Bhd þ /À and Bhd þ / þ mice and Bhd levels were measured cysts and tumors were observed in the kidneys. Eleven using real-time RT–PCR, targeting the junction of exons Bhd þ /À mice (45%) had cysts (range 1–11 cysts per 9 and 10. As expected, Bhd þ /À mice had about 50% as mouse; Table 1), with an average diameter of 0.8 mm much Bhd mRNA as wild-type mice (0.52±0.13 relative (range 0.1–2.1 mm). The cysts had oncocytic cells similar mRNA levels in the Bhd þ /À mice compared to to those in the younger mice. One 13-month-old mouse 0.99±0.20 in the Bhd þ / þ mice). had multiple oncocytic cysts and two solid renal tumors The Bhd heterozygous mice had no obvious physical (0.5 and 1.5 mm), both in the same kidney, the larger of or behavioral abnormalities relative to their wild-type which was associated with a cyst (Figure 4d). Both littermates. Seven Bhd þ /À mice were killed between 3 tumors in this mouse were composed of nests of cells and 6months of age. No macroscopic abnormalities with foamy cytoplasm and perinuclear clearing, similar were observed in any organ. Microscopic examination to the oncocytoma-like cells observed in the 5-month- of the kidneys revealed cysts and solid tumors. Four of old mouse, but with more pleomorphism and nuclear these mice (57%) had a single renal cyst, with an average irregularity, resembling a chromophobe renal carcinoma diameter of 0.4 mm (range 0.1–0.6mm), and one, which (Figure 4e). A small oncocytic tumor (0.4 mm) asso- was 5 months old, had a 0.3 mm solid tumor (Table 1). ciated with an oncocytic cyst with papillary projections Both the solid tumor (Figure 4b) and the cysts was also observed in a 16-month-old Bhd þ /À mouse,

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1597

Figure 2 Birt–Hogg–Dube´ (BHD) downregulation causes a reduction in phospho-S6levels. ( a) U251 cells were treated with TSC2 siRNA, BHD siRNA or control siRNA in 10% fetal bovine serum (FBS) for 24 h, then serum starved for 24 h (serum À) or grown in 10% FBS for 24 h (serum þ ). Blot was probed with anti-TSC2 (tuberin), anti-folliculin mouse mAb, anti-phospho-S6(Ser 235/236) and anti-b-actin for a loading control. (b) HEK293 cells were treated with indicated siRNA in 10% FBS for 48 h. Cells were either lysed (serum þ ) or serum starved for 24 additional hours (serum À). Both long and short exposures are shown for phospho-S6. (c) Human kidney proximal tubule epithelial cells (HK-2) were treated with indicated siRNA in 10% FBS for 24 h. Cells were either lysed (serum 24 h) or serum starved for 24 additional hours and lysed (serum À) or incubated for 15 min with 20% FBS (serum 15 m) to stimulate the mTOR pathway. Both long and short exposures are shown for folliculin and phospho-S6. (d) U251 cells were treated with BHD or control siRNA in 10% FBS for 24 h. Cells were either lysed (serum 24 h) or serum starved for 24 additional hours and lysed (serum À) or incubated for 15 min of 20% FBS (serum 15 m) to stimulate the mammalian target of Rapamycin (mTOR) pathway, as indicated. Blot is probed with anti-folliculin rabbit polyclonal and anti-b-actin. Similar results were seen in at least two independent experiments for each cell line shown.

Figure 3 Gene trap disruption of the Bhd gene in mouse embryonic stem (ES) cells. (a) ES cell line RRX115 was produced by BayGenomics. The bgeo cassette (green box) is inserted into intron 8 of the Bhd gene and causes a truncated folliculin-b-galactosidase/ neomycin fusion protein. (b) Primers were designed to confirm the presence of the bgeo cassette in processed Bhd mRNA. (c) RT–PCR confirmation of Bhd-bgeo spliced mRNA in the ES cells. Lane 1, water control (forward primer exon 7, reverse primer bgeo); lane 2, gene trap ES cell RNA (forward primer exon 7, reverse primer bgeo); lane 3, normal mouse fibroblast RNA (forward primer exon 7, reverse primer exon 9); lane 4, myc-BHD plasmid cDNA wild-type positive control (forward primer exon 7, reverse primer exon 9).

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1598 Table 1 Renal cysts and tumors in Bhd+/À and Bhd+/+ mice along with three other oncocytic cysts. As a control, þ / þ Age Age kidneys from 5 Bhd mice age 3–6months and 10 þ / þ 3–6 months 9–17 months Bhd mice age 9–17 months were analysed. No n ¼ 7 n ¼ 24 tumors were identified. One mouse, age 17 months, had a single 0.3 mm cyst that lacked the oncocytic +/À Bhd þ /À Mice with tumors 1 (14%) 2 (8%) features noted in the Bhd cysts (Table 1). Mice with cysts 4 (57%) 11 (45%) The lung, skin, heart, liver and spleen of eleven 9 to Average number of cysts per mouse 0.61.3 17-month-old Bhd þ /À mice were examined microscopi- Average cyst diameter, mm (range) 0.4 (0.1–0.6) 0.8 (0.1–2.1) cally. Five of eleven mice (age 13–17 months) showed diffuse lymphoproliferative disease in several organs. n ¼ 5 n ¼ 10 Three of these mice had abundant small cell lymphoma Wild type infiltrates in lungs, thymus and lymph nodes, whereas Mice with tumors 0 0 two mice had moderate to intense lymphoid hyperplasia Mice with cysts 0 1 (10%) in lymph nodes, lungs and kidneys. These latter cases Average number of cysts per mouse 0 0.1 Average cyst diameter, mm (range) 0 0.3 were interpreted as consistent with early lymphoma. One mouse (17 months) exhibited splenomegaly and

Figure 4 Bhd heterozygotes develop spontaneous oncocytic cysts and tumors. (a) A typical kidney with small visible cysts (arrows) from a 14-month-old Bhd þ /À mouse is shown next to a kidney from an age-matched wild-type mouse. (b) An oncocytic tumor and (c) an oncocytic cyst from a 5-month-old Bhd þ /À mouse,  40 magnification. (d and e) A large tumor associated with a cyst from a 13-month-old Bhd þ /À mouse,  10 (d) and  40 (e) magnification. Normal kidney tubules can be seen adjacent to altered tissue in (b–d). All panels are hematoxylin and eosin (H&E) stained.

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1599 microscopically showed mild lymphoid hyperplasia also (Robb et al., 2007). The spontaneous tumor from the in lymph nodes. One mouse (16months) had an 8 mm untreated 13-month-old Bhd þ /À mouse was completely primary lung adenocarcinoma, and another mouse negative for phospho-S6staining (Figures 6dand e). (10 months) had a 0.3 mm lung adenoma. All of these Adjacent normal tubules with moderate positivity again lesions are frequently seen in aging mice and may provided an internal positive control (Figure 6d). represent background incidence for this strain (Stutman, Phospho-S6staining in the tumor from the 3.5-month- 1975; Harvey et al., 1993; Mohr et al., 1996; Babbitt old ENU-treated Bhd þ /À mouse was almost entirely et al., 2000). For example, Babbitt et al. found a 34.7% negative (Figure 6f), although occasionally cells showed incidence of lymphoma in control C57BL/6at the time 1 þ positivity. Phospho-S6staining of wild-type mice of killing at 30 months of age. treated with ENU showed similar positive staining of a fraction of normal tubules (Figures 6g–i). ENU enhances renal cyst formation in Bhd þ /À mice Exposure of Tsc1 þ /À and Tsc2 þ /À mice and rats to the mutagen ENU (N-ethyl-N-nitrosourea) has been shown Discussion to accelerate renal tumor and cyst development (Hino et al., 1993; Kobayashi et al., 1999, 2001). To determine We report here that downregulation of folliculin leads to if exposure to ENU accelerates tumor formation in this lower levels of phospho-ribosomal protein S6both in Bhd model, we intraperitoneally injected pregnant cell culture and in mice with heterozygous inactivation Bhd þ /À mice with 50 mg/kg body weight ENU at of Bhd. This is both an expected and a surprising result. embryonic day E14. Progeny mice were killed at 3–5 We had previously found that in S. pombe, the Bhd and months of age. A roughly Mendelian ratio of Bhd Tsc homologs function in opposing pathways. Because genotypes was observed, with B33% wild-type pups, downregulation of the TSC proteins leads to Tor B67% heterozygous pups and no homozygous Bhd null activation in S. pombe as well as in mammalian cells, pups. Two of the ENU-treated wild-type mice developed our finding that downregulation of BHD leads to a single renal cyst, and neither had an oncocytic mTOR inhibition was consistent with the S. pombe appearance (Figure 5a). Sixteen of 17 (94%) of the results. However, the fact that BHD downregulation ENU-treated Bhd þ /À mice developed between 1 and 14 leads to TORC1 inhibition in vitro and in mice carrying cysts (average 5.5 cysts per mouse) ranging from 0.1 to targeted inactivation of Bhd is surprising for multiple 1.9 mm in diameter (Table 2). The cysts in the ENU- reasons. First, because TSC and BHD patients develop treated Bhd þ /À mice were lined by oncocytic cells with hamartomatous skin tumors, lung cysts and renal uniform, round nuclei (Figures 5b–e). tumors, one might expect the TSC proteins and Two of the ENU-treated Bhd þ /À mice developed solid folliculin to have similar effects on mTOR activity. renal tumors. One mouse had two small oncocytic Second, Baba et al. (2006) found that BHD patient- tumors (0.2 and 0.4 mm in diameter) at age 3 months, derived UOK257 cells lacking folliculin had higher levels one of which was associated with a small oncocytic cyst of mTOR activation than UOK257 cells reexpressing (Figure 5f). This mouse had eight other oncocytic cysts folliculin. Third, two groups have reported that mice elsewhere in the kidneys. The other mouse, which was with targeted inactivation of Bhd in the kidney 3.5 months old, had a 0.5 mm tumor (Figure 5g) that (by a cadherin 16(KSP)-Cre transgene) develop resembled the spontaneous chromophobe-like tumors in massively cystic kidneys with increased phosphorylation the 13-month-old untreated mouse (Figures 4d and e). of ribosomal protein S6(Baba et al., 2008; Chen et al., The 3.5-month-old ENU-treated mouse also had five 2008), and that renal carcinomas from BHD patients oncocytic cysts (Figures 5h and i). Interestingly, a 14- have weak to moderate staining of phospho-mTOR month-old female that was injected with ENU during (Baba et al., 2008). pregnancy at 8 months of age had one enlarged kidney However, alongside these data suggesting that loss of (18 Â 13 Â 12 mm) containing a 0.4 mm oncocytic tumor BHD results in mTOR activation, there are other and four oncocytic cysts (range 0.1–12 mm), and a indications that the relationship between folliculin and second normal-sized kidney (12 Â 5 Â 4 mm) with four mTOR is complex. For example, Baba et al. found that oncocytic cysts and a 0.1 mm oncocytic tumor. No cells lacking BHD had lower levels of mTOR activation ENU-treated wild-type littermates developed solid renal than cells reexpressing BHD in a specific cellular tumors. condition (deprivation of amino acids), and Takagi et al. (2008) recently reported that in HeLa cells, siRNA Bhd þ /À tumors have low phospho-S6 levels downregulation of BHD causes a reduction in phospho- To determine whether the mTOR pathway is dysregu- S6kinase (Thr 389), although the serum conditions for lated in renal cysts and tumors in Bhd þ /À mice, we these experiments were not specified. stained paraffin-embedded tissue specimens from three The reasons for these different effects of folliculin on separate mice with cysts and tumors with anti-phospho- mTOR are not yet clear. One possibility is that the S6ribosomal protein (Ser 235/236).The oncocytic cells earliest impact of BHD loss results in mTOR inhibition, lining the cysts were uniformly negative for phospho-S6 consistent with the S. pombe data, and that subsequent (Figures 6a–c). Some normal tubules showed moderate events associated with tumor progression activate positivity, providing an internal control for the antibody mTOR. Consistent with this notion, Baba et al. (2008)

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1600

Figure 5 ENU (N-ethyl-N-nitrosourea) accelerates development of oncocytic cysts and tumors in Birt–Hogg–Dube´ (BHD) heterozygotes. Pregnant Bhd þ /À mice were injected with ENU at embryonic day E14. Both wild-type and Bhd þ /À littermates were killed between 3 and 5 months of age. (a) An example of a non-oncocytic cyst that developed in an ENU-injected wild-type mouse. (b–e) Examples of oncocytic cysts from four different ENU-treated Bhd þ /À mice. (f) A small oncocytic tumor associated with a small cyst from an ENU-treated Bhd þ /Àmouse. (g) An oncocytic tumor from an ENU-treated Bhd þ /À mouse. (h–i) Oncocytic cysts from the same mouse shown in (g). All images are hematoxylin and eosin (H&E) stained and at  40 magnification.

Table 2 Renal cysts and tumors in ENU-treated mice activation (Manning et al., 2002; Tee et al., 2003a; Wild type Bhd+/À Roux et al., 2004; Ballif et al., 2005; Ma et al., 2005; n ¼ 6 n ¼ 17 Rolfe et al., 2005; Cai et al., 2006). Therefore, it is possible that in vivo, mTOR activation occurs not as a Mice with tumors 0 (0%) 2 (12%) primary consequence of BHD mutations, but due to Mice with cysts 2 (33%) 16(94%) a Average number of cysts per mouse 0.3 5.5b activation of kinases upstream of the TSC/Rheb/mTOR Average cyst diameter, mm (range) 0.7 (0.4–0.9) 0.5 (0.1–1.9) pathway. This could explain the low levels of phospho- S6in the cysts and tumors in our mouse model, all aThe number of ENU-treated Bhd+/À mice with cysts compared to of which were relatively small and likely represent ENU-treated wild-type mice with cysts was statistically significant early lesions. (Po0.05), two-tailed Fisher’s exact test. It is also possible that the observed differences in bThe number of cysts per mouse in ENU-treated Bhd+/À mice compared to both ENU-treated wild-type mice, and untreated mTOR activation reflect differences between loss of Bhd+/À mice 3–6months of age, was statistically significant BHD expression in siRNA-treated cells and mutational (Po0.005), likelihood ratio test. inactivation of BHD in UOK257 cells, which have an insertion at 1733, leading to a (Yang et al., 2008). Interestingly, nearly all found that multiple kinase cascades, including Raf/ identified BHD mutations are predicted to truncate MEK/Erk, p90Rsk and Akt, were hyperactive in the prematurely the folliculin protein (Khoo et al., 2002; cystic kidneys of mice with KSP-Cre-driven inactivation Nickerson et al., 2002; Schmidt et al., 2005; Vocke et al., of Bhd. ERK, Rsk1 and Akt all directly phosphorylate 2005; Leter et al., 2008). Most identified germ-line and inactivate tuberin, thereby leading to mTOR mutations are either frameshift or nonsense mutations,

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1601

Figure 6 Oncocytic cysts and tumors from Bhd þ /À mice have no or weak phospho-S6levels. ( a) Oncocytic cyst lining epithelial cells (white arrow) from a 13-month-old untreated Bhd þ /À mouse were negative for anti-phospho-S6ribosomal protein (Ser 235/236) staining. (b) Oncocytic cyst lining cells (white arrow) from a 3.5-month-old ENU-treated Bhd þ /À mouse were negative for phospho-S6 staining. (c) Oncocytic cyst lining cells (white arrow) from a second 3.5-month-old ENU-treated Bhd þ /À mouse were negative for phospho-S6staining. ( d and e) Oncocytic tumor cells (black asterisks) from a 13-month-old untreated Bhd þ /À mouse were negative for phospho-S6staining. ( f) Oncocytic tumor cells (black asterisk) from an ENU-treated 3.5-month-old Bhd þ /À mouse were primarily negative for phospho-S6staining, with a few weakly positive (1 þ ) staining cells. (g and h) Normal kidney tissue from N-ethyl-N- nitrosourea (ENU)-treated Bhd þ / þ mice stained with anti-phospho-S6ribosomal protein (Ser 235/236).( i) Normal kidney tissue from an ENU-treated Bhd þ / þ mouse using phosphate-buffered saline in place of phospho-S6primary antibody. In panels ( a–d) black arrowheads indicate moderately (2 þ ,3þ ) phospho-S6positive normal tubules adjacent to the cysts and tumors in Bhd þ /À mice. Slides were counterstained with hematoxylin. All images shown at  40 magnification.

and roughly 44% occur in a ‘hot spot’ for BHD adenomas and carcinomas, both cystic and solid in frameshift mutations in a poly-C tract within exon 11 nature (Hino et al., 2001; Okimoto et al., 2004a, b; (Schmidt et al., 2005). It is possible that truncation of Kouchi et al., 2006), and the previously discussed mice folliculin removes a C-terminal region involved in with Cre-driven inactivation of Bhd in the kidney, which mTOR inhibition or leads to a dominant negative effect, die of renal failure by 3 weeks of age, some of which whereas loss of the entire protein removes an N-terminal develop cystic RCCs (Baba et al., 2008; Chen et al., region required for mTOR activation. Finally, it is 2008). The solid tumors in our Bhd þ /À mice were possible that folliculin regulates mTOR in a cell type- or composed of oncocytic cells with variable nuclear context-specific manner that is not fully recapitulated by pleomorphism, reminiscent of the spectrum of tumors either the cell culture experiments or the targeted found in BHD patients, which include chromophobe inactivation in vivo. However, we found that BHD renal tumors and oncocytic hybrid tumors (Pavlovich knockdown resulted in a reduction of phospho-S6(Ser et al., 2002, 2005; Murakami et al., 2007). In addition to 235/236) in three cell types: U251, HEK293 and HK-2. providing a model of BHD renal carcinogenesis, this is The mice reported here are the first Bhd mouse model also, to our knowledge, the first mouse model of that develops solid renal tumors. Previous animal oncocytic and chromophobe renal tumors. Although models of BHD include the German Shepherd dog, BHD mutations are not frequent in sporadic chromo- which develops hereditary multifocal renal cystadeno- phobe renal carcinomas (Nagy et al., 2004), Warren carcinoma and nodular dermatofibrosis (Lingaas et al., et al. (2004) observed weak to absent BHD mRNA 2003), the Nihon rat, which develops renal cell levels in renal oncocytomas, chromophobe renal cancer

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1602 and oncocytic hybrid tumors, which is in contrast to the (in arbitrary units) using standard curve (five points, fivefold high BHD mRNA levels observed in other cancers dilutions) established with a calibrator sample. Quantitative including breast, ovarian and prostate. Intriguingly, we real-time RT–PCR results were normalized to TATA-binding have observed weak or absent phospho-S6levels in protein mRNA levels. For each sample, the two values of sporadic oncocytomas and chromophobe RCCs, in relative quantity (from two PCR assays) were averaged. contrast to high levels of phospho-S6in the majority of sporadic clear cell renal carcinomas (Robb et al., Small-interfering RNA transfections 2007). To knockdown TSC2 and BHD mRNA levels, cells were In conclusion, we report here that loss of folliculin treated with 100 nmol TSC2, BHD or control siRNA leads to mTOR pathway inhibition both in vitro and (Dharmacon, Lafayette, CO, USA) for 24 or 48 h and either in vivo, consistent with data from S. pombe in which the serum starved, grown in 10% FBS or serum starved overnight Bhd and Tsc pathways have opposing roles in amino- followed by 15 min of 20% FBS to stimulate the mTOR acid homeostasis. We speculate that the most ‘ancient’ pathway. evolutionarily conserved function of folliculin is related to mTOR activation. In more complex organisms Western blotting folliculin may have additional context-specific regula- Cells were lysed in radioimmunoprecipitation assay buffer tory functions, highlighting the potential complexity of (50 mM Tris (pH 8), 150 mM NaCl, 2 mM PMSF, 0.1% SDS, mTOR regulation in cells carrying BHD mutations. 1% Triton-X and 1% deoxycholic acid), and 20 mg of total Understanding the function of folliculin in mTOR protein was loaded on 4–20% SDS–PAGE gels (Bio-Rad regulation may elucidate not only the pathogenesis of Laboratories, Hercules, CA, USA). Proteins were transferred BHD syndrome, but also the pathogenesis of sporadic onto Immobilon membranes (Millipore, Billerica, MA, USA) chromophobe and oncocytic-hybrid renal carcinomas for western blotting with the following antibodies: anti- and the normal homeostatic regulation of mTOR phospho-S6(Ser 235/236)(Cell Signaling Technology, Dan- activity. Finally, these studies may contribute to the vers, MA, USA), anti-tuberin (TSC2 (Abcam, Cambridge, MA, USA)), anti-FLCN (a rabbit polyclonal antibody development of targeted therapeutic strategies for BHD generated against a 20 amino-acid peptide comprising the patients, which are urgently needed. C terminus of folliculin: Ac-CYKSHLMSTVRSPTASESRN- OH), anti-FLCN (a mouse monoclonal antibody against folliculin (Baba et al., 2006), the generous gift of Dr Laura Schmidt, National Cancer Institute, Bethesda, MD, USA) and Materials and methods anti-b-actin (Cell Signaling Technology). Western blots were developed using horseradish peroxidase-conjugated secondary Cell culture and reagents antibodies and enhanced chemiluminescence (Amersham Human glioblastoma cells (U251) and HEK293 cells were Biosciences, Piscataway, NJ, USA). maintained in Dulbecco’s modified Eagle’s medium (DMEM) plus 10% fetal bovine serum (FBS). Immortalized human kidney proximal tubule epithelial (HK-2) cells (American Type Development of mice with targeted disruption of Bhd Culture Collection, Manassas, VA, USA) were maintained in DMEM plus 10% FBS and ITS liquid media supplement The ES cell line RRX115 from BayGenomics containing a (Sigma, St Louis, MO, USA). gene trap bgeo (b-galactosidase/neomycin) cassette that has integrated between exons 8 and 9 in the Bhd gene was used. The bgeo cassette contains a splice acceptor site upstream of Real-time quantitative reverse transcription-PCR analysis the bgeo sequences. During mRNA processing, this splice 50-Nuclease assays using TaqMan chemistry or SYBR Green acceptor site is recognized and results in fusion of the bgeo assays were run on a 7900 HT sequence detection system cassette with exon 8 of Bhd, creating a fusion protein and (Applied Biosystems, Foster City, CA, USA) using universal truncation of the folliculin protein. Disruption of Bhd mRNA PCR master mix or Power SYBR Green (Applied Biosystems). by the gene trap vector was confirmed in the ES cells FirstChoice Human Total RNA Survey Panel (Ambion) was by RT–PCR using an upstream primer in Bhd exon 7 used for qRT–PCR comparison of BHD mRNA levels in 20 (50-CCAGATGGAGAAGCTTGCTG-30), and a downstream human tissues. Each pool is composed of RNA from at least 3 primer in the b-gal sequence of the gene trap (50-CGATTAA tissue donors. For comparison of BHD mRNA levels in GTTGGGTAACGCC-30). The Bhd mutant RT–PCR band human cancer cell lines, mRNA was isolated using Trizol (366 bp) can be distinguished from wild-type Bhd using a (Invitrogen, Carlsbad, CA, USA). The sequences of the downstream primer in exon 9 (50-CATGCCAAGCCAACAT primers and probe for human BHD were: forward: ACGG-30), which results in a 250-bp RT–PCR product. The 50-CAAGGCGCTCAAGGTGTTT-30; reverse: 50-AATGGC ES cells were injected into blastocysts from C57BL/6mice GTGAAGGCTGTGT-30; probe: 6FAM-AGTTTGGATGC and implanted into pseudopregnant female C57BL/6mice. CCACAGCGTGCT-BHQ1. Chimeric pups were backcrossed with C57BL/6mice. Mice For comparison of Bhd mRNA levels in Bhd þ / þ and Bhd þ /À were genotyped using the following primers designed to mice, RNA was prepared from each kidney of two mice of recognize either wild-type or gene-trapped BHD DNA: each genotype (RNeasy kit; Qiagen, Valencia, CA, USA). wild-type and mutant forward primer in intron 8: 50-GTG Real-time RT–PCR was performed using primers amplifying TGGAGGTACATACATGTGTGCC-30; wild-type reverse over the junction of exons 9 and 10 (forward: 50-TGGTCCAT primer in intron 8: 50-AGTACCACCCCCCGTCAGTAA TCAGCGTTTGAA-30; reverse: 50-AGGGATAGGCACG TTCC-30 (533-bp PCR product); mutant reverse primer in GGAGG-30). For each sample, two reverse transcription bgeo: 50-CCTGGCCTCCAGACAAGTAGATCC-30 (460-bp reactions were performed with 100 and 20 ng of input RNA. PCR product). All mice were subsequently crossed into a Cycle threshold values were converted to quantities C57BL/6background.

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1603 For ENU experiments, Bhd heterozygous mice were crossed Statistical analysis in timed matings and E14 pregnant mice were injected with A single hematoxylin and eosin (H&E)-stained cross section 50 mg/kg ENU (Sigma) intraperitoneally. After birth, pups were through the largest dorsoventral section of each kidney was genotyped and analysed between 3 and 5 months of age. scored for renal cysts and tumors. A two-tailed Fisher’s exact All animal work was performed in accordance to protocols test was used for statistical analysis to compare the number of approved by the Institutional Animal Care and Use Commit- mice with one or more cysts between wild-type and Bhd þ /À tee of Fox Chase Cancer Center. mice treated with ENU, and to compare the number of mice with one or more cysts between untreated and ENU-treated þ /À Bhd mice. A likelihood ratio test was used to compare the Paraffin sections were deparaffinized, rehydrated, boiled in means of two Poisson distributions to determine the statistical 10 mM sodium citric buffer (pH 6.0; Sigma) and blocked with significance of the numbers of cysts per mouse between the þ /À þ /À 3% hydrogen peroxide in methanol. The sections were ENU-treated Bhd mice and the untreated Bhd mice of incubated with rabbit polyclonal antibody against phospho- similar age (o6months) and to compare the numbers of cysts þ /À S6(Ser 235/236)(Cell Signaling Technology) overnight at 4 1C, per mouse between the ENU-treated Bhd and ENU-treated developed using the Histostain-Plus kit (Invitrogen) and lightly wild-type mice. Significance was achieved at Pp0.05. counterstained with hematoxylin (Biomeda, Burlingame, CA, USA). For mouse tissues, EnVision þ System-HRP labeled Acknowledgements polymer (DakoCytomation, Carpinteria, CA, USA) was used in place of the rabbit secondary antibody and Histostain-Plus We thank Victoria Robb for critical reading of this paper. kit HRP-Streptavidin was used to reduce background staining We also thank Dr Laura Schmidt for the FLCN mAb, of the kidney. Negative control included substitution of Dr Samuel Litwin for statistical analysis and the Fox Chase phosphate-buffered saline for the primary antibody. Cysts Cancer Center Lab Animal Facility for technical assistance and tumors were scored for phospho-S6staining using a scale with the development and maintenance of the Bhd mutant of negative (À), weakly positive ( þ ), moderately positive mice. This work was supported by NIH RO1 (DK51052). (2 þ ) or strongly positive (3 þ ) (Robb et al., 2007). Dr Hartman was supported by NIH F32 (DK076443-01).

References

Baba M, Furihata M, Hong SB, Tessarollo L, Haines DC, Southon E Crino PB, Nathanson KL, Henske EP. (2006). The tuberous sclerosis et al. (2008). Kidney-targeted Birt–Hogg–Dube gene inactivation in complex. N Engl J Med 355: 1345–1356. a mouse model: Erk1/2 and Akt-mTOR activation, cell hyper- European Chromosome 16Tuberous Sclerosis Consortium (1993). proliferation, and polycystic kidneys. J Natl Cancer Inst 100: Identification and characterization of the tuberous sclerosis gene on 140–154. chromosome 16. Cell 75: 1305–1315. Baba M, Hong SB, Sharma N, Warren MB, Nickerson ML, Iwamatsu Garami A, Zwartkruis FJ, Nobukuni T, Joaquin M, Roccio M, Stocker A et al. (2006). Folliculin encoded by the BHD gene interacts with a H et al. (2003). Insulin activation of Rheb, a mediator of mTOR/ binding protein, FNIP1, and AMPK, and is involved in AMPK S6K/4E-BP signaling, is inhibited by TSC1 and 2. Mol Cell 11: and mTOR signaling. Proc Natl Acad Sci USA 103: 1457–1466. 15552–15557. Harvey M, McArthur MJ, Montgomery Jr CA, Bradley A, Babbitt JT, Kharazi AI, Taylor JM, Bonds CB, Mirell SG, Frumkin E Donehower LA. (1993). Genetic background alters the et al. (2000). Hematopoietic neoplasia in C57BL/6mice exposed to spectrum of tumors that develop in p53-deficient mice. FASEB J split-dose ionizing radiation and circularly polarized 60 Hz magnetic 7: 938–943. fields. Carcinogenesis 21: 1379–1389. Hasumi H, Baba M, Hong SB, Hasumi Y, Huang Y, Yao M et al. Ballif BA, Roux PP, Gerber SA, MacKeigan JP, Blenis J, Gygi SP. (2008). Identification and characterization of a novel folliculin- (2005). Quantitative phosphorylation profiling of the ERK/p90 interacting protein FNIP2. Gene 415: 60–67. ribosomal S6kinase-signaling cassette and its targets, the tuberous Henske EP, Neumann HP, Scheithauer BW, Herbst EW, Short MP, sclerosis tumor suppressors. Proc Natl Acad Sci USA 102: Kwiatkowski DJ. (1995). Loss of heterozygosity in the tuberous 667–672. sclerosis (TSC2) region of chromosome band 16p13 occurs in Barbour GL, Casali RE. (1978). Bilateral angiomyolipomas and renal sporadic as well as TSC-associated renal angiomyolipomas. Genes cell carcinoma in polycystic kidney. Urology 12: 694–698. Cancer 13: 295–298. Birt AR, Hogg GR, Dube WJ. (1977). Hereditary multiple fibrofolli- Henske EP, Scheithauer BW, Short MP, Wollmann R, Nahmias J, culomas with trichodiscomas and acrochordons. Arch Dermatol 113: Hornigold N et al. (1996). Allelic loss is frequent in tuberous 1674–1677. sclerosis kidney lesions but rare in brain lesions. Am J Hum Genet Cai SL, Tee AR, Short JD, Bergeron JM, Kim J, Shen J et al. (2006). 59: 400–406. Activity of TSC2 is inhibited by AKT-mediated phosphorylation Hino O, Mitani H, Knudson AG. (1993). Genetic predisposition to and membrane partitioning. J Cell Biol 173: 279–289. transplacentally induced renal cell carcinomas in the Eker rat. Carsillo T, Astrinidis A, Henske EP. (2000). Mutations in the tuberous Cancer Res 53: 5856–5858. sclerosis complex gene TSC2 are a cause of sporadic pulmo- Hino O, Okimoto K, Kouchi M, Sakurai J. (2001). A novel renal nary lymphangioleiomyomatosis. Proc Natl Acad Sci USA 97: 6085–6090. carcinoma predisposing gene of the Nihon rat maps on chromosome Castro AF, Rebhun JF, Clark GJ, Quilliam LA. (2003). Rheb binds 10. Jpn J Cancer Res 92: 1147–1149. tuberous sclerosis complex 2 (TSC2) and promotes S6kinase Inoki K, Li Y, Xu T, Guan KL. (2003). Rheb GTPase is a direct target activation in a rapamycin- and farnesylation-dependent manner. of TSC2 GAP activity and regulates mTOR signaling. Genes Dev 17: J Biol Chem 278: 32493–32496. 1829–1834. Chen J, Futami K, Petillo D, Peng J, Wang P, Knol J et al. (2008). Khoo SK, Giraud S, Kahnoski K, Chen J, Motorna O, Nickolov R Deficiency of FLCN in mouse kidney led to development of et al. (2002). Clinical and genetic studies of Birt–Hogg–Dube polycystic kidneys and renal neoplasia. PLoS ONE 3: e3581. syndrome. J Med Genet 39: 906–912.

Oncogene BHD, mTOR regulation and renal tumorigenesis TR Hartman et al 1604 Kobayashi T, Minowa O, Kuno J, Mitani H, Hino O, Noda T. (1999). phosphorylation of tuberous sclerosis complex 2 (TSC2). Biochem Renal carcinogenesis, hepatic hemangiomatosis, and embryonic J 388: 973–984. lethality caused by a germ-line Tsc2 mutation in mice. Cancer Res Roux PP, Ballif BA, Anjum R, Gygi SP, Blenis J. (2004). Tumor- 59: 1206–1211. promoting phorbol esters and activated Ras inactivate the tuberous Kobayashi T, Minowa O, Sugitani Y, Takai S, Mitani H, Kobayashi E sclerosis tumor suppressor complex via p90 ribosomal S6kinase. et al. (2001). A germ-line Tsc1 mutation causes tumor development Proc Natl Acad Sci USA 101: 13489–13494. and embryonic lethality that are similar, but not identical to, Saucedo LJ, Gao X, Chiarelli DA, Li L, Pan D, Edgar BA. (2003). those caused by Tsc2 mutation in mice. Proc Natl Acad Sci USA 98: Rheb promotes cell growth as a component of the insulin/TOR 8762–8767. signalling network. Nat Cell Biol 5: 566–571. Kouchi M, Okimoto K, Matsumoto I, Tanaka K, Yasuba M, Hino O. Schmidt LS, Nickerson ML, Warren MB, Glenn GM, Toro JR, (2006). Natural history of the Nihon (Bhd gene mutant) rat, a novel Merino MJ et al. (2005). Germline BHD-mutation spectrum and model for human Birt–Hogg–Dube syndrome. Virchows Arch 448: analysis of a large cohort of families with Birt–Hogg– 463–471. Dube syndrome. Am J Hum Genet 76: 1023–1033. Leter EM, Koopmans AK, Gille JJ, van Os TA, Vittoz GG, David EF Smolarek TA, Wessner LL, McCormack FX, Mylet JC, Menon AG, et al. (2008). Birt–Hogg–Dube syndrome: clinical and genetic studies Henske EP. (1998). Evidence that lymphangiomyomatosis is caused of 20 families. J Invest Dermatol 128: 45–49. by TSC2 mutations: chromosome 16p13 loss of heterozygosity in Lingaas F, Comstock KE, Kirkness EF, Sorensen A, Aarskaug T, angiomyolipomas and lymph nodes from women with lymphangio- Hitte C et al. (2003). A mutation in the canine BHD gene is myomatosis. Am J Hum Genet 62: 810–815. associated with hereditary multifocal renal cystadenocarcinoma and Stocker H, Radimerski T, Schindelholz B, Wittwer F, Belawat P, nodular dermatofibrosis in the German Shepherd dog. Hum Mol Daram P et al. (2003). Rheb is an essential regulator of S6K in Genet 12: 3043–3053. controlling cell growth in Drosophila. Nat Cell Biol 5: 559–565. Ma L, Chen Z, Erdjument-Bromage H, Tempst P, Pandolfi PP. (2005). Stutman O. (1975). Immunodepression and malignancy. Adv Cancer Phosphorylation and functional inactivation of TSC2 by Erk Res 22: 261–422. implications for tuberous sclerosis and cancer pathogenesis. Cell Takagi Y, Kobayashi T, Shiono M, Wang L, Piao X, Sun G et al. 121: 179–193. (2008). Interaction of folliculin (Birt–Hogg–Dube gene product) with Manning BD, Tee AR, Logsdon MN, Blenis J, Cantley LC. (2002). a novel Fnip1-like (FnipL/Fnip2) protein. Oncogene 27: 5339–5347. Identification of the tuberous sclerosis complex-2 tumor suppressor Tee AR, Anjum R, Blenis J. (2003a). Inactivation of the tuberous gene product tuberin as a target of the phosphoinositide 3-kinase/ sclerosis complex-1 and -2 gene products occurs by phosphoinositide akt pathway. Mol Cell 10: 151–162. 3-kinase/Akt-dependent and -independent phosphorylation of tu- Mohr U, Dungworth DL, Capen CC, Carlton WW, Sundberg JP, berin. J Biol Chem 278: 37288–37296. Ward JM. (1996). Pathobiology of the Aging Mouse, vol. 1. Tee AR, Manning BD, Roux PP, Cantley LC, Blenis J. (2003b). International Life Sciences Institute: Washington, DC, 527pp. Tuberous sclerosis complex gene products, Tuberin and Hamartin, Murakami T, Sano F, Huang Y, Komiya A, Baba M, Osada Y et al. control mTOR signaling by acting as a GTPase-activating protein (2007). Identification and characterization of Birt–Hogg–Dube complex toward Rheb. Curr Biol 13: 1259–1268. associated renal carcinoma. J Pathol 211: 524–531. Toro JR, Glenn G, Duray P, Darling T, Weirich G, Zbar B et al. Nagy A, Zoubakov D, Stupar Z, Kovacs G. (2004). Lack of mutation (1999). Birt–Hogg–Dube syndrome: a novel marker of kidney of the folliculin gene in sporadic chromophobe renal cell carcinoma neoplasia. Arch Dermatol 135: 1195–1202. and . Int J Cancer 109: 472–475. Urban T, Lazor R, Lacronique J, Murris M, Labrune S, Valeyre D Nickerson ML, Warren MB, Toro JR, Matrosova V, Glenn G, et al. (1999). Pulmonary lymphangioleiomyomatosis. A study of 69 Turner ML et al. (2002). Mutations in a novel gene lead to kidney patients. Groupe d’Etudes et de Recherche sur les Maladies tumors, lung wall defects, and benign tumors of the hair ‘Orphelines’ Pulmonaires (GERM‘O’P). Medicine (Baltimore) 78: follicle in patients with the Birt–Hogg–Dube syndrome. Cancer Cell 321–337. 2: 157–164. van Slegtenhorst M, de Hoogt R, Hermans C, Nellist M, Janssen B, O’Callaghan FJK, Shiell AW, Osborne JP, Martyn CN. (1998). Verhoef S et al. (1997). Identification of the tuberous sclerosis gene Prevalence of tuberous sclerosis estimated by capture-recapture TSC1 on chromosome 9q34. Science 277: 805–808. analysis. Lancet 351: 1490. van Slegtenhorst M, Khabibullin D, Hartman TR, Nicolas E, Kruger Okimoto K, Kouchi M, Matsumoto I, Sakurai J, Kobayashi T, Hino WD, Henske EP. (2007). The Birt–Hogg–Dube and tuberous O. (2004a). Natural history of the Nihon rat model of BHD. Curr sclerosis complex homologs have opposing roles in Mol Med 4: 887–893. homeostasis in Schizosaccharomyces pombe. J Biol Chem 282: Okimoto K, Sakurai J, Kobayashi T, Mitani H, Hirayama Y, 24583–24590. Nickerson ML et al. (2004b). A germ-line insertion in the Birt– Vocke CD, Yang Y, Pavlovich CP, Schmidt LS, Nickerson ML, Hogg–Dube (BHD) gene gives rise to the Nihon rat model of Torres-Cabala CA et al. (2005). High frequency of somatic inherited renal cancer. Proc Natl Acad Sci USA 101: 2023–2027. frameshift BHD gene mutations in Birt–Hogg–Dube-associated Painter JN, Tapanainen H, Somer M, Tukiainen P, Aittomaki K. renal tumors. J Natl Cancer Inst 97: 931–935. (2005). A 4-bp deletion in the Birt–Hogg–Dube gene (FLCN) causes Warren MB, Torres-Cabala CA, Turner ML, Merino MJ, Matrosova dominantly inherited spontaneous pneumothorax. Am J Hum Genet VY, Nickerson ML et al. (2004). Expression of Birt–Hogg–Dube 76: 522–527. gene mRNA in normal and neoplastic human tissues. Mod Pathol Pavlovich CP, Grubb III RL, Hurley K, Glenn GM, Toro J, 17: 998–1011. Schmidt LS et al. (2005). Evaluation and management of renal Yang Y, Padilla-Nash HM, Vira MA, Abu-Asab MS, Val D, Worrell tumors in the Birt–Hogg–Dube syndrome. J Urol 173: 1482–1486. R et al. (2008). The UOK 257 cell line: a novel model for studies of Pavlovich CP, Walther MM, Eyler RA, Hewitt SM, Zbar B, Linehan the human Birt–Hogg–Dube gene pathway. Cancer Genet Cytogenet WM et al. (2002). Renal tumors in the Birt–Hogg–Dube syndrome. 180: 100–109. Am J Surg Pathol 26: 1542–1552. Zbar B, Alvord WG, Glenn G, Turner M, Pavlovich CP, Schmidt L Robb VA, Karbowniczek M, Klein-Szanto AJ, Henske EP. (2007). et al. (2002). Risk of renal and colonic and spontaneous Activation of the mTOR signaling pathway in renal clear cell pneumothorax in the Birt–Hogg–Dube syndrome. Cancer Epidemiol carcinoma. J Urol 177: 346–352. Biomarkers Prev 11: 393–400. Rolfe M, McLeod LE, Pratt PF, Proud CG. (2005). Activation of Zhang Y, Gao X, Saucedo LJ, Ru B, Edgar BA, Pan D. (2003). Rheb protein synthesis in cardiomyocytes by the hypertrophic agent is a direct target of the tuberous sclerosis tumour suppressor phenylephrine requires the activation of ERK and involves proteins. Nat Cell Biol 5: 578–581.

Oncogene