You are here

Article Text

Article menu
Original article
Renal tumour suppressor function of the Birt–Hogg–Dubé syndrome gene product folliculin
  1. V Hudon1,
  2. S Sabourin1,
  3. A B Dydensborg1,
  4. V Kottis1,
  5. A Ghazi1,
  6. M Paquet2,
  7. K Crosby3,
  8. V Pomerleau1,
  9. N Uetani1,
  10. A Pause1
  1. 1Goodman Cancer Centre and Department of Biochemistry, McGill University, Montréal, Québec, Canada
  2. 2Comparative Medicine and Animal Resources Centre, McGill University, Montreal, Quebec, Canada
  3. 3Cell Signaling Technologies, Danvers, Massachusetts, USA
  1. Correspondence to Arnim Pause, Goodman Cancer Centre, McGill University, Room 707A, McIntyre Building, 3655 Sir William Osler Promenade, Montréal, Québec, Canada H3G1Y6; arnim.pause{at}mcgill.ca

Abstract

Background Renal cell carcinoma (RCC) comprises five major molecular and histological subtypes. The Birt–Hogg–Dubé (BHD) syndrome is a hereditary human cancer syndrome that predisposes affected individuals to develop renal carcinoma of nearly all subtypes, in addition to benign fibrofolliculomas, and pulmonary and renal cysts. BHD is caused by loss-of-function mutations in the folliculin (FLCN) protein. The molecular function of FLCN is still largely unknown; opposite and conflicting evidence of the role of FLCN in mammalian target of rapamycin signalling/phosphorylated ribosomal protein S6 (p-S6) activation had recently been reported.

Results and Methods Here, the expression pattern of murine Flcn was described, and it was observed that homozygous disruption of Flcn results in embryonic lethality early during development. Importantly, heterozygous animals manifest early preneoplastic kidney lesions, devoid of Flcn expression, that progress towards malignancy, including cystopapillary adenomas. A bona fide tumour suppressor activity of FLCN was confirmed by nude mouse xenograft assays of two human RCC cell lines with either diminished or re-expressed FLCN. It was observed that loss of FLCN expression leads to context-dependent effects on S6 activation. Indeed, solid tumours and normal kidneys show decreased p-S6 upon diminished FLCN expression. Conversely, p-S6 is found to be elevated or absent in FLCN-negative renal cysts.

Conclusion In accordance with clinical data showing distinct renal malignancies arising in BHD patients, in this study FLCN is shown as a general tumour suppressor in the kidney.

  • BHD syndrome
  • kidney cysts
  • renal adenocarcinoma
  • folliculin
  • mTORC1
  • molecular genetics
  • oncology
  • cancer: urological
  • cell biology
  • renal medicine
  • BHD
  • Birt–Hogg–Dubé
  • β-geo
  • β-galactosidase-neomycin-phosphotransferase II
  • E
  • embryonic day
  • FLCN
  • folliculin
  • LOH
  • loss of heterozygosity
  • mTOR
  • mammalian target of rapamycin
  • RCC
  • renal cell carcinoma
  • RCND
  • renal cysadenocarcinoma and nodular dermatofibrosis
  • SEM
  • standard error of the mean
  • TSC
  • tuberous sclerosis complex
  • VHL
  • von Hippel Lindau

http://dx.doi.org/10.1136/jmg.2009.072009

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Renal cell carcinoma (RCC) is a diverse cancer type that comprises five distinct histological subtypes: clear cell, papillary (type I and II), chromophobe, oncocytoma and collecting duct carcinoma.1 These subtypes of renal cancer all follow distinct clinical courses and involve disturbance of discrete biochemical pathways,2 such as hypoxia sensing (von Hippel Lindau (VHL) and Hypoxia Inducible Factor 1a (HIFα) in clear cell RCC), signalling (c-Met in papillary type I) and metabolism (fumarate hydratase in papillary type II).3 The majority of renal cancer incidences are sporadic; however, an estimated 1% to 4% of cases are hereditary in nature. Indeed, seven different autosomal-dominant inherited renal cancer syndromes have been described, one of which is the Birt–Hogg–Dubé (BHD) syndrome. BHD patients are predisposed to develop bilateral multifocal tumours of the kidneys, benign tumours of the hair follicle (fibrofolliculoma) as well as pulmonary and renal cysts.4–8 Single-allele germline frameshift and nonsense mutations in the FLCN gene are the cause of the BHD syndrome.9–11 FLCN encodes a 64 kDa protein, folliculin (FLCN), which presents no significant homology to any known protein and is highly conserved across species, suggesting an important biological role.11 Interestingly, BHD patients can develop renal tumours of several histological subtypes of RCC within one kidney (clear cell, chromophobe and oncocytoma),12 and somatic mutations or loss of heterozygosity of FLCN have been identified in sporadic RCC of the papillary, clear cell, chromophobe and oncocytoma subtypes.13 The association of loss of FLCN expression with nearly all histological subtypes of RCC suggests that FLCN is involved in molecular functions that are near universal in the pathogenesis of RCC. Somatic second-hit mutations and loss of heterozygosity identified in several sporadic and BHD-associated kidney tumours12 13 imply that both copies of the FLCN gene are inactivated for tumourigenesis in the kidneys, thus conforming to the Knudson “two-hit” tumour suppressor model.14

    Conditional homozygous inactivation of Flcn in the mouse kidney triggers the development of highly enlarged polycystic kidneys, which led to renal failure and resulted in early death 3 weeks after birth.15 16 However, the usefulness of these models is limited due to the early lethality. Recently, a heterozygous mutant Flcn mouse model was generated, and the mice developed cysts and tumours in the kidneys.17

    The molecular functions of FLCN are poorly understood; however, indirect interactions between FLCN and AMPK (5′ AMP-activated protein kinase) of the mTOR (mammalian target of rapamycin complex 1 (mTORC1)) signalling networks have been firmly established and are mediated by two novel proteins, FNIP1 and FNIP2 (folliculin-interacting proteins 1 and 2).18–20 AMPK is a negative regulator of mTORC1 and a key protein for energy sensing in cells.21–25 Thus, the interaction of FLCN with FNIP1/2 and AMPK suggests a role for FLCN in nutrient/energy-sensing through the AMPK/mTORC1 signalling pathways. The functional role of FLCN in mTORC1 signalling is, however, controversial since several recent publications have reported opposite impacts on phosphorylated ribosomal protein S6 (p-S6) signalling as a consequence of FLCN downregulation. Two studies recently reported that transient downregulation of FLCN by siRNA in human cell lines results in reduction of p-S6.17 20 Reduction of p-S6 was also observed in renal cysts developing in mice heterozygous for Flcn.17 In contrast, kidney-specific homozygous knockout of Flcn results in an increase of p-S6, which contributed to the development of polycystic kidneys.15 16 Furthermore, knockout of the FLCN homolog in Schizosaccharomyces pombe showed that FLCN, which is thought to act at the level of AMPK, has an opposite biochemical role of the AMPK target TSC2 in amino acid transport.26 Such opposite roles are surprising given the overlapping clinical characteristics (skin harmatomas, lung and kidney cysts as well as renal carcinomas) associated with BHD syndrome and tuberous sclerosis complex (TSC) disease, and further indicate that the function of FLCN in AMPK/mTORC1 signalling is complex.

    To clarify the functions of FLCN during tumour formation, we generated a mouse model of BHD. Our results demonstrate that homozygous disruption of the Flcn gene in mice is lethal before embryonic day (E) 8.5. We also extensively characterise the expression pattern of Flcn in adult tissues and cell types, in which it was widely expressed but differed in certain key tissues when compared to the human expression. In addition, heterozygous Flcn knockout mice developed renal cysts and tumours upon loss of Flcn expression, and we confirmed a tumour suppressor function of FLCN in two independent human RCC cell lines using a xenograft assay. Our data demonstrate that loss of Flcn is associated with an opposing impact on S6 activation, which may lead to formation of several molecular types of RCC. The data also show that FLCN is a bona fide renal tumour suppressor gene with a general function in renal oncogenesis.

    Results

    Generation and characterisation of a BHD mouse model

    We generated a transgenic mouse line carrying an in-frame β-galactosidase–neomycin–phosphotransferase II (β-geo) insertion between exons 8 and 9 of Flcn (Supplemental figure 1A). Such an insertion generates a mutant FLCN protein lacking the last 226 amino acids of the wild type FLCN protein and thus mimics the majority of the loss-of-function mutations found in BHD patients.27 Homozygous deletion of Flcn was observed to cause embryonic lethality before E8.5 (Supplemental figure 1B–C, Supplemental table 1). Semiquantitative reverse transcription-PCR (RT-PCR), and northern and western blots confirmed that heterozygous carriers of the disrupted allele had reduced expression levels of Flcn mRNA and protein (Supplemental figure 1D–F). To analyse the expression pattern of Flcn in adult tissues, we performed semiquantitative RT-PCR on mRNA from various organs as well as X-gal staining of tissue slides from Flcn+/− animals (Supplemental figure 2A–G). Interestingly, we did not observe any expression of Flcn in the skin of the mouse model in contrast with the frequent dermal manifestations of BHD syndrome in humans. Table 1 provides a detailed overview of the observed expression pattern.

    View this table:
    Table 1

    Expression of folliculin (Flcn) in adult mouse tissue

    Flcn+/− mice display preneoplastic kidney lesions and tumours

    BHD syndrome predisposes patients to develop cutaneous fibrofolliculoma and lung cysts, and increases their risk to develop renal cysts and neoplasia.7 8 11 To determine if the Flcn mouse model developed BHD-related kidney lesions, histological analysis was performed on renal tissue sections of Flcn+/− mice between 1 and 24 months of age (n=72) (figure 1A–D). Flcn+/+ littermates presented with normal kidneys (figure 1A, table 2), whereas Flcn+/− animals presented with sporadic renal tubule hyperplasia in young mice (Table 2), single cysts and multiocular polycystic kidneys (figure 1B). The incidence rate increased with age and ranged from 29% to 51% (table 2). The cystic lesions possessed oncocytic features such as abundant eosinophilic and granular cytoplasm and papillary projections (figure 1C). Rare cases of cystopapillary adenomas were observed (figure 1D, table 2). None of the Flcn+/− mice investigated (n=42) presented with lung cysts or any skin lesions (n=18) consistent with the disparities of the expression pattern in the lung compared to humans and the lack of specific endogenous Flcn expression in the murine skin. These data demonstrate that heterozygotic loss of Flcn in the mouse strongly predisposes to the formation of premalignant kidney lesions that has the capacity to progress to adenomas.

    Figure 1
    Figure 1

    Spectrum of kidney lesions arising in Flcn+/− mice. (A) Flcn+/+ and (B–D) Flcn+/− kidney paraffin sections were stained with H&E for histopathological analysis (10× magnification). (A′–D′) 40× magnification of boxed section in A–D. Kidney lesions observed in Flcn+/− mice included epithelial cell hyperplasia with cystic dilation of tubules (B), tubular hyperplasia cystic with papillary projections (C) and adenoma (D). Cy, cyst; Cy pap, cyst with papillary projections; ad, adenoma.

    View this table:
    Table 2

    Histopathological analysis of 38 Flcn+/+ and 72 Flcn+/− pairs of kidneys

    Renal cysts in Flcn+/− mice are derived from proximal tubules and lose Flcn expression

    To investigate the origin of the cysts, we stained sequential Flcn+/− kidney sections for the proximal tubule marker Lotus tetragonolobus agglutinin (LTA) 28 29 and for FLCN (figure 2A and B). Colocalisation of LTA (figure 2A) and FLCN (figure 2B) staining in normal tissue confirmed our previous observations that FLCN is expressed in the proximal tubules of the kidney (figure 2A and B, arrows). However, while cysts expressed the marker of proximal tubules LTA (figure 2C, arrowhead), FLCN expression was lost in the epithelial lining of the cysts and in the surrounding cyst-associated tissues (figure 2D). Loss of FLCN expression thus leads to formation of renal cysts that are derived from proximal tubules.

    Figure 2
    Figure 2

    Renal cysts lose folliculin (Flcn) expression and arise from proximal tubules. Flcn+/− kidney sections were stained and analysed for L tetragonolobus agglutinin (LTA; A and C) and FLCN (B and D). Arrows indicate colocalisation of FLCN and LTA, and arrowheads point to the expression of LTA in the cyst. Scale bar, 100 μm.

    Increase in proliferation contributes to renal cystogenesis

    To verify whether loss of Flcn induces abnormal cell proliferation and/or apoptosis during kidney cyst development, we assessed by immunohistochemical staining the expression of the cellular proliferation marker Ki-67 to evaluate any difference in proliferating cells between Flcn+/+ and Flcn+/− kidneys (figure 3A, C and E). Although no significant difference was observed between normal Flcn+/+ and Flcn+/− kidney tissues (figure 3A and C, respectively), staining for Ki-67 was increased in the epithelial cell lining of cysts and the surrounding tissue (figure 3E). Quantification of Ki67-positive nuclei in the epithelia lining the cysts of Flcn+/− animals compared to normal kidney tissues from Flcn+/+ and Flcn+/− animals revealed a dramatic increase in the cysts (p<0.05, Tukey one-way ANOVA) (figure 3G). Immunohistochemical staining for caspase-3, an apoptosis marker, demonstrated an absence of apoptotic cells within normal kidney tissues of Flcn+/+ and Flcn+/− animals (figure 3B and D, respectively), and was not a striking feature in cystic kidneys (figure 3F). Thus, increased proliferation is a prominent feature of the cysts and correlated with loss of Flcn expression.

    Figure 3
    Figure 3

    Cystic kidneys exhibit increase in proliferation. Kidney sections from Flcn+/+ (A and B) and Flcn+/− mice (C–F) were stained for the proliferation marker Ki-67 (A, C and E) as well as for the apoptotic marker caspase-3 (B, D and F) and counterstained with hematoxylin. Arrows indicate positive staining for both markers in cyst, and arrowheads indicate Ki-67 staining in the tissue. Scale bar, 100 μm. (G) Quantification of Ki-67-positive nuclei. Data represent the mean percentage of Ki-67-positive cells/total cells per field (SEM). *p<0.05, one-way ANOVA. FLCN, folliculin; LTA, L tetragonolobus agglutinin.

    FLCN acts as a general tumour suppressor in human RCC cells in vivo

    BHD patients develop RCC of several different histological subtypes, including the most abundant type: clear cell renal cell carcinoma (CC-RCC). These observations suggests that the function of FLCN in tumour biology is independent of additional molecular insults to the cells, such as in the case of CC-RCC loss of function of the von Hippel Lindau (VHL) tumour suppressor gene. To directly investigate the hypothesis that FLCN might exert kidney tumour-suppressive functions regardless of the molecular subtype of the tumour, we modified FLCN levels in two human RCC cell lines with different VHL status. 786-0 cells are classic models of CC-RCC and correspondingly are VHL negative, whereas ACHN cells are VHL positive RCC cells. We downregulated and overexpressed FLCN in the ACHN and 786-0 cells, respectively (figure 4A). Next, we verified the growth potential of these cells in vitro (figure 4B) since we observed an increase in proliferation in renal cysts in the mouse (figure 3G). Neither downregulation of FLCN in ACHN cells nor re-expression of FLCN in 786-0 cells had any effect on their growth potential in vitro (figure 4B). In addition, we did not observe any FLCN-dependent changes in levels of the VHL target gene HIF2α in these two cell lines, indicating that the two pathways might be independent (Supplemental figure S3). However, FLCN expression levels had a strong effect on tumour outgrowth of ACHN and 786-0 cells following subcutaneous injection into athymic nude mice. Indeed, knockdown of FLCN in ACHN cells led to the formation of significantly larger tumours (figure 4C, left panel; p<0.05, t test), while reintroduction of FLCN in 786-0 cells led to a strong decrease in tumour growth (figure 4C, right panel; p<0.01, t test). Using a reciprocal genetic engineering strategy in human RCC cells, these data demonstrate that FLCN indeed functions as a tumour suppressor in vivo; however, FLCN expression levels do not affect in vitro proliferation. These results are similar to previous studies on VHL demonstrating its tumour suppressor function in a xenograft setting but not in tissue culture conditions.30 31 In addition, the histology of the tumours was not affected by FLCN status (data not shown). Furthermore, since the tumour suppressor role of FLCN was observed regardless of VHL status, it might suggest that FLCN is involved in a VHL-independent tumour suppressor pathway.

    Figure 4
    Figure 4

    Tumor suppressor function of folliculin (FLCN) in human renal cell carcinoma cells. (A) FLCN expression was verified in ACHN-derived (left panel) cell lysates and 786-0-derived (right panel) cell lysates by western blot analysis using an anti-FLCN antibody. β-Actin was used as loading control. (B) Cell proliferation assay of ACHN-derived cells (left panel) and 786-0-derived cells (right panel) in 10% serum. Proliferation was plotted as cell number (SEM) versus days after plating. (C) Growth curves of xenografts illustrating mean tumour volume (SEM) following the subcutaneous injection of ACHN-derived (left panel) and 786-0-derived (right panel) cell lines into both flanks of nude mice. Left graph, n=8 for each cell line; right graph, n=6 for the control and n=8 for 786-0-FLCN cells. FLCN, folliculin.

    Complex and context-dependent activation of S6 and ERK1/2 in kidney cysts

    Although FLCN function has been linked to mTORC1/S6 activity in several recent studies, highly conflicting evidence of the role of FLCN in the mTORC1 pathways had been described. Indeed, loss of FLCN function was associated with elevated 16 18 or decreased 17 levels of mTORC1 and S6 activity during renal cystogenesis and tumourigenesis. To verify the status of mTORC1 activity in our Flcn mouse model, we evaluated by western blot analysis the level of a downstream effector of mTORC1, p-S6 (figure 5A). We observed a decrease in p-S6 (Ser235/236) in tissue extracts from Flcn+/− E12.5 embryos and adult (3 months) kidney compared to wild type, while the level of total S6 remained unchanged (figure 5A). Additionally, we investigated the levels of p-S6 in the extracts from solid tumours derived from subcutaneous injections of ACHN and 786-0 cells into nude mice. Again, we observed that FLCN positively regulates p-S6 levels as seen by the overall loss and induction of p-S6 in the knockdown and overexpressing cell lines, respectively (figure 5B, left and right panel, respectively; quantification in Supplemental figure S4). However, the activation level of S6 in ACHN and 786-0 cells grown in vitro were not affected by FLCN expression levels regardless of growth conditions (serum starvation and serum shock) (figure 5C). Furthermore, p-S6 levels displayed a complex and differential pattern in renal cystogenesis. Surprisingly, immunohistochemical staining for p-S6 in Flcn+/− kidney sections revealed high levels of p-S6 in the epithelial cells lining multilocular and large cysts from polycystic kidneys (figure 5D), while small single cysts were negative for p-S6 (figure 5E and F). The intensity of the p-S6 staining was very low in the surrounding normal tissue (Supplemental figure S5A).

    Figure 5
    Figure 5

    Multifaceted activation of S6 in Flcn+/− cystic kidneys and human renal cell carcinoma cells. (A) Western blot analysis of S6 and p-S6 (Ser235/236) using E12.5 embryos and kidney lysates from Flcn+/+ and Flcn+/− mice. Actin was used as a loading control. Western blot analysis of ACHN-derived (left panel) and 786-0-derived (right panel) tumours (B) and cells (C) for FLCN, S6 and p-S6. β-Actin (B) and α-tubulin (C) were used as loading controls. Representative data of two tumours are shown for the various types of tumours (B). (D–I) Kidney sections from Flcn+/− mice were evaluated by immunohistochemistry for p-S6 (D–F) and pERK1/2 (G–I). Scale bar, 100 μm. Arrowhead in I, sporadic strong staining for pERK1/2. FBS, fetal bovine serum; FLCN, folliculin; pERK1/2, phospho mitogen-activated protein kinase 3 & -1.

    Staining for p-ERK1/2 on sequential slides revealed a complex pattern with similarities to p-S6 staining. Strikingly, strong staining was detected in large cysts from polycystic kidneys (figure 5G), while p-ERK1/2 was almost absent in small cysts from monocystic kidneys (figure 5H). However, some pockets of strongly positive p-ERK1/2 cells were found in the epithelium, lining several cysts that stained negative for p-S6 (figure 5I, arrowheads). Similar to the p-S6 levels, sporadic p-ERK1/2 staining was observed in normal kidney tubules (Supplemental figure S3B). Loss of Flcn expression thus results in improperly elevated and diminished activation of S6, dependent on cellular context.

    Discussion

    Tissue distribution of FLCN

    In the present study, we characterised in great detail the tissue distribution of Flcn. Similar to its human homolog,32 we observed a broad distribution of mouse Flcn, suggesting a widespread biological function. The Flcn expression appeared to be significant in various mouse adult tissues, including the kidneys, lungs, heart, spleen, submandibular salivary glands and testes. However, the murine expression pattern presented some disparities with the human expression pattern. For instance, we found no expression of Flcn in the murine skin, whereas humans express FLCN in the skin.32 This, in all probability, accounts for the lack of skin lesions observed in our model. Furthermore, although BHD-associated kidney tumours are considered to arise from distal tubules in the human disease,33 LTA expression in the cells lining the cysts supports the notion that cyst developed from proximal tubules. This is in accordance with data from another study showing lack of expression of the distal tubule marker NaCl cotransporter in cysts, arising after conditional deletion of both FLCN alleles.16 Finally, we observed that homozygous disruption of Flcn results in embryonic lethality in the early stages of development (<E8.5), as was reported in other canine and rodent models.16 17 34 35 Early Flcn expression is therefore undoubtedly essential for embryonic development; however, it is presently not clear what role Flcn plays in the early stages of embryogenesis.

    Role of FLCN in tumour formation

    Familial genetic linkage analysis has firmly established that the BHD syndrome is caused by loss-of-function mutations in a single allele of the FLCN gene.11 However, insights into the molecular functions of FLCN are still largely undeveloped and, consequently, little knowledge of the pathogenesis of BHD-syndrome manifestations is available. Recent studies have produced kidney-specific homozygous disruption of murine Flcn15 16; however, the disparity between the human condition and the extreme phenotype reported makes it problematic to extend the associated molecular analysis to the human condition. As such, the pathology associated with the kidney in the human condition is characterised by a late onset (>40 years of age) and comparatively few lesions, whereas the murine conditional renal knockout models die from renal failure at 3 weeks of age and did not develop tumours.15 16 In the present study, we utilised another experimental approach, which allowed us to more closely mimic the human BHD syndrome. Thus, deleting a single allele in all tissues led to later onset of cystic kidney lesions (as compared to the conditional FLCN knockout model), which progressed towards kidney adenomas. This is in accordance with the commonly accepted model of formation of proliferative lesions in the renal tubules, in which renal tubular hyperplasia, cysts, adenoma and adenocarcinoma represent a progressive continuum.36

    That a complete loss of FLCN is required for the formation of renal malignancies and premalignacies to occur is supported by the non-complete penetrance of kidney cancer associated with the BHD syndrome in humans. Indeed, 25% to 40% of human BHD patients develop renal tumours, which is somewhat in good agreement with the penetrance of 50% associated with the present mouse model. Furthermore, renal carcinomas in BHD patients are generally devoid of FLCN expression in accordance with the Knudson two-hit tumour suppressor paradigm.2 Our finding of a complete absence of the FLCN protein in the cystic lesions therefore indicates that a second-hit mutation of FLNC is an early event in the multistep development of the renal carcinomas of the BHD syndrome. Interestingly, the non-cystic renal epithelium surrounding the cysts was also found to be FLCN negative, suggesting that even though complete loss of FLCN appears essential for the formation of premalignant lesions to form, loss of FLCN expression in itself might not be sufficient to initiate uncontrolled proliferation. During kidney organogenesis, a population of expanding pluripotent renal cap cells gives rise to all epithelial cell lineages within discrete nephritic units and correspond to mesodermal cells around the distal end of the ureteric bud that gives rise to nephrons in the kidney.37 It is plausible that such a second-hit mutation, leading to a complete loss of FLCN expression, occurs early in the developing kidney, while the pluripotent renal cap cell population expands. Thus, a single pluripotent cap cell suffering a second-hit mutation could give rise to a finite number of daughter cells that produces small clusters of FLCN-negative nephric tubule units, which eventually develop discrete mutations, leading to the formation of multiple renal tumours of different histological and molecular nature. In such a model, the timing of the second-hit FLCN mutation determines the number of FLCN-negative nephritic units (early mutation yields more daughter units compared with late mutations) and thus also determines the predisposition of the individual BHD patient to develop multiple tumours. Such a slow process dependent on several individual mutations is in contrast to the extremely proliferative response seen in kidney conditional homozygous deletion of Flcn, in which knockout mice dies at 3 weeks of age due to cystic, dilated non-functional kidneys.15 16 However, while these models have demonstrated the importance of FLCN as a player in nephritic tissue homeostasis, the extreme nature of these models does not mimic the human condition very well. Indeed, BHD patients experience the onset of kidney lesions in the fourth decade of life with a penetrance of 25% to 40%.27

    Complex modulations of S6 activity by FLCN

    In respect to the impact of FLCN on mTORC1/S6 signalling, we observed two categories of renal cysts: one type that had elevated expression of p-S6 which correlated with large polycystic lesions and another that was negative for p-S6 which correlated with smaller single cysts. While this seems somewhat ambiguous, it is in overall agreement with several recently published studies. Indeed, kidney-specific deletions of Flcn led to increased levels of p-mTORC1 and p-S6 in FLCN-deficient cystic tissues.15 16 On the other hand, Hartman et al described renal cysts and tumours negative for p-S6 as a consequence of heterozygotic loss of Flcn,17 suggesting that FLCN positively affects the mTORC1 signalling pathway. Of further support of this notion is our observation of diminished and increased activation of S6 as a consequence of diminished and increased expression levels of FLCN in human RCC cells when grown subcutaneously as solid tumours. However, when grown in vitro regardless of culture conditions and FLCN expression status, we observed no differences in S6 activation. It is also noteworthy that the mTORC1 inhibitor rapamycin can delay, but not prevent, the formation of polycystic kidneys in the conditional knockout models.15 16 On the basis of these conflicting studies,15–17 as well as our observations of differential activation of S6, the role of FLCN in S6 activation appears complex and very likely context dependent. Such context dependency can be speculated to be caused by additional genetic events in other pathways, leading to the formation of a specific histological and molecular type of renal cancer. Indeed, activation of mTORC1 and S6 in RCC has been demonstrated to be dependent on the histology of the renal tumour. Thus, mTORC1 and S6 are generally highly activated in clear cell and papillary renal carcinoma, but low or absent in oncocytoma and chromophobe RCC.38 In addition, it is also possible that the complex levels of S6 activation could be caused by the mesenchymal–epithelial interactions, which are frequently involved in tumourigenesis.39 40 It has recently been reported that fibroblast-like cells in TSC skin tumours, but not epidermal cells, showed bi-allelic deletion of TSC2. This event led to overexpression of epiregulin, an epidermal growth factor-related mitogen, which induced epidermal changes in proliferation as measured by an increase in pS6 levels.41

    Altogether, the conflicting studies of the impact of loss of Flcn on mTORC1 and S6 activities,15–17 in combination with our findings of p-S6-positive and p-S6-negative cysts, led us to propose that the role of loss of FLCN in renal cysts formation is more complex than solely activating the mTORC1 pathway and that FLCN may affect other cellular signalling pathways and/or biochemical mechanisms. This is further supported by our finding of elevated p-ERK1/2 in cystic epithelia devoid of p-S6, as well as the emergence of renal carcinomas of essentially all subtypes in BHD patients.12

    In conclusion, we suggest that FLCN inactivation occurs as an early event in the formation of kidney cysts and tumours. Furthermore, depending on the secondary molecular events occurring during tumourigenesis after loss of FLCN, malignancies of different molecular signatures will arise. Our observations and suggestions reflect the human BHD disease in which patients can develop renal tumours of several histologies, even within the same kidney.1 12

    Materials and methods

    Generation of a Flcn mouse model

    All procedures for generating the Flcn mouse model were performed at the transgenic facility of McGill University. Maintenance and experimental manipulation of animals were performed according to the guidelines and regulations of the McGill University Research and Ethic Animal Committee and the Canadian Council on Animal Care. See Supplemental materials and methods for details on Embryonic Stemcells cell clones, backgrounds and genotyping of mutant mice.

    Northern blot and semiquantitative RT-PCR analysis

    Total RNA was isolated from embryonic mice and adult mouse kidneys using RNA PureLink Micro-to-Midi (Invitrogen, Burlington, ON, Canada) according to the manufacturer's instructions. Single-stranded cDNAs were synthesised using SuperScript II-RNas H-reverse transcriptase (Invitrogen) with random primers. Northern blot analysis was done with a 900 bp cDNA probe for Flcn. A GAPDH probe was used as a loading control. See Supplemental materials and methods for detailed description of northern blot and semiquantitative RT-PCR procedures.

    Western blot analysis

    Protein extracts were resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes (Bio-Rad, Mississauga, ON, Canada). After blocking in 5% nonfat milk/TBST, the membranes were probed with monoclonal and polyclonal antibodies against FLCN (the monoclonal antibody was kindly provided to us by Dr Laura Schmidt (National Cancer Institute, Bethesda, Maryland, 21702 USA), and the polyclonal antibody was raised against full-length human protein in our laboratory), β-actin (Sigma-Aldrich, Oakville, ON, Canada), β-tubulin (Abcam, CedarLane, Burington, ON, Canada), S6 ribosomal protein (Cell Signaling Technology) or p-S6 ribosomal protein (Ser235/236) (Cell Signaling Technology, New England Biolabs, Danvers, MA, USA). See Supplemental materials and methods for details.

    X-gal staining and immunohistochemistry of adult mouse tissues

    See Supplemental materials and methods.

    Human RCC cell lines

    RCC cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were maintained in DMEM supplemented with 10% fetal bovine serum (Hyclone) and antibiotics at 37°C in a humidified 5% CO2 atmosphere.

    We developed a stable FLCN-knockdown RCC cell line by infecting ACHN cells with five different FLCN shRNA-expressing lentiviruses (MISSION, Sigma-Aldrich) according to the instructions of the manufacturer. A scrambled shRNA vector was included as control. Stable lentiviral cell lines were established using puromycin selection. To produce a stable FLCN-restored cell line, we infected the 786-0 RCC cell line with an FLCN-expressing lentiviral vector using the ViraPower Lentiviral Expression System (Invitrogen) and selected expressing cells with blasticidin.

    In vitro and in vivo growth analysis of RCC cell lines and in vitro S6 activation assay

    See supplemental materials and methods.

    Statistical analysis

    Statistical analyses were done by Student t test and one-way ANOVA with a Tukey post hoc test. Statistical significance was determined as p<0.05.

    Acknowledgments

    We are grateful to Dr M-C Gingras and S Welbourn for critical reading of the manuscript. We also thank A. Jimenez for her assistance with the mouse xenograft assays, M Narlis for her excellent histology technical assistance and advice, and D Grote for discussion and histology advice. Finally, we are thankful to Dr A Haggarty for her help with the development of the polyclonal anti-FLCN antibody.

    References

    1. ↵
      1. Pavlovich CP,
      2. Schmidt LS
      . Searching for the hereditary causes of renal-cell carcinoma. Nat Rev Cancer 2004;4(5):381–93.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Linehan WM,
      2. Pinto PA,
      3. Srinivasan R,
      4. et al
      . Identification of the genes for kidney cancer: opportunity for disease-specific targeted therapeutics. Clin Cancer Res 2007;13(2 Pt 2):671s–9s.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Linehan WM,
      2. Vasselli J,
      3. Srinivasan R,
      4. et al
      . Genetic basis of cancer of the kidney: disease-specific approaches to therapy. Clin Cancer Res 2004;10(18 Pt 2):6282S–9S.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Birt AR,
      2. Hogg GR,
      3. Dube WJ
      . Hereditary multiple fibrofolliculomas with trichodiscomas and acrochordons. Arch Dermatol 1977;113(12):1674–7.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Toro JR,
      2. Glenn G,
      3. Duray P,
      4. et al
      . Birt–Hogg–Dube syndrome: a novel marker of kidney neoplasia. Arch Dermatol 1999;135(10):1195–202.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Zbar B,
      2. Alvord WG,
      3. Glenn G,
      4. et al
      . Risk of renal and colonic neoplasms and spontaneous pneumothorax in the Birt–Hogg–Dube syndrome. Cancer Epidemiol Biomarkers Prev 2002;11(4):393–400.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Lindor NM,
      2. Hand J,
      3. Burch PA,
      4. et al
      . Birt–Hogg–Dube syndrome: an autosomal dominant disorder with predisposition to cancers of the kidney, fibrofolliculomas, and focal cutaneous mucinosis. Int J Dermatol 2001;40(10):653–6.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Welsch MJ,
      2. Krunic A,
      3. Medenica MM
      . Birt–Hogg–Dube syndrome. Int J Dermatol 2005;44(8):668–73.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Khoo SK,
      2. Bradley M,
      3. Wong FK,
      4. et al
      . Birt–Hogg–Dube syndrome: mapping of a novel hereditary neoplasia gene to chromosome 17p12-q11.2. Oncogene 2001;20(37):5239–42.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Schmidt LS,
      2. Warren MB,
      3. Nickerson ML,
      4. et al
      . Birt–Hogg–Dube syndrome, a genodermatosis associated with spontaneous pneumothorax and kidney neoplasia, maps to chromosome 17p11.2. Am J Hum Genet 2001;69(4):876–82.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Nickerson ML,
      2. Warren MB,
      3. Toro JR,
      4. et al
      . Mutations in a novel gene lead to kidney tumors, lung wall defects, and benign tumors of the hair follicle in patients with the Birt–Hogg–Dube syndrome. Cancer Cell 2002;2(2):157–64.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Vocke CD,
      2. Yang Y,
      3. Pavlovich CP,
      4. et al
      . High frequency of somatic frameshift BHD gene mutations in Birt–Hogg–Dube-associated renal tumors. J Natl Cancer Inst 2005;97(12):931–5.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Khoo SK,
      2. Kahnoski K,
      3. Sugimura J,
      4. et al
      . Inactivation of BHD in sporadic renal tumors. Cancer Res 2003;63(15):4583–7.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Knudson AG Jr
      . Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 1971;68(4):820–3.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Baba M,
      2. Furihata M,
      3. Hong SB,
      4. et al
      . Kidney-targeted Birt–Hogg–Dube gene inactivation in a mouse model: Erk1/2 and Akt-mTOR activation, cell hyperproliferation, and polycystic kidneys. J Natl Cancer Inst 2008;100(2):140–54.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Chen J,
      2. Futami K,
      3. Petillo D,
      4. et al
      . Deficiency of FLCN in mouse kidney led to development of polycystic kidneys and renal neoplasia. PLoS ONE 2008;3(10):e3581.
      OpenUrlCrossRefPubMed
    17. ↵
      1. Hartman TR,
      2. Nicolas E,
      3. Klein-Szanto A,
      4. et al
      . The role of the Birt–Hogg–Dube protein in mTOR activation and renal tumorigenesis. Oncogene 2009.
    18. ↵
      1. Baba M,
      2. Hong SB,
      3. Sharma N,
      4. et al
      . Folliculin encoded by the BHD gene interacts with a binding protein, FNIP1, and AMPK, and is involved in AMPK and mTOR signaling. Proc Natl Acad Sci U S A 2006;103(42):15552–7.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Hasumi H,
      2. Baba M,
      3. Hong SB,
      4. et al
      . Identification and characterization of a novel folliculin-interacting protein FNIP2. Gene 2008;415(1–2):60–7.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Takagi Y,
      2. Kobayashi T,
      3. Shiono M,
      4. et al
      . Interaction of folliculin (Birt–Hogg–Dube gene product) with a novel Fnip1-like (FnipL/Fnip2) protein. Oncogene 2008;27(40):5339–47.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Inoki K,
      2. Zhu T,
      3. Guan KL
      . TSC2 mediates cellular energy response to control cell growth and survival. Cell 2003;115(5):577–90.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Corradetti MN,
      2. Inoki K,
      3. Bardeesy N,
      4. et al
      . Regulation of the TSC pathway by LKB1: evidence of a molecular link between tuberous sclerosis complex and Peutz–Jeghers syndrome. Genes Dev 2004;18(13):1533–8.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Shaw RJ,
      2. Bardeesy N,
      3. Manning BD,
      4. et al
      . The LKB1 tumor suppressor negatively regulates mTOR signaling. Cancer Cell 2004;6(1):91–9.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Liu L,
      2. Cash TP,
      3. Jones RG,
      4. et al
      . Hypoxia-induced energy stress regulates mRNA translation and cell growth. Mol Cell 2006;21(4):521–31.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Gwinn DM,
      2. Shackelford DB,
      3. Egan DF,
      4. et al
      . AMPK phosphorylation of raptor mediates a metabolic checkpoint. Mol Cell 2008;30(2):214–26.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. van Slegtenhorst M,
      2. Khabibullin D,
      3. Hartman TR,
      4. et al
      . The Birt–Hogg–Dube and tuberous sclerosis complex homologs have opposing roles in amino acid homeostasis in Schizosaccharomyces pombe. J Biol Chem 2007;282(34):24583–90.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Toro JR,
      2. Wei MH,
      3. Glenn GM,
      4. et al
      . BHD mutations, clinical and molecular genetic investigations of Birt–Hogg–Dube syndrome: a new series of 50 families and a review of published reports. J Med Genet 2008;45(6):321–31.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Holthofer H,
      2. Virtanen I,
      3. Pettersson E,
      4. et al
      . Lectins as fluorescence microscopic markers for saccharides in the human kidney. Lab Invest 1981;45(5):391–9.
      OpenUrlPubMedWeb of Science
    29. ↵
      1. Holthofer H,
      2. Miettinen A,
      3. Paasivuo R,
      4. et al
      . Cellular origin and differentiation of renal carcinomas. A fluorescence microscopic study with kidney-specific antibodies, antiintermediate filament antibodies, and lectins. Lab Invest 1983;49(3):317–26.
      OpenUrlPubMedWeb of Science
    30. ↵
      1. Lieubeau-Teillet B,
      2. Rak J,
      3. Jothy S,
      4. et al
      . von Hippel-Lindau gene-mediated growth suppression and induction of differentiation in renal cell carcinoma cells grown as multicellular tumor spheroids. Cancer Res 1998;58(21):4957–62.
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Kondo K,
      2. Kim WY,
      3. Lechpammer M,
      4. et al
      . Inhibition of HIF2alpha is sufficient to suppress pVHL-defective tumor growth. PLoS Biol 2003;1(3):E83.
      OpenUrlPubMed
    32. ↵
      1. Warren MB,
      2. Torres-Cabala CA,
      3. Turner ML,
      4. et al
      . Expression of Birt–Hogg–Dube gene mRNA in normal and neoplastic human tissues. Mod Pathol 2004;17(8):998–1011.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Pavlovich CP,
      2. Walther MM,
      3. Eyler RA,
      4. et al
      . Renal tumors in the Birt–Hogg–Dube syndrome. Am J Surg Pathol 2002;26(12):1542–52.
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Okimoto K,
      2. Sakurai J,
      3. Kobayashi T,
      4. et al
      . A germ-line insertion in the Birt–Hogg–Dube (BHD) gene gives rise to the Nihon rat model of inherited renal cancer. Proc Natl Acad Sci U S A 2004;101(7):2023–7.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Lingaas F,
      2. Comstock KE,
      3. Kirkness EF,
      4. et al
      . A mutation in the canine BHD gene is associated with hereditary multifocal renal cystadenocarcinoma and nodular dermatofibrosis in the German Shepherd dog. Hum Mol Genet 2003;12(23):3043–53.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. Bernstein J,
      2. Evan AP,
      3. Gardner KD Jr
      . Epithelial hyperplasia in human polycystic kidney diseases. Its role in pathogenesis and risk of neoplasia. Am J Pathol 1987;129(1):92–101.
      OpenUrlPubMedWeb of Science
    37. ↵
      1. Kobayashi A,
      2. Valerius MT,
      3. Mugford JW,
      4. et al
      . Six2 defines and regulates a multipotent self-renewing nephron progenitor population throughout mammalian kidney development. Cell Stem Cell 2008;3(2):169–81.
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Robb VA,
      2. Karbowniczek M,
      3. Klein-Szanto AJ,
      4. et al
      . Activation of the mTOR signaling pathway in renal clear cell carcinoma. J Urol 2007;177(1):346–52.
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      1. van Kempen LC,
      2. Ruiter DJ,
      3. van Muijen GN,
      4. et al
      . The tumor microenvironment: a critical determinant of neoplastic evolution. Eur J Cell Biol 2003;82(11):539–48.
      OpenUrlCrossRefPubMedWeb of Science
    40. ↵
      1. Li S,
      2. Takeuchi F,
      3. Wang JA,
      4. et al
      . MCP-1 overexpressed in tuberous sclerosis lesions acts as a paracrine factor for tumor development. J Exp Med 2005;202(5):617–24.
      OpenUrlAbstract/FREE Full Text
    41. ↵
      1. Li S,
      2. Takeuchi F,
      3. Wang JA,
      4. et al
      . Mesenchymal-epithelial interactions involving epiregulin in tuberous sclerosis complex hamartomas. Proc Natl Acad Sci U S A 2008;105(9):3539–44.
      OpenUrlAbstract/FREE Full Text
      1. Skarnes WC
      . Entrapment vectors: a new tool for mammalian genetics. Biotechnology (N Y) 1990;8(9):827–31.
      OpenUrlCrossRefPubMed
      1. Nord AS,
      2. Chang PJ,
      3. Conklin BR,
      4. et al
      . The International Gene Trap Consortium website: a portal to all publicly available gene trap cell lines in mouse. Nucleic Acids Res 2006;34(Database issue):D642–8.
      OpenUrlAbstract/FREE Full Text
      1. Kurban G,
      2. Hudon V,
      3. Duplan E,
      4. et al
      . Characterization of a von Hippel Lindau pathway involved in extracellular matrix remodeling, cell invasion, and angiogenesis. Cancer Res 2006;66(3):1313–9.
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    Footnotes

    • SS, ABD and VK are equal contributors.

    • Funding This work was supported by grant from the Kidney Foundation of Canada and a Miriam and Saul Goldberg internal Goodman Cancer Center operating grant. VH is supported by a studentship from Terry Fox Foundation, SS is a recipient of a Canadian Institutes of Health Research Cancer Training Grant Award and ABD is a recipient of a Gerry Price Fellowship from the Cancer Research Society, Inc. AP is a recipient of the Canada Research Chair in Molecular Oncology.

    • Competing interests None to declare.

    • Provenance and peer review Not commissioned; externally peer reviewed.