TEXT

Top Abstract Text References

Mycobacterium bovis is the main causative agent of bovine tuberculosis (TB), a disease of great concern in developing countries such as Mexico for its detrimental economic impact and its effect on human beings (7). To control bovine TB, most programs rely on tuberculin tests to identify infected cattle, which are then sacrificed. Tuberculin tests, however, are not entirely reliable, and false-negative and false-positive results are not uncommon (8, 19, 20, 22). To fight this disease, it would be a great asset to develop resistant breeds (17), a goal at present unattainable, as little is known about genetic factors of bovine TB (18, 28).

During the last decade, the gene coding for the so-called natural resistance-associated macrophage protein (NRAMP1) has been the center of much attention. Nramp1 was first described for the mouse as a gene conferring resistance to Salmonella enterica serovar Typhimurium, Leishmania donovani, and M. bovis BCG. A single recessive point mutation in Nramp1 was shown previously to occur in BCG-susceptible mouse strains (11). In human beings, eight NRAMP1 alleles have been identified (13, 24), some of them apparently associated with susceptibility to TB (4, 14). In bovines, two Nramp1 alleles have been cloned (16), the predicted polypeptide being a transmembrane molecule with a conserved transporter motif and multiple phosphorylation sites (9).

To gain information on the role of NRAMP1 in bovine TB, we investigated the expression of NRAMP1 in peripheral blood cells (PBC) and tuberculous granulomas of M. bovis-infected bovines. A polyclonal antiserum was kindly donated by Jennifer Blackwell (University of Cambridge Medical School) for this purpose. The antiserum was raised in rabbits against a recombinant polypeptide spanning 35 residues of the carboxyl end of mouse NRAMP1 (1). Tested in our laboratory, this antiserum reacted with a 65-kDa band on phorbol myristate acetate-activated cells of the murine monocyte line J774A (data not shown). It showed no reactivity with M. bovis proteins by Western blotting (data not shown). PBC were obtained from 20 animals from a herd with 50% TB prevalence; 10 cattle were purified protein derivative (PPD) positive, and 10 were PPD negative. For a control, PBC from five PPD-negative cattle belonging to a TB-free herd were employed. Tracheobronchial and mediastinal lymph nodes (LN) with TB granulomas were isolated at autopsy from five cattle with M. bovis infection confirmed by histology, Ziehl-Neelsen stains, and bacteriological cultures (data not shown). For the Western blot, cells and tissues were treated with RIPA buffer (50 mM Tris, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% sodium dodecyl sulfate), containing a cocktail of protease inhibitors (pepstatin, leupeptin, aprotinin, quimostatin, antipain, and phenylmethylsulfonyl fluoride) and phosphatase inhibitors (Na₃VO₄ and NaF; Sigma Chemical Co., St. Louis, Mo.). Proteins separated by sodium dodecyl sulfate-10% polyacrylamide reducing gels were transferred to polyvinylidene difluoride membranes (Bio-Rad Laboratories, Hercules, Calif.). After blocking with 5% skim milk and 1% bovine serum albumin, membranes were incubated with the anti-NRAMP1 antiserum diluted 1:500 for 3 h at room temperature (1). Bound antibodies were revealed with a goat anti-rabbit immunoglobulin G (IgG) peroxidase-labeled antibody (Zymed Laboratories, South San Francisco, Calif.). A chemiluminescence detection system (Amersham Pharmacia Biotech, Piscataway, N.J.) was used to reveal peroxidase. Controls were carried out with an anti-actin monoclonal antibody (MAb; Boehringer Mannheim, Indianapolis, Ind.) diluted 1:5,000 followed by an anti-mouse IgG peroxidase-labeled antibody diluted 1:4,000 (Zymed). The above studies show overexpression of NRAMP1 in PBC and granulomas from M. bovis-infected bovines. In PBC, the anti-NRAMP1 mouse antiserum revealed a major 65-kDa band and a weak 45-kDa band. Results were similar in both PPD-negative (Fig. 1A) and PPD-positive (Fig. 1B) animals. In PBC of cattle in which TB was ruled out at postmortem examination, the 65-kDa NRAMP1 band was not observed, although occasionally there was weak labeling of 60- and 45-kDa bands (Fig. 1C). The high expression of NRAMP1 was associated with cell activation as shown by Western blotting with an anti-major histocompatibility complex class II MAb, MCA904 (Serotec Ltd., Oxford, United Kingdom), which revealed the expected 32- to 34-kDa band in PBC of tuberculous cattle but not in PBC of healthy cattle (data not shown).

View larger version (18K): [in this window] [in a new window]   FIG. 1.   Western blotting of PBC from M. bovis-infected bovines to detect NRAMP1. Representative results for five cattle for each group are shown. (A) Infected herd, PPD-negative cattle; (B) infected herd, PPD-positive cattle; (C) healthy cattle. The 43-kDa band corresponds to actin control.

Studies of LN with tuberculous granulomas from five cattle with autopsy-confirmed M. bovis infection revealed a more complex Western blot pattern. As seen with PBC, there was a 65-kDa NRAMP1 band and additional bands of 95, 90, 35, and 32 kDa (Fig. 2, lanes 1 and 2). NRAMP1 expression was higher in tissue adjacent to the caseous necrosis areas (Fig. 2, lane 3) than in samples taken from the necrosis itself (lane 4) or from normal-looking distal tissue (lane 5). LN from TB-free animals exhibited only a 12-kDa spurious band (Fig. 2, lane 6). The variety of NRAMP1 bands observed in this work is in keeping with published data. For murine macrophages, several molecular sizes for NRAMP1 polypeptides have been demonstrated previously, including proteins of 100, 90, 65, and 45 kDa (2). As proposed in the literature, the high-molecular-mass proteins (90 and 95 kDa) detected in granulomas could well represent mature phosphorylated and/or highly glycosylated NRAMP1 isoforms induced by cell activation taking place locally in the lesions. However, the possibility that they are cross-reactive bands cannot be ruled out. Expression of the NRAMP1 bands was much stronger in tissue adjacent to the caseous necrosis areas than in samples taken from the necrosis itself and from normal-looking distal tissue. This finding is perhaps related to the abundance of macrophages in these areas (Fig. 3B).

View larger version (38K): [in this window] [in a new window]   FIG. 2.   Western blotting of granulomas from tuberculous cattle to detect NRAMP1. Lanes 1 and 2, tuberculous granulomas; lane 3, tuberculous granuloma, tissue adjacent to the necrotic center; lane 4, tuberculous granuloma, caseous necrosis; lane 5, tuberculous granuloma, distal tissue; lane 6, normal LN from a healthy bovine; lane 7, PBC from a tuberculous bovine. The 43-kDa band corresponds to actin control.

View larger version (147K): [in this window] [in a new window]   FIG. 3.   Immunolabeling of granulomas with the anti-NRAMP1 antiserum. (A) Many NRAMP1-labeled cells are seen around the caseous necrosis area (cn) of the granuloma. (B) Immunolabeling with an antimacrophage MAb showing many labeled cells around the necrotic area (cn). (C) Coarse cytoplasmic granules are seen in many macrophages. (D) Control in which the anti-NRAMP1 antiserum was omitted; labeling is not observed. (E) Fine NRAMP1 granules are seen in the cytoplasm of a multinucleated giant cell of the Langhans type. Granules are seen outlining cytoplasmic vacuoles (arrows). (F) Normal LN taken from a healthy bovine. A few scattered macrophages display membrane labeling with the anti-NRAMP1 antiserum (arrow). Magnifications: A and B, ×40; C and D, ×80; E and F, ×200.

The expression of NRAMP1 was also analyzed by immunohistochemistry of LN with granulomas of five cattle with confirmed M. bovis infection. The procedure was as follows: 5-µm sections obtained from formalin-fixed paraffin-embedded LN were placed on poly-L-lysine-treated slides (Sigma). After deparaffinization and rehydration, sections were immersed in 10 mM sodium citrate (pH 6.0) solution and submitted to two rounds of 5 min of heating at 95°C with cooling at room temperature. Endogenous peroxidase was eliminated with H₂O₂ for 10 min. To diminish nonspecific binding, slides were treated with a solution of 10% horse serum and 1% bovine serum albumin. After rinsing, slides were incubated with the anti-NRAMP1 antiserum diluted 1:50 in the blocking solution for 3 h at room temperature. Bound antibodies were revealed with a goat anti-rabbit IgG peroxidase-labeled antibody (Zymed) diluted 1:200 for 90 min. These studies showed strong NRAMP1 labeling of many cells within the granuloma, particularly in areas adjacent to foci of caseous necrosis (Fig. 3A). In control slides in which the anti-NRAMP1 antiserum was omitted, there was no cell labeling (Fig. 3D). In LN obtained from healthy cattle, only occasional cells were weakly labeled at the plasma membrane (Fig. 3F, arrows). Immunolabeling with rabbit anti-M. bovis antiserum revealed very rare macrophages with few labeled rods (data not shown).

In mice, pigs, cattle, and humans, monocytes/macrophages are the main cells expressing NRAMP1 (5, 9, 26, 29). There are few studies that have analyzed the expression of NRAMP1 in mammalian tissues, all of them using RNA (6), and to the best of our knowledge, there are no studies of NRAMP1 protein expression in granulomas. In this study, the NRAMP1-expressing cells within the granuloma showed abundant cytoplasm and vesicular nuclei compatible with epithelioid macrophages. This was confirmed by immunolabeling with the anti-macrophage MAb MCA874G (Serotec) which showed labeled cells (Fig. 3B) with a topographical distribution similar to that observed with the anti-NRAMP1 antibody (Fig. 3A). Moreover, in several granulomas there were multinucleated cells of the Langhans type with numerous fine NRAMP1 granules within the cytoplasm and outlining vacuoles (Fig. 3E, arrows). Langhans cells result from macrophage fusion induced by T-cell cytokines such as gamma interferon (IFN-) (10) and are characteristic of tuberculous granulomas (21). The coarse cytoplasmic granules observed in macrophages and the vacuolar pattern of Langhans cells (Fig. 3C and E) suggest that NRAMP1 is located in phagosomes and/or phagolysosomes, which are the subcellular location assigned to NRAMP1 in recent studies (15).

The production of NRAMP1 by macrophages is up-regulated by activation factors such as IFN-, tumor necrosis factor alpha, interleukin 1, and lipopolysaccharide or by intracellular infection (12, 25, 26). The action of one or more of these factors could explain the high expression of NRAMP1 in granulomas of M. bovis-infected cattle. Granulomas are the battlefield where host defense cells and mycobacteria confront one another. In these reactions, considered the tissue expression of antimycobacterial cell-mediated immunity, T cells and macrophages activated by mycobacterial products release IFN- and tumor necrosis factor alpha. A high level of IFN- production has been documented in granulomas of M. bovis-infected cattle (23). The high expression of the NRAMP1 protein in granuloma cells of tuberculous cattle contrasts with its lack of expression in mouse strains susceptible to M. bovis BCG infection (27).

There is little information about the role of NRAMP1 in bovine TB. A recent genetic study failed to show evidence of an association between this gene and resistance or susceptibility to TB in cattle (3). However, further analyses are needed to elucidate the role of this protein in resistance to M. bovis infection. In conclusion, the present work shows that active M. bovis infection in cattle can occur despite high-level expression of NRAMP1 in granulomas and PBC.