SEARCH

SEARCH BY CITATION

Keywords:

  • renal cell carcinoma;
  • agmatinase;
  • 2D-PAGE;
  • kidney;
  • proteome;
  • arginine metabolism;
  • nitric oxide;
  • polyamine

Abstract

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The proteome of RCC was analyzed by 2D PAGE to search for tumor-associated proteins. Agmatinase, which hydrolyzes agmatine to putrescine and urea, was identified by mass spectrometry and database searches and shown to be downregulated in tumor cells. Additionally, RT-PCR and Northern blot analyses demonstrated a clearly decreased amount of agmatinase mRNA in tumor cells. The differential expression of agmatinase mRNA was confirmed at the protein level. Western blot analysis showed almost no detectable agmatinase protein in tumor cells compared to corresponding normal renal tissue. Agmatinase mRNA is most abundant in human liver and kidney but expressed to a lesser extent in several other tissues, including skeletal muscle and small intestine. The human agmatinase gene encodes a 352-residue protein with a putative mitochondrial targeting sequence at the N-terminus. Using transfection and immunohistochemical studies, we show that agmatinase is localized in the mitochondria. Immunohistochemical studies revealed that agmatinase in the normal kidney is restricted to tubulus epithelial cells, while in tumors staining was low and heterogeneous. Thus, expression of human agmatinase is altered in RCC. We discuss the consequences of these findings in terms of polyamine, NO metabolism and macrophage function. © 2003 Wiley-Liss, Inc.

Histopathologically, 75–80% of all RCCs can be classified as the clear cell type. Human RCC is characterized by abnormally high glycogen and lipid deposition caused by metabolic and biochemical alterations.1, 2 This type of renal tumor is associated with the occurrence of highly specific deletion of chromosome 3p and mutation of the VHL gene.3

Prognosis of RCC patients is dependent mainly on the time of diagnosis. Detailed knowledge of the molecular changes within tumor cells may provide a better understanding of this disease and may help to develop new diagnostic markers and specific therapies. Analysis of the proteome provides a method to identify tumor-associated alterations. This method demands standardized conditions for tissue preparation to ensure a clear distinction between tumor and normal tissue. We compared tumor samples to corresponding normal kidney parenchyma of 20 patients using 2D PAGE. As we reported previously, 12 proteins were diminished in the tumor.4 Database searches revealed that one of these proteins identified by mass spectrometry was identical to a hypothetical UP (GenBank accession AK 027037). A repeated database search resulted in identification of the UP as human agmatinase, which plays an important role in arginine metabolism. This identification was made possible by the isolation and cloning of human agmatinase in March 2002.5

Here, we demonstrate that expression of human agmatinase is reduced in RCC. This observation was verified by RT-PCR, Northern blot and Western blot analyses. The distribution of the agmatinase protein within the tumor and normal renal tissue was studied by immunohistochemistry, and we provide evidence of the subcellular localization of human agmatinase.

MATERIAL AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Clinical samples

Malignant and benign renal tissues were obtained immediately after radical nephrectomy from 20 patients. Following nephrectomy, the kidney was sliced longitudinally along the largest diameter of the tumor. Tissue samples were taken from the tumor and from the macroscopically normal surrounding renal tissue. One-half of each probe was frozen in liquid nitrogen and stored at −80°C until use. The second half was taken for histologic examination.

Preparation of protein samples

For proteome analysis and Western blotting, tissues were homogenized by addition of a 10-fold volume of lysis buffer (8 M urea, 4% CHAPS, 65 mM DTT and 40 mM TRIS), followed by centrifugation at 100,000g for 60 min at 4°C. The protein concentration in the supernatant was determined according to the method of Bradford.6 Samples were stored at −80°C.

Analytic and preparative 2D-PAGE, gel analysis and protein identification

Analyses and 2D-PAGE were performed as described previously.4

Database searches

Comparisons of the sequences of the UP and agmatinase with nucleotide and protein sequences in the GenBank database were done using the NCBI BLAST program.

To analyze the N-terminal mitochondrial targeting sequence of agmatinase, we used the ExPASy iPSORT Prediction (www.expasy.ch) and MitoProtII 1.0a4 program (http://bioinformer.ebi.ac.uk/newsletter/archives/2/mitoprotii.html).

RNA isolation and measurement of gene transcripts by semiquantitative RT-PCR

RNA was isolated from 100 mg of normal and malignant renal tissue using Trizol reagent (Life Technologies, Gaithersburg, MD) according to the manufacturer's protocol. RNA concentration was determined spectrophotometrically. cDNA was prepared by reverse transcription of 1.5 μg of total RNA using oligo(dT)15 primer and MMLV reverse transcriptase (Promega, Madison, WI). Human agmatinase was amplified by 30 cycles of PCR using Taq Polymerase (Pharmacia, Gaithersburg, MD) and the following primers: RT-hAgm, 5′-GAATCAGTGATGCTTCGGACAG-3′ and 5′-CCATCACGTTCAGGCCTTGACA-3′, which give a product of 646 bp. In a comparable assay, the ribosomal protein gene ribosomal protein large p0 (RPLP0) was amplified by 25 cycles of PCR using Taq polymerase and the following primers: RPLP0, 5′-TTGTGTTCACCAAGGAGGAC-3′ and 5′-GACTCTTCCTTGGCTTCAAC-3′, which yield a product of 649 bp.

Amplified products were of the size expected from the mRNA sequence. The identity of the products was verified by restriction digestion (data not shown). PCR products were separated by electrophoresis on 1.0% agarose gels. Ethidium bromide–stained bands were visualized by UV illumination and quantified using Gel Pro Analyzer software (Media Cybernetics, Carlsbad, CA).

The data represent expression of the indicated gene product in relation to that of RPLP0.

Western blot

Agmatinase was detected with a polyclonal antibody directed against the human agmatinase peptide: PGTGTPEIAGLTPSQ. Proteins (25 μg) were separated by 12.5% SDS-PAGE and transferred onto nitrocellulose membranes using a semidry blotter. The blot was blocked for 1 hr with a blocking buffer (Roth, Karlsrühe, Germany) at room temperature and subsequently incubated in a 1:500 dilution of the polyclonal rabbit anti-agmatinase peptide antibody overnight at 4°C. After washing the blot with PBS + 0.05% Tween-20, an alkaline phosphatase–conjugated goat antirabbit IgG (1:20,000) was used as secondary antibody. Bound antibody was detected with BCIP and NBT as substrate.

Construction of pEGFP-Agm

Agmatinase (GenBank accession NP 079034) was amplified with PWO polymerase in a DNA thermal cycler (Eppendorf) using the following gene-specific primer pair containing cleavage sites for restriction enzymes (bold letters): pEGFP-hAgm: 5′-CCGCTAGCTAGTCGACGGATGCTGAGGCTGGCG-3′ (Sal I) and 5′-CCGATCACTTCCGCGGTGTCAGACGGTTGTCACTTT-3′ (Sac I), yielding an expected product length of 1,070 bp.

pEGFP-hAgm, a eukaryotic expression vector for human agmatinase–GFP fusion protein, was constructed by inserting the Sal I–Sac I 1,070 bp coding region fragment of agmatinase into the Sal I–Sac I site of the plasmid pEGFP-C2 (Clontech, Palo Alto, CA). The pEGFP-C2 vector thus contains the human agmatinase cDNA downstream from EGFP, and the expressed fusion protein has the EGFP tag as an N-terminal extension to agmatinase. This vector was transformed into Escherichia coli XL1blue cells. Selected clones were completely sequenced (MWG Biotech, Ebersberg, Germany) and compared to the human agmatinase sequence in the GenBank database. Expression of the EGFP tag alone does not result in its transfer to the mitochondria.7

Cell culture and transfection

HepG2 and COS-7 cells were propagated in DMEM supplemented with 10% (v/v) heat-inactivated FCS and 1% penicillin/streptomycin at 37°C in a humidified atmosphere containing 95% air and 5% CO2. Transfection of 5 × 105 COS-7 cells was carried out using calcium phosphate together with 5 μg of the human agmatinase expression plasmid pEGFP-hAgm. After transfection, cells were cultured for 48 hr before being examined.

Immunocytochemistry

HepG2 cells were cultured as described above. After washing with PBS, cells were fixed on coverslips with ice-cold isopropanol for 20 min at 4°C. Air-dried cells were washed with IF buffer (0.2% w/v BSA, 0.05% v/v saponin, 0.1% w/v sodium acid in PBS, pH 7.4), followed by treatment with antiserum against the mitochondrial marker protein HSP60 (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 hr at 4°C. Polyclonal goat anti-HSP antiserum was diluted 1:100 in IF buffer. After washing with IF buffer, cells were incubated for 1 hr at room temperature with red fluorescent Cy3-conjugated or green fluorescent Cy2-conjugated rabbit antigoat IgG (1:1,000). Stained cells were observed using a fluorescence microscope (IX-70, ×500 magnification; Olympus, Tokyo, Japan).

Immunohistochemistry

Formalin-fixed specimens were dehydrated and embedded in paraffin according to standard protocols. Paraffin sections were deparaffinized using xylene (2 × 10 min) and a decreasing ethanol series. Immunohistochemistry was carried out using the 4plus Universal Immunperoxidase Detection System (Biocarta, Hamburg, Germany) according to the manufacturer's protocol. The primary antibody (anti-agmatinase, 1:12,800 in PBS) was incubated overnight at 4°C and then with the secondary antibody dilution (Biocarta) for 10 min at room temperature. Subsequently, sections were incubated with diaminobenzidine as substrate. Control reactions to demonstrate specificity of antibody binding were done by either omitting the primary antibody or using preimmune sera. Control staining with preimmune sera gave no immune-reactive signals. Slides were counterstained with hematoxylin and mounted in glycerol-gelatine (Merck, Darmstadt, Germany). Pictures were taken on an Olympus B ×50 microscope equipped with the Olympus DP 10 digital camera.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Clinical samples

Of the 20 patients, 6 had metastatic disease. Tumor stages ranged from pT1 through pT3b. None of the patients had lymph node metastases. Histologic examination showed that all tissue probes were without necrosis. All tumors were classified as clear cell-type RCC.

Altered expression of UP in malignant kidney

We analyzed the proteome from dissected malignant and normal kidney areas from 20 patients with clear cell RCC. Using 2D-PAGE, several proteins were found to be diminished in the tumor.4 One of these altered proteins was characterized by mass spectrometry and shown to be identical to a so far hypothetical UP (GenBank accession AK 027037) (Fig. 1).

thumbnail image

Figure 1. Coomassie-stained acidic standard protein map of the human kidney. Protein profile from malignant (a) and normal (b) kidney areas. The position of the UP that matched GenBank accession number AK027037 is shown in a magnified view of sections of silver nitrate–stained 2D gels. A total of 500 μg protein were separated using 24 cm immobilized pH gradient dry strips (pH 4–7) and 12.5% SDS polyacrylamide gels.

Download figure to PowerPoint

Identification of the UP

Using the UP sequence as the query in a basic local alignment search tool (BLAST) of the EST database at the NCBI, the UP was identified as human agmatinase. Its sequence has 99% identity to that of human agmatinase.

Agmatinase is decreased in tumor cells

Comparing the proteome of malignant and normal renal tissues, we found human agmatinase to be markedly diminished in malignant samples. The differential expression of agmatinase mRNA was analyzed by Northern blot (data not shown) and RT-PCR. Agmatinase mRNA was reduced in all analyzed tumor samples 2- to 70-fold (Fig. 2). In some tumor samples, almost no mRNA expression was detectable. The RPLP0 mRNA was used as a reference.

thumbnail image

Figure 2. Expression of human agmatinase in normal and tumor kidney tissue. Malignant and nonneoplastic tissues were obtained from 8 patients immediately after radical nephrectomy. (a) RT-PCR from corresponding tumor and nontumor samples. PCR fragments were subjected to gel electrophoresis. T, tumor; N, normal. Normalization was done using the human RPLPO gene. (b) Relative abundance of agmatinase mRNA in malignant and normal renal tissues.

Download figure to PowerPoint

The differential expression of agmatinase mRNA was also confirmed at the protein level by Western blot analysis (Fig. 3). We found nearly no agmatinase expression in the tumor samples, though it was detected in all corresponding normal samples. Immunohistochemical studies verified the altered expression of agmatinase in tumor cells compared to normal ones (Fig. 4). In normal tissue, the proximal tubular system stained intensely for agmatinase, distal tubulus cells stained to a far lesser extent while glomeruli and Bowman's capsule were negative. In tumor samples, only occasional cells were agmatinase-positive.

thumbnail image

Figure 3. Western blot analysis of human agmatinase in malignant (T) and normal (N) renal tissues. From the tissue samples analyzed in Figure 2, the amount of agmatinase protein was determined using an agmatinase peptide antibody. Protein (25 μg) protein was separated on 12.5% SDS polyacrylamide gels. The agmatinase peptide antibody was used at a 1:500 dilution.

Download figure to PowerPoint

thumbnail image

Figure 4. Immunohistochemical detection of agmatinase on paraffin sections of renal tissue. The primary antibody (rabbit anti-agmatinase, 1:12,800 in PBS) was incubated overnight at 4°C and then with the peroxidase-conjugated secondary antibody (Biocarta system) for 10 min at room temperature. (a) Normal kidney, (b) control (preimmune serum), (c) tumor. Slides were counterstained with hematoxylin. Scale bars = 100 μm. Solid arrow with solid point, proximal tubuli; dotted arrow, distal tubuli; solid arrow with open point, glomeruli.

Download figure to PowerPoint

Intracellular localization of agmatinase

To analyze the intracellular localization of the protein, an expression plasmid coding for agmatinase–GFP was transfected into COS-7 cells. Fluorescence was observed in mitochondria-like particulate structures in cells that were apparently transfected with the plasmid (Fig. 5). Immunostaining with an antibody against HSP60 (red) located in the mitochondrial matrix revealed that the agmatinase–GFP protein (green) colocalized with HSP60 (yellow), demonstrating that agmatinase is a mitochondrial enzyme. The mitochondria of the transfected cells appeared swollen, probably because of accumulation of a large amount of agmatinase in these organelles. These results confirm the predicted mitochondrial localization of agmatinase in human cells.

thumbnail image

Figure 5. Mitochondrial localization of agmatinase. The pEGFP–agmatinase fusion protein (green, a), the mitochondrial HSP60 (red, b) and colocalization of both proteins (yellow, c) in the mitochondria of COS-7 cells (×500) are shown after transfecting the respective expression plasmid.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The proteome from dissected malignant and normal kidney areas of patients with clear cell RCC shows diminished amounts of agmatinase in the tumor cells. RT-PCR, Northern blot and Western blot investigations confirmed the decreased expression of this mitochondrial protein.

Agmatinase cDNA has been cloned from human kidney.5 Agmatinase belongs to the arginase family, which also includes arginases, formiminoglutamases and PAH. Its cDNA encodes a protein containing 352 amino acid residues, including a putative N-terminal mitochondrial targeting sequence. Based on the size difference between the theoretical (37.66 kDa) and apparent (32–34 kDa) m.w., as determined by SDS-PAGE, a cleavage site for processing has been postulated. As with other mitochondrial matrix proteins8 the N-terminus of human agmatinase contains several arginine, leucine and serine residues and lacks acidic amino acids. Cleavage sites within mitochondrial signal sequences occur between 2 small, nonpolar amino acid residues.9 Analysis of the amino acid sequence at the N-terminus of agmatinase (MitoProt II 1.0a4) suggested the existence of one cleavage site, located between glutamine-35 and alanine-36. Its use would result in the formation of a mature agmatinase with a predicted m.w. of about 32–34 kDa, which is close to what we observed.

The intracellular localization of agmatinase has not been previously demonstrated. Mistry et al.5 suggested a mitochondrial localization of human agmatinase because of the similarity of its N-terminal sequence to that of the precursor of mitochondrial type II arginase. By transfection and colocalization studies, we here show that agmatinase is indeed localized in mitochondria.

No cases of agmatinase deficiency in humans have been described, so the pathophysiologic consequences of reduced enzyme expression, if any, are unknown. Little is known about arginine metabolism in tumor cells, so a discussion of the physiologic consequences of downregulated agmatinase on polyamine metabolism or on production of NO is necessarily speculative. Arginine is essential for the synthesis of urea, protein, creatine, NO and polyamines. Agmatinase is involved in the biosynthesis of polyamines (Fig. 6). In mammals, 2 pathways for polyamine biosynthesis exist, the classical pathway, where arginase produces ornithine from arginine, which in turn is converted to putrescine by ODC, and an alternate pathway via agmatinase.10 In this second pathway, decarboxylation of L-arginine by the enzyme ADC generates agmatine, which is hydrolyzed by agmatinase (agmatine urea hydrolase) to form urea and putrescine. Putrescine is a metabolic precursor for the biosynthesis of higher polyamines.11 Inhibitors of polyamine biosynthesis have significant potential to inhibit tumor growth.12 The importance of putrescine biosynthesis by the ADC-agmatinase pathway for the growth of some cancers is supported by the analysis of Iyer et al.10 This suggests that agmatinase activity may be advantageous for tumor growth. The polyamines putrescine, spermidine and spermine are required for progression of the cell cycle and, as such, play an important role in cell proliferation13 and most probably in cancer cell growth. In several cases, polyamine levels and polyamine biosynthetic enzymes, e.g., arginase, have been significantly upregulated in tumors compared to normal tissue.10, 12, 14, 15 With this in mind, it is surprising that in clear cell RCC the amount of agmatinase is diminished. However, agmatinase has other biologic functions besides its role in generating the polyamine precursor agmatine. It regulates the intracellular polyamine amount through inhibition of ODC by induction of antizyme synthesis.16 In normal cells, loss of this function may lead to uncontrolled and cytotoxic build-up of polyamines. In contrast, tumor cells have an increased demand for polyamines, which may be satisfied by increased uptake. Extracellular polyamines can be provided by macrophages that are present in tumors. In this way, macrophages support tumor proliferation.17, 18

thumbnail image

Figure 6. Metabolic pathways of arginine. The main possible catabolic routes for arginine are shown. OTC, ornithine-carbamoyltransferase; OAT, ornithine aminotransferase.

Download figure to PowerPoint

In tumor cells, a low level of NO is produced by NOS3, which is probably activated by agmatine. This NO may stimulate angiogenesis and dilation of arteriolar vessels, decrease leukocyte–endothelial cell interaction and increase vascular permeability.19, 20, 21 All of this would be expected to promote tumor growth. In contrast, cytokines such as TNF-α, IL-1β and IFN-γ released by macrophages induce NOS2 activity in tumor cells. NOS2 synthesizes a large and cytotoxic amount of NO. However, IFN-γ additionally activates ADC and, thus, increases the intracellular level of agmatine,22, 23 which is a competitive inhibitor of NOS2. By disabling NOS2-directed NO synthesis, tumor cells may be partially protected against the antitumor activity of macrophages.

The present study shows that human agmatinase expression is diminished in clear cell RCC. We observed in all analyzed RCC patients diminished expression of agmatinase protein and mRNA. Further studies will be required to trace arginine metabolism in tumor cells and to determine how diminished agmatinase activity alters their growth.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

We thank Mr. R. Jack for critical reading of the manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES