Evidence that MIG-6 is a tumor-suppressor gene - Nature

Oncogene (2007) 26, 269–276

& 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc

ORIGINAL ARTICLE

Evidence that MIG-6 is a tumor-suppressor gene Y-W Zhang1, B Staal1, Y Su1, P Swiatek2, P Zhao3, B Cao3, J Resau4, R Sigler5, R Bronson6 and GF Vande Woude1 1 Laboratory of Molecular Oncology, Van Andel Research Institute, Grand Rapids, MI, USA; 2Laboratory of Germline Modification, Van Andel Research Institute, Grand Rapids, MI, USA; 3Laboratory of Antibody Technology, Van Andel Research Institute, Grand Rapids, MI, USA; 4Laboratory of Analytical, Cellular, and Molecular Microscopy, Van Andel Research Institute, Grand Rapids, MI, USA; 5Esperion Therapeutics, Division of Pfizer, Ann Arbor, MI, USA and 6Department of Pathology, Harvard Medical School, Boston, MA, USA

Mitogen-inducible gene 6 (MIG-6) is located in human chromosome 1p36, a locus frequently associated with human lung cancer. MIG-6 is a negative regulator of epidermal growth factor (EGF) signaling, and we show that Mig-6 – like EGF – is induced by hepatocyte growth factor/scatter factor (HGF/SF) in human lung cancer cell lines. Frequently, the receptors for both factors, EGFR and Met, are expressed in same lung cancer cell line, and MIG-6 is induced by both factors in a mitogen-activated protein kinase-dependent fashion. However, not all tumor lines express MIG-6 in response to either EGF or HGF/ SF. In these cases, we find missense and nonsense mutations in the MIG-6 coding region, as well as evidence for MIG-6 transcriptional silencing. Moreover, germline disruption of Mig-6 in mice leads to the development of animals with epithelial hyperplasia, adenoma, and adenocarcinoma in organs like the lung, gallbladder, and bile duct. These data suggests that MIG-6 is a tumorsuppressor gene and is therefore a candidate gene for the frequent 1p36 genetic alterations found in lung cancer. Oncogene (2007) 26, 269–276. doi:10.1038/sj.onc.1209790; published online 3 July 2006 Keywords: MIG-6; mutation; lung carcinogenesis; signal transduction; EGF; HGF/SF

Introduction Mitogen-inducible gene 6 (MIG-6), also known as gene 33 or RALT (Fiorentino et al., 2000; Makkinje et al., 2000), has been mapped to human chromosome 1p36. MIG-6 is an immediate early response gene that can be induced by stressful stimuli and growth factors, as well as by the oncoprotein Ras (Fiorentino et al., 2000; Makkinje et al., 2000; Tsunoda et al., 2002). MIG-6 protein can directly interact with all four members of the

ErbB family, including EGFR and ErbB2–4, and it acts as a negative feedback regulator of the ErbB RTK pathway (Fiorentino et al., 2000; Anastasi et al., 2003; Xu et al., 2005a). Recently, it has been reported that downregulated expression of the MIG-6 gene is observed in human breast carcinomas, which correlates with reduced overall survival of breast cancer patients (Amatschek et al., 2004; Anastasi et al., 2005). However, no mutations in MIG-6 have been detected in human breast carcinomas (Anastasi et al., 2005). Indeed, no mutations have been reported in MIG-6 to date, and the role of Mig-6 in human lung, gallbladder, and bile duct carcinogenesis has not been assessed. Allelic loss of chromosome 1p36 is among the most prominent genetic abnormalities observed in human lung cancer (Girard et al., 2000; Nomoto et al., 2000; Fujii et al., 2002), indicating that a critical tumorsuppressor gene(s) exists in this locus. Moreover, loss of heterozygosity (LOH) of the distal region of mouse chromosome 4, a region syntenic with human chromosome 1p36, is also frequently observed in mouse lung carcinogenesis (Herzog et al., 1995; Herzog et al., 2002). The p53 tumor-suppressor gene homolog, p73, is located in 1p36, but no mutations have been identified in human lung cancers (Nomoto et al., 1998), excluding it as the responsible tumor-suppressor gene. In this study, we report that the MIG-6 gene is mutated in the human non-small-cell lung cancer (NSCLC) cell lines NCI-H226 and NCI-H322M, as well as in one primary human lung cancer. We also show that loss of function of MIG-6 can result from dysregulation of its expression by RTK signaling. To this end, we show that several animals with disruption of Mig-6 by gene targeting develop epithelial hyperplasia as well as adenoma or adenocarcinoma in the lung, gallbladder, and bile duct. These data point to MIG-6 as a candidate tumor-suppressor gene. Results

Correspondence: Dr GF Vande Woude or Dr Y-W Zhang, Laboratory of Molecular Oncology, Van Andel Research Institute, 333 Bostwick NE, Grand Rapids, MI 49503, USA. E-mails: [email protected] or [email protected] Received 24 March 2006; revised 15 May 2006; accepted 15 May 2006; published online 3 July 2006

EGFR and Met signaling regulate MIG-6 expression in lung cancer cells Inappropriate activation of EGFR and Met receptor tyrosine kinase signaling by overexpression or mutation

MIG-6 mutation in lung cancer Y-W Zhang et al

270

is involved in lung carcinogenesis (Zochbauer-Muller et al., 2002). MIG-6 has been shown to be a negative feedback inhibitor of EGFR signaling in other cell types (Fiorentino et al., 2000; Anastasi et al., 2003; Xu et al., 2005a). In addition, we have observed that Mig-6 expression is strongly induced by HGF/SF in a sarcoma cell line (Zhang and Vande Woude, unpublished results). Here, we show that EGF or HGF/SF can also regulate MIG-6 expression in lung cancer cells. Both EGFR and Met are highly expressed in several of nine lung cancer cell lines tested (such as EKVX and HOP62) by Western blot analysis (Figure 1a). We demonstrate that in lung cancer cells, as in other cell types, the MIG-6 protein increases with EGF treatment in HOP62 cells (Figure 1b) and with HGF/SF treatment in EKVX cells (Figure 1c). Notably, no MIG-6 protein was detected in NCI-H322M and NCI-H226 cells (Figure 1a).

expressed significant levels of EGFR and Met (Figure 1a). We found that either EGF or HGF/SF induces MIG-6 expression in these two cell lines (Figure 1b–d). However, pretreating cells with MAP kinase inhibitor U0126 compared to PI3 kinase inhibitor LY294002 markedly diminished EGF- or HGF/SFinduced MIG-6 expression (Figure 1d). These data indicate that the regulation of MIG-6 expression by EGFR or Met signaling is mediated at least partially through the MAP kinase pathway (Keeton et al., 2004). The level of MIG-6 protein is very high in NCI-H23 cells, which carry an activating mutant RAS allele (Koo et al., 1999) but have barely detectable EGFR or Met (Figure 1a). RAS, a downstream molecule of RTK, EGFR, and Met, is required for the activation of the mitogen-activated protein kinases (MAPK)/ERK pathway, and in fact the high level expression of Mig-6 in NCI-H23 cells is in part due to the constitutive activation of the Ras-MAPK pathway (data not shown). MIG-6 feedback regulation by EGFR and Met is lost in NCI-H226 cells In EKVX cells, MIG-6 protein rapidly increased in response to EGF treatment (Figure 2a). Yet, even after

b A549

NCI-H522

NCI-H460

NCI-H322

NCI-H226

NCI-H23

HOP92

EKVX

a

HOP62

Regulation of MIG-6 by EGFR and Met is mediated through the MAP kinase pathway To determine the downstream pathway involved in EGFR- and Met-mediated MIG-6 regulation in lung cancer cells, we treated HOP62 and EKVX cells with various pathway inhibitors for 1 h before EGF or HGF/ SF induction for 4 h. Both HOP62 and EKVX cells

HOP62 (+)EGF Hours 0 0.5 1 2

4

6 MIG-6

MIG-6 β-actin EGFR

c Met

EKVX (+)HGF Hours 0 0.5 1 2

4

6 MIG-6

β-actin

β-actin

LY294002

U0126

DMSO

(+) HGF

DMSO

LY294002

U0126

DMSO

LY294002

U0126

EKVX

(+) HGF

(+) EGF

DMSO

U0126 LY294002

DMSO

(+) EGF

DMSO

HOP62

EKVX

DMSO

HOP62

DMSO

d

MIG-6 β-actin

Figure 1 MIG-6 expression is regulated by EGF and HGF/SF through the MAP kinase pathway in human lung cancer cells. (a) The expression of MIG-6, EGFR, and Met proteins in human lung cancer cell lines. Whole-cell extracts were prepared from lung cancer cell lines and subjected to Western blot analyses probed with anti-Mig-6, anti-EGFR, and anti-Met antibodies. The b-actin serves as an internal control for visualizing the amount of protein loaded in each lane. (b) Upregulation of MIG-6 by EGF. The cell lysates derived from HOP62 cells with or without EGF (50 ng/ml) treatment for the indicated times were subjected to Western blot analysis using antiMig-6 antibody. (c) Upregulation of MIG-6 by HGF/SF. EKVX cells were treated with HGF/SF (200 units/ml) for the indicated times, followed by Western blot analysis. (d) MAPK pathway mediates EGF- and HGF/SF-induced MIG-6 expression. HOP62 and EKVX cells were treated with U0126 or LY294002 for 1 h, respectively, followed by EGF (50 ng/ml) or HGF/SF (200 units/ml) treatment for additional 4 h, respectively. Cells treated with dimethylsulphoxide were used as controls. Western blotting was performed as described above. Oncogene


271

a

EKVX (+)EGF Hours 0 0.5

1

NCI-H322

2 4

6 0 0.5 1

2

NCI-H226 4

0 0.5 1

2

4

6

MIG-6

p-EGFR EGFR

p-ERK

ERK β-actin

b

NCI-H226 0 0.5 1

2

HOP62 4

0 0.5 1

2

4

(+)HGF Hours

c

NCI-H322 0 1

2

4

NCI-H226 0

1

2

4

(+)HGF Hours

MIG-6 p-ERK MIG-6 ERK

Met

GAPDH

β-actin

Figure 2 The regulation of MIG-6 expression by EGF or HGF/SF is defective in NCI-H226 cells. (a) EGF failed to induce MIG-6 protein expression in NCI-H226 cells. EKVX, NCI-H322, and NCI-H226 cells were serum-starved and then treated with EGF (50 ng/ml) for the indicated times. At each time point, the cell lysates were prepared and subjected to Western blot analyses using the indicated antibodies. (b) No induction of MIG-6 protein by HGF/SF was detected in NCI-H226 cells. Serum-starved NCI-H226 and HOP62 cells were treated with HGF/SF (200 units/ml) for the indicated times. Western blotting was performed as described above using the indicated antibodies. (c) EGF failed to induce MIG-6 mRNA transcription. NCI H322 and NCI-H226 cells were serum-starved overnight and treated with EGF (50 ng/ml) at the indicated times. At each time point, RNA was isolated and subjected to Northern blot analysis with a 32 P-labeled MIG-6 probe. As a control, GAPDH was also analysed.

EGF treatment for 4–6 h, no MIG-6 was detected in NCI-H322M or NCI-H226 cells using an antibody directed against the C-terminal 14 amino acids (Figure 2a). EGF-induced EGFR tyrosine phosphorylation and downstream ERK activation in EKVX, NCIH322M, and NCI-H226 cells (Figure 2a). HGF/ SF-induced MIG-6 expression in EKVX and HOP62 cells (Figures 1 and 2b), but like EGF, HGF/SF did not induce MIG-6 in NCI-H226 cells that express high levels of Met and respond to HGF/SF (Figures 1 and 2b). Interestingly, the duration of ERK phosphorylation by EGFR and Met is more sustained in MIG-6-deficient NCI-H226 and NCI-H322M cells than in EKVX and HOP62 cells expressing a MIG-6 product (Figures 2a and b). We asked if MIG-6 is expressed in NCI-H322M and NCI-H226 cells at the transcriptional level. We performed Northern blot analysis using total RNAs prepared from NCI-H322M and NCI-H226 cells with or without EGF treatment. The MIG-6 mRNA level is dramatically increased in NCI-H322M cells within 1 h

(Figure 2c). However, almost no MIG-6 mRNA expression was detected in NCI-H226 cells even after EGF treatment for 1–4 h (Figure 2c). These data strongly suggest that feedback upregulation of MIG-6 by EGFR or Met is defective in NCI-H226 lung cancer cells, even though the MAPK/ERK pathway that mediates the RTK-induced upregulation of MIG-6 is intact (Figure 2a and b). This implies that the promoter regulatory regions of MIG-6 in NCI-H226 are either genetically or epigenetically altered. MIG-6 protein in NCI-H322M cells is not detectable (Figure 2a), despite the fact that EGF can induce MIG-6 mRNA expression (Figure 2c), suggesting a potential alteration of MIG-6 protein properties in these cells. The MIG-6 gene is mutated in human lung cancer cell lines and primary lung cancer The above data prompted us to examine whether MIG-6 is genetically altered in human lung cancer. From nine NSCLC cell lines derived from NCI 60 cell lines, we Oncogene


272

identified two point mutations in the coding region of MIG-6. Even when RTK-induced MIG-6 transcription is silenced (Figure 2b and c), we found that the gene in NCI-H226 cells bears a homozygous missense mutation leading to the replacement of Asp with Asn at codon 109. We also found that the MIG-6 product in NCIH322M carries a homozygous nonsense mutation, resulting in a truncation after codon 83 (Figure 3 and Table 1). This alteration prevented its protein detection by the antibody directed against the MIG-6 C terminus (Figures 1 and 2a). We also examined 41 cases of

a

Wild Type

primary human lung cancers and found a germline mutation in one patient, an alteration of Ala to Val at codon 373 (Figure 3 and Table 1). Polymorphisms in MIG-6 were also observed in lung cancer cell lines and primary lung cancers (Table 1). Disruption of Mig-6 in mice causes lung, gallbladder, and bile duct carcinogenesis Evidence supporting the tumor-suppressor function of Mig-6 was also derived from Mig-6-deficient mice.

b

Wild Type

Wild Type

NCI-H322 (GAA TAA)

NCI-H226 (GAT AAT) E2 E3

E4 EGFR BD

CRIB

SH3 BD

14-3-3 BD

E1

SH3 BD

c

1041190A (Cancer tissue) (GCC GTC)

PEST

AH

1041190B (Normal tissue) (GCC GTC)

459

1 PEST

Asp109 Glu38

Asn Ala373 Val

STOP

Figure 3 Identification of MIG-6 mutations in human lung cancer cell lines and primary lung cancer. (a) The MIG-6 gene is mutated in the NCI-H226 and NCI-H322M NSCLC cancer cell lines. The upper panels show the wild-type sequences and the lower panels show the mutant sequences in MIG-6. The arrows mark the mutated nucleotides derived from the two cell lines. (b) A heterozygous germline mutation of MIG-6 is identified in one primary lung cancer. The top panel shows the wild-type sequence, the middle panel shows the mutant sequence in MIG-6 derived from the primary lung cancer tissue, and the lower panel shows the sequence derived from the normal control tissue from the same patient. The arrows indicate the mutated nucleotide. (c) Schematic representation of the MIG-6 genomic structure, the protein, and the location of the identified mutations.

Table 1

Summary of MIG-6 mutations identified in human lung cancer cell lines and primary lung cancer

Diagnosis

Nucleotide

Adenocarcinoma Squamous cell carcinoma Adenocarcinoma

nt. 942 C-A nt. 537 G-A nt. 459 G-T

Human lung tissues 1041190A Squamous cell carcinoma 1041190B Normal lung tissue

Cell lines NCI-H23 NCI-H226 NCI-H322M

4030373A

Adenocarcinoma

4030373B

Normal lung tissue

4030422A 4030422B

Squamous cell carcinoma Normal lung tissue

Abbreviation: MIG-6, Mitogen-inducible gene 6. Oncogene

Exon

Protein

Mutation

Genotype

4 4 4

R244R (CGA-AGA) D109N (GAT-AAT) E83Stop (GAA-TAA)

Polymorphism Missense Nonsense

Homozygous Homozygous Homozygous

nt. 1118 C-T nt. 1118 C-T

4 4

A373 V (GCC-GTC) A373 V (GCC-GTC)

Missense Missense

Heterozygous Heterozygous

nt. nt. nt. nt.

60 A-G; 730 C-A 60 A-G; 730 C-A

2 4 2 4

L20L (CTA-CTG); R244R (CGA-AGA) L20L (CTA-CTG); R244R (CGA-AGA)

Polymorphism

Heterozygous

Polymorphism

Heterozygous

nt. 730 C-A nt. 730 C-A

4 4

R244R (CGA-AGA) R244R (CGA-AGA)

Polymorphism Polymorphism

Heterozygous Heterozygous


273 Mig-6 +/+

a

e

b

Mig-6 +/-

f

Mig-6 -/-

Mig-6 -/-

c

g

Mig-6 -/-

d

h

Mig-6 -/-

Figure 4 Mig-6 deficiency in mice causes lung carcinogenesis. (a) Normal lung of a 1-year-old Mig-6 þ / þ mouse. (b) Normal lung of a 1-year-old Mig-6 þ / mouse. (c) Representative image showing the bronchi and bronchiole epithelial hyperplasia observed in the lung of a 9-month-old Mig-6/ mouse. (d) Proliferation of round cells in the alveoli of Mig-6/ lung (1-year-old). (e) Development of lung adenomas in a Mig-6/ mouse (9.5-months old). The adenoma shows distinct borders and a monomorphic population of alveolar/ bronchiolar cells. Note: this mouse had two adenomas on two different lobes. (f) High magnification of the image in the square shown in (e). (g) Development of lung adenocarcinoma in an 8.5-month-old Mig-6/ mouse. The large early alveolar/bronchiolar carcinoma shows an indistinct border and central necrosis with cholesterol clefts. (h) High magnification of the image in the square shown in (g).

We disrupted the Mig-6 gene in mice by gene targeting technology and previously reported that loss of Mig-6 leads to early onset degenerative joint disease (Zhang et al., 2005). Although Mig-6 is expressed in mouse lung tissue, we did not observe developmental lung defects at early stages of development in Mig-6-deficient mice (data not shown). However, from 5 to 13 months we observed that Mig-6 mutant mice develop lung abnormalities, ranging from epithelial hyperplasia of bronchi and bronchiole to the formation of adenomas or adenocarcinomas (Figure 4 and Table 2). A majority of Mig-6 mutant mice die within 6 months due to the joint abnormality (Zhang et al., 2005), and out of a total of 29 Mig-6/ mice between 5 and 13 months, we observed four lung cancers; one animal had an adenocarcinoma, a second animal had two adenomas in two different lobes, and two mice with a single adenoma each. In addition, in the 29 homozygous animals we observed 11 cases of bronchi or bronchiole epithelial hyperplasia (Table 2). Statistical analysis (Fisher’s exact test) reveals that the lungs derived from Mig-6-deficient mice have significant pathological changes, including hyperplasia and neoplasia, relative to those from the control wild-type (P-value: 0.001398) and heterozygous mice (P-value: 0.000017). In addition to lung cancer, we also observed gallbladder and/or bile duct neoplasms, which ranged from epithelial hyperplasia to carcinoma in several other Mig-6/ animals (Figure 5a and b; Table 3). Immunohistochemical staining of the tissues with proliferating cell nuclear antigen (PCNA) revealed increased numbers of proliferating cells within the mutant gallbladder (Figure 5c). These data implicate a loss of Mig-6 function in lung, gallbladder, and bile duct carcinogenesis and are consistent with the evidence that Mig-6 is a tumorsuppressor gene.

Table 2

Lung pathologies in Mig-6-deficient and control mice between 5 and 13 months of age

Lung adenoma or adenocarcinoma Bronchi and bronchiole epithelial hyperplasia

Mig-6+/+ (n ¼ 17)

Mig-6+/ (n ¼ 31)

Mig-6/ (n ¼ 29)

0 1

1 0

4 11

Abbreviation: MIG-6, Mitogen-inducible gene 6.

Discussion MIG-6 localizes in human chromosome 1p36, a locus that has been widely suggested to harbor putative tumor-suppressor genes. Allelic imbalance of chromosome 1p36 is one of the most frequent genetic alterations observed in various human cancers (Ragnarsson et al., 1999; Thiagalingam et al., 2002). Linkage analyses using microsatellite markers revealed deletions of 1p36 in nearly 50% of primary human lung cancers (Nomoto et al., 2000). Similar results were also reported in human lung cancer cell lines, including both NSCLC and small cell lung cancer (Virmani et al., 1998; Girard et al., 2000; Fujii et al., 2002). The evidence indicating the presence of a tumor-suppressor gene in 1p36 also comes from studies of mouse lung cancer. Loss of heterozygosity in the region of mouse chromosome 4, which is syntenic to human chromosome 1p36, have been frequently observed in spontaneous and carcinogen-induced mouse lung adenocarcinomas (Herzog et al., 1995; Herzog et al., 2002; Sargent et al., 2002). The search for the responsible gene in 1p36 has not been successful. The p53 homolog, p73, is in this locus and has been rigorously tested. However, no mutations in the p73 gene have been identified thus far, although frequent Oncogene


274

a

Mig-6 +/-

Mig-6 -/-

b

Mig-6 +/-

Mig-6 -/-

c Mig-6 +/-

Mig-6 +/-

Mig-6 -/-

Mig-6 -/-

Figure 5 Gallbladder and bile duct cancers in Mig-6-deficient mice. (a) Gallbladder hyperplasia in a Mig-6/ mouse. H&E sections of gallbladders derived from age-matched (2 months) Mig-6 þ / and Mig-6/ mice are shown. (b) Bile duct adenocarcinoma in a Mig-6/ mouse. Bile duct sections derived from 10-month-old Mig-6 þ / and Mig-6/ mice are shown. (c) Hyperproliferation of epithelial cells in Mig-6/ gallbladder. The sections of gallbladder tissues (as shown in a) were IMC stained with anti-PCNA antibody. The brownstained cells are PCNA positive. Table 3

Cases of gallbladder and bile duct carcinogenesis in Mig-6/ mice

Genotype

Age Pathology (months)

Mig-6/ Mig-6+/ Mig-6+/+

12 12 6.5 12 12 6.5 12 12 6.5

Adenocarcinoma of bile ducts (low grade) One adenocarcinoma and one adenoma Adenocarcinoma of bile ducts (low grade) Dilatation of gallbladder Cystic hyperplasia Mild dilatation and hyperplasia of gallbladder Gallbladder a bit dilated and a little hyperplasia Normal Normal

Abbreviation: MIG-6, Mitogen-inducible gene 6.

allelic imbalances have been observed at this locus (Nomoto et al., 1998). In addition, p73 expression has been found to increase in lung cancer rather than decrease (Mai et al., 1998; Tokuchi et al., 1999), and no spontaneous tumor has been observed in p73-deficient mice (Yang et al., 2000). All these suggest the presence of other unidentified tumor-suppressor genes in 1p36. For many reasons, it is plausible to consider MIG-6 as a possible 1p36 lung cancer tumor-suppressor gene. First, it resides at 1p36.12–36.33, in the locus that is considered a hot spot of allelic imbalance for lung cancer (Girard et al., 2000; Nomoto et al., 2000; Fujii et al., 2002). Further, we show that disruption of the Oncogene

mouse Mig-6 gene, whereas localizes to the 1p36 syntenic region in mouse chromosome 4, results in lung carcinogenesis (Figure 4). Importantly, we identify lossof-function mutations in the MIG-6 gene in human lung cancer cell lines. Mig-6 is normally expressed in lung (data not shown) and plays a role in mechanical stress pulmonary ventilation (Makkinje et al., 2000). Also, MIG-6 is a negative regulator of receptor tyrosine kinase signaling from factors like EGF (Fiorentino et al., 2000) and HGF/SF (Figures 1 and 2; Pante et al., 2005), whose receptors have been shown to play important roles in lung malignancy (Zochbauer-Muller et al., 2002; Birchmeier et al., 2003; Ma et al., 2003, 2005; Paez et al., 2004; Stephens et al., 2004). All this evidence supports MIG-6 as a tumor-suppressor gene. Likewise, Mig-6 may also function as a tumor-suppressor in other organs, since animals with a Mig-6 deficiency also develop gallbladder and bile duct cancers (Figure 5 and Table 3). Very recently, another group also reported that Mig-6 deficiency in mice caused skin cancer in addition to the joint phenotype that we have previously reported (Zhang et al., 2005; Ferby et al., 2006). Moreover, it has recently been reported that MIG-6 expression is lost in ErbB2-amplified breast carcinomas (Anastasi et al., 2005). However, it has also been reported that exogenous expression of Mig-6 in MCF7 breast cancer cells inhibited apoptosis (Xu et al., 2005b).


275

Like other tumor-suppressor genes involved in lung carcinogenesis (Kohno and Yokota, 1999; ZochbauerMuller et al., 2002), the inactivation of MIG-6 may be due to either genetic or epigenetic changes. LOH seems to be the case for the NCI-H322M human lung adenocarcinoma cell line, in which a single nonsense point mutation was identified in one allele of the MIG-6 gene, while the other allele appeared to be deleted (http://www.ncbi.nlm.nih.gov/sky/skyweb.cgi). Inactivation of MIG-6 appears to involve another mechanism in NCI-H226 human lung squamous cell carcinoma cells. In addition to the missense mutation identified in the Mig-6 coding region, regulation of MIG-6 gene expression by either EGFR or Met was defective. There are at least two explanations for this dysregulation of MIG-6 by receptor signaling: either a deletion or mutation occurs in the promoter regulatory region or the promoter silencing is caused epigenetically. A similar mechanism could also be involved in the loss of MIG-6 expression in ErbB2-amplified breast carcinomas (Anastasi et al., 2005). What might be the role of MIG-6 in normal lung function and during lung carcinogenesis? MIG-6 is a scaffolding protein involved in receptor signal transduction. The expression of MIG-6 is induced by EGF, whose signaling plays an important role in normal lung development (Miettinen et al., 1995, 1997). Like many other tyrosine kinase receptors, EGF receptor signaling needs to be attenuated after activation; constitutive activation is deleterious to normal lung epithelial cells and can lead to carcinogenesis (Zochbauer-Muller et al., 2002; Paez et al., 2004; Stephens et al., 2004). MIG-6 interacts with the ErbB receptor family and negatively regulates EGF signaling (Fiorentino et al., 2000; Anastasi et al., 2003; Xu et al., 2005a), thereby providing through negative feedback a fine tuning of EGF signaling shortly after its activation. MIG-6 deficiency caused by mutation or failure of feedback regulation may lead to inappropriate activation of EGF signaling and other signaling as well (such as HGF/SFMet signaling). We do observe that in MIG-6-deficient cells there is prolonged receptor tyrosine kinasemediated MAPK activation (Figure 2a and b). It has been shown that overexpression of Mig-6 inhibits ErbB2-mediated NIH 3T3 cell transformation (Fiorentino et al., 2000). Mig-6 may provide a checkpoint for normal cell proliferation in certain tissues; disruption of Mig-6 leads to uncontrolled proliferation of cells as revealed by PCNA staining in gallbladder epithelium (Figure 5) and in joint tissues (Zhang et al., 2005). A role for MIG-6 in cell cycle regulation has also been implied, as its expression is regulated during the normal cell cycle progression (Wick et al., 1995). Moreover, many stressful stimuli also induce the expression of Mig-6, which activates SAPK/JNK (Makkinje et al., 2000; Keeton and Messina, 2005). SAPK/JNK activity is usually suppressed in order for transformed cells to escape SAPK/JNK-dependent apoptosis and become tumorigenic (Davis, 2000; Benhar et al., 2002). Inactivation of Mig-6 may result in an inability to induce SAPK/JNK-dependent apoptosis

and lead to immortalization of cells. In addition, there are several well-known protein–protein interaction motifs residing in MIG-6, including a Cdc42/Rac interactive binding (CRIB) domain, a Src homology 3 (SH3) domain binding motif, and a 14-3-3 interacting motif (Makkinje et al., 2000). It is still not clear how MIG-6 interacts with its partner proteins and exerts its function during various cellular processes, but taken together, our data show that abnormal regulation of MIG-6 reveals its activity as a tumor-suppressor gene, and loss of its activity contributes to the initiation of lung carcinogenesis as well as other cancers.

Materials and methods Human lung cancer cell lines and cell culture The nine NSCLC cell lines EKVX, HOP62, HOP92, NCIH23, NCI-H226, NCI-H322M, NCI-H446, NCI-H522, and A549 were derived from the NCI 60 cell lines. The cells were cultured in Rosewell’s Park Memorial Institute media medium 1640 supplemented with 10% fetal bovine serum. Mutational analysis of MIG-6 Human lung cancer and normal control tissues were obtained through the Cooperative Human Tissue Network (CHTN). Genomic DNAs were isolated from human cell lines and tissues. Polymerase chain reaction (PCR) was performed to amplify the entire coding regions of MIG-6 (exons 2, 3, and partial exon 4) using three primer pairs (the primer sequences are available upon request). The PCR products were purified by QIAquick PCR Purification Kit (QIAGEN, Valencia, CA, USA) and sequenced using an ABI7000 sequencer. Western blot analysis Western blotting was performed as described previously (Zhang et al., 2003). Briefly, the total cell lysates were extracted from lung cancer cells and resolved by Tris-glycine gel (Invitrogen, Carlsbad, CA, USA). The proteins were then transferred to a polyvinylidene difluoride membrane (Invitrogen) and detected by immunoblotting with the indicated antibody. The anti-EGFR, anti-p-EGFR, and anti-Met were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA); the anti-b-actin was from Sigma (St Louis, MO, USA); the anti-p-ERK and anti-ERK were from Cell Signaling Technology (Beverly, MA, USA); and the anti-Mig-6 was produced by immunizing rabbits with the synthesized peptides derived from the C-terminal 14 amino acids of Mig-6. Northern blot analysis Total RNA (20 mg per sample) was subjected to Northern blot analysis as described (Zhang et al., 2003). The DNA fragment used for probing MIG-6 was amplified from the region between nucleotides 213 and 1601 of human MIG-6 (accession no. NM_018948) by reverse transcriptase PCR. The probe for human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) has been described previously (Zhang et al., 2003). Mouse histology and immunohistochemistry The generation and genotyping of Mig-6 knockout mice has been described elsewhere (Zhang et al., 2005). The mice analysed in this study are on a B6/129 background. Mouse tissues were fixed in formalin and embedded in paraffin. Sections (5 mm) were stained with hematoxylin and eosin Oncogene


276 (H&E) for pathological examination. immunohistochemistry (IHC) staining of PCNA was performed as previously described (Zhang et al., 2005). Acknowledgements We thank Bryn Eagleson, Jason Martin, and Kellie Sisson for helping with the mice, Matt VanBrocklin and Han-Mo Koo

for providing the lung cancer cell lines, Bree Berghuis and JC Goolsby for histological assistance, David Petillo for sequencing, Eric Kort for statistic analysis, David Nadziejka for technical editing of the manuscript, and Michelle Reed and Troy Carrigan for manuscript preparation. This work was supported by the Van Andel Foundation and in part by the Michigan Economic Development Corporation and Michigan Technology Tri-Corridor Grant MAMC 085P1000815.

References Amatschek S, Koenig U, Auer H, Steinlein P, Pacher M, Gruenfelder A et al. (2004). Cancer Res 64: 844–856. Anastasi S, Fiorentino L, Fiorini M, Fraioli R, Sala G, Castellani L et al. (2003). Oncogene 22: 4221–4234. Anastasi S, Sala G, Huiping C, Caprini E, Russo G, Iacovelli S et al. (2005). Oncogene 24: 4540–4548. Benhar M, Engelberg D, Levitzki A. (2002). EMBO Rep 3: 420–425. Birchmeier C, Birchmeier W, Gherardi E, Vande Woude GF. (2003). Nat Rev Mol Cell Biol 4: 915–925. Davis RJ. (2000). Cell 103: 239–252. Ferby I, Reschke M, Kudlacek O, Knyazev P, Pante G, Amann K et al. (2006). Nature Med 12: 568–573. Fiorentino L, Pertica C, Fiorini M, Talora C, Crescenzi M, Castellani L et al. (2000). Mol Cell Biol 20: 7735–7750. Fujii T, Dracheva T, Player A, Chacko S, Clifford R, Strausberg RL et al. (2002). Cancer Res 62: 3340–3346. Girard L, Zochbauer-Muller S, Virmani AK, Gazdar AF, Minna JD. (2000). Cancer Res 60: 4894–4906. Herzog CR, Devereux TR, Pittman B, You M. (2002). Cancer Res 62: 6424–6429. Herzog CR, Wang Y, You M. (1995). Oncogene 11: 1811–1815. Keeton AB, Messina JL. (2005). J Cell Biochem 94: 1190–1198. Keeton AB, Xu J, Franklin JL, Messina JL. (2004). Biochim Biophys Acta 1679: 248–255. Kohno T, Yokota J. (1999). Carcinogenesis 20: 1403–1410. Koo HM, McWilliams MJ, Alvord WG, Vande Woude GF. (1999). Cancer Res 59: 6057–6062. Ma PC, Jagadeeswaran R, Jagadeesh S, Tretiakova MS, Nallasura V, Fox EA et al. (2005). Cancer Res 65: 1479–1488. Ma PC, Kijima T, Maulik G, Fox EA, Sattler M, Griffin JD et al. (2003). Cancer Res 63: 6272–6281. Mai M, Yokomizo A, Qian C, Yang P, Tindall DJ, Smith DI et al. (1998). Cancer Res 58: 2347–2349. Makkinje A, Quinn DA, Chen A, Cadilla CL, Force T, Bonventre JV et al. (2000). J Biol Chem 275: 17838–17847. Miettinen PJ, Berger JE, Meneses J, Phung Y, Pederson RA, Werb Z et al. (1995). Nature 376: 337–341.

Oncogene

Miettinen PJ, Warburton D, Bu D, Zhao JS, Berger JE, Minoo P et al. (1997). Dev Biol 186: 224–236. Nomoto S, Haruki N, Kondo M, Konishi H, Takahashi T, Takahashi T et al. (1998). Cancer Res 58: 1380–1383. Nomoto S, Haruki N, Tatematsu Y, Konishi H, Mitsudomi T, Takahashi T et al. (2000). Genes Chromosomes Cancer 28: 342–346. Paez JG, Janne PA, Lee JC, Tracy S, Greulich H, Gabriel S et al. (2004). Science 304: 1497–1500. Pante G, Thompson J, Lamballe F, Iwata T, Ferby I, Barr FA et al. (2005). J Cell Biol 171: 337–348. Ragnarsson G, Eiriksdottir G, Johannsdottir JT, Jonasson JG, Egilsson V, Ingvarsson S. (1999). Br J Cancer 79: 1468–1474. Sargent LM, Senft JR, Lowry DT, Jefferson AM, Tyson FL, Malkinson AM et al. (2002). Cancer Res 62: 1152–1157. Stephens P, Hunter C, Bignell G, Edkins S, Davies H, Teague J et al. (2004). Nature 431: 525–526. Thiagalingam S, Foy RL, Cheng KH, Lee HJ, Thiagalingam A, Ponte JF. (2002). Curr Opin Oncol 14: 65–72. Tokuchi Y, Hashimoto T, Kobayashi M, Hayashi M, Nishida S, Imai K et al. (1999). Br J Cancer 80: 1623–1629. Tsunoda T, Inokuchi J, Baba I, Okumura K, Naito S, Sasazuki T et al. (2002). Cancer Res 62: 5668–5671. Virmani AK, Fong KM, Kodagoda D, McIntire D, Hung J, Tonk V et al. (1998). Genes Chromosomes Cancer 21: 308–319. Wick M, Burger C, Funk M, Muller R. (1995). Exp Cell Res 219: 527–535. Xu D, Makkinje A, Kyriakis JM. (2005a). J Biol Chem 280: 2924–2933. Xu J, Keeton AB, Wu L, Franklin JL, Cao X, Messina JL. (2005b). Breast Cancer Res Treat 91: 207–215. Yang A, Walker N, Bronson R, Kaghad M, Oosterwegel M, Bonnin J et al. (2000). Nature 404: 99–103. Zhang YW, Su Y, Lanning N, Swiatek PJ, Bronson RT, Sigler R et al. (2005). Proc Natl Acad Sci USA 102: 11740–11745. Zhang YW, Su Y, Volpert OV, Vande Woude GF. (2003). Proc Natl Acad Sci USA 100: 12718–12723. Zochbauer-Muller S, Gazdar AF, Minna JD. (2002). Annu Rev Physiol 64: 681–708.