Gene expression alterations during HGF-induced dedifferentiation of a renal tubu



免疫细胞治疗专区	欢迎关注干细胞微信公众号

12 3 4 5 6 7 8 9 10 ... 24 下一页

返回列表

查看: 414272\|回复: 236	go Gene expression alterations during HGF-induced dedifferentiation of a renal tubu

轻羽

新手上路

Rank: 1

积分: 0
威望: 0
包包: 0

发表于 2009-4-22 08:15 |显示全部帖子

作者：Daniel F. Balkovetz, Edward R. Gerrard, Jr., Shixiong Li, David Johnson, James Lee, John W. Tobias, Katherine K. Rogers, Richard W. Snyder, and Joshua H. Lipschutz作者单位：1 Departments of Medicine and Cell Biology, University of Alabama at Birmingham, and Veterans Administration Medical Center, and 2 Department of Surgery, University of Alabama at Birmingham, Birmingham, Alabama 35294; 3 Department of Medicine, 4 Genomics Institute, Bioinformatics Core, and 5 Departm
   . t% {  A1 H7 G- D1 E
   0 |) [, g4 P& B. Q4 U

                     / \% i* b+ p6 O, r" o% k; I
         6 e# x9 A! q9 w' S# g( _% m
         & }. l; y. C) \" T& _; S# i
         % `+ \; }: t9 ^7 I! L

                  ! p. p' z+ S( Y& u
      ! T8 Z, j3 z9 F2 e
      ' \" E4 X3 D6 ]5 i( ?4 D& [/ [2 g

      【摘要】
   Hepatocyte growth factor (HGF) elicits a broad spectrum of biological activities, including epithelial cell dedifferentiation. One of the most widely used and best-studied polarized epithelial cell lines is the Madin-Darby canine kidney (MDCK) cell line. Here, we describe and validate the early response of polarized monolayers of MDCK cells stimulated with recombinant HGF using a novel canine DNA microarray designed to query 12,473 gene sequences. In our survey, eight genes previously implicated in the HGF signaling pathway were differentially regulated, demonstrating that the system was responsive to HGF. Also identified were 117 genes not previously known to be involved in the HGF pathway. The results were confirmed by real-time PCR or Western blot analysis for 38 genes. Of particular interest were the large number of differentially regulated genes encoding small GTPases, proteins involved in endoplasmic reticulum translation, proteins involved in the cytoskeleton, the extracellular matrix, and the hematopoietic and prostaglandin systems.
      【关键词】 hepatocyte growth factor MadinDarby canine kidney cells1 X2 C9 r9 T% o9 P% b9 R2 B
   WE HAVE BEEN INTERESTED IN the mechanisms that lead to epithelial cell dedifferentiation. Epithelial cell dedifferentiation is a prerequisite for both epithelial cell transformation to carcinoma and epithelial organ regeneration. Previous work has shown that transgenic animals overexpressing HGF develop carcinomas in epithelial organs such as the liver, skin, and mammary gland as well as skeletal muscle ( 35, 37 ). Epithelial dedifferentiation is also an important phase in the process of epithelial organ regeneration after injury, and HGF has been shown to be important for the regeneration of epithelial organs such as the liver, kidney, lung, and pancreas ( 2 ).
' d. o) O1 \* l6 S* h
The Madin-Darby canine kidney (MDCK) cell line, derived from the kidney tubules of a normal cocker spaniel in 1958 ( 16 ), is one of the best-characterized and most widely used epithelial cell lines ( 38 ). In vitro, treatment of previously well-polarized MDCK cells with HGF leads to a loss of cell polarity (dedifferentiation) ( 2 ). In addition, HGF treatment of well-polarized monolayers results in detachment of E-cadherin from the actin cytoskeleton ( 3 ), loss of contact inhibition of mitosis ( 1 ), and increased susceptibility to bacterial pathogens such as Escherichia coli and Pseudomonas aeruginosa ( 7, 44 ). When MDCK cells are placed in a three-dimensional collagen matrix and allowed to form cysts, HGF-induced tubulogenesis involves the eight-protein exocyst complex and the Rho family of small GTPases ( 17, 31 ). Elucidation of the genomic pathways activated by HGF would provide important clues about the mechanisms of epithelial transformation to carcinoma and epithelial organ regeneration. Some of the signaling pathways initiated by HGF activation of the c-met receptor tyrosine kinase have been identified ( 32 ). However, the picture is not complete.
8 R8 p6 Y1 E$ y! a; X
Recent advancements in microarray technology have revolutionized genomics by allowing researchers to study the expression of thousands of genes simultaneously during highly complex processes ( 13 ). A microarray analysis of early gene expression changes induced by HGF treatment of polarized MDCK cell monolayers would provide important information regarding the signaling processes that are modulated during epithelial dedifferentiation. Until recently, the only mammalian DNA chips available for analysis have been derived from humans, mice, and rats. Because MDCK cells are derived from Canis familiaris, the genomic responses to various stimuli by MDCK cells have not previously been assayed using DNA microarray technology, as DNA chips spotted with thousands of C. familiaris gene sequences did not exist. Here, we describe and validate the response of MDCK cells grown on Transwell filters for 4 days and stimulated with recombinant HGF for 0, 3, or 24 h using a novel canine DNA microarray designed to query 12,473 gene sequences from C. familiaris.8 o6 o' j. W$ h9 z2 X
: J- y0 M: Y& F$ V- F) R. r
This newly developed canine DNA microarray chip should be extremely useful not only for canine research but also for the many epithelial cell biologists who utilize the MDCK epithelial cell model (a PubMed search using "MDCK" as a keyword identified 3,384 published manuscripts). For example, Boccaccio and colleagues ( 4 ) describe multiple signaling pathways from HGF. Some of these HGF signaling pathways can be inhibited; i.e., use of a tyrosine-phosphorylated peptide interferes with both STAT dimerization and the association of STAT to the receptor, and hence it is possible to see which pathways regulate these genes. An additional application, which we are pursuing, compares HGF signaling in MDCK cells grown in a two-dimensional monolayer with a three-dimensional matrix.
3 G' q; t. l( f
MATERIALS AND METHODS
( C9 C* {) \% A! M! h+ c  q
MDCK culture, HGF treatment, and RNA isolation. Low-passage type II MDCK cells were obtained from K. Mostov [Univ. of California San Francisco (UCSF)] and used between passages 3 and 10 as previously described ( 2, 3 ). These cells were originally cloned by Daniel Louvard at EMBL and came to Keith Mostov via Karl Matlin. Cells were cultured in modified Eagle's MEM containing Earl's balanced salt solution and glutamine supplemented with 5% fetal calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml amphotericin B. MDCK cells were seeded at confluency on 24-mm Transwell filter units (Costar, Cambridge, MA). Pore size on all filters was 0.4 µm. Cell monolayers were used for experiments after 4 days of culture, with daily changes of medium. HGF at 100 ng/ml was added to the basolateral compartment of MDCK cell monolayers for a 3- and 24-h pretreatment. This dose has been shown to dedifferentiate MDCK cell monolayers ( 2 ). Recombinant human HGF was generously provided by R. Schwall (Genentech, South San Francisco, CA). Total RNA was isolated from MDCK cell monolayers by following the Rneasy Mini Protocol (Qiagen) for the isolation of total RNA from animal cells. RNA samples were visualized by agarose RNA electrophoresis. RNA concentration and purity were measured by determining the 260 1.8.

Microarray analysis. For our experiments, all protocols were conducted as described in the Affymetrix GeneChip Expression Analysis Technical Manual. Fifteen micrograms of total RNA, collected from MDCK cells grown for 4 days on Transwell filters and exposed to 0, 3, or 24 h of HGF, were converted to first-strand cDNA. Second-strand cDNA synthesis was followed by in vitro transcription for linear amplification of each transcript and incorporation of biotinylated CTP and UTP. The cRNA products were fragmented to 200 nucleotides or less, heated at 99°C for 5 min, and hybridized for 16 h at 45°C to the ToxExpress GeneChip Canine Array, which was developed as a custom array by Gene Logic and Pfizer, using Affymetrix standard design parameters and manufactured by Affymetrix. The ToxExpress GeneChip Canine Array contains probe sets corresponding to 12,473 gene sequences from C. familiaris. In the ToxExpress GeneChip Canine Array, oligonucleotide probes are synthesized in situ complimentary to each corresponding sequence. Sixteen pairs of oligonucleotide probes are used to measure the expression level of each gene sequence represented on the array.2 i. }; w% L8 e# ]; ^; u

The ToxExpress GeneChip Canine Array chips were then washed at low and high stringency and stained with streptavidin-phycoerythrin. Fluorescence was amplified by adding biotinylated anti-streptavidin and an additional aliquot of streptavidin-phycoerythrin stain. A confocal scanner was used to collect fluorescence signal at 3-nm resolution after excitation at 570 nm. The average signal from two sequential scans was calculated for each microarray feature. A complete microarray expression data set has been submitted to the Gene Expression Omnibus (GEO) database at the National Center for Biotechnology Information's (NCBI; GEO submissions GPL374 , GSM8880  -GSM8888, GSE578 ; NCBI tracking system no.15020840).% R4 n+ m7 @3 L! T

Real-time PCR. Fifteen micrograms of total RNA, collected from MDCK cells grown for 4 days on Transwell filters and exposed to 0, 3, or 24 h of HGF, were converted to first-strand cDNA. cDNA and the Taq Man primer/probe system, individualized for each mRNA, were used in conjunction with the 7700 PRISM Sequence Detection Instrument (both Applied Biosystems) as described in the Applied Biosystems Technical Manual. When the reaction product amplification exceeded the threshold value (denoted by the solid line in Fig. 4 ), the corresponding cycle number was termed C T (arrows). Fold-change between conditions was calculated through an exponential function of the observed difference in C T as previously described ( 19 ). The values were normalized to a control mRNA, the 18S ribosome, and all real-time PCR studies were performed in at least triplicate.* l6 p. y' x: P9 T1 K' |

Fig. 4. Real-time PCR targeting matrix metalloproteinase-13 (MMP-13). RNA was harvested from MDCK cells grown on filters for 4 days and exposed to 0 (red), 3 (blue), or 24 h of HGF (yellow). RNA was converted to cDNA and amplified using the Taq Man primer/probe fluorescence assay on a 7700 PRISM instrument (Applied Biosystems). The primer and probe were designed to specifically recognize MMP-13 cDNA. When the reaction product amplification exceeds the threshold value (horizontal black line), the corresponding cycle number is termed C T (arrows). Fold-change between conditions is calculated through an exponential function of the observed difference in C T. Green line, negative control (lacking template).
  h% E& Y- H. ]
Cell lysate preparation and Western blotting. MDCK II cells grown on 24-mm-diameter Costar filters with 0.4-µm-size pores were used for preparation of cell lysates for Western blot analysis. Control and HGF-treated cells on filters were exposed to 0.2 ml of 20 mM Tris·HCl (pH 7.4), 150 mM NaCl, 0.1% SDS, 1% Triton X-100, 1% deoxycholic acid, and 5 mM EDTA (RIPA buffer) containing inhibitors of proteases (2 mM phenylmethylsulfonyl fluoride, 50 µg/ml pepstatin, 50 µg/ml chymostatin, and 10 µg/ml antipain) for 15 min on ice. The protein concentration of each cell lysate was determined by using the bicinchoninic acid determination assay (Pierce Chemical, Rockford, IL). Western blot analysis was performed as previously described ( 2, 3 ). Briefly, equal amounts of protein in lysates of MDCK cells were run on 7.5% acrylamide gels at 160 V for 45 min. Proteins were transferred to Immobilon P membranes (Millipore, Bedford, MA). Membranes were blocked with for 30 min with PBS - (PBS without Ca 2  and Mg 2  ), 5% milk, and 0.1% Tween 20 (block solution). Filters were probed with antibodies against claudin-2 (rabbit polyclonal IgG, 1:250, Zymed), cyclooxygenase-2 (COX-2; mouse monoclonal IgG, 1:250, Transduction Laboratories), or Sec61 (a gift from V. Lingappa, UCSF, rabbit polyclonal IgG, 1:500) for 1 h and then washed with PBS -, 0.1% Tween 20 (4 times, 5 min). Filters were then probed with either horseradish peroxidase-labeled goat anti-mouse (1:2,500 for COX-2) or goat anti-rabbit (1:5,000 for claudin-2 and Sec61 ) secondary antibodies diluted in block solution for 1 h. Filters were again washed (4 times, 5 min) with PBS-0.1% Tween 20. Filters were developed using enhanced chemiluminescence kit (Amersham) and visualized on Kodak X-OMAT film.: t4 e- k8 g. q+ K/ c1 a1 d

RESULTS AND DISCUSSION4 ]% c, ?: M9 Z) \# o+ o

MDCK cells were grown on a two-dimensional Transwell filter for 4 days and exposed to recombinant HGF (100 ng/ml) for 0, 3, or 24 h. The 3-h time point was chosen to identify the immediate gene expression changes that occur during HGF-induced epithelial dedifferentiation. Phenotypically, the changes of epithelial cell polarity are first observed at 24 h of HGF stimulation ( 2 ). Accordingly, we were interested in identifying gene expression changes induced by HGF that occurred both before and during the period of phenotypic change. Exposure of MDCK cell monolayers to HGF for periods extending to 48 and 72 h has demonstrated loss of polarity (as manifested by a distribution of E-cadherin along the entire surface of MDCK cells) and a pseudostratified appearance ( 1, 2 ).

To test for early changes in gene expression levels after exposure of filter-grown MDCK cells to 0, 3, and 24 h of HGF, we harvested RNA, which was converted to cDNA and used to query the 12,473 gene sequences on the DNA microarray chip. Gene intensity values, corresponding to the mRNA levels, were calculated by Affymetrix MAS5 software. GeneSpring (Silicon Genetics) was used to scale the gene intensity values so that the median of each chip was equivalent, and then a per gene normalization to the median of each gene's expression across the nine samples (each of the 3 conditions in triplicate) was performed to facilitate comparison of the data. We then generated a list of genes that were flagged as present (P) by Affymetrix MAS5 software in all three replicates of at least one of our conditions. There were 5,573 genes in this list. We used this gene list to hierarchically cluster all nine of our individual samples. The resulting dendogram (tree) is shown with branches for each of our three conditions similarly colored ( Fig. 1 A ).2 V3 y) t# {( I6 u# y4 k
1 u) l& t' S& O7 ]
Fig. 1. Replicates are similar in each condition. A : entire set of 5,573 genes that were present in all 3 replicates of at least 1 condition was used to cluster the samples. The resulting dendogram is shown at the top. The fact that all 3 replicates in each condition [blue for control, green for 3 h of hepatocyte growth factor (HGF) exposure, and red for 24 h of HGF exposure] cluster validates the similarity of the replicates. B : columns represent the 667 genes, ordered in rows across the columns, that were differentially regulated between conditions. Different colors refer to levels of regulation, with blue indicating the gene was relatively downregulated and red indicating the gene was relatively upregulated.
) {( p9 W% d) T% d+ `' V
We then exported the normalized data for all genes to SAM (Statistical Analysis of Microarrays, Stanford University) and performed a multiclass analysis to test for significantly changed genes in any of three conditions. The three conditions were MDCK cells grown for 4 days on Transwell filters and exposed to recombinant HGF for 0, 3, or 24 h. We chose to accept a 10% false-positive rate (termed the False Discovery Rate or FDR). Of the 12,473 genes on the array, there were 1,672 that met our criteria. These 1,672 genes were imported into GeneSpring, and all six control, HGF 24 h, HGF 3 h. For instance, control means MDCK cells exposed to HGF for 3 h that are upregulated compared with control MDCK cells (no HGF exposure). We defined "significantly regulated" to be a twofold or greater change (based on an average of three normalized replicates) and present in all replicates of the condition in which higher mRNA expression occurred. We then combined the pairwise lists to generate a final list of 667 unique genes ( Fig. 1 B ).: `; L1 D! |: D. q2 }

Of the 667 genes, 125 were annotated to known C. familiaris 90%) homology to other known mammalian genes ( Fig. 2 ). We found 8 genes among the 125 annotated genes that had previously been implicated in the HGF signaling pathway, thereby demonstrating that the system was responsive to HGF ( Table 1 ). This group of regulated genes has already expanded the list of genes known to be involved in the various classes of HGF/c-met signaling "families." The remaining 542 regulated genes are not currently annotated. Nevertheless, many important genes were assayed. The number of genes that are "annotatable" on this DNA microarray chip will undoubtedly increase over time as a partial canine sequence (1.5 x of the genome) was completed last year by The Institute for Genomic Research (Rockville, MD) ( 14 ), and a more complete, National Institutes of Health-funded project (6 x sequence) has just started and has been predicted, by those involved, to take 1 yr to complete (Dr. Ewen Kirkness, TIGR, personal communication).

Fig. 2. Changes in gene expression in filter-grown Madin-Darby canine kidney (MDCK) cells in response to HGF. After 4 days of growth on a Transwell filter, MDCK cells were exposed to 0, 3, or 24 h of recombinant HGF, and mRNA was harvested as described in MATERIALS AND METHODS. This mRNA was converted to cDNA and used to probe a novel canine microarray developed by Gene Logic, Pfizer, and Affymetrix. Column 1 gives the names of the annotated genes that were differentially regulated, column 2 gives the fold-change and directionality, whereas column 3 gives the GenBank sequence numbers. In column 3, % refers to the similarity between the canine DNA sequence used in the ToxExpress GeneChip Canine Array and the mammalian homolog referenced. ER, endoplasmic reticulum.

Fig. 2 - Continued9 \! x" K' p: {! Y  ]

Fig. 2 - Continued' y! f8 k+ u: R: L0 b9 F' O
$ Q! j* a+ {# y/ a, w# d& L: ]8 {
Table 1. Genes previously linked to the HGF pathway

The microarray gene expression results were confirmed by testing mRNA levels using real-time PCR for all 35 genes in which we had complete mRNA sequence information ( Fig. 3 ). Complete sequence information was necessary to design the primers and probes for use with the Taq Man System and the 7700 PRISM Sequence Detection Instrument (both Applied Biosystems). For example, several extracellular matrix proteins, including matrix metalloproteinase-13 (MMP-13), were dramatically upregulated after HGF exposure. By real-time PCR, the calculated fold-increase for MMP-13, normalized to the 18S ribosome (a control mRNA), was 27.86 ± 0.19 and 15.18 ± 0.52 for 3 and 24 h of HGF exposure, respectively, compared with 0 h of HGF exposure (control) ( Fig. 4 ). This was very similar to the fold-increase determined by microarray analysis ( Fig. 2 ). With respect to the other 34 genes tested by real-time PCR, in all cases the direction of fold-change in both real-time PCR and microarray analysis was the same. The absolute fold-change values were also remarkably consistent, thereby validating the microarray results, although there were a few differences. For example, the Na-K-ATPase showed a 2.34-fold change by microarray analysis, although almost no change by real-time PCR (1.1 fold) ( Fig. 3 ). This may be due to the fact that, as noted above, we chose to accept a 10% false-positive rate in our microarray analysis, and therefore real-time PCR may be more specific. Other extracellular matrix proteins that were dramatically upregulated were integrin -6 and fibronectin ED-A. Fibronectin is particularly interesting in that it was very recently shown to be required for branching morphogenesis ( 34 ) and, in a three-dimensional matrix, HGF induces tubulogenesis in MDCK cells ( 17, 26, 31 ). Another gene involved in branching is sprouty, which is transiently upregulated and is an antagonist of fibroblast growth factor (FGF) signaling in Drosophila airways ( 9 ). In our array, sprouty homolog 2 was also transiently upregulated.0 U0 `0 y& e, `8 ~; f5 c
! e# T' d, R- m
Fig. 3. mRNA expression levels as determined by real-time PCR validate the microarray results. The mRNA expression levels, using real-time (quantitative) PCR, for all 35 genes in which we had complete sequence information and could, therefore, design the appropriate primers/probes are shown. Fold-changes for the control samples (no HGF exposure) were normalized and are listed as 1. In columns 1-3, the mRNA expression levels are listed for 0, 3, and 24 h of HGF exposure. These values were then converted to fold-changes ( column 4 ) and compared with the significant fold-changes as determined by microarray analysis ( column 5 ). All the real-time PCR data were normalized to a control mRNA, the 18S ribosome, and all real-time PCR studies were performed in at least triplicate. inhib., Inhibitor; precurs., precursor.
0 Z& i: [! J- \) @: ~. o
Another way to confirm the array results is by examining the protein expression levels by Western blot analysis. For example, claudin-2 is a component of the tight junction ( 41 ) and claudin-2 gene expression was downregulated 23.8-fold after 24 h of HGF exposure compared with controls. We confirmed this dramatic decrease in claudin-2 expression by HGF at the protein level ( Fig. 5 A ). Interestingly, two other components of the tight junction, claudin-3 and occludin 1B, were upregulated by 24 h of HGF exposure compared with 3 h of exposure. We hypothesize that differential expression of claudin-2 plays a key role in mediating the cellular response of MDCK cells to HGF. Recently, HGF has been shown to modulate transepithelial resistance (TER) in MDCK cells ( 28 ). Our microarray data suggest that HGF-induced changes in claudin-2 expression contribute to these changes in MDCK cell TER. We have completed a detailed analysis of the effects of HGF on tight junction associated proteins to help understand how HGF alters MDCK cell TER, and this analysis supports our hypothesis regarding the importance of claudin-2 regulation (unpublished observations). Other cytoskeletal proteins were also upregulated, including actinin -1 and microtubule-associated proteins.( b) G: D" k, w& b4 q
6 Z, W# r+ u/ g. \. n: G4 B" j
Fig. 5. Protein levels of claudin-2, cyclooxygenase 2 (COX-2), and Sec61 in MDCK cells grown on Transwell filters and exposed to HGF. To validate the gene expression results at the protein level, Western blot analysis was performed on lysate from MDCK cells grown on a Transwell filter for 4 days and exposed to control medium (C) or medium containing HGF. A : per the microarray results, claudin-2 was downregulated 23.8-fold at the gene level and had a similar decrease at the protein level after 24 h (24') of exposure to HGF. B : COX-2 was upregulated 2.51-fold at the gene level and had a similar increase at the protein level after 24 h of exposure to HGF. Western blotting was performed on triplicate samples. C : expression of the Sec61 component of the Sec61 complex was upregulated 2.11-fold after 24 h (24') of HGF exposure compared with controls. A similar increase at the protein level after 24 h of exposure to HGF was seen compared with 3 h (3') of exposure to HGF and controls (C).

Mesenchymal production of HGF has been described to be regulated by numerous factors including acidic and basic fibroblast growth factor, epidermal growth factor, and prostaglandins ( 8, 21, 22 ). COX-2 is overexpressed in multiple types of carcinoma including colon, lung, bladder, breast, melanocytes, head and neck, and pancreas ( 15 ). Several epidemiological studies have demonstrated that inhibition of COX-2 activity has a chemoprotective effect on carcinoma tumor growth and metastasis ( 15 ). Our analysis shows that COX-2 gene expression was upregulated 2.51-fold after 24 h of HGF exposure compared with controls. Western blot analysis confirmed a similar increase in COX-2 protein expression after 24 h of HGF treatment ( Fig. 5 B ). A major metabolite of COX-2 activity is prostaglandin E 2 ( 5 ). Interestingly, the prostaglandin E 2 receptor, which increases intracellular cAMP after activation, was also upregulated by HGF treatment. In what may be an autocrine response to HGF, analysis of the array results showed that both epidermal growth factor and FGF receptor 2 were downregulated. It has also been reported that elevation of intracellular cAMP levels is an important signal for HGF production ( 24 ), and adrenomedullin precursor, which is involved in activation of cAMP, was dramatically downregulated after 3 h of HGF, but even more dramatically upregulated from 3 to 24 h of HGF. Conversely, transforming growth factor- 1 and glucocorticoids have been shown to inhibit HGF ( 8, 23 ), and the activin A receptor was downregulated.! A$ q3 n& ]: M; D1 s7 K9 z
& ~" H3 p0 i" U" n5 L, f) F6 c2 q
Another group of genes that changed in response to HGF were those encoding small GTPases. Most, like the Ras homolog Rac2, GDP dissociation inhibitor, and Rab10, were upregulated, as might be expected in the first phase (scattering) of the response to HGF ( 4 ). However, some, such as Rab3AIP and GEF1, were downregulated.

HGF induction of transformation, but not cell motility, requires stimulation of the Ras-MAP kinase cascade ( 29 ). HGF has also been recently shown to modulate invasiveness and motility of ovarian carcinomas by a Ras-mediated pathway ( 42 ). As noted, the Ras homolog was upregulated as was mitogen-activated protein kinase kinase kinase kinase 4; however, mitogen-activated protein kinase kinase kinase 5 was only transiently upregulated.7 t& |  J/ Q, m9 g2 v

Stat-3 is also activated by HGF signaling during MDCK cell tubulogenesis ( 4 ); however, per a reverse BLAST search of the entire chip using Stat-3 cDNA, Stat-3 was not one of the 12,473 gene sequences on the chip. Other genes known to be in the HGF signaling pathway, including C-Fos, cyclin-D1, GAB1, phosphatidylinositol 3-kinase, PLC-, and Ron ( 2 ), were also not represented on the chip. In fact, of several dozen cDNAs used in a reverse BLAST search, we could only annotate an additional two, -actin and the exocyst component Exo70, and neither of these genes met the criteria for a twofold change in expression. The c-met receptor was also not represented on the chip. While a canine c-met has now been deposited in GenBank (AB096697  ), it was submitted on November 22, 2002, after the ToxExpress GeneChip Canine Array, described here, was already constructed.& x' ]5 b# K$ h; _

HGF has been identified in the endothelia ( 33 ), and the human serine protease (HGF-converting enzyme) that converts proHGF to the mature heterodimeric form exhibits structural similarity to blood coagulation factor XII ( 25 ). HGF has also been shown to initiate angiogenesis ( 43 ). HGF itself has 38% homology to the serine protease plasminogen (another kringle domain protein) ( 27 ). Several hematopoietic proteins were upregulated, including acetylhydrolase, which is involved in platelet activation ( 39 ), glycoprotein GPIIIA, and vascular endothelial growth factor receptor-1. Some were also downregulated, including endothelin-1 precursor and thrombospondin.4 G. R* x- ^, K3 l
; t' o# I" }3 k; q. a% b) C
Not surprisingly for a growth factor, in response to HGF there were numerous transcription factors that were either up- or downregulated as well as nuclear factors. Interestingly, there were also a number of genes encoding proteins involved in translation and translocation of proteins across the endoplasmic reticulum (ER) that were all upregulated, including the Sec61 homologue ( Fig. 5 C ), signal receptor protein -subunit, eukaryotic translation initiation factor 4E, and eukaryotic translation termination factor. We and others have recently shown that the exocyst complex is involved in protein synthesis, by interacting with the Sec61 component of the translational machinery of the ER ( 18, 40 ). As expected from the microarray results, no changes in Sec61 protein levels were detected in MDCK cell lysates after 3 h of HGF treatment ( Fig. 5 B ). Of the eight exocyst components, only Exo70 was found on the chip (by a reverse BLAST search) and, as noted above, it did not quite meet the twofold criterion for change.
& @9 X+ e% N# A7 ~0 @5 R
Conclusion. Using a novel canine DNA microarray designed to query 12,473 gene sequences from C. familiaris, we have evaluated HGF signaling in MDCK cells grown on two-dimensional Transwell filters and stimulated with recombinant HGF. We have identified known genes, as well as new genes, that are significantly up- or downregulated after exposure to HGF. In addition, we believe that this microarray will be of great interest to the general epithelial cell biology community who rely on the MDCK cell culture model.

ACKNOWLEDGMENTS
. r  S( B6 T- a
GRANTS6 a: i: }. w6 A" V6 i

This work was supported in part by grants from the Polycystic Kidney Disease Foundation (J. H. Lipschutz), National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-58090 and DK-02509 (J. H. Lipschutz), Cystic Fibrosis Pilot Grant CFF SORSCH98SO (D. F. Balkovetz), and a Veterans Administration Merit Award (D. F. Balkovetz).4 J7 E+ m  @8 q- a# a/ A# @8 T) z& q
      【参考文献】) v. s/ h' d7 X5 H% N: t* X/ s+ d
Balkovetz DF. Evidence that hepatocyte growth factor abrogates contact inhibition of mitosis in Madin-Darby canine kidney cell monolayers. Life Sci 64: 1393-1401, 1999.: y. S# y0 O9 U. k9 c

7 c  V$ V- v/ `0 G* i/ \1 ?6 M5 L

Balkovetz DF, Pollack AL, and Mostov KE. Hepatocyte growth factor alters the polarity of Madin-Darby canine kidney cell monolayers. J Biol Chem 272: 3471-3477, 1997.* t) a7 j: ?! N1 w2 Z+ s- p

" |* {+ B8 U, |3 w6 u$ I, p

Balkovetz DF and Sambandam V. Dynamics of E-cadherin and gamma-catenin complexes during dedifferentiation of polarized MDCK cells. Kidney Int 56: 910-921, 1999.5 K+ d# |4 v/ y2 V, J8 K; h
7 @% f. X0 K! t$ \0 i

Boccaccio C, Ando M, Tamagnone L, Bardelli A, Michieli P, Battistini C, and Comoglio PM. Induction of epithelial tubules by growth factor HGF depends on the STAT pathway. Nature 391: 285-288, 1998./ K/ X* W5 ~3 ?; R' e: R5 v

9 B8 r" U0 G# `6 ^* R1 E

Breyer MD and Breyer RM. Prostaglandin E receptors and the kidney. Am J Physiol Renal Physiol 279: F12-F23, 2000.2 O- j+ B0 B/ U7 l* f
8 w3 g* o( s) ~' ~
% Y4 k1 `1 k7 L# T' D# ^/ \

Donato MT, Gomez-Lechon MJ, Jover R, Nakamura T, and Castell JV. Human hepatocyte growth factor down-regulates the expression of cytochrome P450 isoenzymes in human hepatocytes in primary culture. J Pharm Exp Ther 284: 760-767, 1998.

: b3 K- J2 }# k( V+ p1 a

Fleiszig SM, Vallas V, Jun CH, Mok L, Balkovetz DF, Roth MG, and Mostov KE. Susceptibility of epithelial cells to Pseudomonas aeruginosa invasion and cytotoxicity is upregulated by hepatocyte growth factor. Inf Immun 66: 3443-3446, 1998.
: j4 Z) \( D$ n6 j; d) h9 }* W

Gohda E, Matsunaga T, Kataoka H, Takebe T, and Yamamoto I. Induction of hepatocyte growth factor in human skin fibroblasts by epidermal growth factor, platelet-derived growth factor and fibroblast growth factor. Cytokine 6: 633-640, 1994.

2 N* v$ }1 e( R+ U0 t4 i2 l
Hacohen N, Kramer S, Sutherland D, Hiromi Y, and Krasnow MA. Sprouty encodes a novel antagonist of FGF signaling that patterns apical branching of the Drosophila airways. Cell 92: 253-263, 1998.
& |: B6 x0 s& `' t6 s$ ~( p& J

Ieraci A, Forni PE, and Ponzetto C. Viable hypomorphic signaling mutant of the Met receptor reveals a role for hepatocyte growth factor in postnatal cerebellar development. Proc Natl Acad Sci USA 99: 15200-15205, 2002.

Inoue T, Nabeshima K, Shimao Y, Meng JY, and Koono M. Regulation of fibronectin expression and splicing in migrating epithelial cells: migrating MDCK cells produce a lesser amount of, but more active fibronectin. Biochem Biophys Res Commun 280: 1262-1268, 2001.. B, w& ]' N, z
; z, j+ \  y) K# t3 b' K2 q* @

* b5 w6 E9 }( e9 s; \! ]
Ishibashi K, Sasaki S, Sakamoto H, Nakamura Y, Hata T, Nakamura T, and Marumo F. Hepatocyte growth factor is a paracrine factor for renal epithelial cells: stimulation of DNA synthesis and NA,K-ATPase activity. Biochem Biophys Res Commun 182: 960-965, 1992.' Z' d: H1 U+ Q% [( q  u* M
; O! Z7 L& E' o

. h  w  y& V2 I  U5 ^
Jordan B. Historical background and anticipated developments. Ann NY Acad Sci 975: 24-32, 2002.
: {. Q8 l6 w! f8 r# V
9 }+ q/ r% c4 b# k8 o
/ B9 r0 {( D  ]
Kirkness EF, Bafna V, Halpern AL, Levy S, Remington K, Rusch DB, Delcher AL, Pop M, Wang W, Fraser CM, and Venter JC. The dog genome: survey sequencing and comparative analysis. Science 301: 1898-1903, 2003.8 D) \9 d9 [0 x5 }
: C, s$ ]$ G# M# N
$ o4 i0 D# z, W/ N' q  w5 |- |
% Y; q7 f& W# ~2 y- F' G
Koki AT, Khan NK, Woerner BM, Seibert K, Harmon JL, Dannenberg AJ, Soslow RA, and Masferrer JL. Characterization of cyclooxygenase-2 (COX-2) during tumorigenesis in human epithelial cancers: evidence for potential clinical utility of COX-2 inhibitors in epithelial cancers. Prostaglandins Leukot Essent Fatty Acids 66: 13-18, 2002.
' H" `7 B; C, \( \: Z' x. N
% [& \# c! \' ]5 P' B
- c! _0 `) s; Q0 j* V
Leighton J, Estes LW, Mansukhani S, and Brada Z. A cell line derived from normal dog kidney (MDCK) exhibiting qualities of papillary adenocarcinoma and of renal tubular epithelium. Cancer 26: 1022-1028, 1970. <a href="/cgi/external_ref?access_num=10.1002/1097-0142(197011)26:5
% A+ L4 m, _/ f) J

Lipschutz JH, Guo W, O'Brien LE, Nguyen YH, Novick P, and Mostov KE. Exocyst is involved in cystogenesis and tubulogenesis and acts by modulating synthesis and delivery of basolateral plasma membrane and secretory proteins. Mol Biol Cell 11: 4259-4275, 2000.% o% C* e( K7 n% F

' z, x0 D4 v* S/ a3 e9 r
& Y8 N6 G3 ?7 w
Lipschutz JH, Lingappa VR, and Mostov KE. The exocyst affects protein synthesis by acting on the translocation machinery of the endoplasmic reticulum. J Biol Chem 278: 20954-20960, 2003.

Livak KJ and Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2-, CT method. Methods 25: 402-408, 2001.! T( Q& Q9 B- ?7 `

4 L- i9 I6 ?) F4 X( B& k% S- [

Maeshima A, Zhang YQ, Furukawa M, Naruse T, and Kojima I. Hepatocyte growth factor induces branching tubulogenesis in MDCK cells by modulating the activin-follistatin pathway. Kidney Int 58: 1511-1522, 2000.1 [: r% ?4 X& I/ ^. f+ a, i
1 b7 Y, @; B& ~! s4 P% X% O

Matsumoto K, Okazaki H, and Nakamura T. Novel function of prostaglandins as inducers of gene expression of HGF and putative mediators of tissue regeneration. J Biochem (Tokyo) 117: 458-464, 1995.
3 k: V& V' q" N. S2 p

Matsumoto K, Okazaki H, and Nakamura T. Up-regulation of hepatocyte growth factor gene expression by interleukin-1 in human skin fibroblasts. Biochem Biophys Res Commun 188: 235-243, 1992.
5 E% n0 Y" J4 g

Matsumoto K, Tajima H, Okazaki H, and Nakamura T. Negative regulation of hepatocyte growth factor gene expression in human lung fibroblasts and leukemic cells by transforming growth factor 1 and glucocorticoids. J Biol Chem 267: 24917-24920, 1992.

Matsunaga T, Gohda E, Takebe T, Wu YL, Iwao M, Kataoka H, and Yamamoto I. Expression of hepatocyte growth factor is up regulated through activation of a cAMP mediated pathway. Exp Cell Res 210: 326-335, 1994.

6 R$ d4 Q, G0 _- |, d" f7 I
Miyazawa K, Shimomura T, Kitamura A, Kondo J, Morimoto Y, and Kitamura N. Molecular cloning and sequence analysis of the cDNA for a human serine protease responsible for activation of hepatocyte growth factor. Structural similarity of the protease precursor to blood coagulation factor XII. J Biol Chem 268: 10024-10028, 1993.
( n0 @% ?9 l. r4 q  G& w

, V4 ~$ S) z0 V$ f: z- I
Montesano R, Schaller G, and Orci L. Induction of epithelial tubular morphogenesis in vitro by fibroblast-derived soluble factors. Cell 66: 697-711, 1991.6 M! |5 n: }9 I2 l! z6 n

, c8 O/ n" v0 v8 B6 z7 x6 ^7 F
Nakamura T, Nishizawa T, Hagiya M, Seki T, Shimonishi M, Sugimura A, Tashiro K, and Shimizu S. Molecular cloning and expression of human hepatocyte growth factor. Nature 342: 440-443, 1989.5 e& I5 j% ?* i8 X8 p& u
2 x9 j+ R* m$ X* O" Z0 Q2 r6 x6 n

Pollack AL, Apodaca G, and Mostov KE. Hepatocyte growth factor induces MDCK cell morphogenesis without causing loss of tight junction functional integrity. Am J Physiol Cell Physiol 286: C482-C494, 2004. First published October 30, 2003; 10.1152/ajpcell.00377.2003.
" u! Y% F: y4 {1 ~6 J

, d, i" b" {$ y  E* F/ N  ~
Ponzetto C, Zhen Z, Audero E, Maina F, Bardelli A, Basile ML, Giordano S, Narsimhan R, and Comoglio P. Specific uncoupling of GRB2 from the Met receptor. Differential effects on transformation and motility. J Biol Chem 271: 14119-14123, 1996./ N0 {1 l/ Y% z# N3 \) G, z0 z  O
/ I) n& Q: u, [8 N6 h8 ^8 T/ q
; d3 a0 _. H$ ]; p8 p/ N" h

Recio JA and Merlino G. Hepatocyte growth factor/scatter factor induces feedback up-regulation of CD44v6 in melanoma cells through Egr-1. Cancer Res 63: 1576-1582, 2003." w. X* s& X" q! c' Q
% T! N! z5 q6 s5 t7 n( V" f
" ]; G. f  R6 z
& q- h9 X* P, D! t- X
Rogers KK, Jou TS, Guo W, and Lipschutz JH. The Rho family of small GTPases is involved in epithelial cystogenesis and tubulogenesis. Kidney Int 63: 1632-1644, 2003.

) h# u; X; m- [4 _' e5 |0 J9 n

Rosario M and Birchmeier W. How to make tubes: signaling by the Met receptor tyrosine kinase. Trends Cell Biol 13: 328-335, 2003.

Rosen EM, Grant DS, Kleinman HK, Goldberg ID, Bhargava MM, Nickoloff BJ, Kinsella JL, and Polverini P. Scatter factor (hepatocyte growth factor) is a potent angiogenesis factor in vivo. Symp Soc Exp Biol 47: 227-234, 1993.
( F" }. }3 |" m5 C9 M

Sakai T, Larsen M, and Yamada KM. Fibronectin requirement in branching morphogenesis. Nature 423: 876-881, 2003.

. y# n* m  o+ G* K+ ?
  T, R% P  c. s7 j3 w! ^# {
Sakata H, Takayama H, Sharp R, Rubin JS, Merlino G, and LaRochelle WJ. Hepatocyte growth factor/scatter factor overexpression induces growth, abnormal development, and tumor formation in transgenic mouse livers. Cell Growth Differ 7: 1513-1523, 1996.' j3 }: Q% h$ R5 H8 }0 x/ T
. p) C0 P5 v) A- ?4 Z6 W5 X( d& ?2 \5 n

. L/ w  {9 ~0 }5 t. ]- i/ Z
Skouteris GG and Schroder CH. Cytosolic phospholipase A 2 is activated by the hepatocyte growth factor receptor-kinase in Madin-Darby canine kidney cells. J Cell Sci 110: 1655-1663, 1997.+ I6 H3 }0 `- P# [' W

- o" W8 t4 g% K1 L9 A0 [! t+ A
" j9 M0 V4 f. z! Q1 r
Takayama H, LaRochelle WJ, Sharp R, Otsuka T, Kriebel P, Anver M, Aaronson SA, and Merlino G. Diverse tumorigenesis associated with aberrant development in mice overexpressing hepatocyte growth factor/scatter factor. Proc Natl Acad Sci USA 94: 701-706, 1997.
& i! f% h5 t- D" x8 n2 n/ v4 Y

Taub M and Saier MH Jr. An established but differentiated kidney epithelial cell line (MDCK). Methods Enzymol 58: 552-560, 1979.
) o* I1 b0 g# n% T
9 G$ \, \. x* y) T; p! }1 ~

Tjoelker LW, Eberhardt C, Unger J, Trong HL, Zimmerman GA, McIntyre TM, Stafforini DM, Prescott SM, and Gray PW. Plasma platelet-activating factor acetylhydrolase is a secreted phospholipase A 2 with a catalytic triad. J Biol Chem 270: 25481-25487, 1995.. P& e" W: t: M6 L) r
* Z  n! n5 o" [* Z' F: c
7 K) F$ K% b; o# l

Toikkanen JH, Miller KJ, Soderlund H, Jantti J, and Keranen S. The subunit of the Sec61p endoplasmic reticulum translocon interacts with the exocyst complex in Saccharomyces cerevisiae. J Biol Chem 278: 20946-20953, 2003.

' D' \5 E' e8 e6 q
Tsukita S and Furuse M. Claudin-based barrier in simple and stratified cellular sheets. Curr Opin Cell Biol 14: 531-536, 2002.

: X; i$ M. u" u* L9 Q) n# C8 w
Ueoka Y, Kato K, and Wake N. Hepatocyte growth factor modulates motility and invasiveness of ovarian carcinomas via ras mediated pathway. Mol Cell Endocrinol 202: 81-88, 2003.

3 W" d. v: |8 z% ?( t5 T

Wojta J, Kaun C, Breuss JM, Koshelnick Y, Beckmann R, Hattey E, Mildner M, Weninger W, Nakamura T, Tschachler E, and Binder BR. Hepatocyte growth factor increases expression of vascular endothelial growth factor and plasminogen activator inhibitor-1 in human keratinocytes and the vascular endothelial growth factor receptor flk-1 in human endothelial cells. Lab Invest 79: 427-438, 1999.

9 }. a: D% f9 f( r

Wu JH, Billings BJ, and Balkovetz DF. Hepatocyte growth factor alters renal epithelial cell susceptibility to uropathogenic Escherichia coli. J Am Soc Nephrol 12: 2543-2553, 2001.

+ L! o+ q+ q% d; H; N
Zeng Q, McCauley LK, and Wang CY. Hepatocyte growth factor inhibits anoikis by induction of activator protein 1-dependent cyclooxygenase-2. Implication in head and neck squamous cell carcinoma progression. J Biol Chem 277: 50137-50142, 2002.