| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online December 19, 2003 as doi:10.1096/fj.03-0584fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,||
* The Scripps Research Institute, Department of Cell Biology, La Jolla California, USA;
Department of Pathology, Vanderbilt University,
Department of Medicine, Vanderbilt University and Veterans Affairs Hospital, Nashville, TN;
Division of Cell Biology, Kihara Institute for Biological Research, Yokohama City University, Totsuka-ku, Yokohama 244-0813, Japan; and
|| Department of Cancer Biology and Center for Matrix Biology, Vanderbilt University, Nashville, Tennessee, USA
2 Correspondence: Department of Cancer Biology and Center for Matrix Biology, Vanderbilt University Medical Center, 771 Preston Bldg., Nashville, TN 37232-6840, USA. E-mail: vito.quaranta{at}vanderbilt.edu
SPECIFIC AIMS
We aimed to define MT1-MMP cleavage sites on Ln-5 in detail; define the correlation between MT1-MMP cleavage of Ln-5 and epithelial cell differentiation; and determine whether MT1-MMP is the physiological proteolytic enzyme for Ln-5 in vivo by analyzing tissues from MT1-MMP KO mice.
PRINCIPAL FINDINGS
1. Identification of MT1-MMP cleavage sites on rat Ln-5 
2 chain
Our previous data demonstrate that MMP2 cleaves Ln-5 2 chain at the boundary of domain III (DIII) with DII, generating a 80 kDa 2x chain (2
2x). We proposed that MT1-MMP used this MMP-2 cleavage site and the conserved 22' site located at the boundary of DIII and DIV domains, giving rise to the 100 kDa 2' chain.
Such a proposal is of interest because MMP cleavage simultaneously at both sites (22x and 22') excises the DIII domain, which has an EGF-like structure. Therefore, we characterized in detail the exact positions of the MT1-MMP cleavage sites on 2, especially the one that generates 2' (22'). We produced a recombinant protein encompassing domains III to V (rDIII-V) of rat Ln-5 2 chain, carrying a carboxyl-terminal His-Tag. This rDIII-V construct contained proteolytic processing sites 22' and 22x.
rDIII-V was incubated with commercial recombinant-soluble MT1-MMP, then analyzed by amino-terminal sequencing and Western blot, using anti-DIII antibody (Ab) 2778 and anti His-Tag Ab. The Ab 2778 reacted with four bands of 80, 75, 35 and 30 kDa. The larger 80 kDa band corresponds to intact rDIII-V; the other three bands should all contain at least parts of DIII. Only the 80 and 35 kDa bands were reactive with anti His-tag Ab. We concluded that the 75 and 30 kDa bands were cleaved in the vicinity of the junction between DIII-V and the carboxyl-terminal His-Tag itself. This cleavage site likely corresponds to the location of the 22x MMP-2 cleavage site identified earlier(Ala586-Leu587). The amino terminus of the 30 kDa fragment was Asp414-Glu415-Asn416-Pro417, which is identical to the amino terminus of 2', the in vivo proteolytically processed form of the 2 subunit. This shows that MT1-MMP also cleaves between Gly413 and Asp414 at the border between DIII and DIV, giving rise to 2' if the 140 kDa 2 chain were cleaved at this site alone or generating free DIII if the 22x site was also cleaved.
2. Direct correlation between MT1-MMP expression levels and 2 cleavage in a cultured cell line
By Western blot with anti-Ln-5-DIII mAb D4B5, 2 and only minimal amounts of 2' were detected in the conditioned medium (CM) of the gastric carcinoma cell line MKN45. Detergent cell lysates of MKN45 contain very low levels, if any, of MT1-MMP by Western blot. FACS analyses confirmed that MKN45 express virtually undetectable levels of MT1-MMP on the cell surface.
To corroborate a link between MT1-MMP expression and 22' processing, we transfected MKN45 cells with full-length transmembrane MT1-MMP cDNA. In MT1-MMP transfectants, the 140 kDa 2 band and the 100 kDa 2' band were readily detectable by Western blot. In parental or mock-transfected MKN45 cells, the 2' band was very faint or undetectable. These results correlate expression of MT1-MMP with processing of 2 to 2'.
3. Effect of MT1-MMP expression on cell spreading morphology and motility of cells plated on Ln-5
MKN45 and mock-transfected MKN45 cells failed to spread on Ln-5, fibronectin, or uncoated plastic wells, whereas MT1-MMP transfectants showed spread morphology on Ln-5 but not on fibronectin or uncoated plastic. In Transwell migration assays, MT1-MMP transfectants migrated well on Ln-5 coated filters whereas little or no migration was detectable in mock-transfected cells. These results suggest that expression of MT1-MMP promotes the ability of cells to spread and migrate on Ln-5, possibly via 2 cleavage to 2'.
4. Deficient proteolytic processing of the Ln-5 2 subunit in MT1-MMP KO mice
In mouse kidney tissue (Fig. 1
), both forms of the Ln-5 2 chain were detected by Western blot with anti-DIII antibody 2778: a 140 kDa band corresponding to full-length 2, and the 100 kDa 2' form, resulting from amino-terminal proteolysis of 2. In tissue extracts from MT1-MMP KO mice the 2 form, but not processed 2' was detectable. The total amount of Ln-5 2 chain was 
threefold higher in KO than WT kidney tissue. This result indicated that in kidney tissue MT1-MMP proteolytically processes Ln-5 2 to 2'. Similar results were obtained with lung and skin.
|
5. Alterations of kidney tubular epithelium in MT1-MMP KO animals
Kidneys from 10-wk-old KO mice (Fig. 2
) were reduced in size along the longitudinal axis compared with WT kidneys, reflecting loss of cortical and medullary mass. Glomeruli were approximately equal in number in KO and WT tissues, suggesting there are probably no abnormalities in ureteric bud branching morphogenesis. The tubular epithelium was considered dysmorphic, suggesting a failure of terminal differentiation (Fig. 2)
. Lectins that bind to distinct segments of terminally differentiated renal tubules were used to confirm these abnormalities. Strong staining of 
-L-fucose residues found on the apical surface of cells in the proximal tubules of the kidney was seen in WT kidney; such staining is markedly decreased in KO kidneys. With Arachis hypogaea lectin, we observed an apical staining pattern predominantly in the distal tubules and collecting ducts in WT kidney. This staining was markedly decreased in the KO animals, consistent with the possibility that the kidney tubules of MT1-MMP KO mice may be dysmorphic because of poor terminal differentiation.
|
Immunohistochemistry revealed the presence of Ln-5 in glomeruli, in proximal and distal convoluted tubules, and in collecting ducts predominantly toward the basolateral aspect of cells, consistent with BM localization. In KO animals, the distribution of Ln-5 staining was similar to WT, but staining intensity was much stronger and often also apical. As expected, MT1-MMP expression was undetectable in KO mice but virtually ubiquitous in WT kidney.
These results point to an increase in quantity and possibly an abnormal cellular distribution of Ln-5 in the tubules of MT1-MMP KO kidneys. This altered expression pattern of Ln-5, confined to tubular structures, correlates well with the tubular abnormalities, suggesting that the inability of the MT1-MMP KO mouse to cleave 2 may contribute to the dysmorphic tubule phenotype observed.
CONCLUSIONS AND SIGNIFICANCE
Having characterized the cleavage sites of the Ln-5 2 short arm used by MT1-MMP, we found that the lack of MT1-MMP in MKN45 cells parallels the lack of Ln-5 processing from its 140 kDa 2 chain to 100 kDa '. Reintroduction of MT1-MMP in MKN45 enhanced proteolytic processing of the 2 chain as well as cell spreading and migration on Ln-5, but not on fibronectin. Although causal relationships between the lack of Ln-5 processing and epithelial phenotypes need to be more firmly established, our data strongly suggest that epithelial morphology and motility are influenced by MT1-MMP expression and subsequent cleavage of Ln-5 2 to 2'. Our in vitro data were corroborated by the finding that in MT1-MMP knockout (KO) mice the processing of Ln-5 2 chain to the amino terminus truncated 2' is absent in the kidney and is reduced in all tissues tested, including lung. In tubular epithelium of the kidney, lack of 2 processing is associated with epithelial morphological changes reminiscent of human congenital mixed renal hypoplasia/dysplasia and with increased amounts and abnormal distribution of Ln-5, possibly as a result of reduced turnover rate. Tubular epithelial cells display poor terminal differentiation in the MT1-MMP KO animals. These findings indicate that MT1-MMP is a physiologically relevant enzyme for proteolytic processing of Ln-5 in vivo.
|
FOOTNOTES
1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.03-0584fje 
This article has been cited by other articles:
|
|
 |
|
  H. H. N. Yan, D. D. Mruk, E. W. P. Wong, W. M. Lee, and C. Y. Cheng An autocrine axis in the testis that coordinates spermiation and blood-testis barrier restructuring during spermatogenesis PNAS, July 1, 2008; 105(26): 8950 - 8955. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Koshikawa, T. Minegishi, K. Nabeshima, and M. Seiki Development of a New Tracking Tool for the Human Monomeric Laminin-{gamma}2 Chain In vitro and In vivo Cancer Res., January 15, 2008; 68(2): 530 - 536. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Baumgartner, G. Radziwill, M. Lorger, A. Weiss, and K. Moelling c-Src-Mediated Epithelial Cell Migration and Invasion Regulated by PDZ Binding Site Mol. Cell. Biol., January 15, 2008; 28(2): 642 - 655. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Oku, E. Sasabe, E. Ueta, T. Yamamoto, and T. Osaki Tight Junction Protein Claudin-1 Enhances the Invasive Activity of Oral Squamous Cell Carcinoma Cells by Promoting Cleavage of Laminin-5 {gamma}2 Chain via Matrix Metalloproteinase (MMP)-2 and Membrane-Type MMP-1. Cancer Res., May 15, 2006; 66(10): 5251 - 5257. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Wang, J. Sun, S. Kitamoto, M. Yang, A. Grubb, H. A. Chapman, R. Kalluri, and G.-P. Shi Cathepsin S Controls Angiogenesis and Tumor Growth via Matrix-derived Angiogenic Factors J. Biol. Chem., March 3, 2006; 281(9): 6020 - 6029. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Lehti, E. Allen, H. Birkedal-Hansen, K. Holmbeck, Y. Miyake, T.-H. Chun, and S. J. Weiss An MT1-MMP-PDGF receptor-{beta} axis regulates mural cell investment of the microvasculature Genes & Dev., April 15, 2005; 19(8): 979 - 991. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
