- Email: [email protected]
Meprin metalloprotease expression and regulation in kidney, intestine, urinary tract infections and cancer
FEBS 29478
FEBS Letters 579 (2005) 3317–3322
Minireview
Meprin metalloprotease expression and regulation in kidney, intestine, urinary tract infections and cancer Judith S. Bond*, Gail L. Matters, Sanjita Banerjee, Renee E. Dusheck Department of Biochemistry and Molecular Biology, The Pennsylvania State University College of Medicine, Hershey, PA 17033, USA Accepted 23 March 2005 Available online 31 March 2005 Edited by Pe´ter Friedrich
Abstract Meprins are unique plasma membrane and secreted metalloproteinases that are highly regulated at the transcriptional and post-translational levels. Meprin a and b subunits are abundantly expressed in kidney and intestinal epithelial cells, are secreted into the urinary tract and intestinal lumen, and are found in leukocytes and cancer cells under certain conditions. Their location and proteolytic activities indicate functions at the interface of the host and the external environment, and in trafficking of macrophages and metastases of cancer cells. These proteases can be detrimental when there is tissue damage or disruption, as in acute renal injury or intestinal inflammation, and there is evidence they are involved in movement of leukocytes and cancer cells to sites of infection or in metastasis, respectively. Ó 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Keywords: Metalloprotease; Kidney; Intestine; Inflammation; Infection; Cancer
1. Introduction Meprins are highly expressed at the brush border membrane of proximal tubule cells of the kidney and epithelial cells of the intestine, particularly in the ileum and colon, and in mesenteric leukocytes [1,2]. For example, meprin proteins were estimated to compose approximately 5% of the brush border membrane protein in rodent kidneys, second only to actin in abundance [3]. These proteases are multidomain, highly glycosylated proteins that act extracellularly, and are either bound to the plasma membrane by transmembrane segments, or are secreted or shed into extracellular spaces [4,5]. Meprin proteases are composed of two evolutionarily related subunits, a and b, that are approximately 50% identical at the amino acid level. The subunits are encoded on two genes: the a gene is on human chromosome
* Corresponding author. Fax: +1 717 531 7072. E-mail address: [email protected] (J.S. Bond).
Abbreviations: DSS, dextran sulfate sodium; IBD, inflammatory bowel disease; UTI, urinary tract infection; RT-PCR, reverse transcriptasepolymerase chain reaction; GAPDH, glyceraldehyde phosphate dehydrogenase
6 (mouse 17) near the histocompatibility complex; the b subunit on chromosome 18 in both the mouse and human genomes [6]. The subunits can be expressed separately, or coordinately, and consequently they form homo or heterocomplexes where the basic unit is a disulfide-linked dimer. Both subunits are synthesized as membrane spanning proteins with small cytoplasmic tails, and the bulk of the protein is extracellular. However, the meprin a subunit is proteolytically cleaved during biosynthesis so that it loses its transmembrane domain. Thus, meprin a exists at the membrane only when it associates with meprin b, which exists as a type I protein at the plasma membrane. Membrane forms of meprins are dimeric (homodimeric b; meprin B) or tetrameric (hetero a2b2 or a3b2 complexes) (see Fig. 1A). The secreted form of meprin a tends to oligomerize into high molecular mass complexes containing 10–100 subunits (dimers that associate noncovalently), forming the largest proteases known to exist (see Fig. 1B). The self-associating homo-oligomeric complexes of meprin A are secreted as latent proteases (containing the prosequence) and can move through extracelluar spaces in a non-destructive manner, and deliver a concentrated form of this metalloproteinase to sites that have activating proteases, such as site of inflammation, infection or cancerous growth [4,7]. Thus, meprin structures provide means to concentrate proteolytic activity at the cell membrane (through homo–hetero-oligomeric complexes) and extracellularly (through homotypic interactions). Meprins are capable of cleaving a wide range of substrates, from bioactive peptides, such as gastrin and angiotensin, to extracellular matrix proteins, such as collagen IV, fibronectin and nidogen [8,9]. The heterooligomeric form of meprin was identified as a major matrix degrading protease in studies of rat kidney [10,11]. The peptide bond specificities of meprins a and b differ substantially, indicating different functions, however, they can both hydrolyze extracellular matrix proteins. Among the best substrates for meprin b are gastrin and osteopontin, which are growth factors and chemokines [8]. Meprin a, but not b, cleaves cytokines such as monocyte chemoattractant protein-1, and this could be important in immune responses [12]. Meprin metalloproteases are highly regulated at the transcriptional and post-translational level and are normally restricted to specific locations, but they can be detrimental to host tissues when there is tissue damage and disruption of cells, e.g., in intestinal inflammation and kidney disruption.
0014-5793/$30.00 Ó 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2005.03.045
3318
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
Fig. 1. (A) Domain and oligomeric structure of meprins a and b. Domains are S, signal sequence; Pro, prosequence; Protease, catalytic domain; MAM, meprin, A5 protein, protein tyrosine phosphatase l; MATH, meprin and TRAF homology; AM, after MATH; I, inserted domain; EGF, epidermal growth factor-like, TM, transmembrane-spanning; C, cytoplasmic domain. The MATH and AM domains are now designated the TRAF domain (4,5). (B) Electron micrograph of various forms of secreted, latent homooligomeric meprin A demonstrating high molecular mass complexes. Row 1, closed rings and crescents (containing about 10 to 12 meprin a subunits); rows 2 and 3 show tubes and spirals containing up to 100 subunits.
2. Expression of meprins in development, in leukocytes, and in germ-free mice In situ hybridization studies of embryonic and adult mice and immunohistochemistry demonstrated the tissue-specific expression of meprin subunits in the epithelial cells of kidney and intestine only [13]. These studies also demonstrated that there was differential regulation of meprins a and b after birth. Meprin a was present in the fetus by day 11, increased expression after birth, and expression was precipitously decreased in the intestine after weaning (21 days). By contrast, meprin b increased during fetal growth and after birth, and became the predominant isoform of the enzyme in the intestine after weaning. The kidney expression of mouse meprin subunits were strain-dependent; all strains express both subunits during fetal stages, some strains increase both subunits after birth (e.g., C57BL/6) while others only express meprin b, and down regulate meprin a (e.g., C3H/He). The consequence of the expression patterns in mice is that meprin b is the predominant form of meprin in the adult intestine; while meprin a is abundantly expressed in kidney of most adult mice, it is not expressed in certain inbred strains of mice (mainly strains of the H-2k haplotype, e.g., CBA, C3H, AKR). Low meprin a expression in the kidney is inherited as an autosomal recessive trait [14]. It has been known for some time that meprin b is highly expressed on the brush border membrane of the ileum (see Fig. 2). Meprin a and b are also expressed in leukocytes of the mouse mesenteric lymph node [2]. When mice were given dextran sulfate sodium (DSS) in their drinking water, to generate an experimental model for inflammatory bowel disease (IBD), only meprin b was present in lymph nodes of DSS-treated mice [2]. In addition, leukocytes from meprin b null mice are deficient in their ability to invade into matrigel. Other functional properties of the meprin b/ leukocytes, such as phagocytosis and superoxide release were not different in the / and +/+ genotypes. The presence of meprin b in the draining lymph node of the intestine, coupled with the decreased ability of meprin b-macrophages to migrate through matrigel implicates
Fig. 2. Immunofluorescent localization of the meprin b metalloprotease on the brush border membrane of the mouse ileum and in macrophages in the lamina propria during inflammation induced by dextran sulfate sodium. Meprin b, green; macrophage-specific protein F4/80, red; colocalization of the latter two proteins appears yellow. Nuclei, blue.
this protease in leukocyte transmigration important to the immune response, and to IBD which is aggravated by leukocyte infiltration. To test whether meprins are induced by the microbial flora of the intestine, RNA samples were obtained (from Dr. Jeffrey GordonÕs laboratory at Washington University) from mice that were either germ-free, colonized with Bacteriodes thetaiotaomicron at day 10, or conventionalized at 10 days. The mice were on two different strain backgrounds, NMRI and C57BL/ 6. Reverse transcriptase-polymerase chain reaction (RT-PCR) analyses showed no difference between the expression of meprin a or b for any of these mice. Thus, the microbial flora are not responsible for induction of intestinal meprins, but rather meprins are innately present in the mouse gut.
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
3319
3. Regulation of meprins in kidney and urinary tract disease High levels of meprin a at the brush border membrane in rodent kidneys have been implicated in tissue damage caused by insults such as ischemia, while low levels of meprin a and b have been associated with fibrosis [15,11,16]. Acute renal failure is a common disorder that affects about 5% of all hospitalized patients, and has a mortality rate of over 50%. An understanding of the cellular and molecular mechanisms of kidney injury is critical to learning how to improve mortality. Studies of inbred strains of mice subjected to ischemia reperfusion have revealed that mice that express low levels of kidney meprin a as adults (C3H/He, CBA) have markedly less kidney damage compared to those that express high levels of kidney meprin a (ICR, C57BL/6) [15]. Studies with rats have demonstrated that high levels of active meprin A can be cytotoxic [11]. The latter authors found active meprin A was cytotoxic to renal cells in culture (LLC-PK1 and MDCK cells) and that actinonin, a potent inhibitor of meprin a, could protect against meprin damage in vitro and in vivo in a model ischemia-reperfusion injury. They also demonstrated that after acute ischemic renal tubule damage there is a shift in meprin location form the brush border membrane to cytoplasmic compartments and to the tubular lumen. One interpretation of these observations is that once tissue is damaged, high levels of active meprins are found in places other than the apical membrane (cytoplasm, basolateral side of the cell, and cell–cell junctions) and are damaging. Inhibitors of meprins can protect against this type of damage. Multiple studies have demonstrated that meprin expression is downregulated in models of renal injury that lead to acute renal failure. For example, renal pathology associated with hydronephrosis (unilateral urethral obstruction) occurs with an early and progressive decline in rat meprin a and b subunit mRNA and protein [17]. Microarray analysis demonstrated marked declines in mouse meprin b expression in acute renal ischemic reperfusion injury and in adriamycin-induced nephropathy [18,16]. The latter authors also reported that meprin b is downregulated during transdifferentiation of renal tubule epithelial cells to a fibroblast-like phenotype in vitro on exposure to TGF-ß and EGF, suggesting a role in fibrogenesis. Meprin b expression is also downregulated in collagen IVA3 knockout mice that develop AlportÕs syndrome and renal pathology [19]. It appears that meprins are downregulated in a variety of situations that lead to cell death through apoptosis or necrosis, and are upregulated in specific highly differentiated cells and in some proliferating cells (such as cancer cells and leukocytes under certain conditions). The location, balance and regulation of meprins are important in kidney disease progression. Urinary tract infections are one of the most common bacterial infections for mankind, and around 50% of women are affected at one point in their lifetime. Recurrent urinary tract infections are also one of the problems after kidney transplantation, and this appears to be related to ischemic conditions and time taken between removal of the kidney and transplantation. There are multiple factors involved in recurrent infections, however, there is a need to understand the mechanisms involved to develop more effective therapies, increase the incidence of renal allograft survival, and prevent renal scarring associated with infection. The secreted form of meprin a is in an ideal location for preventing or lessening
Fig. 3. Western blot of meprin a in human urine from healthy women and women with acute urinary tract infections. Concentrated urine samples (40 ll each) were immunoblotted with a human recombinant meprin a antibody. Low or no meprin immunoreactivity was observed in samples from healthy subjects (sample I.D. 31, 32, 33, and 34) while high meprin immunoreactivity was observed in samples from subjects suffering from acute UTI (sample I.D. 52, 53, 56, and 57).
infections of the urinary tract. Experimental models of urinary tract infections have established that there are mouse strain differences in susceptibility to bacterial infection, and that they are related to the major histocompatibility complex, the area of the genome that encodes meprin a. There is clear evidence that C3H/He mice are more susceptible to infections by uropathic bacteria than Balb/c mice [20]. C3H/He mice express low meprin a in adult kidney, while Balb/c mice express high levels. Moreover, susceptibility is recessive, as is low meprin a expression. Our hypothesis is that meprin a is one of the factors that determine the susceptibility to urinary tract diseases. Meprin a is poised to be activated in the urinary tract; it is secreted from the proximal tubule of the kidney into the lumen as an inactive complex with high proteolytic potential, and can be readily activated by trypsin-like proteases (e.g., subtilisin). Meprin a and b knockout mice will be valuable tools to determine whether in fact meprins are important factors for susceptibility to urinary tract infections by uropathic bacteria. Little is known about meprin in human urinary tract infections (UTIs). To examine meprin expression in human urine, voided urine samples from premenopausal women with a history of frequent UTIs, without a history of frequent UTIs, and with an acute UTI were collected for analysis. The urine samples were concentrated 10-fold, and their total protein and creatinine concentrations were determined. Quantitative Western blotting showed that meprin a protein is detectable in the urine of premenopausal women at low concentrations, and that concentrations are variable from subject to subject. No correlation between the urinary levels of meprin protein and the susceptibility of women to UTIs was observed. Eighty-nine percent of healthy women had low or undetectable levels of urinary meprin, similar to certain inbred strains of mice (e.g., C3H/He mice). In contrast, the majority (86%) of women suffering from an acute UTIs had high (greater than 5.2 ng) or very high (greater than 9.1 ng) levels of meprin protein in the urine (Fig. 3). These results are in concordance with studies of meprin expression in the intestine, where meprin subunit expression is markedly increased during inflammation as a result of leukocyte infiltration.
4. Meprins in intestinal disease Meprin expression in the intestinal tract is highest in the ileum and large intestine where host and microorganisms are in contact, and where intestinal inflammatory diseases develop.
3320
Inflammatory bowel diseases (IBDs) are a collection of clinically heterogeneous intestinal diseases that result in damage to intestinal epithelium and affect 1 million people in the USA [21]. The two major forms of IBD, ulcerative colitis and CrohnÕs disease, involve chronic inflammation of the colon or the ileum plus colon, respectively, and are thought to be caused by genetic and environmental factors that affect host-microbe interactions and production of inflammatory mediators. Candidate genes for susceptibility to IBD have been identified, and these include the NOD2 gene on chromosome 16 (for CrohnÕs), and susceptibility loci on chromosome 19p13 and the HLA region on chromosome 6 for both ulcerative colitis and CrohnÕs disease [22]. Experimental models of IBD and clinical studies point to a dysregulation of the intestinal immune response, but no suitable therapy has proven effective for most. For example, treatments of patients aimed at decreasing activity of TNFa or a4b7 integrins, that decrease leukocyte trafficking to the intestine, have had beneficial effects, however, there are side effects and a significant percentage do not respond. There is consensus, however, that blockage of macrophage infiltration into the intestine is a good therapeutic strategy for IBD. Our results and those of others have led to the hypothesis that meprins play a role in the pathogenesis of IBDs. Data indicate that meprins influence IBD by affecting intestinal leukocyte dissemination to inflammatory sites in the gut, by interacting with bacteria at the epithelial surface, by degradation of compounds such as defensins that kill bacteria, or by exacerbating host tissue damage in the inflamed gut. Previous studies had demonstrated high expression of meprin subunits in leukocytes of the lamina propria of human inflammatory sites [23]. This observation, plus the known ability of meprins to hydrolyze extracellular matrix proteins, led to the speculation that meprins play a role in the movement of macrophages to inflammatory sites. Our studies with meprin b knockout mice, and macrophages from the lymph nodes of mice with an experimentally induced form of IBD, have strengthened our proposal that meprin b plays an important role in intestinal leukocyte dissemination to the gut [2]. These observations need to be extended to show that macrophages of meprin b/ mice do in fact move less efficiently to sites of inflammation in vivo, and once there do less damage. The implication of meprin b in the disease process warrants investigation of administration of inhibitors of the protease in vivo to determine whether pathology is lessened. The contribution of the bacterial flora in the gut to IBD is unclear. However, it is known that germ-free mice do not develop IBD, thus bacteria likely have a determinative role in the process. In the course of studying the meprin b knockout mice we discovered that the aerobic intestinal bacterial flora of the null mice were distinct from the wild-type (unpublished observations). Thus, there is a difference in the bacterial profile for the genotypes. This difference was worsened upon administration of DSS, an intestinal irritant. In general, the adverse effects of the insult correlated with the ÔloadÕ of aerobic gramnegative bacteria. The fact that the bacterial profile is altered in mice deficient in meprin b provides a clue that the effects of meprins in the progression of IBD involve interactions of meprin proteases with bacteria or their products, or in the modulation of host molecules, such as defensins that kill bacteria.
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
5. Meprin upregulation in cancer cells Cultured human cancer cell lines from a variety of tissues express a unique isoform of meprin b [24–26]. In cancer cells, the human MEP1B gene is expressed from a transcription start site about 70 bp upstream from the normal site used in the kidney [26]. Use of the upstream start site results in a slightly longer 5 0 untranslated region of the mRNA, but no changes to the protein coding regions. Meprin b mRNA expression is increased in cancer cells by tumor promoters, consistent with the presence of multiple AP-1 and PEA3 elements (Fos/Jun and Ets binding sites) in the MEP1B promoter, whereas these elements are rarely found in the MEP1A promoter [27]. Both meprin a and b are expressed in mesenteric lymph nodes, where meprin b affects the migration of intestinal leukocytes [2]. It has also been shown that meprin a is secreted from both the apical and basolateral surfaces of the human colon cancer cell line Caco-2, supplying proteolytic activity to the stromal compartment [23]. Active meprin a was also detected in a majority of colorectal tumor samples. MeprinÕs potential role in cell migration and extracellular matrix remodeling is further supported by in vitro observations that meprin a cleaves extracellular matrix components, such laminin-1 and 5, gelatin, and fibronectin [28,9,8]. To examine the proposition that meprins are involved in the progression of tumor cells from a non-invasive /non-metastatic phenotype to a more aggressive phenotype, meprin expression was studied in cells with differing metastatic potential. One pair of cells examined was human colon cancer cells SW480 and SW620. SW480 cells were established from a primary colon tumor, while the SW620 cells were isolated from a lymph node metastasis of colon tumor cells in the same individual a year later [29]. The SW480/SW620 cell system has been validated as a model for colon cancer progression [30]. SW620 cells are more locally invasive and form more liver metastases from intrasplenic injection sites. In addition, SW620 cells produce more of the secreted matrix metalloprotease MMP-7 [31]. In addition, the endogenous expression of meprin subunits in the human hepatoma cell lines, HepG2 and Huh7, was investigated since colon cancer cells frequently metastasize to the liver. Using the constitutive glyceraldehyde phosphate dehydrogenase (GAPDH) mRNA as a template loading control, the amount of meprin a and b mRNA in SW480, SW620, Huh7, and HepG2 cells were compared by end-point RT-PCR (Fig. 4). The metastatic SW620 cells express 3-fold more meprin a mRNA than the nonmetastatic SW480 cells, while meprin b levels in SW480 and SW620 are similar. These results indicate that the increase in metastatic potential of SW620 cells correlates with the levels of meprin a expression. The amount of meprin a mRNA in HepG2 cells were at least 100-fold more than in SW620, while meprin b mRNA levels in HepG2 cells were less than either SW480 or SW620. Direct comparison of HepG2 and Huh7 for expression of meprin a indicated that these hepatocarcinoma cell lines differed in their level of meprin a expression. HepG2 cells expressed significantly more meprin a mRNA than Huh7 cells. In addition to RNA, the presence of secreted meprin a protein in media collected from HepG2 and Huh7 cell cultures was examined by Western blot analysis and activity assays. Data from mouse and human systems indicate that overexpression of meprin a relative to meprin b affects the extracellular localization of the meprin a subunit. The covalent
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
3321
tion of secreted meprin a in hepatocarcinomas requires serine proteases provided by the stromal compartment. Release of meprin a from the cell surface to the extracellular mileu has implications for cancer cell invasion and metastasis. The high level of meprin a in HepG2 cell line presents a model for further studies on the role of secreted meprin a in cancer cells.
6. Conclusions
Fig. 4. End-point RT-PCR analysis of meprin expression in human cultured liver cancer cells (HepG2 and Huh7), and in human cultured colon cancer cells SW480 (480) and SW620 (620). When normalized to the GAPDH loading controls, meprin a expression is higher in the metastatic SW620 cells compared to the SW480 primary tumor cells. Likewise meprin a mRNA levels are higher in HepG2 cells compared to the Huh7 liver cancer cells. GAPDH, glyceraldehyde phosphate dehydrogenase.
association of meprin a and b subunits results in the retention of meprin a at the apical surface of kidney and intestinal brush border epithelial cells [32]. However, in tissues with low meprin b levels, or in meprin b knock-out mice, the meprin a subunit is released from the cell surface, indicating that meprin a does not bind to other cell surface proteins [33,34]. Based on the differential expression of meprin a and b mRNA determined by RT-PCR, it is predicted that the overexpression of meprin a mRNA relative to meprin b mRNA in HepG2 cells would result in the release of meprin a protein into the culture media. When conditioned media from HepG2 and Huh7 cells was tested for the presence of meprin a protein by Western blot, only HepG2 cells produced detectable meprin a. The lower level of meprin a mRNA in Huh7 cells in comparison the HepG2 implies that Huh7 cells produce less meprin a protein and it is retained at the cell surface in a/b complexes. The lack of detectable meprin a in Huh7 conditioned media supports this hypothesis. Meprin a is secreted in a latent form from Caco-2 colon cancer cells and activated by the serine protease plasmin, which can be provided by stromal fibroblasts [7]. Plasmin activation of human meprin is specific for meprin a and does not occur with meprin b [35]. Proteolytic activity assays indicate that the meprin a protein secreted by HepG2 cells is in the latent form. Conditioned HepG2 media alone had no meprin a proteolytic activity in a standard bradykinin assay [8]. Proteolytic activity was detected, however, upon pretreatment of the media with trypsin, a serine protease that removes the propeptide of human meprin a, indicating that the HepG2 secreted meprin a protein was latent. As in the colon cancer system, the activa-
Meprins are highly regulated metalloproteinases, as indicated by both tissue-specific expression patterns and differential expression patterns observed during birth and weaning. The data indicate an important function for meprins in several disease states, including acute renal failure, cancer, IBD, and UTI. Numerous studies have demonstrated that the location and downregulation of meprins in the kidney play a key role in the severity and progression of renal injury, with meprins exacerbating the diseased condition. Upregulation of meprins in cancer cells along with the ability of meprins to cleave extracellular matrix proteins implicates a role in the growth, invasion, and metastasis of cancer cells. Meprins may also have a role in inflammation, since meprin subunit expression is increased in both an in vivo model of IBD as a result of leukocyte infiltration and also in the urine from women with acute UTI. The diverse expression pattern of meprins among a variety of disease conditions warrants further investigation of this unique protease. A more thorough understanding of how meprins are regulated will provide insight into progression of and ultimately better therapies for these conditions.
References [1] Bertenshaw, G.P. and Bond, J.S. (2004) Meprin A; Meprin B in: Handbook of Proteolytic Enzymes (Barrett, A.J., Woessner, F. and Rawlings, N., Eds.), pp. 599–605, Academic Press, London, 605–609. [2] Crisman, J.M., Zhang, B., Norman, L.P. and Bond, J.S. (2004) Deletion of the mouse meprin b metalloproteinase gene diminishes the ability of leukocytes to disseminate through extracellular matrix. J. Immunol. 172, 4510–4519. [3] Bond, J.S. and Beynon, R.J. (1986) Meprin: a membrane-bound metalloendopeptidase. Curr. Top. Cell Regul. 28, 263–290. [4] Villa, J.P., Bertenshaw, G.P., Bylander, J.E. and Bond, J.S. (2003) Meprin proteolytic complexes at the cell surface and in extracellular spaces. Biochem. Soc. Symp. 70, 53–63. [5] Hengst, J.A. and Bond, J.B. (2004) Transport of meprin subunits through the secretory pathway: role of the transmembrane and cytoplasmic domains and oligomerization. J. Biol. Chem. 279, 34856–34864. [6] Bond, J.S., Rojas, K., Overhauser, J., Zoghbi, H.Y. and Jiang, W. (1995) The structural genes MEP1A and MEP1B, for the a and b subunits of the metalloendopeptidase meprin map to human chromosome 6p and 18q, respectively. Genomics 25, 300–303. [7] Rosmann, S., Hahn, D., Lottaz, D., Kruse, M.N., Stocker, W. and Sterchi, E.E. (2002) Activation of human meprin-alpha in a cell culture model of colorectal cancer is triggered by the plasminogen-activating system. J. Biol. Chem. 277, 40650–40658. [8] Bertenshaw, G.P., Turk, B.E., Hubbard, S.J., Matters, G.L., Bylander, J.E., Crisman, J.M., Cantley, L.C. and Bond, J.S. (2001) Marked differences between metalloproteases meprin A and B in substrate and peptide bond specificity. J. Biol. Chem. 276, 13248–13255. [9] Kruse, M.N., Becker, C., Lottaz, D., Kohler, D., Yiallouros, I., Krell, H.W., Sterchi, E.E. and Stocker, W. (2004) Human meprin alpha and beta homo-oligomers; cleavage of basement membrane
3322
[10] [11] [12]
[13]
[14]
[15] [16]
[17]
[18] [19] [20]
[21] [22]
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322 proteins and sensitivity to metalloprotease inhibitors. Biochem. J. 378, 383–389. Walker, P.D., Kaushal, G.P. and Shah, S.V. (1998) Meprin A, the major matrix degrading enzyme in renal tubules, produces a novel nidogen fragment in vitro and in vivo. Kidney Int. 53, 1673–1680. Carmago, S., Shah, S.V. and Walker, P.D. (2002) Meprin, a brush-border enzyme, plays an important role in hypoxic/ischemic acute renal tubular injury in rats. Kidney Int. 61, 959–966. Norman, L.P., Matters, G.L., Crisman, J.M and Bond, J.S. (2003) Expression of meprins in health and disease. In: Cell Surface Proteases. Current Topics in Developmental Biology, 54, 145– 166. Kumar, J.M. and Bond, J.S. (2001) Developmental expression of meprin metalloprotease subunits in ICR and C3H/He mouse kidney and intestine in the embryo, postnatally and after weaning. Biochem. Biophys. Acta 1518, 106–114. Bond, J.S., Beynon, R.J., Reckelhoff, J.F. and David, C.S. (1984) Mep-1 gene controlling a kidney metalloendopeptidase is linked to the major histocompatibility complex in mice. Proc. Natl. Acad. Sci. USA 81, 5542–5545. Trachtman, H., Valderrama, E., Dietrich, J.M. and Bond, J.S. (1995) The role of meprin A in the pathogenesis of acute renal failure. Biochem. Biophys. Res. Commun. 208, 498–505. Sadlier, D.M., Connolly, S.B. and Kieran, N.E. (2004) Sequential extracelluar matrix-focused and baited-global cluster analysis of serial transcriptiomic profiles identifies candidate modulators of renal tubulointerstitial fibrosis in murine adriamycin induced nephropathy. J. Biol. Chem. 279, 29670–29680. Ricardo, J.S., Bond, J.S., Johnson, G.D., Kaspar, J. and Diamond, J.R. (1996) Expression of subunits of the metalloendopeptidase meprin in the rat renal cortex during experimental hydronephrosis. Am. J. Physiol. (Renal) 270, F669–F676. Kieran, N.E., Doran, P.P. and Connolly, S.B. (2003) Modification of the transcriptomic response to renal ischemia/reperfusion injury by lipoxin analog. Kidney Int. 64, 480–492. Sampson, N.S., Ryan, S.T. and Enke, D.A. (2001) Global gene expression analysis reveals a role for the alpha 1 integrin in renal pathogenesis. J. Biol. Chem. 276, 34182–34188. Hopkins, W.J., Elkahwaji, J.E., Heisey, D.M. and Ott, C.J. (2003) Inheritance of susceptibility to induced Escherichia coli bladder and kidney infections in female C3H/HeJ mice. J. Infect. Dis. 187, 418–423. Bouma, G. and Strober, W. (2003) The immunological and genetic basis of inflammatory bowel disease. Nat. Rev. Immunol. 3, 521–533. Yap, L.M., Ahmad, T. and Jewell, D.P. (2004) The contribution of HLA genes to IBD susceptibility and phenotype. Best Pract. Res. Clin. Gastroenterol. 18, 577–596.
[23] Lottaz, D., Maurer, C.A., Hahn, D., Buchler, M.W. and Sterchi, E.E. (1999) Nonpolarized secretion of human meprin alpha in colorectal cancer generates and increased proteolytic potential in stroma. Cancer Res. 59, 1127–1133. [24] Dietrich, J.M., Jiang, W. and Bond, J.S. (1996) A novel meprin b mRNA in mouse embryonal and human colon carcinoma cells. J. Biol. Chem. 271, 2271–2278. [25] Matters, G.L. and Bond, J.S. (1999) Expression and regulation of the meprin b gene in human cancer cells. Mol. Carcinogen. 25, 169–178. [26] Matters, G.L. and Bond, J.S. (1999) Meprin B: transcriptional and posttranscriptional regulation of the meprin beta metalloproteinase subunit in human and mouse cancer cells. APMIS 107, 19–27. [27] Hahn, D., Illisson, R., Metspalu, A. and Sterchi, E.E. (2000) Human N-benzoyl-L-tyrosyl-p-aminobenzoic acid hydrolase (human meprin): genomic structure of the alpha and beta subunits. Biochem. J. 346, 83–91. [28] Kohler, D., Kruse, M., Stocker, W. and Sterchi, E.E. (2000) Heterologously overexpressed, affinity-purified human meprin a is functionally active and cleaves components of the basement membrane in vitro. FEBS Lett. 465, 2–7. [29] Leibovitz, A., Stinson, J.C., McCombs, W.B. and McCoy, C.E. (1976) Classification of human colorectal adenocarcinoma cell lines. Cancer Res. 36, 4562–4569. [30] Hewitt, R.E., McMarlin, A., Kleiner, D., Wersto, R., Martin, P., Tsoskas, M., Stamp, G.W.H. and Stetler-Stevenson, W.G. (2000) Validation of a model of colon cancer progression. J. Pathol. 192, 446–454. [31] Witty, J.P., McDonnell, S., Newell, K.J., Cannon, P., Nare, M., Tressler, R.J. and Matrisian, L.M. (1994) Modulation of matrilysin levels in colon carcinoma cell lines affects tumorigenicity in vivo. Cancer Res. 54, 4805–4812. [32] Marchand, P., Tang, J. and Bond, J.S. (1994) Membrane association and oligomeric organization of the alpha and beta subunits of mouse meprin A. J. Biol. Chem. 269, 15388–15393. [33] Norman, L.P., Jiang, W., Han, X., Saunders, T.L. and Bond, J.S. (2003) Targeted disruption of the meprin b gene in mice leads to under representation of knock-out mice and changes in renal gene expression profiles. Mol. Cell. Biol. 23, 1221–1230. [34] Lottaz, D., Hahn, D., Muller, S., Muller, C. and Sterchi, E.E. (1999) Secretion of human meprin from intestinal epithelial cells depends on differential expression of the alpha and beta subunits. Eur. J. Biochem. 259, 496–504. [35] Becker, C., Kruse, M.N., Slotty, K.A., Kohler, D., Harris, J.R., Rosmann, S., Sterchi, E.E. and Stocker, W. (2003) Differences in the activation mechanism between the a and b subunits of human meprin. Biol. Chem. 384, 825–831.
FEBS Letters 579 (2005) 3317–3322
Minireview
Meprin metalloprotease expression and regulation in kidney, intestine, urinary tract infections and cancer Judith S. Bond*, Gail L. Matters, Sanjita Banerjee, Renee E. Dusheck Department of Biochemistry and Molecular Biology, The Pennsylvania State University College of Medicine, Hershey, PA 17033, USA Accepted 23 March 2005 Available online 31 March 2005 Edited by Pe´ter Friedrich
Abstract Meprins are unique plasma membrane and secreted metalloproteinases that are highly regulated at the transcriptional and post-translational levels. Meprin a and b subunits are abundantly expressed in kidney and intestinal epithelial cells, are secreted into the urinary tract and intestinal lumen, and are found in leukocytes and cancer cells under certain conditions. Their location and proteolytic activities indicate functions at the interface of the host and the external environment, and in trafficking of macrophages and metastases of cancer cells. These proteases can be detrimental when there is tissue damage or disruption, as in acute renal injury or intestinal inflammation, and there is evidence they are involved in movement of leukocytes and cancer cells to sites of infection or in metastasis, respectively. Ó 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Keywords: Metalloprotease; Kidney; Intestine; Inflammation; Infection; Cancer
1. Introduction Meprins are highly expressed at the brush border membrane of proximal tubule cells of the kidney and epithelial cells of the intestine, particularly in the ileum and colon, and in mesenteric leukocytes [1,2]. For example, meprin proteins were estimated to compose approximately 5% of the brush border membrane protein in rodent kidneys, second only to actin in abundance [3]. These proteases are multidomain, highly glycosylated proteins that act extracellularly, and are either bound to the plasma membrane by transmembrane segments, or are secreted or shed into extracellular spaces [4,5]. Meprin proteases are composed of two evolutionarily related subunits, a and b, that are approximately 50% identical at the amino acid level. The subunits are encoded on two genes: the a gene is on human chromosome
* Corresponding author. Fax: +1 717 531 7072. E-mail address: [email protected] (J.S. Bond).
Abbreviations: DSS, dextran sulfate sodium; IBD, inflammatory bowel disease; UTI, urinary tract infection; RT-PCR, reverse transcriptasepolymerase chain reaction; GAPDH, glyceraldehyde phosphate dehydrogenase
6 (mouse 17) near the histocompatibility complex; the b subunit on chromosome 18 in both the mouse and human genomes [6]. The subunits can be expressed separately, or coordinately, and consequently they form homo or heterocomplexes where the basic unit is a disulfide-linked dimer. Both subunits are synthesized as membrane spanning proteins with small cytoplasmic tails, and the bulk of the protein is extracellular. However, the meprin a subunit is proteolytically cleaved during biosynthesis so that it loses its transmembrane domain. Thus, meprin a exists at the membrane only when it associates with meprin b, which exists as a type I protein at the plasma membrane. Membrane forms of meprins are dimeric (homodimeric b; meprin B) or tetrameric (hetero a2b2 or a3b2 complexes) (see Fig. 1A). The secreted form of meprin a tends to oligomerize into high molecular mass complexes containing 10–100 subunits (dimers that associate noncovalently), forming the largest proteases known to exist (see Fig. 1B). The self-associating homo-oligomeric complexes of meprin A are secreted as latent proteases (containing the prosequence) and can move through extracelluar spaces in a non-destructive manner, and deliver a concentrated form of this metalloproteinase to sites that have activating proteases, such as site of inflammation, infection or cancerous growth [4,7]. Thus, meprin structures provide means to concentrate proteolytic activity at the cell membrane (through homo–hetero-oligomeric complexes) and extracellularly (through homotypic interactions). Meprins are capable of cleaving a wide range of substrates, from bioactive peptides, such as gastrin and angiotensin, to extracellular matrix proteins, such as collagen IV, fibronectin and nidogen [8,9]. The heterooligomeric form of meprin was identified as a major matrix degrading protease in studies of rat kidney [10,11]. The peptide bond specificities of meprins a and b differ substantially, indicating different functions, however, they can both hydrolyze extracellular matrix proteins. Among the best substrates for meprin b are gastrin and osteopontin, which are growth factors and chemokines [8]. Meprin a, but not b, cleaves cytokines such as monocyte chemoattractant protein-1, and this could be important in immune responses [12]. Meprin metalloproteases are highly regulated at the transcriptional and post-translational level and are normally restricted to specific locations, but they can be detrimental to host tissues when there is tissue damage and disruption of cells, e.g., in intestinal inflammation and kidney disruption.
0014-5793/$30.00 Ó 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2005.03.045
3318
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
Fig. 1. (A) Domain and oligomeric structure of meprins a and b. Domains are S, signal sequence; Pro, prosequence; Protease, catalytic domain; MAM, meprin, A5 protein, protein tyrosine phosphatase l; MATH, meprin and TRAF homology; AM, after MATH; I, inserted domain; EGF, epidermal growth factor-like, TM, transmembrane-spanning; C, cytoplasmic domain. The MATH and AM domains are now designated the TRAF domain (4,5). (B) Electron micrograph of various forms of secreted, latent homooligomeric meprin A demonstrating high molecular mass complexes. Row 1, closed rings and crescents (containing about 10 to 12 meprin a subunits); rows 2 and 3 show tubes and spirals containing up to 100 subunits.
2. Expression of meprins in development, in leukocytes, and in germ-free mice In situ hybridization studies of embryonic and adult mice and immunohistochemistry demonstrated the tissue-specific expression of meprin subunits in the epithelial cells of kidney and intestine only [13]. These studies also demonstrated that there was differential regulation of meprins a and b after birth. Meprin a was present in the fetus by day 11, increased expression after birth, and expression was precipitously decreased in the intestine after weaning (21 days). By contrast, meprin b increased during fetal growth and after birth, and became the predominant isoform of the enzyme in the intestine after weaning. The kidney expression of mouse meprin subunits were strain-dependent; all strains express both subunits during fetal stages, some strains increase both subunits after birth (e.g., C57BL/6) while others only express meprin b, and down regulate meprin a (e.g., C3H/He). The consequence of the expression patterns in mice is that meprin b is the predominant form of meprin in the adult intestine; while meprin a is abundantly expressed in kidney of most adult mice, it is not expressed in certain inbred strains of mice (mainly strains of the H-2k haplotype, e.g., CBA, C3H, AKR). Low meprin a expression in the kidney is inherited as an autosomal recessive trait [14]. It has been known for some time that meprin b is highly expressed on the brush border membrane of the ileum (see Fig. 2). Meprin a and b are also expressed in leukocytes of the mouse mesenteric lymph node [2]. When mice were given dextran sulfate sodium (DSS) in their drinking water, to generate an experimental model for inflammatory bowel disease (IBD), only meprin b was present in lymph nodes of DSS-treated mice [2]. In addition, leukocytes from meprin b null mice are deficient in their ability to invade into matrigel. Other functional properties of the meprin b/ leukocytes, such as phagocytosis and superoxide release were not different in the / and +/+ genotypes. The presence of meprin b in the draining lymph node of the intestine, coupled with the decreased ability of meprin b-macrophages to migrate through matrigel implicates
Fig. 2. Immunofluorescent localization of the meprin b metalloprotease on the brush border membrane of the mouse ileum and in macrophages in the lamina propria during inflammation induced by dextran sulfate sodium. Meprin b, green; macrophage-specific protein F4/80, red; colocalization of the latter two proteins appears yellow. Nuclei, blue.
this protease in leukocyte transmigration important to the immune response, and to IBD which is aggravated by leukocyte infiltration. To test whether meprins are induced by the microbial flora of the intestine, RNA samples were obtained (from Dr. Jeffrey GordonÕs laboratory at Washington University) from mice that were either germ-free, colonized with Bacteriodes thetaiotaomicron at day 10, or conventionalized at 10 days. The mice were on two different strain backgrounds, NMRI and C57BL/ 6. Reverse transcriptase-polymerase chain reaction (RT-PCR) analyses showed no difference between the expression of meprin a or b for any of these mice. Thus, the microbial flora are not responsible for induction of intestinal meprins, but rather meprins are innately present in the mouse gut.
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
3319
3. Regulation of meprins in kidney and urinary tract disease High levels of meprin a at the brush border membrane in rodent kidneys have been implicated in tissue damage caused by insults such as ischemia, while low levels of meprin a and b have been associated with fibrosis [15,11,16]. Acute renal failure is a common disorder that affects about 5% of all hospitalized patients, and has a mortality rate of over 50%. An understanding of the cellular and molecular mechanisms of kidney injury is critical to learning how to improve mortality. Studies of inbred strains of mice subjected to ischemia reperfusion have revealed that mice that express low levels of kidney meprin a as adults (C3H/He, CBA) have markedly less kidney damage compared to those that express high levels of kidney meprin a (ICR, C57BL/6) [15]. Studies with rats have demonstrated that high levels of active meprin A can be cytotoxic [11]. The latter authors found active meprin A was cytotoxic to renal cells in culture (LLC-PK1 and MDCK cells) and that actinonin, a potent inhibitor of meprin a, could protect against meprin damage in vitro and in vivo in a model ischemia-reperfusion injury. They also demonstrated that after acute ischemic renal tubule damage there is a shift in meprin location form the brush border membrane to cytoplasmic compartments and to the tubular lumen. One interpretation of these observations is that once tissue is damaged, high levels of active meprins are found in places other than the apical membrane (cytoplasm, basolateral side of the cell, and cell–cell junctions) and are damaging. Inhibitors of meprins can protect against this type of damage. Multiple studies have demonstrated that meprin expression is downregulated in models of renal injury that lead to acute renal failure. For example, renal pathology associated with hydronephrosis (unilateral urethral obstruction) occurs with an early and progressive decline in rat meprin a and b subunit mRNA and protein [17]. Microarray analysis demonstrated marked declines in mouse meprin b expression in acute renal ischemic reperfusion injury and in adriamycin-induced nephropathy [18,16]. The latter authors also reported that meprin b is downregulated during transdifferentiation of renal tubule epithelial cells to a fibroblast-like phenotype in vitro on exposure to TGF-ß and EGF, suggesting a role in fibrogenesis. Meprin b expression is also downregulated in collagen IVA3 knockout mice that develop AlportÕs syndrome and renal pathology [19]. It appears that meprins are downregulated in a variety of situations that lead to cell death through apoptosis or necrosis, and are upregulated in specific highly differentiated cells and in some proliferating cells (such as cancer cells and leukocytes under certain conditions). The location, balance and regulation of meprins are important in kidney disease progression. Urinary tract infections are one of the most common bacterial infections for mankind, and around 50% of women are affected at one point in their lifetime. Recurrent urinary tract infections are also one of the problems after kidney transplantation, and this appears to be related to ischemic conditions and time taken between removal of the kidney and transplantation. There are multiple factors involved in recurrent infections, however, there is a need to understand the mechanisms involved to develop more effective therapies, increase the incidence of renal allograft survival, and prevent renal scarring associated with infection. The secreted form of meprin a is in an ideal location for preventing or lessening
Fig. 3. Western blot of meprin a in human urine from healthy women and women with acute urinary tract infections. Concentrated urine samples (40 ll each) were immunoblotted with a human recombinant meprin a antibody. Low or no meprin immunoreactivity was observed in samples from healthy subjects (sample I.D. 31, 32, 33, and 34) while high meprin immunoreactivity was observed in samples from subjects suffering from acute UTI (sample I.D. 52, 53, 56, and 57).
infections of the urinary tract. Experimental models of urinary tract infections have established that there are mouse strain differences in susceptibility to bacterial infection, and that they are related to the major histocompatibility complex, the area of the genome that encodes meprin a. There is clear evidence that C3H/He mice are more susceptible to infections by uropathic bacteria than Balb/c mice [20]. C3H/He mice express low meprin a in adult kidney, while Balb/c mice express high levels. Moreover, susceptibility is recessive, as is low meprin a expression. Our hypothesis is that meprin a is one of the factors that determine the susceptibility to urinary tract diseases. Meprin a is poised to be activated in the urinary tract; it is secreted from the proximal tubule of the kidney into the lumen as an inactive complex with high proteolytic potential, and can be readily activated by trypsin-like proteases (e.g., subtilisin). Meprin a and b knockout mice will be valuable tools to determine whether in fact meprins are important factors for susceptibility to urinary tract infections by uropathic bacteria. Little is known about meprin in human urinary tract infections (UTIs). To examine meprin expression in human urine, voided urine samples from premenopausal women with a history of frequent UTIs, without a history of frequent UTIs, and with an acute UTI were collected for analysis. The urine samples were concentrated 10-fold, and their total protein and creatinine concentrations were determined. Quantitative Western blotting showed that meprin a protein is detectable in the urine of premenopausal women at low concentrations, and that concentrations are variable from subject to subject. No correlation between the urinary levels of meprin protein and the susceptibility of women to UTIs was observed. Eighty-nine percent of healthy women had low or undetectable levels of urinary meprin, similar to certain inbred strains of mice (e.g., C3H/He mice). In contrast, the majority (86%) of women suffering from an acute UTIs had high (greater than 5.2 ng) or very high (greater than 9.1 ng) levels of meprin protein in the urine (Fig. 3). These results are in concordance with studies of meprin expression in the intestine, where meprin subunit expression is markedly increased during inflammation as a result of leukocyte infiltration.
4. Meprins in intestinal disease Meprin expression in the intestinal tract is highest in the ileum and large intestine where host and microorganisms are in contact, and where intestinal inflammatory diseases develop.
3320
Inflammatory bowel diseases (IBDs) are a collection of clinically heterogeneous intestinal diseases that result in damage to intestinal epithelium and affect 1 million people in the USA [21]. The two major forms of IBD, ulcerative colitis and CrohnÕs disease, involve chronic inflammation of the colon or the ileum plus colon, respectively, and are thought to be caused by genetic and environmental factors that affect host-microbe interactions and production of inflammatory mediators. Candidate genes for susceptibility to IBD have been identified, and these include the NOD2 gene on chromosome 16 (for CrohnÕs), and susceptibility loci on chromosome 19p13 and the HLA region on chromosome 6 for both ulcerative colitis and CrohnÕs disease [22]. Experimental models of IBD and clinical studies point to a dysregulation of the intestinal immune response, but no suitable therapy has proven effective for most. For example, treatments of patients aimed at decreasing activity of TNFa or a4b7 integrins, that decrease leukocyte trafficking to the intestine, have had beneficial effects, however, there are side effects and a significant percentage do not respond. There is consensus, however, that blockage of macrophage infiltration into the intestine is a good therapeutic strategy for IBD. Our results and those of others have led to the hypothesis that meprins play a role in the pathogenesis of IBDs. Data indicate that meprins influence IBD by affecting intestinal leukocyte dissemination to inflammatory sites in the gut, by interacting with bacteria at the epithelial surface, by degradation of compounds such as defensins that kill bacteria, or by exacerbating host tissue damage in the inflamed gut. Previous studies had demonstrated high expression of meprin subunits in leukocytes of the lamina propria of human inflammatory sites [23]. This observation, plus the known ability of meprins to hydrolyze extracellular matrix proteins, led to the speculation that meprins play a role in the movement of macrophages to inflammatory sites. Our studies with meprin b knockout mice, and macrophages from the lymph nodes of mice with an experimentally induced form of IBD, have strengthened our proposal that meprin b plays an important role in intestinal leukocyte dissemination to the gut [2]. These observations need to be extended to show that macrophages of meprin b/ mice do in fact move less efficiently to sites of inflammation in vivo, and once there do less damage. The implication of meprin b in the disease process warrants investigation of administration of inhibitors of the protease in vivo to determine whether pathology is lessened. The contribution of the bacterial flora in the gut to IBD is unclear. However, it is known that germ-free mice do not develop IBD, thus bacteria likely have a determinative role in the process. In the course of studying the meprin b knockout mice we discovered that the aerobic intestinal bacterial flora of the null mice were distinct from the wild-type (unpublished observations). Thus, there is a difference in the bacterial profile for the genotypes. This difference was worsened upon administration of DSS, an intestinal irritant. In general, the adverse effects of the insult correlated with the ÔloadÕ of aerobic gramnegative bacteria. The fact that the bacterial profile is altered in mice deficient in meprin b provides a clue that the effects of meprins in the progression of IBD involve interactions of meprin proteases with bacteria or their products, or in the modulation of host molecules, such as defensins that kill bacteria.
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
5. Meprin upregulation in cancer cells Cultured human cancer cell lines from a variety of tissues express a unique isoform of meprin b [24–26]. In cancer cells, the human MEP1B gene is expressed from a transcription start site about 70 bp upstream from the normal site used in the kidney [26]. Use of the upstream start site results in a slightly longer 5 0 untranslated region of the mRNA, but no changes to the protein coding regions. Meprin b mRNA expression is increased in cancer cells by tumor promoters, consistent with the presence of multiple AP-1 and PEA3 elements (Fos/Jun and Ets binding sites) in the MEP1B promoter, whereas these elements are rarely found in the MEP1A promoter [27]. Both meprin a and b are expressed in mesenteric lymph nodes, where meprin b affects the migration of intestinal leukocytes [2]. It has also been shown that meprin a is secreted from both the apical and basolateral surfaces of the human colon cancer cell line Caco-2, supplying proteolytic activity to the stromal compartment [23]. Active meprin a was also detected in a majority of colorectal tumor samples. MeprinÕs potential role in cell migration and extracellular matrix remodeling is further supported by in vitro observations that meprin a cleaves extracellular matrix components, such laminin-1 and 5, gelatin, and fibronectin [28,9,8]. To examine the proposition that meprins are involved in the progression of tumor cells from a non-invasive /non-metastatic phenotype to a more aggressive phenotype, meprin expression was studied in cells with differing metastatic potential. One pair of cells examined was human colon cancer cells SW480 and SW620. SW480 cells were established from a primary colon tumor, while the SW620 cells were isolated from a lymph node metastasis of colon tumor cells in the same individual a year later [29]. The SW480/SW620 cell system has been validated as a model for colon cancer progression [30]. SW620 cells are more locally invasive and form more liver metastases from intrasplenic injection sites. In addition, SW620 cells produce more of the secreted matrix metalloprotease MMP-7 [31]. In addition, the endogenous expression of meprin subunits in the human hepatoma cell lines, HepG2 and Huh7, was investigated since colon cancer cells frequently metastasize to the liver. Using the constitutive glyceraldehyde phosphate dehydrogenase (GAPDH) mRNA as a template loading control, the amount of meprin a and b mRNA in SW480, SW620, Huh7, and HepG2 cells were compared by end-point RT-PCR (Fig. 4). The metastatic SW620 cells express 3-fold more meprin a mRNA than the nonmetastatic SW480 cells, while meprin b levels in SW480 and SW620 are similar. These results indicate that the increase in metastatic potential of SW620 cells correlates with the levels of meprin a expression. The amount of meprin a mRNA in HepG2 cells were at least 100-fold more than in SW620, while meprin b mRNA levels in HepG2 cells were less than either SW480 or SW620. Direct comparison of HepG2 and Huh7 for expression of meprin a indicated that these hepatocarcinoma cell lines differed in their level of meprin a expression. HepG2 cells expressed significantly more meprin a mRNA than Huh7 cells. In addition to RNA, the presence of secreted meprin a protein in media collected from HepG2 and Huh7 cell cultures was examined by Western blot analysis and activity assays. Data from mouse and human systems indicate that overexpression of meprin a relative to meprin b affects the extracellular localization of the meprin a subunit. The covalent
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322
3321
tion of secreted meprin a in hepatocarcinomas requires serine proteases provided by the stromal compartment. Release of meprin a from the cell surface to the extracellular mileu has implications for cancer cell invasion and metastasis. The high level of meprin a in HepG2 cell line presents a model for further studies on the role of secreted meprin a in cancer cells.
6. Conclusions
Fig. 4. End-point RT-PCR analysis of meprin expression in human cultured liver cancer cells (HepG2 and Huh7), and in human cultured colon cancer cells SW480 (480) and SW620 (620). When normalized to the GAPDH loading controls, meprin a expression is higher in the metastatic SW620 cells compared to the SW480 primary tumor cells. Likewise meprin a mRNA levels are higher in HepG2 cells compared to the Huh7 liver cancer cells. GAPDH, glyceraldehyde phosphate dehydrogenase.
association of meprin a and b subunits results in the retention of meprin a at the apical surface of kidney and intestinal brush border epithelial cells [32]. However, in tissues with low meprin b levels, or in meprin b knock-out mice, the meprin a subunit is released from the cell surface, indicating that meprin a does not bind to other cell surface proteins [33,34]. Based on the differential expression of meprin a and b mRNA determined by RT-PCR, it is predicted that the overexpression of meprin a mRNA relative to meprin b mRNA in HepG2 cells would result in the release of meprin a protein into the culture media. When conditioned media from HepG2 and Huh7 cells was tested for the presence of meprin a protein by Western blot, only HepG2 cells produced detectable meprin a. The lower level of meprin a mRNA in Huh7 cells in comparison the HepG2 implies that Huh7 cells produce less meprin a protein and it is retained at the cell surface in a/b complexes. The lack of detectable meprin a in Huh7 conditioned media supports this hypothesis. Meprin a is secreted in a latent form from Caco-2 colon cancer cells and activated by the serine protease plasmin, which can be provided by stromal fibroblasts [7]. Plasmin activation of human meprin is specific for meprin a and does not occur with meprin b [35]. Proteolytic activity assays indicate that the meprin a protein secreted by HepG2 cells is in the latent form. Conditioned HepG2 media alone had no meprin a proteolytic activity in a standard bradykinin assay [8]. Proteolytic activity was detected, however, upon pretreatment of the media with trypsin, a serine protease that removes the propeptide of human meprin a, indicating that the HepG2 secreted meprin a protein was latent. As in the colon cancer system, the activa-
Meprins are highly regulated metalloproteinases, as indicated by both tissue-specific expression patterns and differential expression patterns observed during birth and weaning. The data indicate an important function for meprins in several disease states, including acute renal failure, cancer, IBD, and UTI. Numerous studies have demonstrated that the location and downregulation of meprins in the kidney play a key role in the severity and progression of renal injury, with meprins exacerbating the diseased condition. Upregulation of meprins in cancer cells along with the ability of meprins to cleave extracellular matrix proteins implicates a role in the growth, invasion, and metastasis of cancer cells. Meprins may also have a role in inflammation, since meprin subunit expression is increased in both an in vivo model of IBD as a result of leukocyte infiltration and also in the urine from women with acute UTI. The diverse expression pattern of meprins among a variety of disease conditions warrants further investigation of this unique protease. A more thorough understanding of how meprins are regulated will provide insight into progression of and ultimately better therapies for these conditions.
References [1] Bertenshaw, G.P. and Bond, J.S. (2004) Meprin A; Meprin B in: Handbook of Proteolytic Enzymes (Barrett, A.J., Woessner, F. and Rawlings, N., Eds.), pp. 599–605, Academic Press, London, 605–609. [2] Crisman, J.M., Zhang, B., Norman, L.P. and Bond, J.S. (2004) Deletion of the mouse meprin b metalloproteinase gene diminishes the ability of leukocytes to disseminate through extracellular matrix. J. Immunol. 172, 4510–4519. [3] Bond, J.S. and Beynon, R.J. (1986) Meprin: a membrane-bound metalloendopeptidase. Curr. Top. Cell Regul. 28, 263–290. [4] Villa, J.P., Bertenshaw, G.P., Bylander, J.E. and Bond, J.S. (2003) Meprin proteolytic complexes at the cell surface and in extracellular spaces. Biochem. Soc. Symp. 70, 53–63. [5] Hengst, J.A. and Bond, J.B. (2004) Transport of meprin subunits through the secretory pathway: role of the transmembrane and cytoplasmic domains and oligomerization. J. Biol. Chem. 279, 34856–34864. [6] Bond, J.S., Rojas, K., Overhauser, J., Zoghbi, H.Y. and Jiang, W. (1995) The structural genes MEP1A and MEP1B, for the a and b subunits of the metalloendopeptidase meprin map to human chromosome 6p and 18q, respectively. Genomics 25, 300–303. [7] Rosmann, S., Hahn, D., Lottaz, D., Kruse, M.N., Stocker, W. and Sterchi, E.E. (2002) Activation of human meprin-alpha in a cell culture model of colorectal cancer is triggered by the plasminogen-activating system. J. Biol. Chem. 277, 40650–40658. [8] Bertenshaw, G.P., Turk, B.E., Hubbard, S.J., Matters, G.L., Bylander, J.E., Crisman, J.M., Cantley, L.C. and Bond, J.S. (2001) Marked differences between metalloproteases meprin A and B in substrate and peptide bond specificity. J. Biol. Chem. 276, 13248–13255. [9] Kruse, M.N., Becker, C., Lottaz, D., Kohler, D., Yiallouros, I., Krell, H.W., Sterchi, E.E. and Stocker, W. (2004) Human meprin alpha and beta homo-oligomers; cleavage of basement membrane
3322
[10] [11] [12]
[13]
[14]
[15] [16]
[17]
[18] [19] [20]
[21] [22]
J.S. Bond et al. / FEBS Letters 579 (2005) 3317–3322 proteins and sensitivity to metalloprotease inhibitors. Biochem. J. 378, 383–389. Walker, P.D., Kaushal, G.P. and Shah, S.V. (1998) Meprin A, the major matrix degrading enzyme in renal tubules, produces a novel nidogen fragment in vitro and in vivo. Kidney Int. 53, 1673–1680. Carmago, S., Shah, S.V. and Walker, P.D. (2002) Meprin, a brush-border enzyme, plays an important role in hypoxic/ischemic acute renal tubular injury in rats. Kidney Int. 61, 959–966. Norman, L.P., Matters, G.L., Crisman, J.M and Bond, J.S. (2003) Expression of meprins in health and disease. In: Cell Surface Proteases. Current Topics in Developmental Biology, 54, 145– 166. Kumar, J.M. and Bond, J.S. (2001) Developmental expression of meprin metalloprotease subunits in ICR and C3H/He mouse kidney and intestine in the embryo, postnatally and after weaning. Biochem. Biophys. Acta 1518, 106–114. Bond, J.S., Beynon, R.J., Reckelhoff, J.F. and David, C.S. (1984) Mep-1 gene controlling a kidney metalloendopeptidase is linked to the major histocompatibility complex in mice. Proc. Natl. Acad. Sci. USA 81, 5542–5545. Trachtman, H., Valderrama, E., Dietrich, J.M. and Bond, J.S. (1995) The role of meprin A in the pathogenesis of acute renal failure. Biochem. Biophys. Res. Commun. 208, 498–505. Sadlier, D.M., Connolly, S.B. and Kieran, N.E. (2004) Sequential extracelluar matrix-focused and baited-global cluster analysis of serial transcriptiomic profiles identifies candidate modulators of renal tubulointerstitial fibrosis in murine adriamycin induced nephropathy. J. Biol. Chem. 279, 29670–29680. Ricardo, J.S., Bond, J.S., Johnson, G.D., Kaspar, J. and Diamond, J.R. (1996) Expression of subunits of the metalloendopeptidase meprin in the rat renal cortex during experimental hydronephrosis. Am. J. Physiol. (Renal) 270, F669–F676. Kieran, N.E., Doran, P.P. and Connolly, S.B. (2003) Modification of the transcriptomic response to renal ischemia/reperfusion injury by lipoxin analog. Kidney Int. 64, 480–492. Sampson, N.S., Ryan, S.T. and Enke, D.A. (2001) Global gene expression analysis reveals a role for the alpha 1 integrin in renal pathogenesis. J. Biol. Chem. 276, 34182–34188. Hopkins, W.J., Elkahwaji, J.E., Heisey, D.M. and Ott, C.J. (2003) Inheritance of susceptibility to induced Escherichia coli bladder and kidney infections in female C3H/HeJ mice. J. Infect. Dis. 187, 418–423. Bouma, G. and Strober, W. (2003) The immunological and genetic basis of inflammatory bowel disease. Nat. Rev. Immunol. 3, 521–533. Yap, L.M., Ahmad, T. and Jewell, D.P. (2004) The contribution of HLA genes to IBD susceptibility and phenotype. Best Pract. Res. Clin. Gastroenterol. 18, 577–596.
[23] Lottaz, D., Maurer, C.A., Hahn, D., Buchler, M.W. and Sterchi, E.E. (1999) Nonpolarized secretion of human meprin alpha in colorectal cancer generates and increased proteolytic potential in stroma. Cancer Res. 59, 1127–1133. [24] Dietrich, J.M., Jiang, W. and Bond, J.S. (1996) A novel meprin b mRNA in mouse embryonal and human colon carcinoma cells. J. Biol. Chem. 271, 2271–2278. [25] Matters, G.L. and Bond, J.S. (1999) Expression and regulation of the meprin b gene in human cancer cells. Mol. Carcinogen. 25, 169–178. [26] Matters, G.L. and Bond, J.S. (1999) Meprin B: transcriptional and posttranscriptional regulation of the meprin beta metalloproteinase subunit in human and mouse cancer cells. APMIS 107, 19–27. [27] Hahn, D., Illisson, R., Metspalu, A. and Sterchi, E.E. (2000) Human N-benzoyl-L-tyrosyl-p-aminobenzoic acid hydrolase (human meprin): genomic structure of the alpha and beta subunits. Biochem. J. 346, 83–91. [28] Kohler, D., Kruse, M., Stocker, W. and Sterchi, E.E. (2000) Heterologously overexpressed, affinity-purified human meprin a is functionally active and cleaves components of the basement membrane in vitro. FEBS Lett. 465, 2–7. [29] Leibovitz, A., Stinson, J.C., McCombs, W.B. and McCoy, C.E. (1976) Classification of human colorectal adenocarcinoma cell lines. Cancer Res. 36, 4562–4569. [30] Hewitt, R.E., McMarlin, A., Kleiner, D., Wersto, R., Martin, P., Tsoskas, M., Stamp, G.W.H. and Stetler-Stevenson, W.G. (2000) Validation of a model of colon cancer progression. J. Pathol. 192, 446–454. [31] Witty, J.P., McDonnell, S., Newell, K.J., Cannon, P., Nare, M., Tressler, R.J. and Matrisian, L.M. (1994) Modulation of matrilysin levels in colon carcinoma cell lines affects tumorigenicity in vivo. Cancer Res. 54, 4805–4812. [32] Marchand, P., Tang, J. and Bond, J.S. (1994) Membrane association and oligomeric organization of the alpha and beta subunits of mouse meprin A. J. Biol. Chem. 269, 15388–15393. [33] Norman, L.P., Jiang, W., Han, X., Saunders, T.L. and Bond, J.S. (2003) Targeted disruption of the meprin b gene in mice leads to under representation of knock-out mice and changes in renal gene expression profiles. Mol. Cell. Biol. 23, 1221–1230. [34] Lottaz, D., Hahn, D., Muller, S., Muller, C. and Sterchi, E.E. (1999) Secretion of human meprin from intestinal epithelial cells depends on differential expression of the alpha and beta subunits. Eur. J. Biochem. 259, 496–504. [35] Becker, C., Kruse, M.N., Slotty, K.A., Kohler, D., Harris, J.R., Rosmann, S., Sterchi, E.E. and Stocker, W. (2003) Differences in the activation mechanism between the a and b subunits of human meprin. Biol. Chem. 384, 825–831.