Linking Epithelial Polarity and Carcinogenesis by Multitasking Helicobacter Pylori Virulence Factor Caga

Home , Epithelial polarity

Oncogene (2008) 27, 7047–7054 & 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $32.00 www.nature.com/onc REVIEW Linking epithelial polarity and carcinogenesis by multitasking Helicobacter pylori virulence factor CagA

M Hatakeyama

Division of Molecular Oncology, Institute for Genetic Medicine, Hokkaido University, Sapporo, Japan

Loss of cell polarity and tissue architecture is a hallmark second and fourth most common form of malignant of carcinomas that arise from epithelial cells. Recent tumors in male and female individuals, respectively, studies on Drosophila tumor suppressors have provided accounting for 876 000 estimated new cases and 405 000 evidence that epithelial polarity and cell proliferation are estimated deaths in the year 2000 (Parkin et al., 2001). functionally coupled, suggesting a function for polarity There isaccumulating evidence that chronic infection defects in the development of carcinomas. This notion is with Helicobacter pylori, a spiral-shaped bacterium that supported by the findings that mammalian orthologs of colonizes the gastric mucosa of more than half of the these Drosophila tumor suppressors are targeted by a world’shuman population, hasa causativefunction in number of viral oncoproteins. Chronic infection with the development of gastric carcinoma (Nomura et al., Helicobacter pylori is causally associated with gastric 1991; Parsonnet et al., 1991; Uemura et al., 2001). carcinoma. H. pylori virulence factor CagA (cytotoxin- Studieson H. pylori–gastric epithelial cell interactions associated gene A), which is delivered into gastric have greatly contributed to our current understanding epithelial cells through a bacterial type IV secretion of H. pylori-triggered mucosal lesions that direct gastric system, has an important function in cell transformation carcinogenesis, pointing to the H. pylori virulence factor through interacting with and deregulating SHP-2 phos- CagA (cytotoxin-associated gene A) as a central player phatase, a bona fide oncoprotein that is associated with in this process (Peek and Blaser, 2002; Hatakeyama, human malignancies. Recent studies have further revealed 2004, 2008). that CagA specifically binds and inhibits PAR1/MARK polarity-regulating kinase, thereby causing junctional and polarity defects in epithelial cells. Thus, the bacterial oncoprotein simultaneously targets the polarity-regulating Translocation of H. pylori CagA into gastric epithelial system and growth-regulatory system. These findings cells indicate that loss of cell polarity underlies the abnormal proliferation of epithelial cells that directs carcinogenesis. CagA isa 120 B145-kDa H. pylori protein that is Oncogene (2008) 27, 7047–7054; doi:10.1038/onc.2008.353 encoded by the cagA gene (Covacci et al., 1993; Tummuru et al., 1993). The cagA gene islocalized at Keywords: Helicobacter pylori; CagA; gastric carcino- one end of the cag pathogenicity island (cag PAI), a ma; epithelial polarity; PAR1/MARK 40-kb DNA segment that is considered to be horizon- tally transferred into the H. pylori genome (Censini et al., 1996; Akopyants et al., 1998). Whereasalmostall of the East Asian H. pylori isolates are cagA-positive, approximately half of the H. pylori strains isolated in Introduction Western countries do not carry cag PAI and thusare cagA negative. In addition to CagA, the cag PAI B Cell polarity isfundamental for cell-fate decisionduring contains 30 putative genes, among which 18 genes development aswell asfor the maintenance of differ- encode proteinsservingasbuilding blocksof a type IV entiated cells that constitute normal tissues. Tumors are secretion system. The type IV secretion system forms a formed by cells that have lost the ability to assemble and syringe-like structure that is capable of penetrating into create normal tissues. It is therefore reasonable to the cytoplasm of gastric epithelial cells. Infection with speculate that defects in cell polarity have an important cagA-positive H. pylori strains is associated with severe function in the development of tumors. The majority of mucosal inflammation that underlies peptic ulceration, human malignant tumorsthat arisefrom epithelial atrophic gastritis and gastric carcinoma (Blaser et al., tissues are termed carcinomas. Gastric carcinoma is the 1995; Kuipers et al., 1995; Parsonnet et al., 1997). Following the attachment of cagA-positive H. pylori to the surface of gastric epithelial cells, CagA is Correspondence: Dr M Hatakeyama, Division of Molecular Onco- delivered from the bacterium into the cytoplasm of host logy, Institute for Genetic Medicine, Hokkaido University, Kita-15, Nishi-7, Kita-ku, Sapporo 060-0815, Japan. cells through the type IV secretion system (Segal et al., E-mail: [email protected] 1999; Asahi et al., 2000; Backert et al., 2000; Odenbreit Linking epithelial polarity and carcinogenesis by H. pylori CagA M Hatakeyama 7048 et al., 2000; Stein et al., 2000). Thisprocessismediated Instead, they have a distinct EPIYA-containing at least partly through an interaction of H. pylori CagL sequence, termed the EPIYA-D segment, which is with integrins(Kwok et al., 2007). CagL, also encoded unique to East Asian CagA. by cag PAI, istargeted to the surfaceof the type IV secretion pili and acts as a specific adhesin that binds to the host integrin a5b1 in an Arg-Gly-Asp (RGD) motif- dependent manner. CagL–integrin interaction triggers Interaction of CagA with SHP-2 oncoprotein type IV injection of CagA into gastric epithelial cells. Within the host cells, CagA localizes to the inner surface Upon tyrosine phosphorylation, CagA acquires the of the plasma membrane, where it undergoes tyrosine ability to interact with the cytoplasmic protein tyrosine phosphorylation by Src family kinases or Abl kinase phosphatase SHP-2, which possesses two tandem- (Selbach et al., 2002; Stein et al., 2002; Poppe et al., repeated Src homology-2 (SH2) domainsat the 2007; Tammer et al., 2007). The tyrosine phosphoryla- N-terminal region (N-SH2 and C-SH2) and a protein tion site of CagA is characterized by the Glu-Pro-Ile- tyrosine phosphatase domain at the C-terminal region Tyr-Ala (EPIYA) motif, which ispresentin multiple (Higashi et al., 2002a, b) (Figure 1). Binding of tyrosine- numbersin the carboxy-terminal polymorphic region phosphorylated CagA to the SH2 domains causes a (EPIYA-repeat region) of the protein (Higashi et al., conformational change in SHP-2 that relievesintramo- 2002a, b). On the basis of sequences flanking the EPIYA lecular inhibition of the phosphatase domain by the N- motifs, four distinct EPIYA segments, EPIYA-A, -B, -C SH2 domain, resulting in aberrant activation of SHP-2 and -D, each of which containsa singleEPIYA phosphatase activity. Western and East Asian CagA motif, have been identified in the EPIYA-repeat species, respectively, bind SHP-2 through the tyrosine- region (Higashi et al., 2002b, 2005; Naito et al., 2006) phosphorylated EPIYA-C and EPIYA-D segments. (Figure 1). The representative CagA proteins of Western Notably, SHP-2-binding activity of the EPIYA-D H. pylori isolates (Western CagA) possess the EPIYA-A segment is significantly stronger than that of the and EPIYA-B segments followed by the EPIYA-C EPIYA-C segment, rendering East Asian CagA biolo- segment. Intriguingly, the EPIYA-C segment, consisting gically more active than Western CagA (Naito et al., of 34 amino-acid residues, variably multiplies in tandem, 2006). mostly from one time to three times, among different SHP-2 has an important function in signal transduc- Western CagA species as a result of homologous tion downstream of growth factor/cytokine receptors recombination of a 102-bp cagA gene segment encoding that promotescell proliferation, morphogenesisand EPIYA-C. The representative CagA proteins of H. motility (Neel et al., 2003). Consequently, expression of pylori isolated in East Asian countries (East Asian CagA in gastric epithelial cells causes morphological CagA) also possess the EPIYA-A and EPIYA-B transformation termed the hummingbird phenotype, segments but not the repeatable EPIYA-C segment. which ischaracterized by elongated cell shapewith

SHP-2 N-SH2 C-SH2 Phosphatase domain

ABC-CagA P P P A B C (Western) CM sequence

PAR-1 Kinase domain

ABD-CagA (East Asian) A B D CM sequence P P P

SHP-2 N-SH2 C-SH2 Phosphatase domain Figure 1 Interaction of CagA with cellular targets. Western CagA contains EPIYA-A, EPIYA-B and EPIYA-C segments, whereas East Asian CagA comprises EPIYA-A, EPIYA-B and EPIYA-D segments. Each of the EPIYA segments contains a tyrosine-phosphorylatable EPIYA motif. Western and East Asian CagA interact with SHP-2 phosphatase in a tyrosine phosphorylation-dependent manner. The interaction involves the two SH2 domains of SHP-2 and tyrosine-phosphorylatable EPIYA-C or EPIYA-D segment of CagA. CagA also interacts with PAR1 kinase through the C-terminal 16-amino-acid residues termed the CM sequence, which locates immediately downstream of the EPIYA-C or EPIYA-D segment. CagA binds to the C-terminal 27-amino-acid residues present in the kinase catalytic domain of PAR1, thereby causing inhibition of the PAR kinase activity. CagA, cytotoxin-associated gene A; CM, CagA multimerization.

Oncogene Linking epithelial polarity and carcinogenesis by H. pylori CagA M Hatakeyama 7049 dramatic cytoskeletal rearrangements (Segal et al., 1999; Disruption of tight junction and the loss of epithelial Higashi et al., 2002a). An important question is the polarity by CagA nature of the substrates that are dephosphorylated by CagA-activated SHP-2 in the morphogenetic activity of The establishment and maintenance of mammalian CagA. Recently, focal adhesion kinase (FAK) has been epithelial apical–basolateral polarity are mediated by identiﬁed as one such substrate (Tsutsumi et al., 2006). complex interplaysamong polarity-regulating proteins FAK hasan important function in the regulation of that include mammalian orthologsof Caenorhabditis focal adhesions that mediate cell–extracellular matrix elegans PAR (partitioning-defective) proteins(from interaction. CagA-activated SHP-2 dephosphorylates at PAR1 to PAR6), which are essential for the formation the activating tyrosine phosphorylation sites of FAK in of anterior–posterior asymmetry in the C. elegans gastric epithelial cells. This in turn leads to the down- embryo, and mammalian orthologsof Drosophila tumor regulation of FAK kinase activity, which is critically suppressors such as Lethal giant larvae (Lgl), Discs large involved in the induction of the hummingbird pheno- (Dlg) and Scribble (Scrib) (Kemphues et al., 1988; type by CagA. Consistent with the function of SHP-2 in Tepass et al., 2001; Macara, 2004; Suzuki and Ohno, potentiating the Erk microtubule-associated protein 2006). In mammalian epithelial cells, atypical protein (MAP) kinase pathway by both Ras-dependent and kinase C (aPKC) complexed with PAR3 and PAR6 Ras-independent mechanisms, CagA also elicits sus- (aPKC/PAR3/PAR6 complex) localizesand determines tained Erk activation, which hasan important function the apical membrane domain (Figure 2a). Their function in G1 to S phase cell-cycle progression, in gastric isantagonized by PAR1 aswell asa setof tumor epithelial cells(Higashi et al., 2004). suppressor proteins (Lgl, Dlg and Scrib) that localize

a Epithelial cell b PAR1 CagA-PAR1 CagA CagA-SHP-2 Apical (dimer) interaction phosphorylation interaction

Tight junction Src CagA CagA CagA CagA CagA CagA PAR1 PAR1

Lateral P P SH2 SH2 PTP PAR1 CM sequence SHP-2

Basal PAR1 inhibition SHP-2 activation

c cagA-positive H. pylori

Apical

CagA

P P SHP-2 P P Lateral

PAR1 PAR1

Disruption of tight junctions Basal Loss of cell polarity Abnormal mitogenic signal Hummingbird phenotype EMT-like change Figure 2 Disruption of epithelial polarity by CagA. (a) Regulation of apical–basolateral polarity. In epithelial cells, PAR1 kinase localizesto the basolateralmembrane, whereasatypical PKC complexed with PAR3 and PAR6 (aPKC complex) localizesto and above the tight junctions. Asymmetric distribution of PAR1 and the aPKC complex maintains the epithelial apical–basolateral polarity. At the tight junctions, aPKC phosphorylates PAR1, which induces dissociation of PAR1 from the membrane, and thereby preventsthe distributionof PAR1 to the apical membrane domain. ( b) Sequential interaction of CagA with PAR1 kinase and SHP-2 phosphatase. Two CagA proteins bind a PAR1 dimer through the CM sequences and speciﬁcally inhibit the PAR1 kinase activity. PAR1-dimerized CagA proteins then undergo tyrosine phosphorylation at the EPIYA segments by Src kinases, and the resulting tyrosine-phosphorylated CagA-PAR1 complex acquires the ability to stably interact with the SHP-2 oncoprotein containing two SH2 domains. (c) Upon delivery into gastric epithelial cells, CagA binds and inhibits PAR1 throughout the basolateral membrane, thereby disrupting tight junctions and inducing loss of epithelial apical–basolateral polarity. In cells that lost epithelial polarity, CagA elicits aberrant mitogenic signalaswell aselevated cell motility upon deregulation of SHP-2. CagA, cytotoxin-associatedgene A; CM, CagA multimerization; PKC, protein kinase C.

Oncogene Linking epithelial polarity and carcinogenesis by H. pylori CagA M Hatakeyama 7050 and determine the basolateral membrane domain (for of the PAR1 kinase activity by CagA (Saadat review see Humbert et al., this issue). The apical and et al., 2007). The observation indicates that the basolateral membrane domains are separated by a phosphorylation-dependent CagA activity is also de- physical barrier termed the apical junctional complex pendent on the phosphorylation-independent CagA comprising tight junctions and adherens junctions. Tight activity. junctionsestablishand maintain cell polarity by limiting the diffusion of integral membrane proteins between apical and basolateral membranes. Tight junctions also generate a seal between the membranes of neighboring Targeting E-cadherin/b-catenin complex by CagA cellsand act asa paracellular barrier of the polarized epithelial monolayer. Epithelial cell–cell adhesion is provided by the homo- CagA colocalizeswith tight junction proteinssuchas philic interaction of E-cadherin at the adherensjunc- ZO-1 and JAM and causes disruption of the tight tions. The cytoplasmic domain of E-cadherin also junction in polarized epithelial cellsin a manner that is interactswith a variety of peripheral membrane independent of CagA tyrosine phosphorylation (Amieva proteins, including p120ctn, b-catenin, a-catenin, vinculin et al., 2003). As a result, epithelial cells expressing CagA and a-actinin, and is associated with the actin cytoske- lose cell polarity, which is concomitantly associated with leton through these proteins. Besides acting as a cell–cell pseudopodia formation and degradation of the base- interaction apparatus, E-cadherin functions as an ment membrane (Bagnoli et al., 2005). Hence, CagA inhibitor of the b-catenin signal by preventing the inducescellular changesthat resemble the epithelial-to- function of b-catenin asa transcriptional factor by mesenchymal transition. This CagA activity is mediated sequestrating it to the plasma membrane. Direct by specific interaction of CagA with PAR1 (Saadat connection between E-cadherin and gastric carcinogen- et al., 2007: Zeaiter et al., 2008) (Figure 1), one of the six esis has been demonstrated through the finding that PAR proteinsisolatedin C. elegans. PAR1 isa serine/ genetic mutationsin CDH1, the gene encoding threonine kinase that has a central function in the E-cadherin, are associated with hereditary diffuse gastric establishment and maintenance of the basolateral cancer (Guilford et al., 1998). It is also well established membrane domain in epithelial cells(Hurov et al., that deregulated b-catenin signaling has an important 2004; Suzuki et al., 2004) (Figure 2a). Mammalian function in the development of colorectal carcinoma PAR1 wasoriginally identified asa microtubule affinity- (Segditsas and Tomlinson, 2006). regulating kinase (MARK) based on its ability to Recent studies indicate a relationship between CagA phosphorylate MAPs such as tau, MAP2 and MAP4 and the E-cadherin/b-catenin complex. A carcinogenic (Drewes et al., 1997). Upon phosphorylation, PAR1 H. pylori strain in Mongolian gerbils shows the ability to destabilizes microtubules by releasing MAPs from the selectively activate b-catenin-dependent transcription in microtubules. There are four PAR1 paralogs in mamma- a CagA-dependent manner (Franco et al., 2005). lian cells, PAR1a/MARK3/C-TAK1, PAR1b/MARK2/ Activation of the b-catenin signal by CagA is EMK, PAR1c/MARK1 and PAR1d/MARK4/MARKL1, independent of CagA tyrosine phosphorylation and is which may have both shared and unique functions. mediated by the interaction of CagA with E-cadherin, Considering the fact that cell polarity is established by which leadsto destabilization of the E-cadherin/ b- the concerted activity of asymmetrically localized catenin complex and thereby causes nuclear transloca- proteins/protein complexes and the fact that microtu- lization of otherwise membranous b-catenin to bules serve as tracks for regulated movement of such activate transcription (Murata-Kamiya et al., 2007). proteins/protein complexes, PAR1 may regulate cell Thus, CagA impairs adherens junctions and at the polarity at least partly through controlling the stability same time elicits deregulated b-catenin signaling. Dereg- of microtubulesby acting asa MARK. ulation of b-catenin by CagA again requiresthe CagA- CagA interactswith the kinasecatalytic domain of multimerization sequence to which PAR1 binds, in- PAR1, from PAR1a to PAR1d, and thereby inhibits dicating that the CagA–E-cadherin interaction isindir- PAR1 kinase activity (Saadat et al., 2007) (Figure 1). ect and ismediated through the CagA–PAR1 complex Conversely, PAR1 binds to the C-terminal 16-amino- (Kurashima et al., 2008). Possibly, the CagA–PAR1 acid sequence of CagA known as the CagA-multi- complex attractsa number of additional cellular merization sequence (Ren et al., 2006) (Figure 1). As proteinsother than SHP-2 (Figure 2b). One such PAR1 is present as a dimer (Panneerselvam et al., 2006), protein may be E-cadherin and the interaction could two CagA proteins passively dimerize upon interacting destabilize the E-cadherin/b-catenin complex. CagA- with a PAR1 dimer and, following tyrosine phosphor- deregulated b-catenin transactivates several genes ylation by Src, the PAR1-mediated CagA dimer including cdx1, which encodes an intestine-specific forms a stable complex with a single SHP-2 molecule transcription factor, Cdx1, that mediates intestinal through the two SH2 domainsof SHP-2 (Figure 2b). differentiation (Murata-Kamiya et al., 2007). This Inhibition of PAR1 by CagA causes junctional and observation raises the possibility that the CagA-deregu- polarity defects, which are followed by disorganization lated b-catenin isinvolved in the development of of the epithelial monolayer (Figure 2c). Interestingly, intestinal metaplasia, a precancerous transdifferentia- induction of the hummingbird phenotype by CagA- tion of gastric epithelial cells to an intestinal phenotype deregulated SHP-2 requiressimultaneousinhibition (Correa, 1992).

Oncogene Linking epithelial polarity and carcinogenesis by H. pylori CagA M Hatakeyama 7051 Function of CagA–SHP-2 interaction in tumorigenesis seem to be sufficient for the development of epithelial tumors (carcinomas), as transgenic expression of a gain- The recent establishment of transgenic mice that systemi- of-function SHP-2 mutant in mice resulted in the cally expressCagAhasprovided convincing evidence for development of hematological malignanciesbut not the in vivo function of CagA asa bacterial oncoprotein solid tumors (Araki et al., 2004). Thisin turn indicates (Ohnishi et al., 2008). The cagA-transgenic mice developed that the development of epithelial tumorsrequiresboth gastric epithelial hyperplasia, gastric hyperplastic polyps tyrosine phosphorylation-dependent and -independent and gastrointestinal carcinomas without showing any signs CagA activities, suggesting a function for the CagA– of inflammation, indicating that the oncogenic potential of PAR1 interaction in thisprocess.Physiologically, PAR1 CagA iscell autonomous.Importantly, the levelsof CagA kinases are activated by phosphorylation of the T-loop in transgenic mice established were far lower than those on a threonine residue. This phosphorylation is carried achieved by infection with H. pylori. Given that deregulated out by another serine/threonine kinase, LKB1 (Lizcano hyperactivation of SHP-2 isembryonically lethal (Araki et al. (2004); for review see Bardeesy et al., this issue). et al., 2004), overexpression of CagA would not be tolerated Germline loss-of-function mutations in the LKB1 gene during mouse development. A portion of the cagA- are responsible for autosomal dominant Peutz–Jeghers transgenic mice also developed hematological malignancies, syndrome, which is characterized by the development of especially myeloid leukemias and B-cell lymphomas. The benign hamartomatouspolypsthroughout the gastro- study using these mice further demonstrated the critical intestinal tract (Hemminki et al., 1998). One of the most function of CagA tyrosine phosphorylation in the develop- important features associated with Peutz–Jeghers syn- ment of tumors in vivo. As CagA tyrosine phosphorylation drome is the increased risk of carcinomas from the is an essential prerequisite for CagA–SHP-2 interaction and stomach, colon, lung, uterus and breast. These observa- subsequent deregulation of SHP-2 by CagA, the results of tions indicate that LKB1 is a tumor suppressor acting the work using these mice suggest a critical function for upstream of PAR1. Indeed, LKB1 is a mammalian CagA-deregulated SHP-2 in tumorigenesis. ortholog of C. elegans PAR4 and isinvolved in polarity Indeed, there isaccumulating evidence that SHP-2 regulation in mammalian cells(Baas et al., 2004). The playsan important role in the development of human causal link between LKB1 and the cancer-prone Peutz– malignancies. Germline gain-of-function mutations of Jeghers syndrome supports the idea that defects in PTPN11, the gene encoding SHP-2, are associated with epithelial polarity are required for the development of Noonan syndrome, a developmental disorder caused by epithelial tumors. PAR1 is also phosphorylated by the the deregulated activation of the Ras–MAP kinase aPKC/PAR3/PAR6 complex (Hurov et al., 2004; pathway (Tartaglia et al., 2001; Bentires-Alj et al., Suzuki et al., 2004). ThisPAR1 phosphorylation creates 2006). Notably, Noonan patients exhibit increased risk a docking site for 14-3-3, the mammalian ortholog of of malignancies, especially juvenile myelomonocytic C. elegans PAR5. Subsequent interaction of PAR1 with leukemia and neuroblastoma. Furthermore, PTPN11 14-3-3 dissociates PAR1 from the plasma membrane, mutations have been found in sporadic cases of juvenile thereby preventing the distribution of PAR1 to the myelomonocytic leukemia, childhood myelodysplastic apical membrane domain. Gene amplification and syndrome, acute B-lymphocytic leukemia and acute elevated activity of aPKC, which downregulatesPAR1 myelocytic leukemia as well as some solid tumors such functions, have also been reported in ovarian, lung and asneuroblastoma(Tartaglia et al., 2003; Bentires-Alj colon carcinomas(Murray et al., 2004; Eder et al., 2005; et al., 2004). Hence, SHP-2 isa bona fide human Regala et al., 2005). oncoprotein like Ras(Mohi and Neel, 2007). Notably, In addition to MAPs, PAR1a (MARK3/C-TAK1) has the spectrum of human hematological malignancies been reported to phosphorylate CDC25c phosphatase, associated with SHP-2 mutation is similar to that which dephosphorylatesand activatescyclin B/CDC2 to observed in cagA-transgenic mice. Together with the promote G2/M transition and subsequent M-phase observation that mice expressing disease-associated progression. Upon phosphorylation by PAR1a, Cdc25c PTPN11 mutants also developed similar hematological createsa docking sitefor 14-3-3 that sequestratesCdc25c abnormalities(Araki et al., 2004; Mohi et al., 2005), to the cytoplasm (Peng et al., 1998). Similarly, PAR1a- CagA-deregulated SHP-2 may have a major function in mediated phosphorylation of the kinase suppressor of the development of hematological malignanciesin cagA- Raf-1, which potentiates the Ras–MAP kinase signaling transgenic mice. It should be noted here that gastric as a scaffold, generates a 14-3-3 binding site and thereby mucosa-associated lymphoid tissue lymphoma, which is inhibits kinase suppressor of Raf-1 activity (Muller et al., also associated with chronic infection with cagA-positive 2001). Thus, PAR1a is capable of inhibiting cell multi- H. pylori, isderived from B lymphocytes,thereby merization at variable cell cycle phases. This in turn suggesting a possible involvement of CagA in the develop- indicatesthat the inhibition of PAR1 kinasesby CagA ment of mucosa-associated lymphoid tissue lymphoma. can directly stimulate cell proliferation.

Function of PAR1 in carcinogenesis Loss of epithelial polarity in cell transformation

Although the CagA–SHP-2 oncoprotein interaction has Growing evidence supports the idea that impaired cell an important function in cell transformation, it does not polarity is not only associated with the acquisition of

Oncogene Linking epithelial polarity and carcinogenesis by H. pylori CagA M Hatakeyama 7052 more malignant phenotypes such as metastatic potential Conclusion: multistep gastric carcinogenesis but also involved in early stages of carcinogenesis. In Drosophila, mutations in the neoplastic tumor suppres- The study of cagA-transgenic mice indicates that CagA- sor genes, lgl,dlg and scrib, disrupt polarity of epithelia mediated tumorigenesis is cell autonomous in that it and simultaneously induce overproliferation of epithe- neither requiresadditional bacterial factorsnor elicits lial cells with malignant characteristics, indicating the host responses such as inflammation (Ohnishi et al., function of the fly tumor suppressors in coupling cell 2008). Why doesCagA have an oncogenic potential? In polarity and cell proliferation (Humbert et al., 2003; other words, what is the advantage for H. pylori to have Bilder, 2004; Humbert et al., this issue). Consistent with the CagA oncoprotein? Making a carcinoma in the host thisidea, decreasedexpressionof Lgl, Dlg or Scrib is stomach may not be the primary purpose of cagA- observed in a variety of human tumors (Cavatorta et al., positive H. pylori. CagA per se doesnot confer growth 2004; Nakagawa et al., 2004; Schimanski et al., 2005; advantage on H. pylori in vitro. Possibly, the presence of Kuphal et al., 2006). Furthermore, Dlg and Scrib are CagA enables H. pylori to colonize more successfully targeted for ubiquitin-mediated degradation by the than cagA-negative H. pylori in the host stomach. In this high-risk human papillomavirus E6 oncoprotein, which context, the bacterium may utilize the oncogenic (abnor- is associated with cervical carcinoma, and also by the mal mitogenic) signal triggered by CagA to gently but human T-cell leukemia virustype 1 tax oncoprotein efficiently induce senescence or apoptosis/programmed (Kiyono et al., 1997; Lee et al., 1997; Gardiol et al., cell death through the mechanism known as the 1999; Suzuki et al., 1999; Nakagawa and Huibregtse, ‘oncogenic stress’ (Ben-Porath and Weinberg, 2005; 2000). How could loss of epithelial polarity promote Prieur and Peeper, 2008). This results in a decreased cell transformation? Physiologically, the establishment number of acid-secreting cellsand thereby lowersgastric of apical junctional complexesisa prerequisitefor acidity to a level that providesa comfortable living contact inhibition. Thisagain indicatesthat coordinated environment for H. pylori. regulation of cell polarity and cell growth hasan Development of gastric adenocarcinoma is a multi- important function in the prevention of uncontrolled step process that requires qualitative as well as cell proliferation. If transformed cells cannot form quantitative alterationsin the expressionof oncogenes apical junctional complexeswith ‘normal’ neighboring and tumor suppressor genes, lasting for several decades. cells, they are excluded from the polarized epithelial During infection with cagA-positive H. pylori, gastric monolayer and acquire the ability to initiate abnormal epithelial cellsare continuouslyexposedto injection of proliferation in the absence of growth inhibitory CagA from the bacterium. The injected CagA binds cues, which may otherwise be generated through and deregulatesSHP-2 oncoprotein, while disrupting epithelial cell–cell interaction. In thisregard, the apical junctional complexesby interacting with PAR1, CagA–PAR1–SHP-2 complex hasa dual function; to induce oncogenic stress and subsequent apoptosis disruption of the epithelial polarity and aberrant in gastric epithelial cells as noted above. This in turn activation of the Ras–MAPK pathway, indicating causes elevated epithelial (stem) cell turnover and that the protein complex coordinatescell polarity thereby increases the chance for gastric epithelial cells defectsand oncogenic signaling to promote epithelial to acquire genetic and/or epigenetic changesin cancer- cell transformation (Saadat et al., 2007). Such functional related genessuchas p53,ras, b-catenin,APC and cooperativity hasalready been indicated by the ob- RUNX3. Although the oncogenic potential of CagA servation that tumors induced by an oncogenic Ras in appearsto be cell autonomous,the processof tumor Drosophila do not have metastatic potential, whereas development may be greatly enhanced in the presence of those induced by oncogenic Ras under the condition chronic inflammation caused by H. pylori infection of loss of heterozygosity for lgl,dlg and scrib show more (Rieder et al., 2005). In gastric epithelial cells that have malignant phenotypesand frequently metastasize acquired mutationsin pro-apoptotic genes,CagA (Brumby and Richardson, 2003; Pagliarini and Xu, injection may provoke abnormal proliferation rather 2003). than induce apoptosis by oncogenic stress, thereby Many of the genesinvolved in epithelial cell polarity selectively and effectively expanding such precancerous are also involved in the establishment and maintenance cells. In this scenario, therefore, gastric cancer is a of polarity in stem cells (Betschlinger and Knoblich, long-term side effect of the sustained oncogenic stress 2004). In particular, the function of PAR1 in the by CagA. Interestingly, chronic infection with cagA- regulation of microtubules suggests that it could play a positive H. pylori inducesaberrant expressionof role in asymmetric cell division during mitosis. It is activation-induced cytidine deaminase, a key enzyme therefore possible that defects in the asymmetric division that generatesantibody diversification, in a manner that of epithelial stem cells by CagA-mediated PAR1 isdependent on nuclear factor-kB (Matsumoto et al., inhibition lead to an increased number of stem cells 2007). Activation-induced cytidine deaminase upregula- and thereby increase the chance of subsequent tumor tion may significantly increase the chance of introducing development due to excess cell proliferation in the somatic mutations in genes such as p53 and other genes epithelial compartment. involved in gastric carcinogenesis.

Oncogene Linking epithelial polarity and carcinogenesis by H. pylori CagA M Hatakeyama 7053 References

AkopyantsNS, Clifton SW, KersulyteD, Crabtree JE, Youree BE, loss of apical–basal polarity and cyclin E overexpression in ovarian Reece CA et al. (1998). Analyses of the cag pathogenicity island of cancer. Proc Natl Acad Sci USA 102: 12519–12524. Helicobacter pylori. Mol Microbiol 28: 37–53. Franco AT, Israel DA, Washington MK, Krishna U, Fox JG, Rogers Amieva MR, Vogelmann R, Covacci A, TompkinsLS, NelsonWJ, AB et al. (2005). Activation of b-catenin by carcinogenic Falkow S. (2003). Disruption of the epithelial apical–junctional Helicobacter pylori. Proc Natl Acad Sci USA 102: 10646–10651. complex by Helicobacter pylori CagA. Science 300: 1430–1434. Gardiol D, Kuhne C, Glaunsinger B, Lee SS, Javier R, Banks L. Araki T, Mohi MG, Ismat FA, Bronson RT, Williams IR, Kutok JL (1999). Oncogenic human papillomavirusE6 proteinstarget the et al. (2004). Mouse model of Noonan syndrome reveals cell type- discs large tumour suppressor for proteasome-mediated degrada- and gene dosage-dependent effects of Ptpn11 mutation. Nat Med 10: tion. Oncogene 18: 5487–5496. 849–857. Guilford P, HopkinsJ, Harraway J, McLeod M, McLeod N, Harawira Asahi M, Azuma T, Ito S, Ito Y, Suto H, Nagai Y et al. (2000). P et al. (1998). E-cadherin germline mutationsin familial gastric Helicobacter pylori CagA protein can be tyrosine phosphorylated in cancer. Nature 392: 402–405. gastric epithelial cells. J Exp Med 191: 593–602. Hatakeyama M. (2004). Oncogenic mechanisms of Helicobacter pylori BaasAF, KuipersJ, van der Wel NN, Batlle E, Koerten HK, PetersPJ CagA protein. Nat Rev Cancer 4: 688–694. et al. (2004). Complete polarization of single intestinal epithelial Hatakeyama M. (2008). SagA of CagA in Helicobacter pylori cellsupon activation of LKB1 by STRAD. Cell 116: 457–466. pathogenesis. Curr Opin Microbiol 11: 30–37. Backert S, Ziska E, Brinkmann V, Zimny-Arndt U, Fauconnier A, Hemminki A, Markie D, Tomlinson I, Avizienyte E, Roth S, Loukola Jungblut PR et al. (2000). Translocation of the Helicobacter pylori A et al. (1998). A serine/threonine kinase gene defective in Peutz– CagA protein in gastric epithelial cells by a type IV secretion Jegherssyndrome. Nature 391: 184–187. apparatus. Cell Microbiol 2: 155–164. Higashi H, Nakaya A, Tsutsumi R, Yokoyama K, Fujii Y, Ishikawa S Bagnoli F, Buti L, TompkinsL, Covacci A, Amieva MR. (2005). et al. (2004). Helicobacter pylori CagA provokesRas-independent Helicobacter pylori CagA inducesa transitionfrom polarized to morphogenetic response through targeting SHP-2. J Biol Chem 279: invasive phenotypes in MDCK cells. Proc Natl Acad Sci USA 102: 17205–17216. 16339–16344. Higashi H, Tsutsumi R, Fujita A, Yamazaki S, Asaka M, Azuma T Ben-Porath I, Weinberg RA. (2005). The signals and pathways et al. (2002b). Biological activity of the Helicobacter pylori virulence activating cellular senescence. Int J Biochem Cell Biol 37: 961–976. factor CagA is determined by variation in the tyrosine phosphor- Bentires-Alj M, Kontaridis MI, Neel BG. (2006). Stops along the RAS ylation sites. Proc Natl Acad Sci USA 99: 14428–14433. pathway in human genetic disease. Nat Med 12: 283–285. Higashi H, Tsutsumi R, Muto S, Sugiyama T, Azuma T, Asaka M Bentires-Alj M, Paez JG, David FS, Keihack H, Halmos B, Naoki K et al. (2002a). SHP-2 tyrosine phosphatase as an intracellular target et al. (2004). Activating mutationsof the noonan syndrome- of Helicobacter pylori CagA protein. Science 295: 683–686. associated SHP2/PTPN11 gene in human solid tumors and adult Higashi H, Yokoyama K, Fujii Y, Ren S, Yuasa H, Saadat I et al. acute myelogenousleukemia. Cancer Res 64: 8816–8820. (2005). EPIYA motif isa membrane targeting signalof Helicobacter Betschlinger J, Knoblich JA. (2004). Dare to be different: asymmetric pylori virulence factor CagA in mammalian cells. J Biol Chem 280: cell division in Drosophila, C. elegans and vertebrates. Curr Biol 14: 23130–23137. R674–R685. Humbert P, Russell S, Richardson H. (2003). Dlg, Scrib and Lgl in cell Bilder D. (2004). Epithelial polarity and proliferation control: linksfrom polarity, cell proliferation and cancer. Bioessays 25: 542–553. the Drosophila neoplastic suppressors. Genes Dev 18: 1909–1925. Hurov JB, WatkinsJL, Piwnica-WormsH. (2004). Atypical PKC Blaser MJ, Perez-Perez GI, Kleanthous H, Cover TL, Peek RM, phosphorylates PAR-1 kinases to regulate localization and activity. Chyou PH et al. (1995). Infection with Helicobacter pylori strains Curr Biol 14: 736–741. possessing cagA is associated with an increased risk of developing KemphuesKJ, PriessJR,Morton DG, Cheng NS. (1988). Identiﬁca- adenocarcinoma of the stomach. Cancer Res 55: 2111–2115. tion of genesrequired for cytoplasmiclocalization in early C. Brumby AM, Richardson HE. (2003). scribble mutantscooperate with elegans embryos. Cell 52: 311–320. oncogenic Ras or Notch to cause neoplastic overgrowth in Kiyono T, Hiraiwa A, Fujita M, Hayashi Y, Akiyama T, Ishibashi M. Drosophila. EMBO J 22: 5769–5779. (1997). Binding of high-risk human palillomavirus E6 oncoproteins Cavatorta AL, Fumero G, Chouhy D, Aguirre R, Nocito AL, Giri AA to the human homologue of the Drosophila discs large tumor et al. (2004). Differential expression of the human homologue of suppressor protein. Proc Natl Acad Sci USA 94: 11612–11616. Drosophila discs large oncosuppressor in histologic samples from Kuipers EJ, Perez-Perez GI, Meuwissen SG, Blaser MJ. (1995). human papillomavirus-associated lesions as a marker for progres- Helicobacter pylori and atrophic gastritis. importance of the cagA sion to malignancy. Int J Cancer 111: 373–380. status. J Natl Cancer Inst 87: 1777–1780. Censini S, Lange C, Xiang Z, Crabtree JE, Ghiara P, Borodovsky M Kuphal S, Wallner S, Schimanski CC, Batalle F, Hofer P, Strand S et al. (1996). cag, a pathogenicity island of Helicobacter pylori, et al. (2006). Expression of Hugl-1 is strongly reduced in malignant encodes type I-speciﬁc and disease-associated virulence factors. melanoma. Oncogene 25: 103–110. Proc Natl Acad Sci USA 93: 14648–14653. Kurashima Y, Murata-Kamiya N, Kikuchi K, Higashi H, Azuma T, Correa P. (1992). Human gastric pathogenesis: a multistep and Kondo S et al. (2008). Deregulation of Wnt/b-catenin signaling by multifactorial process—First American cancer Society Award Helicobacter pylori CagA requiresthe CagA-multimerization Lecture on Cancer Epidemiology and Prevention. Cancer Res 52: sequence. Int J Cancer 122: 823–831. 6735–6740. Kwok T, Urman S, Rohde M, Hartig M, Wessler S, Misselwitz R et al. Covacci A, Censini S, Bugnoli M, Petracca R, Burroni D, Macchia G (2007). Helicobacter exploitsintegrin for type IV secretionand et al. (1993). Molecular characterization of the 128-kDa kinase activation. Nature 449: 862–866. immunodominant antigen of Helicobacter pylori associated Lee SS, Weiss RS, Javier RT. (1997). Binding of human oncoproteins with cytotoxicity and duodenal ulcer. Proc Natl Acad Sci USA 90: to hDlg/SAP97, a mammalian homolog of the Drosophila discs 5791–5795. large tumor suppressor protein. Proc Natl Acad Sci USA 94: DrewesG, Ebneth A, PreussU,Mandelkow EM, Mandelkow E. 6670–6675. (1997). MARK, a novel family of protein kinases that phosphor- Lizcano JM, Goransson O, Toth R, Deak M, Morrice NA, Boudeau J ylate microtubule-associated proteins and trigger microtubule et al. (2004). LKB1 is a master kinase that activates 13 kinases of the disruption. Cell 89: 297–308. AMPK subfamily, including MARK/PAR-1. EMBO J 25: 833–843. Eder A, Sui X, Rosen DG, Nolden LK, Cheng KW, Lahad JP et al. Macara IG. (2004). Parsing the polarity code. Nat Rev Mol Biol 5: (2005). Atypical PKCi contributesto poor prognosisthrough the 220–231.

Oncogene Linking epithelial polarity and carcinogenesis by H. pylori CagA M Hatakeyama 7054 Matsumoto Y, Marusawa H, Kinoshita K, Endo Y, Kou T, Morisawa Regala RP, Weems C, Jamieson L, Copeland JA, Thompson EA, T et al. (2007). Helicobacter pylori infection triggersaberrant FieldsAP. (2005). Atypical protein kinaseC i isan oncogene in expression of activation-induced cytidine deaminase in gastric human non-small cell lung cancer. Cancer Res 65: 8905–8911. epithelium. Nat Med 13: 470–476. Ren S, Higashi H, Lu H, Azuma T, Hatakeyama M. (2006). Structural Mohi MG, Neel BG. (2007). The role of Shp2 (PTPN11) in cancer. basis and functional consequence of Helicobacter pylori CagA Curr Opin Genet Dev 17: 23–30. multimerization in cells. J Biol Chem 281: 32344–32352. Mohi MG, WilliamsIR, Dearolf CR, Chan G, Kutok JL, Cohen S Rieder G, Merchant JL, HaasR. (2005). Helicobacter pylori cag-Type et al. (2005). Prognostic, therapeutic, and mechanistic implications IV secretion system facilitates corpus colonization to induce of a mouse model of leukemia evoked by Shp2 (PTPN11) precancerousconditionsin Mongolian gerbils. Gastroenterology mutations. Cancer Cell 7: 179–191. 128: 1229–1242. Muller J, Ory S, Copeland T, Piwnica-WormsH, MorrisonDK. Saadat I, Higashi H, Obuse C, Umeda M, Murata-Kamiya N, Saito Y (2001). C-TAK1 regulates Ras signaling by phosphorylating the et al. (2007). Helicobacter pylori CagA targetsPAR1/MARK kinase MAPK scaffold, KSR1. Mol Cell 8: 983–993. to disrupt epithelial cell polarity. Nature 447: 330–333. Murata-Kamiya N, Kurashima Y, Teishikata Y, Yamahashi Y, Saito Schimanski CC, Schmitz G, Kashyap A, Bosserhoff AK, Bataille F, Y, Higashi H et al. (2007). Helicobacter pylori CagA interactswith Schafer SC et al. (2005). Reduced expression of Hugl-1, the human E-cadherin and deregulatesthe b-catenin signal that promotes homologue of Drosophila tumor suppressor gene lgl, contributesto intestinal transdifferentiation in gastric epithelial cells. Oncogene 26: progression of colorectal cancer. Oncogene 24: 3100–3109. 4617–4626. Segal ED, Cha J, Lo J, Falkow S, Tompkins LS. (1999). Altered states: Murray NR, Jamieson L, Yu W, Zhang J, Gokmen-Polar Y, Sier D involvement of phosphorylated CagA in the induction of host et al. (2004). Protein kinase Ci isrequired for Rastransformation cellular growth changesby Helicobacter pylori. Proc Natl Acad Sci and colon carcinogenesis in vivo. J Cell Biol 164: 797–802. USA 96: 14559–14564. Naito M, Yamazaki T, Tsutsumi R, Higashi H, Onoe K, Yamazaki S Segditsas S, Tomlinson I. (2006). Colorectal cancer and genetic et al. (2006). Inﬂuence of EPIYA-repeat polymorphism on the alterationsin the Wnt pathway. Oncogene 25: 7531–7537. phosphorylation-dependent biological activity of Helicobacter pylori Selbach M, Moese S, Hauck CR, Meyer TF, Backert S. (2002). Src is CagA. Gastroenterology 130: 1181–1190. the kinase of the Helicobacter pylori CagA protein in vitro and Nakagawa S, Huibregtse JM. (2000). Human scribble (Vartul) is in vivo. J Biol Chem 277: 6775–6778. targeted for ubiquitin-mediated degradation by the high-risk Stein M, Bagnoli F, Halenbeck R, Rappuoli R, Fantl WJ, Covacci A. papillomavirusE6 proteinsand the E6AP ubiquitin ligase. Mol (2002). c-Src/Lyn kinases activate Helicobacter pylori CagA through Cell Biol 20: 8244–8253. tyrosine phosphorylation of the EPIYA motifs. Mol Microbiol 43: Nakagawa S, Yano T, Nakagawa K, Takizawa S, Suzuki Y, Yasugi T 971–980. et al. (2004). Analysis of the expression and localisation of a LAP Stein M, Rappuoli R, Covacci A. (2000). Tyrosine phosphorylation of protein, human scribble, in the normal and neoplastic epithelium of the Helicobacter pylori CagA antigen after cag-driven host cell uterine cervix. Br J Cancer 90: 194–199. translocation. Proc Natl Acad Sci USA 97: 1263–1268. Neel BG, Gu H, Pao L. (2003). The ‘Shp’ing news: SH2 domain- Suzuki A, Hirata M, Kamimura K, Maniwa R, Yamanaka T, Mizuno containing tyrosine phosphatases in cell signaling. Trends Biochem K et al. (2004). aPKC actsupstreamof PAR-1b in both the Sci 28: 284–293. establishment and maintenance of mammalian epithelial polarity. Nomura A, Stemmermann GN, Chyou PH, Kato I, Perez-Perez GI, Curr Biol 14: 1425–1435. Blaser MJ. (1991). Helicobacter pylori infection and gastric carcinoma Suzuki A, Ohno S. (2006). The PAR-aPKC system: lessons in polarity. among Japanese Americans in Hawaii. N Engl J Med 325: 1132–1136. J Cell Sci 119: 979–987. Odenbreit S, PulsJ, Sedlmaier B, Gerland E, FischerW, HaasR. Suzuki T, Ohsugi Y, Uchida-Toita M, Akiyama T, Yoshida M. (1999). (2000). Translocation of Helicobacter pylori CagA into gastric Tax oncoprotein of HTLV-1 bindsto the human homologue of epithelial cellsby type IV secretion. Science 287: 1497–1500. Drosophila discs large tumor suppressor protein hDLG, and perturbs Ohnishi N, Yuasa H, Tanaka S, Sawa H, Miura M, Matsui A et al. itsfunction in cell growth control. Oncogene 18: 5967–6972. (2008). Transgenic expression of Helicobacter pylori CagA induces Tammer I, Brandt S, Hartig R, Konig W, Backert S. (2007). Activation gastrointestinal and hematopoietic neoplasms in mouse. Proc Natl of Abl by Helicobacter pylori: A novel kinase for CagA and crucial Acad Sci USA 105: 1003–1008. mediator of host cell scattering. Gastroenterology 132: 1309–1319. Pagliarini RA, Xu T. (2003). A genetic screen in Drosophila for Tartaglia M, Mehler EL, Goldberg R, Zampino G, Brunner HG, metastatic behavior. Science 302: 1227–1231. Kremer H et al. (2001). Mutationsin PTPN11, encoding the protein Panneerselvam S, Marx A, Mandelkow EM, Mandelkow E. (2006). tyrosine phosphatase SHP-2, cause Noonan syndrome. Nat Genet Structure of the catalytic and ubiquitin-associated domains of the 111: 421–427. protein kinase MARK/Par-1. Structure 14: 173–183. Tartaglia M, Niemeyer CM, Fragale A, Song X, Buechner J, Jung A Parkin DM, Bray FI, Devesa SS. (2001). Cancer burden in the year et al. (2003). Somatic mutationsin PTPN11 in juvenile myelomo- 2000. The global picture. Eur J Cancer 37: S4–S66. nocytic leukemia, myelodysplastic syndromes and acute myeloid Parsonnet J, Friedman GD, Orentreich N, Vogelman H. (1997). Risk leukemia. Nat Genet 34: 148–150. for gastric cancer in people with CagA positive or CagA negative Tepass U, Tanentzapf G, Ward R, Fehon R. (2001). Epithelial cell Helicobacter pylori infection. Gut 40: 297–301. polarity and cell junctionsin Drosophila. Annu Rev Genet 35: 747–781. Parsonnet J, Friedman GD, Vandersteen DP, Chang Y, Vogelman JH, Tsutsumi R, Takahashi A, Azuma T, Higashi H, Hatakeyama M. (2006). Orentreich N et al. (1991). Helicobacter pylori and the risk of gastric FAK is a substrate and downstream effector of SHP-2 complexed with carcinoma. N Engl J Med 325: 1127–1131. Helicobacter pylori CagA. MolCellBiol26: 261–276. Peek Jr RM, Blaser MJ. (2002). Helicobacter pylori and gastrointest- Tummuru MK, Cover TL, Blaser MJ. (1993). Cloning and expression of inal tract adenocarcinomas. Nat Rev Cancer 2: 28–37. a high-molecular-mass major antigen of Helicobacter pylori: evidence Peng CY, GravesPR, Ogg S, Thoma RS, ByrnesIII MJ, Wu Z et al. of linkage to Cytotoxin production. Infect Immun 61: 1799–1809. (1998). C-TAK1 protein kinase phosphorylates human Cdc25C on Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S, serine 216 and promotes 14-3-3 protein binding. Cell Growth Differ Yamakido M et al. (2001). Helicobacter pylori infection and the 9: 197–208. development of gastric cancer. N Engl J Med 345: 784–789. Poppe M, Feller SM, Romer G, Wessler S. (2007). Phosphorylation of Zeaiter Z, Cohen D, Musch A, Bagnoli F, Covacci A, Stein M. (2008). Helicobacter pylori CagA by c-Abl leadsto cell motility. Oncogene Analysis of detergent-resistant membranes of Helicobacter pylori 26: 3462–3472. infected gastric adenocarcinoma cells reveals a role for MARK2/ Prieur A, Peeper DS. (2008). Cellular senescence in vivo: a barrier to Par1b in CagA-mediated disruption of cellular polarity. Cell tumorigenesis. Curr Opin Cell Biol 20: 150–155. Microbiol 10: 781–794.

Oncogene