Identification of Tumor-associated Autoantigens for the Diagnosis of Colorectal Cancer in Serum Using High Density Protein Microarrays

There is a mounting evidence of the existence of autoantibodies associated to cancer progression. Antibodies are the target of choice for serum screening because of their stability and suitability for sensitive immunoassays. By using commercial protein microarrays containing 8000 human proteins, we examined 20 sera from colorectal cancer (CRC) patients and healthy subjects to identify autoantibody patterns and associated antigens. Forty-three proteins were differentially recognized by tumoral and reference sera (p value <0.04) in the protein microarrays. Five immunoreactive antigens, PIM1, MAPKAPK3, STK4, SRC, and FGFR4, showed the highest prevalence in cancer samples, whereas ACVR2B was more abundant in normal sera. Three of them, PIM1, MAPKAPK3, and ACVR2B, were used for further validation. A significant increase in the expression level of these antigens on CRC cell lines and colonic mucosa was confirmed by immunoblotting and immunohistochemistry on tissue microarrays. A diagnostic ELISA based on the combination of MAPKAPK3 and ACVR2B proteins yielded specificity and sensitivity values of 73.9 and 83.3% (area under the curve, 0.85), respectively, for CRC discrimination after using an independent sample set containing 94 sera representative of different stages of progression and control subjects. In summary, these studies confirmed the presence of specific autoantibodies for CRC and revealed new individual markers of disease (PIM1, MAPKAPK3, and ACVR2B) with the potential to diagnose CRC with higher specificity and sensitivity than previously reported serum biomarkers.Colorectal cancer (CRC)¹ is the second most prevalent cancer in the western world. The development of this disease takes decades and involves multiple genetic events. CRC remains a major cause of mortality in developed countries because most of the patients are diagnosed at advanced stages because of the reluctance to use highly invasive diagnostic tools like colonoscopy. Actually only a few proteins have been described as biomarkers in CRC (carcinoembryonic antigen (CEA), CA19.9, and CA125 (1–3)), although none of them is recommended for clinical screening (4). Proteomics analysis is actively used for the identification of new biomarkers. In previous studies, the use of two-dimensional DIGE and antibody microarrays allowed the identification of differentially expressed proteins in CRC tissue, including isoforms and post-translational modifications responsible for modifications in signaling pathways (5–8). Both approaches resulted in the identification of a collection of potential tumoral tissue biomarkers that is currently being investigated.However, the implementation of simpler, non-invasive methods for the early detection of CRC should be based on the identification of proteins or antibodies in serum or plasma (9–13). There is ample evidence of the existence of an immune response to cancer in humans as demonstrated by the presence of autoantibodies in cancer sera. Self-proteins (autoantigens) altered before or during tumor formation can elicit an immune response (13–19). These tumor-specific autoantibodies can be detected at early cancer stages and prior to cancer diagnosis revealing a great potential as biomarkers (14, 15, 20). Tumor proteins can be affected by specific point mutations, misfolding, overexpression, aberrant glycosylation, truncation, or aberrant degradation (e.g. p53, HER2, NY-ESO1, or MUC1 (16, 21–25)). In fact, a number of tumor-associated autoantigens (TAAs) were identified previously in different studies involving autoantibody screening in CRC (26–28).Several approaches have been used to identify TAAs in cancer, including natural protein arrays prepared with fractions obtained from two-dimensional LC separations of tumoral samples (29, 30) or protein extracts from cancer cells or tissue (9, 31) probed by Western blot with patient sera, cancer tissue peptide libraries expressed as cDNA expression libraries for serological screening (serological analysis of recombinant cDNA expression libraries (SEREX)) (22, 32), or peptides expressed on the surface of phages in combination with microarrays (17, 18, 33, 34). However, these approaches suffer from several drawbacks. In some cases TAAs have to be isolated and identified from the reactive protein lysate by LC-MS techniques, or in the phage display approach, the reactive TAA could be a mimotope without a corresponding linear amino acid sequence. Moreover, cDNA libraries might not be representative of the protein expression levels in tumors as there is a poor correspondence between mRNA and protein levels.Protein arrays provide a novel platform for the identification of both autoantibodies and their respective TAAs for diagnostic purposes in cancer serum patients. They present some advantages. Proteins printed on the microarray are known “a priori,” avoiding the need for later identifications and the discovery of mimotopes. There is no bias in protein selection as the proteins are printed at a similar concentration. This should result in a higher sensitivity for biomarker identification (13, 35, 36).In this study, we used commercially available high density protein microarrays for the identification of autoantibody signatures and tumor-associated antigens in colorectal cancer. We screened 20 CRC patient and control sera with protein microarrays containing 8000 human proteins to identify the CRC-associated autoantibody repertoire and the corresponding TAAs. Autoantibody profiles that discriminated the different types of CRC metastasis were identified. Moreover a set of TAAs showing increased or decreased expression in cancer cell lines and paired tumoral tissues was found. Finally an ELISA was set up to test the ability of the most immunoreactive proteins to detect colorectal adenocarcinoma. On the basis of the antibody response, combinations of three antigens, PIM1, MAPKAPK3, and ACVR2B, showed a great potential for diagnosis.     

Stable Isotope Labeling in Zebrafish Allows in Vivo Monitoring of Cardiac Morphogenesis

Quantitative proteomics is an important tool to study biological processes, but so far it has been challenging to apply to zebrafish. Here, we describe a large scale quantitative analysis of the zebrafish proteome using a combination of stable isotope labeling and liquid chromatography-mass spectrometry (LC-MS). Proteins derived from the fully labeled fish were used as a standard to quantify changes during embryonic heart development. LC-MS-assisted analysis of the proteome of activated leukocyte cell adhesion molecule zebrafish morphants revealed a down-regulation of components of the network required for cell adhesion and maintenance of cell shape as well as secondary changes due to arrest of cellular differentiation. Quantitative proteomics in zebrafish using the stable isotope-labeling technique provides an unprecedented resource to study developmental processes in zebrafish.Over the past years, mass spectrometry-based proteomics has been widely used to analyze complex biological samples (1). Although the latest generation of MS instrumentation allows proteome-wide analysis, protein quantitation is still a challenge (2, 3). Metabolic labeling using stable isotopes has been used for almost a century. Today, the most commonly used techniques for relative protein quantification are based on ¹⁵N labeling and stable isotope labeling by amino acids in cell culture (SILAC)¹ (4, 5). SILAC was initially developed for cell culture experiments, and recent approaches extended labeling to living organisms, including bacteria (6), yeast (7), flies (8), worms (9), and rodents (10, 11). In addition, several pulsed SILAC (also known as dynamic SILAC) experiments were performed to assess protein dynamics in cell culture and living animals (12–15).The zebrafish (Danio rerio) has proved to be an important model organism to study developmental processes. It also serves as a valuable tool to investigate basic pathogenic principles of human diseases such as cardiovascular disorders and tissue regeneration (16). So far, most researchers rely on immunohistochemistry and Western blots for semi-quantitative protein analysis, an approach that is hampered by the paucity of reliable antibodies in zebrafish. Proteomics approaches that depend on two-dimensional gel approaches (17–19) have not gained wide popularity because of issues with workload, reproducibility, and sensitivity (20, 21).Another approach for protein quantitation is the chemical modification of peptides, and so far several isobaric tagging methods, including ICAT (22), iTRAQ (23), ¹⁸O (24), and dimethyl labeling (25), have been proven to be successful methods.Recently, a quantitative phosphopeptide study based on dimethyl labeling in zebrafish showed the consequences of a morpholino-based kinase knockdown (26). However, each chemical modification bears the risk of nonspecific and incomplete labeling, which complicates mass spectrometric data interpretation.Alternatively, a metabolic labeling study with stable isotopes was recently performed on adult zebrafish by the administration of a mouse diet containing [¹³C₆]lysine (Lys-6) (27). Feeding adult zebrafish with the Lys-6-containing mouse chow leads to an incorporation rate of 40%, and SILAC labeling was used to investigate protein and tissue turnover.Here, we have developed a SILAC fish diet made in-house for the complete SILAC labeling of zebrafish. We established a Lys-6-containing diet as a universal fish food for larval and adult zebrafish. The method allows accurate quantitation of large numbers of proteins, and we proved our approach by the analysis of embryonic heart development. In addition, we investigated the consequences of the morpholino-based activated leukocyte cell adhesion molecule (ALCAM) knockdown during development and identified the lipid anchor protein Paralemmin as a down-regulated protein during heart development. Our approach yielded a huge resource of protein expression data for zebrafish development and provided the basis for more refined studies depending on accurate SILAC protein quantification.     

A Genetic Engineering Solution to the ��Arginine Conversion Problem�� in Stable Isotope Labeling by Amino Acids in Cell Culture (SILAC)

Stable isotope labeling by amino acids in cell culture (SILAC) provides a straightforward tool for quantitation in proteomics. However, one problem associated with SILAC is the in vivo conversion of labeled arginine to other amino acids, typically proline. We found that arginine conversion in the fission yeast Schizosaccharomyces pombe occurred at extremely high levels, such that labeling cells with heavy arginine led to undesired incorporation of label into essentially all of the proline pool as well as a substantial portion of glutamate, glutamine, and lysine pools. We found that this can be prevented by deleting genes involved in arginine catabolism using methods that are highly robust yet simple to implement. Deletion of both fission yeast arginase genes or of the single ornithine transaminase gene, together with a small modification to growth medium that improves arginine uptake in mutant strains, was sufficient to abolish essentially all arginine conversion. We demonstrated the usefulness of our approach in a large scale quantitative analysis of proteins before and after cell division; both up- and down-regulated proteins, including a novel protein involved in septation, were successfully identified. This strategy for addressing the “arginine conversion problem” may be more broadly applicable to organisms amenable to genetic manipulation.Stable isotope labeling by amino acids in cell culture (SILAC)¹ (1) is one of the key methods for large scale quantitative proteomics (2, 3). In SILAC experiments, proteins are metabolically labeled by culturing cells in media containing either normal (“light”) or heavy isotope-labeled amino acids, typically lysine and arginine. Peptides derived from the light and heavy cells are thus distinguishable by mass spectrometry and can be mixed for accurate quantitation. SILAC is also possible at the whole-organism level (4).An inherent problem in SILAC is the metabolic conversion of labeled arginine to other amino acids, as this complicates quantitative analysis of peptides containing these amino acids. Arginine conversion to proline is well described in mammalian cells, although the extent of conversion varies among cell types (5). When conversion is observed, typically 10–25% of the total proline pool is found to contain label (6–11). Arginine conversion has also been reported in SILAC experiments with budding yeast Saccharomyces cerevisiae (3, 12, 13).Because more than 50% of tryptic peptides in large data sets contain proline (7), it is not practical simply to disregard proline-containing peptides during quantitation. Several methods have been proposed to either reduce arginine conversion or correct for its effects on quantitation. In some cell types, arginine conversion can be prevented by lowering the concentration of exogenous arginine (6, 14–16) or by adding exogenous proline (9). However, these methods can involve significant changes to growth media and may need to be tested for each experimental condition used. Given the importance of arginine in many metabolic pathways, careful empirical titration of exogenous arginine concentration is required to minimize negative effects on cell growth (14). In addition, low arginine medium can lead to incomplete arginine labeling, although the reasons for this are not entirely clear (7). An alternative strategy is to omit labeled arginine altogether (3, 13, 17), but this reduces the number of quantifiable peptides. Correction methods include using two different forms of labeled arginine (7) or computationally compensating for proline-containing peptides (11, 12, 18). Ultimately, none of these methods address the problem at its root, the utilization of arginine in cellular metabolism.To develop a differential proteomics work flow for the fission yeast Schizosaccharomyces pombe, we sought to adapt SILAC for use in this organism, a widely used model eukaryote with excellent classical and reverse genetics. Here we describe extremely high conversion of labeled arginine to other amino acids in fission yeast as well as a novel general solution to the problem that should be applicable to other organisms. As proof of principle, we quantitated changes in protein levels before and after cell division on a proteome-wide scale. We identified both up- and down-regulated proteins, including a novel protein involved in septation.     

The SILAC Fly Allows for Accurate Protein Quantification in Vivo

Stable isotope labeling by amino acids in cell culture (SILAC) is widely used to quantify protein abundance in tissue culture cells. Until now, the only multicellular organism completely labeled at the amino acid level was the laboratory mouse. The fruit fly Drosophila melanogaster is one of the most widely used small animal models in biology. Here, we show that feeding flies with SILAC-labeled yeast leads to almost complete labeling in the first filial generation. We used these “SILAC flies” to investigate sexual dimorphism of protein abundance in D. melanogaster. Quantitative proteome comparison of adult male and female flies revealed distinct biological processes specific for each sex. Using a tudor mutant that is defective for germ cell generation allowed us to differentiate between sex-specific protein expression in the germ line and somatic tissue. We identified many proteins with known sex-specific expression bias. In addition, several new proteins with a potential role in sexual dimorphism were identified. Collectively, our data show that the SILAC fly can be used to accurately quantify protein abundance in vivo. The approach is simple, fast, and cost-effective, making SILAC flies an attractive model system for the emerging field of in vivo quantitative proteomics.Mass spectrometry-based quantitative proteomics has emerged as a highly successful approach to study biological processes in health and disease (1–3). Most studies have so far been limited to in vitro systems such as cell culture models. Although tremendously useful, these models cannot appropriately reflect relevant regulatory mechanisms of multicellular eukaryotes in vivo. This is particularly relevant for complex processes involving interactions between different cell types such as differentiation and development (4).Relative changes in protein abundance are most accurately measured by comparing the natural form of a peptide with its stable isotope-labeled analog. Several different approaches enable stable isotope labeling of peptides either by chemical reactions or metabolic incorporation of the label (5, 6). Metabolic labeling has several advantages such as high labeling efficiency and intrinsically higher precision. For example, metabolically labeled samples can be combined before further processing steps so that protein quantification is not affected by differences in sample preparation. Labeling of organisms with stable isotope tracers was pioneered by Rudolf Schoenheimer 75 years ago (7, 8). Since then, several model organisms ranging from prokaryotes to mammals have been labeled metabolically (for an excellent review, see Ref. 9). For example, Caenorhabditis elegans and Drosophila melanogaster have successfully been labeled with ¹⁵N (10), and ¹⁵N-labeled flies were recently used to study maternal-to-zygotic transition (11) and seminal fluid proteins (sfps)¹ transferred at mating (12). ¹⁵N has also been used to label entire rats, particularly for quantitative brain proteomics (13, 14). Despite its usefulness, ¹⁵N labeling also has several disadvantages. Because most peptides contain dozens of nitrogen atoms, labeling with highly enriched ¹⁵N still results in only partial peptide labeling and therefore complex isotope clusters. In addition, the mass shift between the labeled (i.e. heavy) and unlabeled (i.e. light) forms of a peptide depends on the number of nitrogen atoms and therefore varies depending on the peptide sequence. This leads to an increase in the number of candidate masses that need to be considered and therefore complicates peptide identification by search algorithms. Both problems result in smaller identification rates and less accurate quantification that can partially be overcome by computational correction (15, 16).Stable isotope labeling by amino acids in cell culture (SILAC) is another metabolic labeling approach with several unique advantages (17): because the label is introduced at the amino acid level, mass spectra can easily be interpreted, and peptides can be quantified with high precision. These features have made SILAC a very popular approach for cell culture-based quantitative and functional proteomics (18). As a potential disadvantage, SILAC is generally thought to be restricted to in vitro cell culture experiments. The only SILAC experiments in the fly model were carried out using cell lines cultivated in vitro (19, 20). However, in 2005, Hayter et al. (21) demonstrated that chicken can be partially labeled at the amino acid level by feeding them with a diet containing stable isotope-labeled valine. Three years later, Krüger et al. (22) achieved essentially complete labeling of the laboratory mouse. Until now, this so-called “SILAC mouse” was the only multicellular organism that has been completely labeled with the SILAC approach, and partial labeling was recently achieved in newts (21, 23).Here, we introduce the fruit fly D. melanogaster in the SILAC zoo. We refer to these animals as SILAC flies because they are obtained by feeding flies on SILAC-labeled yeast. D. melanogaster is one of the best characterized model organisms and has been used to address many fundamental questions in biology (24). Until now, most studies in D. melanogaster have focused on genetic aspects (25). However, proteins are the key actors in most biological processes. It is therefore highly desirable to obtain quantitative information at the protein level in D. melanogaster. We demonstrate in the present study that raising fly larvae on a diet of heavy lysine-labeled yeast cells results in virtually complete heavy labeling in the first filial (F₁) generation. Furthermore, we show that the SILAC fly enables proteome-wide quantification with higher precision than a label-free method. In a series of proof-of-principle experiments, we used the SILAC fly to investigate sexually dimorphic protein expression in D. melanogaster, thus providing the first systematic comparison of male and female flies at the protein level.     

Analysis of Free Energy Signals Arising from Nucleotide Hybridization Between rRNA and mRNA Sequences during Translation in Eubacteria

A decoding algorithm is tested that mechanistically models the progressive alignments that arise as the mRNA moves past the rRNA tail during translation elongation. Each of these alignments provides an opportunity for hybridization between the single-stranded, -terminal nucleotides of the 16S rRNA and the spatially accessible window of mRNA sequence, from which a free energy value can be calculated. Using this algorithm we show that a periodic, energetic pattern of frequency 1/3 is revealed. This periodic signal exists in the majority of coding regions of eubacterial genes, but not in the non-coding regions encoding the 16S and 23S rRNAs. Signal analysis reveals that the population of coding regions of each bacterial species has a mean phase that is correlated in a statistically significant way with species () content. These results suggest that the periodic signal could function as a synchronization signal for the maintenance of reading frame and that codon usage provides a mechanism for manipulation of signal phase.[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32]     

Refined Preparation and Use of Anti-diglycine Remnant (K-ε-GG) Antibody Enables Routine Quantification of 10,000s of Ubiquitination Sites in Single Proteomics Experiments

Detection of endogenous ubiquitination sites by mass spectrometry has dramatically improved with the commercialization of anti-di-glycine remnant (K-ε-GG) antibodies. Here, we describe a number of improvements to the K-ε-GG enrichment workflow, including optimized antibody and peptide input requirements, antibody cross-linking, and improved off-line fractionation prior to enrichment. This refined and practical workflow enables routine identification and quantification of ∼20,000 distinct endogenous ubiquitination sites in a single SILAC experiment using moderate amounts of protein input.The commercialization of antibodies that recognize lysine residues modified with a di-glycine remnant (K-ε-GG)¹ has significantly transformed the detection of endogenous protein ubiquitination sites by mass spectrometry (1–5). Prior to the development of these highly specific reagents, proteomics experiments were limited to identification of up to only several hundred ubiquitination sites, which severely limited the scope of global ubiquitination studies (6). Recent proteomic studies employing anti-K-ε-GG antibodies have enhanced our understanding of ubiquitin biology through the identification of thousands of ubiquitination sites and the analysis of the change in relative abundance of these sites after chemical or biological perturbation (1–3, 5, 7). Use of stable isotope labeling by amino acids in cell culture (SILAC) for quantification has enabled researchers to better understand the extent of ubiquitin regulation upon proteasome inhibition and precisely identify those protein classes, such as newly synthesized proteins or chromatin-related proteins, that see overt changes in their ubiquitination levels upon drug treatment (2, 3, 5). Emanuel et al. (1) have combined genetic and proteomics assays implementing the anti-K-ε-GG antibody to identify hundreds of known and putative Cullin-RING ligase substrates, which has clearly demonstrated the extensive role of Cullin-RING ligase ubiquitination on cellular protein regulation.Despite the successes recently achieved with the use of the anti-K-ε-GG antibody, increased sample input (up to ∼35 mg) and/or the completion of numerous experimental replicates have been necessary to achieve large numbers of K-ε-GG sites (>5,000) in a single SILAC-based experiment (1–3, 5). For example, it has been recently shown that detection of more than 20,000 unique ubiquitination sites is possible from the analysis of five different murine tissues (8). However, as the authors indicate, only a few thousands sites are detected in any single analysis of an individual tissue sample (8). It is recognized that there is need for further improvements in global ubiquitin technology to increase the depth-of-coverage attainable in quantitative proteomic experiments using moderate amounts of protein input (9). Through systematic study and optimization of key pre-analytical variables in the preparation and use of the anti-K-ε-GG antibody as well as the proteomic workflow, we have now achieved, for the first time, routine quantification of ∼20,000 nonredundant K-ε-GG sites in a single SILAC triple encoded experiment starting with 5 mg of protein per SILAC channel. This represents a 10-fold improvement over our previously published method (3).     

Splitting the BLOSUM Score into Numbers of Biological Significance

Mathematical tools developed in the context of Shannon information theory were used to analyze the meaning of the BLOSUM score, which was split into three components termed as the BLOSUM spectrum (or BLOSpectrum). These relate respectively to the sequence convergence (the stochastic similarity of the two protein sequences), to the background frequency divergence (typicality of the amino acid probability distribution in each sequence), and to the target frequency divergence (compliance of the amino acid variations between the two sequences to the protein model implicit in the BLOCKS database). This treatment sharpens the protein sequence comparison, providing a rationale for the biological significance of the obtained score, and helps to identify weakly related sequences. Moreover, the BLOSpectrum can guide the choice of the most appropriate scoring matrix, tailoring it to the evolutionary divergence associated with the two sequences, or indicate if a compositionally adjusted matrix could perform better.[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29]     

Algorithms for Finding Small Attractors in Boolean Networks

A Boolean network is a model used to study the interactions between different genes in genetic regulatory networks. In this paper, we present several algorithms using gene ordering and feedback vertex sets to identify singleton attractors and small attractors in Boolean networks. We analyze the average case time complexities of some of the proposed algorithms. For instance, it is shown that the outdegree-based ordering algorithm for finding singleton attractors works in time for , which is much faster than the naive time algorithm, where is the number of genes and is the maximum indegree. We performed extensive computational experiments on these algorithms, which resulted in good agreement with theoretical results. In contrast, we give a simple and complete proof for showing that finding an attractor with the shortest period is NP-hard.[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32]     

Role of Partitioning-defective 1/Microtubule Affinity-regulating Kinases in the Morphogenetic Activity of Helicobacter pylori CagA

Helicobacter pylori CagA plays a key role in gastric carcinogenesis. Upon delivery into gastric epithelial cells, CagA binds and deregulates SHP-2 phosphatase, a bona fide oncoprotein, thereby causing sustained ERK activation and impaired focal adhesions. CagA also binds and inhibits PAR1b/MARK2, one of the four members of the PAR1 family of kinases, to elicit epithelial polarity defect. In nonpolarized gastric epithelial cells, CagA induces the hummingbird phenotype, an extremely elongated cell shape characterized by a rear retraction defect. This morphological change is dependent on CagA-deregulated SHP-2 and is thus thought to reflect the oncogenic potential of CagA. In this study, we investigated the role of the PAR1 family of kinases in the hummingbird phenotype. We found that CagA binds not only PAR1b but also other PAR1 isoforms, with order of strength as follows: PAR1b > PAR1d ≥ PAR1a > PAR1c. Binding of CagA with PAR1 isoforms inhibits the kinase activity. This abolishes the ability of PAR1 to destabilize microtubules and thereby promotes disassembly of focal adhesions, which contributes to the hummingbird phenotype. Consistently, PAR1 knockdown potentiates induction of the hummingbird phenotype by CagA. The morphogenetic activity of CagA was also found to be augmented through inhibition of non-muscle myosin II. Because myosin II is functionally associated with PAR1, perturbation of PAR1-regulated myosin II by CagA may underlie the defect of rear retraction in the hummingbird phenotype. Our findings reveal that CagA systemically inhibits PAR1 family kinases and indicate that malfunctioning of microtubules and myosin II by CagA-mediated PAR1 inhibition cooperates with deregulated SHP-2 in the morphogenetic activity of CagA.Infection with Helicobacter pylori strains bearing cagA (cytotoxin-associated gene A)-positive strains is the strongest risk factor for the development of gastric carcinoma, the second leading cause of cancer-related death worldwide (1–3). The cagA gene is located within a 40-kb DNA fragment, termed the cag pathogenicity island, which is specifically present in the genome of cagA-positive H. pylori strains (4–6). In addition to cagA, there are ∼30 genes in the cag pathogenicity island, many of which encode a bacterial type IV secretion system that delivers the cagA-encoded CagA protein into gastric epithelial cells (7–10). Upon delivery into gastric epithelial cells, CagA is localized to the plasma membrane, where it undergoes tyrosine phosphorylation at the C-terminal Glu-Pro-Ile-Tyr-Ala motifs by Src family kinases or c-Abl kinase (11–14). The C-terminal Glu-Pro-Ile-Tyr-Ala-containing region of CagA is noted for the structural diversity among distinct H. pylori isolates. Oncogenic potential of CagA has recently been confirmed by a study showing that systemic expression of CagA in mice induces gastrointestinal and hematological malignancies (15).When expressed in gastric epithelial cells, CagA induces morphological transformation termed the hummingbird phenotype, which is characterized by the development of one or two long and thin protrusions resembling the beak of the hummingbird. It has been thought that the hummingbird phenotype is related to the oncogenic action of CagA (7, 16–19). Pathophysiological relevance for the hummingbird phenotype in gastric carcinogenesis has recently been provided by the observation that infection with H. pylori carrying CagA with greater ability to induce the hummingbird phenotype is more closely associated with gastric carcinoma (20–23). Elevated motility of hummingbird cells (cells showing the hummingbird phenotype) may also contribute to invasion and metastasis of gastric carcinoma.In host cells, CagA interacts with the SHP-2 phosphatase, C-terminal Src kinase, and Crk adaptor in a tyrosine phosphorylation-dependent manner (16, 24, 25) and also associates with Grb2 adaptor and c-Met in a phosphorylation-independent manner (26, 27). Among these CagA targets, much attention has been focused on SHP-2 because the phosphatase has been recognized as a bona fide oncoprotein, gain-of-function mutations of which are found in various human malignancies (17, 18, 28). Stable interaction of CagA with SHP-2 requires CagA dimerization, which is mediated by a 16-amino acid CagA-multimerization (CM)² sequence present in the C-terminal region of CagA (29). Upon complex formation, CagA aberrantly activates SHP-2 and thereby elicits sustained ERK MAP kinase activation that promotes mitogenesis (30). Also, CagA-activated SHP-2 dephosphorylates and inhibits focal adhesion kinase (FAK), causing impaired focal adhesions. It has been shown previously that both aberrant ERK activation and FAK inhibition by CagA-deregulated SHP-2 are involved in induction of the hummingbird phenotype (31).Partitioning-defective 1 (PAR1)/microtubule affinity-regulating kinase (MARK) is an evolutionally conserved serine/threonine kinase originally isolated in C. elegans (32–34). Mammalian cells possess four structurally related PAR1 isoforms, PAR1a/MARK3, PAR1b/MARK2, PAR1c/MARK1, and PAR1d/MARK4 (35–37). Among these, PAR1a, PAR1b, and PAR1c are expressed in a variety of cells, whereas PAR1d is predominantly expressed in neural cells (35, 37). These PAR1 isoforms phosphorylate microtubule-associated proteins (MAPs) and thereby destabilize microtubules (35, 38), allowing asymmetric distribution of molecules that are involved in the establishment and maintenance of cell polarity.In polarized epithelial cells, CagA disrupts the tight junctions and causes loss of apical-basolateral polarity (39, 40). This CagA activity involves the interaction of CagA with PAR1b/MARK2 (19, 41). CagA directly binds to the kinase domain of PAR1b in a tyrosine phosphorylation-independent manner and inhibits the kinase activity. Notably, CagA binds to PAR1b via the CM sequence (19). Because PAR1b is present as a dimer in cells (42), CagA may passively homodimerize upon complex formation with the PAR1 dimer via the CM sequence, and this PAR1-directed CagA dimer would form a stable complex with SHP-2 through its two SH2 domains.Because of the critical role of CagA in gastric carcinogenesis (7, 16–19), it is important to elucidate the molecular basis underlying the morphogenetic activity of CagA. In this study, we investigated the role of PAR1 isoforms in induction of the hummingbird phenotype by CagA, and we obtained evidence that CagA-mediated inhibition of PAR1 kinases contributes to the development of the morphological change by perturbing microtubules and non-muscle myosin II.