A Mutation in the Carbohydrate Recognition Domain Drives a Phenotypic Switch in the Role of Galectin-7 in Prostate Cancer

The observation that galectin-7 (gal-7) is specifically expressed in mammary myoepithelial (basal) cells prompted us to investigate whether this protein is expressed in the basal cells of other tissues. Given that breast and prostate cancer have remarkable underlying biological similarities and given the important roles of basal cells in prostate cancer, we examined the expression patterns and role of gal-7 in human prostate cancer. Using tissue microarray, we found that although gal-7 is readily expressed in basal cells in normal prostate tissue, it is downregulated in prostate cancer (PCa) cells. De novo expression of gal-7 in prostate cancer cells increases their sensitivity to apoptosis in response to etoposide and cisplatin. The assessment of a carbohydrate-recognition domain (CRD)-defective mutant form of gal-7 (R7S) showed that the ability of this protein to modulate apoptosis was independent of its CRD activity. This activity was also independent of its ability to translocate to the mitochondrial and nuclear compartments. However, CRD activity was necessary to inhibit the invasive behaviors of prostate cancer cells. In vivo, gal-7 overexpression in PCa cells led to a modest yet significant reduction in tumor size, while its CRD-defective mutant form significantly increased tumor growth compared to controls. Taken together, these results suggest that although de novo expression of gal-7 may be an interesting means of increasing the tumorigenic phenotypes of PCa cells, alterations in the CRD activity of this protein drive a phenotypic switch in its role in PCa cells. This CRD-independent activity represents a paradigm shift in our understanding of the functions of galectin. The R74S model will be useful to distinguish CRD-dependent and CRD-independent functions of gal-7 in cancer progression.


Introduction
Galectin-7 (gal-7) is a p53-induced gene that is mainly expressed in stratified epithelial cells [1,2]. Its expression can also be induced by other transcription factors, including mutant

Materials and Methods
Cell lines and animals PC3 [27] and DU-145 [28] cell lines were a generous gift from Dr. Benoit Ochietti (McGill University, Montréal, QC), and the LNCaP cells [29] were kindly provided by Dr. Thomas Sandersons (INRS-Institut Armand-Frappier, Laval, QC. The HACAT cell line [30] was provided by Dr. Thierry Magnaldo (Génétique et Physiopathologie des Cancers Épidermiques, Faculté de Médecine, Nice, France). All cell lines used in this study were originally obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). The DU-145 and HACAT cell lines were maintained in Dulbecco's modified Eagle's medium. The PC3 and LNCaP cell lines were maintained in an F-12k nutrient mixture and RPMI 1640 medium, respectively. Culture media were supplemented with 10% (v/v] fetal bovine serum, 2 mmol/L L-glutamine, 10 mmol/L HEPES buffer, and 1 mmol/L sodium pyruvate. All cell culture products were purchased from Life Technologies (Burlington, ON, Canada). NOD/SCID mice were obtained from The Jackson Laboratory. Animals were housed under sterile conditions with ad libitum access to food and water. All animal studies were approved by the Institutional Animal Care and Use Committee of the Centre de Recherche du Centre Hospitalier de l'Université de Montréal Immunohistochemistry Tissue microarrays were used to assay 32 tissue samples of prostate adenocarcinoma, 20 hyperplasia, 5 saccular ectasia and 3 cancer-adjacent normal prostate tissues (US Biomax, Rockville, MD, USA and LifeSpan BioSciences, Seattle, WA, USA). Immunostaining reactions for gal-7 were performed using a Discovery XT automated immunostainer (Ventana Medical Systems). Deparaffinized sections were incubated in cell conditioning solution, pH 8.0 (Ventana Medical Systems), for antigen retrieval and then stained for 60 min with an anti-human gal-7 polyclonal antibody (R&D Systems, Minneapolis, MN, USA) diluted 1:150. The slides were counterstained with hematoxylin. The sections were scanned at a high resolution using a Nanozoomer Digital Pathology scanner (Hamamatsu, Bridgewater, NJ).

RNA isolation and RT-PCR
Total cellular RNA was isolated from cells using TRIzol reagent (Life Technologies) according to the manufacturer's instructions. First-strand cDNA was prepared from 2 μg of cellular RNA in a total reaction volume of 20 μL using Omniscript reverse transcriptase (QIAGEN, Mississauga, ON, Canada). After reverse transcription, human gal-7 (gene ID 3963, sense primer: 5'-TCC CAA TGC CAG CAG GTT CCA TGT-3' and antisense primer: 5'-GAA GCC GTC GTC  TGA CGC GAT GAT-3') and GAPDH (gene ID 2597, sense primer: 5'-CGG AGT CAA CGG  ATT TGG TCG TAT-3' and antisense primer: 5'-CAG AAG TGG TGG TAC CTC TTC CGA-3') cDNAs were amplified under the following conditions: 94°C for 3 min, followed by 35  cycles at 94°C for 40 sec, 60°C for 40 seconds, and 72°C for 40 seconds and then a final extension step at 72°C for 10 min. PCR was performed in a thermal cycler (MJ Research, Watertown,  MA). The amplified products were analyzed on 1% (w/v) agarose gels by electrophoresis followed by gel staining with SYBR Safe (Life Technologies).

Generation of stable transfectants expressing gal-7
To obtain stable DU-145 transfectants expressing gal-7 wt or gal-7 R74S , cDNAs encoding the wild-type or mutated (R74S) human gal-7 gene were cloned into the Srα eukaryotic expression vector (kindly provided by Dr. François Denis) as previously described [17] Control cells were generated using an empty Srα vector. Transfections were performed with Lipofectamine 2000 according to the manufacturer's instructions (Life Technologies). After 48 h of culturing, transfected cells were allowed to grow in complete medium containing 2 μg/ml puromycin. Individual colonies were expanded, and gal-7 expression was monitored by western blot analysis. A minimum of two clones of each type was used to confirm the results.

Electron microscopy
Cells were fixed in a 0.1% (v/v) glutaraldehyde and 4% (w/v) paraformaldehyde solution and embedded in Spurr's resin. Ultrathin sections were placed on nickel grids and incubated in sodium metaperiodate. Samples were then blocked in 1% PBA for 5 min and incubated for 60 min with a goat anti-human gal-7 polyclonal antibody (1:150) followed by incubation with a rabbit anti-goat 10-nm gold-conjugated secondary antibody (1:20, Electron Microscopy Sciences, Hatfield, PA, USA). The samples were counterstained with uranyl acetate and lead citrate before visualization under a Hitachi H-7100 transmission electron microscope.

[ 3 H]-thymidine incorporation
Proliferation of cells was determined by measuring the incorporation of [ 3 H]-thymidine. Cells were seeded in triplicate at a density of 2 x 10 3 cells/well into a 96-well plate and subsequently incubated with or without 5 μM cisplatin for 96 h. After 80 h of incubation, 1 μCi of [ 3 H]-thymidine was added to each well. At the end of the incubation period, the cells were harvested with a semiautomatic cell harvester (Skatron Instruments, Lier, Norway) and transferred onto a Printed Filtermat A (Wallak, Turky, Finland). Incorporated radioactivity was determined using a RackBeta (LKB, Turky, Finland) scintillation counter.

Invasion assay
Serum-induced cell invasion was examined using a 24-well Matrigel invasion chamber (BD Biosciences, Mississauga, ON, Canada) with an 8 μm-pore membrane. A total of 5 x 10 4 cells were incubated within the upper chamber in serum-free medium. The lower chamber contained medium supplemented with 10% fetal bovine serum. After 24 h of incubation, the upper surface of the insert was wiped gently with a cotton swab to remove the non-migrating cells. Cells that had migrated to the lower surface of the membrane were stained with toluidine blue and counted separately by microscopy.

Scratch wound healing assay
Confluent monolayers were obtained by seeding 1 x 10 5 cells onto a 24-well plate the day before the experiment. A scratch was made with a pipet tip in the cell monolayer, followed by washing with PBS to remove cell debris. Immediately after and 24 h after the PBS wash step, the microscopic fields were photographed, and the scratch width was measured using Image J software. For live cell imaging, one day prior to the experiment, 4 x 10 5 cells were seeded onto a 6-well glass-bottom culture plate (MatTek Corporation, Ashland, MA, USA). After the scratch was made, the plate was moved to a PM S1 incubator, and the migration was visualized under a Carl Zeiss LSM780 confocal microscope (Carl Zeiss, Toronto, ON, Canada). Images were captured every 10 min for 2 h. For each cell type, the movements of 30 separate cells were measured. Cell movement was analyzed using the following Image J plugins: manual tracking and chemotaxis tool.

Cell proliferation assay
As a control for cell proliferation during the invasion assay and scratch wound healing test, a total of 2.5 x 10 4 cells were seeded onto a 12-well plate (Fisher Scientific). At the indicated times, the cells were washed with PBS, trypsinized, stained with trypan blue and counted using a hemocytometer.

Production of recombinant proteins
Gal-7 cDNA was cloned into pET-22b(+) using the NdeI and HindIII restriction enzymes. The protein was produced in E. coli BL21 (DE3) at 37°C. Isopropyl β-D-1-thiogalactopyranoside (IPTG) (1 mM) was added to the bacterial culture at an OD 600nm of 0.6-0.7, and the bacteria were further incubated for 4 h. Bacterial pellets were resuspended in lysis buffer (0.7 mg/mL lysozyme, 10 mM Tris, pH 8, 100 mM NaCl, 1 mM EDTA, 1 mM DTT and protease inhibitor cocktail), incubated for 1 h at 37°C and centrifuged for 30 min at 15,000 rpm (4°C). The supernatant was then filtered and applied to a lactose-agarose column, and the protein was eluted in 1-mL fractions with a 150-mM lactose solution. Purified fractions were analyzed by SDS-PAGE. Gal-7 was dialyzed against 20 mM potassium phosphate at pH 7.2 for all subsequent experiments.

Glycan array
A mammalian glycan array (V5.2) was performed by the Consortium for Functional Glycomics (CFG). Briefly, recombinant gal-7 wt and gal-7 R74S proteins were conjugated to FITC and tested against version 5.2 of the printed array. This array consisted of 609 glycans in replicates of 6. The lists of the glycans and their linkers used in the different versions of the array can be found at http://www.functionalglycomics.org/static/consortium/resources/resourcecoreh.shtml. FITC-conjugated gal-7 was incubated with the sugars, and relative fluorescence units (RFUs) were measured. To eliminate some of the false hits that contained a single very high or low point, the highest and lowest points from each set of six replicates were removed. Consequently, the averages include 4 values rather than 6.

Binding assay
Briefly, a fluorescein isothiocyanate (FITC)/DMSO solution was added to recombinant gal-7 in a 0.1-M NaHCO 3 (pH 9.2) solution and then incubated for 2 h at room temperature on a roller. FITC-conjugated gal-7 was then purified using a PD-10 sepharose column (GE Healthcare) and eluted with PBS containing 0.01% (v/v] sodium azide. To measure FITC-gal-7 binding to the cell surface, 2.5 x 10 5 cells were incubated for 30 min with the indicated concentrations of gal-7 and then washed twice with PBS and resuspended in 500 μl PBS. For competition assays, 0.1 M β-lactose was added to cells, which were then incubated with FITC-conjugated gal-7. Samples were analyzed by FACSCalibur (BD Biosciences) and Flowing Software.

In vivo experiments
A mix of 3 independent clones (2 x 10 6 cells) for each transfectant was injected subcutaneously into 6-week-old NOD/SCID mice. Tumor measurements were obtained twice a week. On day 61, the mice were sacrificed, and the primary tumors were harvested and snap-frozen in liquid nitrogen.

Statistical analysis
Statistical significance of the experiments was evaluated using unpaired Student's t-tests. The results were considered statistically significant at P 0.05.

Gal-7 expression in human prostate tissues and cancer cell lines
Gal-7 expression in human prostate tissues has not previously been reported. Using immunohistochemistry (IHC), we first investigated the expression pattern of gal-7 in normal prostate tissues. Our results revealed strong nuclear and cytoplasmic expression in the basal cell layers of normal prostate glands, with no staining observed in the luminal epithelial cells (Fig 1A). We then investigated the presence of gal-7 in various types of prostate malignancies (including 32 prostate adenocarcinomas) using commercial tissue microarrays and found negligible gal-7 protein expression in the tumor tissues (Fig 1B and 1C). The low expression in PCa tissues was consistent with expression patterns found during our investigations of gal-7 expression at the mRNA and protein levels in the most common prostate cancer cell line models. Immunoblotting experiments demonstrated that none of these cell lines expressed readily detectable levels of gal-7; however, mRNA was expressed at a low but detectable level in PC3 cells (Fig 1). Taken together, these results showed that although gal-7 is readily expressed in basal cells within normal prostate tissue, it is completely absent in prostate cancer cells.

Production and characterization of wild-type and CRD-defective gal-7 protein
To investigate the functions of gal-7 in prostate cancer cells and to determine the importance of its CRD, we generated a series of stable transfectants expressing wild-type gal-7 and a mutated form (R74S) of the protein (Fig 2A). This mutation is known to inhibit the ability of gal-7 to bind lactose and to reduce its translocation from the cytosol to the mitochondria and/ or the nucleus [17]. Our previous analysis using solution NMR spectroscopy showed that the R74S mutation induced only limited and local changes in gal-7 folding. To further determine the extent to which the CRD of gal-7 is disrupted by the R74S mutation, we compared its binding properties to those of wild-type gal-7 using a CFG glycan array (version 5.2) [31]. The entire list of glycans tested and the results of the binding assays can be found online (http:// functionalglycomics.org/). Our results confirmed that the R74S mutation suppressed CRD activity independent of the glycan motifs assessed (Fig 2B, S1 Table). This suppression of CRD activity was confirmed by flow cytometric analysis of the binding of FITC-labeled recombinant gal-7 wt and gal-7 R74S at the surfaces of DU-145 prostate cancer cells (Fig 2C and 2D). Taken together, these results demonstrate that R74S abolishes the CRD activity of human gal-7.

Characterization of intracellular localization of wild-type gal-7 and gal-7 R74S
To investigate the role of gal-7 and its CRD in prostate cancer, DU-145 transfectants expressing either gal-7 wt or gal-7 R74S were generated. Control transfectants (generated using empty expression vectors) did not express detectable levels of gal-7 (Fig 3A and 3B). Immunoblotting of mitochondrial and nuclear enriched fractions showed detectable expression of gal-7 wt in the cytosol, mitochondria and nuclei of DU-145 cells (Fig 3C and 3D). In contrast, there was no detectable expression of gal-7 in the mitochondria of transfectants expressing gal-7 R74S . Both proteins were found in the extracellular media of the cells (Fig 3E). Electron microscopy analysis of the DU-145 cells confirmed the expression of gal-7 wt in the cytosol, nucleus, and mitochondrial outer membrane, while the expression of gal-7 R74S was restricted to the cytosol (S1 Fig). Interestingly, electron microscopic analysis revealed the presence of gal-7 wt -and gal-7 R74S -rich protrusions at the cytoplasmic membrane (S1 Fig) consistent with previous reports, suggesting that gal-7 associates with the actin cytoskeleton to regulate cell motility [10,32,33].

Gal-7 promotes apoptosis in prostate cancer cells
Several studies have reported that ectopic expression of gal-7 renders cervical, gastric and colon cancer cells more sensitive to apoptosis induced by pro-apoptotic drugs [15]. To determine whether this finding is also true in prostate cancer cells, DU-145 transfectants were treated with increasing doses of pro-apoptotic drugs and analyzed for cleavage of Parp-1, which is a commonly used marker of apoptosis [34]. Our results showed that ectopic expression of gal-7 increased the cleavage of Parp-1 induced by etoposide compared to the control cells ( Fig 4A). Similar results were obtained when the fragmentation of nuclei in apoptotic cells was visualized with DAPI staining (Fig 4B). The ability of gal-7 to increase the sensitivity of DU-145 cells to apoptosis was also observed using cisplatin as a pro-apoptotic drug (Fig 4C). Interestingly, gal-7 R74S was as effective as gal-7 wt in increasing the sensitivity of DU-145 to both pro-apoptotic drugs. In both cases, the intracellular localization of gal-7 remained  -thymidine incorporation assay, we also measured proliferation of the transfectants in the absence or presence of cisplatin. We found that both gal-7 wt -and gal-7 R74S -expressing DU-145 cells proliferated at the same rates compared to control cells under normal conditions but proliferated more slowly in presence of cisplatin (Fig 4D), which is consistent with the ability of gal-7 to promote druginduced apoptosis. Taken together, these results show that gal-7 sensitizes DU-145 cells to proapoptotic agents independent of its CRD activity and its intracellular compartmentalization.

Gal-7 but not R74S reduces invasive behaviors of DU-145 cells
We next investigated whether gal-7 can modulate the invasive behaviors of DU-145 cells using a standard in vitro Matrigel invasion assay. We found that ectopic expression of gal-7 wt significantly reduced the invasive behaviors of DU-145 cells compared to control cells lacking gal-7 (Fig 5A). A similar difference was not observed for DU-145 cells expressing gal-7 R74S . Because cell invasive behaviors might be affected by cell motility, we used live cell imaging to measure cell movements during a scratch wound healing assay (Fig 5B). Our results showed that DU-145 cells expressing gal-7 wt significantly reduced cell velocity compared to control cells lacking gal-7 or cells expressing gal-7 R74S (Fig 5C). These gal-7 wt -expressing cells also had lower accumulated and Euclidean distances of migration (Fig 5D and 5E). The directionality of the DU-145 cells was not affected by expression of the gal-7 wt or gal-7 R74S proteins (Fig 5F). The use of a standard scratch wound healing assay further confirmed that gal-7 wt reduced the cell motility of the DU-145 cells. Again, a similar effect was not observed for the gal-7 R74S mutant (S3 Fig).
No differences in cell proliferation were observed between the cells expressing gal-7 wt or gal-7 R74S and the control cells (S4 Fig). The addition of recombinant gal-7 wt and gal-7 R74S to DU-145 cells had no effect, suggesting that extracellular gal-7 is not involved in reducing invasive behaviors (S5 Fig). Taken together, these results indicate that intracellular gal-7 reduces the invasive behaviors of prostate cancer cells by impairing cell motility in a CRD-dependent manner.

Gal-7 CRD disruption increases tumor growth in vivo
We next investigated whether gal-7 wt and gal-7 R74S impact tumor growth using adult male NOD/SCID mice. Mice were injected subcutaneously with DU-145 transfectants, and tumor size was measured twice a week for 61 days, at which time the tumors were harvested to confirm the expression of gal-7 in the gal-7 wt -and gal-7 R74S -expressing cells (S6 Fig). Our results showed that the overexpression of gal-7 wt led to a modest yet significant (p 0.05) reduction in tumor size (Fig 6). Interestingly, the expression of gal-7 R74S caused a significant (p 0.001) increase in tumor growth compared to both the control and gal-7 wt -expressing cells.

Discussion
Our previous results showing that gal-7 is specifically expressed in mammary myoepithelial (basal) cells but not in mammary luminal cells prompted us to investigate whether this molecule is expressed in the basal cells of prostate tissues. Using an anti-gal-7-specific Ab, we found that gal-7 immunostaining in human prostate tissues was consistently strong in the nuclei and cytoplasm of prostate basal cells, with the luminal cells showing no detectable staining. This pattern of expression is thus clearly distinct from those reported for gal-1 and gal-3. Indeed, gal-3 is expressed in luminal cells but not in basal cells, while gal-1 is expressed in the endothelial and stromal fibromuscular cells of the prostate [35]. This distinct expression pattern for gal-7 is also observed in prostate cancer cell lines. Although we found no detectable expression of gal-7 in the prostate cancer cell lines tested, gal-3 has been shown to be readily expressed in both PC-3 and DU-145 cells [36]. Gal-3 expression is also reduced in PCa cells compared with normal prostate cells but is still detectable by IHC in a significant number of samples [36,37]. Our data, however, clearly showed that gal-3 and gal-7 had distinct properties in PCa cells. For example, in contrast with gal-7, cytoplasmic gal-3 increased Matrigel invasion and cell growth while decreasing apoptosis induction, and nuclear expression had a completely opposite effect [38]. Thus, gal-3 and gal-7 possess completely opposite biological activities in PCa cells. Although future experiments will be needed to confirm these results in other prostatic cell lines (including benign cell lines) and other preclinical PCa models. It is important to note that we cannot not use the PC3 cells (another classical cell model of human prostate cancer) because these cells express low but significant endogenous galectin-7 (as shown in Fig 1). We cannot use the LNCaP cell model (another commonly used model) because this model in androgen-dependent, in contrast to DU-145. Nevertheless, our findings may have important implications in the development of CRD-specific inhibitors against gal-3 and emphasize the need to develop inhibitors that are highly specific for a given galectin.
Our results showed that gal-7 reduced the invasive behaviors of prostate cancer cells by inhibiting their motility. This phenotype is consistent with the localization of this molecule in lamellipodia and filopodia. In normal cells, gal-7 is also found in motility structures, such as podosomes and primary cilia [32,33]. Interestingly, a mutation at position 74 completely abolished this activity. In contrast, this mutation did not affect the ability of gal-7 to induce apoptosis, indicating that both functions are mediated by distinct gal-7 sites. One possibility is that gal-7 regulates the stability and/or localization of proteins, such as β-catenin, and that the mutation at position 74 abolishes this interaction. In the cytoplasm, β-catenin is either ubiquitinated for proteasomal degradation or localized at cell-cell contact sites, stabilizing E-cadherin and affecting motility [39]. This interaction between galectins and β-catenin has been reported previously [40][41][42]. We found that gal-7 overexpression sensitized DU-145 prostate cancer cells to apoptosis induced by cisplatin or etoposide. The dual role of gal-7 in apoptosis has been well documented. Because gal-7 binds to bcl-2, our initial hypothesis was that mitochondrial gal-7 could be responsible for this dual role. However, this assumption is clearly not true because cytoplasmic gal-7 R74S displays similar pro-apoptotic functions as the wild-type protein. This similarity has also been reported for the anti-apoptotic functions of gal-7 in breast cancer cells [17]. It is indeed quite clear that although gal-7 is expressed in basal cells in normal prostatic and mammary tissues, it plays a completely different role in prostate and breast cancer. The mechanisms leading to the anti-and pro-tumorigenic functions of cytoplasmic gal-7 remain unknown. One possible mechanism involves modulation of the JNK1 pathway, as suggested by Kuwabara et al. [14], who showed that the induction of apoptosis by gal-7 in HeLa cells is correlated with activation of this signaling pathway. It is also possible that gal-7 regulates apoptosis by interacting with bcl-2. Other studies have indeed shown that gal-7 interacts directly with Bcl-2 [16]. Cytoplasmic gal-7 could sequester bcl-2 in the cytoplasm, thereby inhibiting its anti-apoptotic function. Alternatively, given the structural homology among members of the bcl-2 family [43], it is possible that gal-7 binds to other bcl-2 structural homologs, thereby altering the delicate balance between pro-and anti-apoptotic proteins. We are currently investigating these possibilities.
The dual role of galectins in modulating tumor progression has been previously noted but is still unclear. Our data showing that gal-7 R74S acts as a tumor suppressor in vitro and as a protumorigenic protein in vivo suggest that the roles of galectins in cancer likely involve a delicate balance between pro-and anti-tumoral interactions occurring within and outside cancer cells, i.e., in the tumor microenvironment. Because gal-7 R74S sensitized DU-145 cells to apoptosis without affecting their invasive behaviors or proliferation, leading to the augmentation of tumor growth in vivo, it is clear that alterations in the CRD of gal-7 that shift the balance towards CRD-independent binding partners not only have a profound effect on its intracellular distribution in cancer cells but also drive a phenotypic switch in its role in cancer. In silico analysis using a publically available dataset, the cBioPortal for Cancer Genomics, shows that the gene encoding human gal-7 is rarely (less than 1% in prostate adenocarcinoma) mutated in PCa and is not mutated within the CRD-coding region. Similar results were obtained with the COSMIC database (one missense mutation out of 528 cases). This indicates that loss of gal-7 expression in PCa cells is probably due to an epigenetic mechanism and/or to the depletion of basal cells. Our data add, however, a new dimension to the role of galectin CRDs in cancer, emphasizing the use of highly specific inhibitors to target members of the galectin family. The identification of important CRD-independent functions represents a paradigm shift in our understanding of galectin functions. Future investigations will be needed to identify in detail the CRD-independent binding partners involved. The availability of our gal-7 R74S model will be useful in this regard. On day 61, the mice were sacrificed, and the primary tumors were harvested and snap-frozen in liquid nitrogen. Tumor mRNA was extracted and gal-7 expression was measured by RT-PCR. N = 6 mice per group. (TIFF) S1 Table. Glycan array. The names of the different sugars used in the glycan array are listed. Only those sugars for which gal-7 wt had an RFU of larger than 10,000 are presented. (TIFF)