- Email: [email protected]
Redox proteomics in the mussel, Mytilus edulis
MARINE ENVIRONMENTAL RESEARCH Marine Environmental Research 62 (2006) S101–S104 www.elsevier.com/locate/marenvrev
Short communication
Redox proteomics in the mussel, Mytilus edulis B. McDonagh, R. Tyther, D. Sheehan
*
Proteomics Research Group, Department of Biochemistry, Environmental Research Institute, University College Cork, Lee Maltings, Prospect Row, Mardyke, Cork, Ireland
Abstract Pollutants (e.g. PAHs, metals) cause oxidative stress (OS) by forming reactive oxygen species. Redox proteomics provides a means for identifying protein-specific OS effects in Mytilus edulis. Groups of mussels were sampled from a clean site in Cork Harbour, Ireland and exposed to 1 mM H2O2 in holding tanks. Protein extracts of gill and digestive gland were separated by two dimensional electrophoresis and similar protein expression profiles were found. Effects of OS on disulphide bridge patterns were investigated in diagonal gels by separating proteins in non-reducing conditions followed by a second reducing dimension. Immunoprecipitation selected carbonylated and glutathionylated proteins. These methodologies can contribute to redox proteomic studies of pollutant responses in marine organisms. 2006 Elsevier Ltd. All rights reserved. Keywords: Carbonylation; Glutathione; Oxidative stress; Proteomics; Mytilus edulis; Actin
Oxidizing changes in redox potential can modify amino acid side chains by carbonylation, glutathionylation or altered formation of disulphide bridges (Stadtman and Levine, 2000). Some modifications cause inactivation, some are protective and others allow the cell to ‘‘sense’’ altered redox status. Aquatic organisms such as mussels are constantly exposed to pro-oxidants in their natural environment since seawater contains appreciable amounts of H2O2 generated by photo-oxidation (O’Sullivan et al., 2005). Some environmental pollutants including polyaromatic hydrocarbons (PAHs) and metals can generate reactive oxygen species (Livingstone et al., 1993). When the antioxidant defences of a marine *
Corresponding author. Tel.: +353 21 4904207; fax: +353 21 4274034. E-mail address: [email protected] (D. Sheehan).
0141-1136/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.marenvres.2006.04.001
S102
B. McDonagh et al. / Marine Environmental Research 62 (2006) S101–S104
organism are overcome by excessive levels of reactive oxygen species, a state of oxidative stress (OS) results. Biochemical markers of OS may provide novel endpoints for exposure to environmental pollutants (Livingstone et al., 1993; McDonagh et al., 2005; Valavanidis et al., in press). Here, we describe approaches to identify specific Mytilus edulis proteins targeted by oxidative stress (OS). Animals (50) approximately 4–6 cm in length were sampled from a reference site in Cork Harbour, Ireland (Lyons et al., 2003). Five groups (five individuals) were acclimated (1 week, 12 h light/dark cycle) with regular feeding (at intervals of 48 h) on PhytoplexTM phytoplankton feed (Kent Marine Inc., Acworth, GA, USA). They were then exposed to 1 mM H2O2 (24 h) and dissected after a recovery period (24 h). Comparison was made with unexposed controls (5 · 5 animals). A relatively short exposure time was used to gain insight into acute toxicity as a result of H2O2-induced OS (McDonagh et al., 2005). The twenty-four hour recovery period was to allow clearance of ‘‘unreacted’’ H2O2 by catalase since residual H2O2 might artefactually oxidize proteins in cell extracts. Gills/digestive glands were dissected, pooled and homogenized in 10 mM Tris/HCl, pH 7.2 containing 500 mM sucrose, 1 mM EDTA and 1 mM PMSF. Extracts were collected by centrifugation at 20,000g for 1 h at 4 C and stored ( 70 C) until required. Similar patterns/expression levels were found for control/exposed samples by two dimensional electrophoresis (2D SDS PAGE), so alternative analytical methods were explored. Proteins (50 lg) were separated by 12% non-reducing SDS-PAGE. The entire lane was excised, incubated in buffer containing 2% DTT (20 min) followed by 2.5% iodoacetamide (20 min). Slices were placed horizontally on a second 12% SDS-PAGE gel which was silver stained after electrophoresis. Proteins lacking disulphide bridges form a diagonal across the reducing gel (Fig. 1). Interchain disulphide bridges produce spots below the diagonal while intrachain disulphides produce spots above the diagonal. OS modifies actin (McDonagh et al., 2005), so we wondered if effects on disulphide bridge patterns of actin could be revealed in diagonal gels. Diagonal separations (Fig. 1) were electroblotted and probed with actin antibodies (Sigma, Poole, Dorset UK). This revealed
Fig. 1. Silver stained gels of gill (A) control (B) H2O2-exposed. Insets: Western blots probed with anti-actin. Protein spots below the diagonal (h) contain reduced intermolecular disulphide bridges while those above the diagonal (O) contain reduced intramolecular disulphide bridges. Location of actin is denoted by dashed boxes.
B. McDonagh et al. / Marine Environmental Research 62 (2006) S101–S104
S103
horizontal actin spots due to interchain disulphides (insets of Fig. 1) suggesting that actin exists in a range of forms depending on disulphide bridge pattern which is consistent with previous results (Brennan et al., 2004). Similarity of OS to control separations suggests that carbonylation/glutathionylation only modestly affect the disulphide status of actin. Identification of up/down regulated spots off the diagonal could provide insights into the redox response mechanisms of M. edulis, especially in haemolymph proteins since secreted (rather than cytosolic) proteins are richer in disulphide bridges. Immunoblotting of glutathionylated/carbonylated proteins in OS has previously been reported (Fratelli et al., 2002; Costa et al., 2002; McDonagh et al., 2005). We combined the selectivity of immunoprecipitation (IP) with the resolving power of 2D SDS PAGE as IP provides pre-separation enrichment. Glutathionylated proteins were incubated with anti-GSH (Virogen, Waterstown, MA, USA) followed by Protein A Sepharose Fast Flow (Amersham-Biosciences, Little Chalfont, Bucks, UK). Beads were separated by centrifugation and protein released by heating before application to 2D SDS PAGE. Carbonylated samples were derivatized with dinitrophenyl (DNP)-hydrazine before incubation with anti-DNP antibody (DakoCytomation). Distinct spots were apparent which were immunodetected as isoforms of actin in the GSH IP (Fig. 2(A)). However, non-actin spots are also visible. The pattern of carbonylated proteins (Fig. 2(B)) is similar to that found by immunodetection (McDonagh et al., 2005). No significant effects of H2O2-exposure on either diagonal gels or carbonylation patterns were observed in digestive gland extracts (not shown). Mytilus digestive gland is rich in Phase II detoxification enzymes, but it has lower levels of Phase I glutathione transferase activity (Fitzpatrick et al., 1995) and is a less important site of glutathionylation and carbonylation on acute H2O2 exposure than gill (McDonagh et al., 2005). We aim to identify specific protein targets for OS which could provide novel biomarkers of stress response in exposed organisms (Valavanidis et al., in press). Our approaches have been used in other research areas (Brennan et al., 2004; England and Cotter, 2004) but are novel in an environmental monitoring context. Several groups have previously used protein expression signatures to reveal changes in levels of specific proteins in bivalve
Fig. 2. (A) Immunoprecipitation with GSH antibodies on gill samples exposed to H2O2. Spots circled represent isoforms of glutathionylated actin. (B) Immunoprecipitation with antibodies to dinitrophenol on exposed gill samples revealing carbonylated proteins.
S104
B. McDonagh et al. / Marine Environmental Research 62 (2006) S101–S104
proteomes (literature cited in McDonagh et al., 2005). However, our experiments were performed against a largely stable proteomic background where little appreciable difference exists in the intensity of specific protein spots. We previously identified actin as a target both for carbonylation and glutathionylation (McDonagh et al., 2005). Use of diagonal gels shown here extends these observations by revealing that a range of disulphide-bridged variants of actin exists. This may provide a subtle means for the gill cytoskeleton to ‘‘sense’’ change in redox status. It is known that endoplasmic reticulum proteins such as protein disulphide isomerase (which forms disulphide bridges in proteins) are targets for OS as this compartment is necessarily partly oxidizing. References Brennan, J.P., Wait, R., Begum, S., Bell, J.R., Dunn, J., Eaton, P., 2004. Journal of Biological Chemistry 279, 41352–41360. Costa, V.M.V., Amorim, M.A., Quintanilha, A., Moradas-Ferreira, P., 2002. Free Radical Biology and Medicine 33, 1507–1515. England, K., Cotter, T., 2004. Biochemical and Biophysical Research Communications 320, 123–130. Fitzpatrick, P.J., Sheehan, D., Livingstone, D.R., 1995. Marine Environmental Research 39, 241–244. Fratelli, M., Demol, H., Puype, M., Casagrande, S., Eberini, I., Salmona, M., Bonetto, V., Mengozzi, M., Duffieux, F., Miclet, E., Bachi, A., Vandekerckhove, J., Gianazza, E., Ghezzi, P., 2002. Proceedings of the Naional Academy of Sciences USA 99, 3505–3510. Livingstone, D.R., Lemarie, P., Matthews, A., Peters, L., Bucke, D., Law, R.J., 1993. Marine Pollution Bulletin 26, 602–606. Lyons, C., Dowling, V., Tedengren, M., Hartl, M.G.J., O’Brien, N.M., van Pelt, F.N.A.M., O’Halloran, J., Sheehan, D., 2003. Marine Environmental Research 56, 585–597. McDonagh, B., Tyther, R., Sheehan, D., 2005. Aquatic Toxicology 73, 315–326. O’Sullivan, D., Neale, P.J., Coffin, R.B., Boyd, T.J., Osburn, C.L., 2005. Marine Chemistry 97, 14–33. Stadtman, E.R., Levine, R.L., 2000. Annual of the New York Academy of Sciences 899, 191–208. Valavanidis, A., Vlahogianni, T., Dassenakis, M., Scoullos, M., in press. Ecotoxicology and Environmental Safety.
Short communication
Redox proteomics in the mussel, Mytilus edulis B. McDonagh, R. Tyther, D. Sheehan
*
Proteomics Research Group, Department of Biochemistry, Environmental Research Institute, University College Cork, Lee Maltings, Prospect Row, Mardyke, Cork, Ireland
Abstract Pollutants (e.g. PAHs, metals) cause oxidative stress (OS) by forming reactive oxygen species. Redox proteomics provides a means for identifying protein-specific OS effects in Mytilus edulis. Groups of mussels were sampled from a clean site in Cork Harbour, Ireland and exposed to 1 mM H2O2 in holding tanks. Protein extracts of gill and digestive gland were separated by two dimensional electrophoresis and similar protein expression profiles were found. Effects of OS on disulphide bridge patterns were investigated in diagonal gels by separating proteins in non-reducing conditions followed by a second reducing dimension. Immunoprecipitation selected carbonylated and glutathionylated proteins. These methodologies can contribute to redox proteomic studies of pollutant responses in marine organisms. 2006 Elsevier Ltd. All rights reserved. Keywords: Carbonylation; Glutathione; Oxidative stress; Proteomics; Mytilus edulis; Actin
Oxidizing changes in redox potential can modify amino acid side chains by carbonylation, glutathionylation or altered formation of disulphide bridges (Stadtman and Levine, 2000). Some modifications cause inactivation, some are protective and others allow the cell to ‘‘sense’’ altered redox status. Aquatic organisms such as mussels are constantly exposed to pro-oxidants in their natural environment since seawater contains appreciable amounts of H2O2 generated by photo-oxidation (O’Sullivan et al., 2005). Some environmental pollutants including polyaromatic hydrocarbons (PAHs) and metals can generate reactive oxygen species (Livingstone et al., 1993). When the antioxidant defences of a marine *
Corresponding author. Tel.: +353 21 4904207; fax: +353 21 4274034. E-mail address: [email protected] (D. Sheehan).
0141-1136/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.marenvres.2006.04.001
S102
B. McDonagh et al. / Marine Environmental Research 62 (2006) S101–S104
organism are overcome by excessive levels of reactive oxygen species, a state of oxidative stress (OS) results. Biochemical markers of OS may provide novel endpoints for exposure to environmental pollutants (Livingstone et al., 1993; McDonagh et al., 2005; Valavanidis et al., in press). Here, we describe approaches to identify specific Mytilus edulis proteins targeted by oxidative stress (OS). Animals (50) approximately 4–6 cm in length were sampled from a reference site in Cork Harbour, Ireland (Lyons et al., 2003). Five groups (five individuals) were acclimated (1 week, 12 h light/dark cycle) with regular feeding (at intervals of 48 h) on PhytoplexTM phytoplankton feed (Kent Marine Inc., Acworth, GA, USA). They were then exposed to 1 mM H2O2 (24 h) and dissected after a recovery period (24 h). Comparison was made with unexposed controls (5 · 5 animals). A relatively short exposure time was used to gain insight into acute toxicity as a result of H2O2-induced OS (McDonagh et al., 2005). The twenty-four hour recovery period was to allow clearance of ‘‘unreacted’’ H2O2 by catalase since residual H2O2 might artefactually oxidize proteins in cell extracts. Gills/digestive glands were dissected, pooled and homogenized in 10 mM Tris/HCl, pH 7.2 containing 500 mM sucrose, 1 mM EDTA and 1 mM PMSF. Extracts were collected by centrifugation at 20,000g for 1 h at 4 C and stored ( 70 C) until required. Similar patterns/expression levels were found for control/exposed samples by two dimensional electrophoresis (2D SDS PAGE), so alternative analytical methods were explored. Proteins (50 lg) were separated by 12% non-reducing SDS-PAGE. The entire lane was excised, incubated in buffer containing 2% DTT (20 min) followed by 2.5% iodoacetamide (20 min). Slices were placed horizontally on a second 12% SDS-PAGE gel which was silver stained after electrophoresis. Proteins lacking disulphide bridges form a diagonal across the reducing gel (Fig. 1). Interchain disulphide bridges produce spots below the diagonal while intrachain disulphides produce spots above the diagonal. OS modifies actin (McDonagh et al., 2005), so we wondered if effects on disulphide bridge patterns of actin could be revealed in diagonal gels. Diagonal separations (Fig. 1) were electroblotted and probed with actin antibodies (Sigma, Poole, Dorset UK). This revealed
Fig. 1. Silver stained gels of gill (A) control (B) H2O2-exposed. Insets: Western blots probed with anti-actin. Protein spots below the diagonal (h) contain reduced intermolecular disulphide bridges while those above the diagonal (O) contain reduced intramolecular disulphide bridges. Location of actin is denoted by dashed boxes.
B. McDonagh et al. / Marine Environmental Research 62 (2006) S101–S104
S103
horizontal actin spots due to interchain disulphides (insets of Fig. 1) suggesting that actin exists in a range of forms depending on disulphide bridge pattern which is consistent with previous results (Brennan et al., 2004). Similarity of OS to control separations suggests that carbonylation/glutathionylation only modestly affect the disulphide status of actin. Identification of up/down regulated spots off the diagonal could provide insights into the redox response mechanisms of M. edulis, especially in haemolymph proteins since secreted (rather than cytosolic) proteins are richer in disulphide bridges. Immunoblotting of glutathionylated/carbonylated proteins in OS has previously been reported (Fratelli et al., 2002; Costa et al., 2002; McDonagh et al., 2005). We combined the selectivity of immunoprecipitation (IP) with the resolving power of 2D SDS PAGE as IP provides pre-separation enrichment. Glutathionylated proteins were incubated with anti-GSH (Virogen, Waterstown, MA, USA) followed by Protein A Sepharose Fast Flow (Amersham-Biosciences, Little Chalfont, Bucks, UK). Beads were separated by centrifugation and protein released by heating before application to 2D SDS PAGE. Carbonylated samples were derivatized with dinitrophenyl (DNP)-hydrazine before incubation with anti-DNP antibody (DakoCytomation). Distinct spots were apparent which were immunodetected as isoforms of actin in the GSH IP (Fig. 2(A)). However, non-actin spots are also visible. The pattern of carbonylated proteins (Fig. 2(B)) is similar to that found by immunodetection (McDonagh et al., 2005). No significant effects of H2O2-exposure on either diagonal gels or carbonylation patterns were observed in digestive gland extracts (not shown). Mytilus digestive gland is rich in Phase II detoxification enzymes, but it has lower levels of Phase I glutathione transferase activity (Fitzpatrick et al., 1995) and is a less important site of glutathionylation and carbonylation on acute H2O2 exposure than gill (McDonagh et al., 2005). We aim to identify specific protein targets for OS which could provide novel biomarkers of stress response in exposed organisms (Valavanidis et al., in press). Our approaches have been used in other research areas (Brennan et al., 2004; England and Cotter, 2004) but are novel in an environmental monitoring context. Several groups have previously used protein expression signatures to reveal changes in levels of specific proteins in bivalve
Fig. 2. (A) Immunoprecipitation with GSH antibodies on gill samples exposed to H2O2. Spots circled represent isoforms of glutathionylated actin. (B) Immunoprecipitation with antibodies to dinitrophenol on exposed gill samples revealing carbonylated proteins.
S104
B. McDonagh et al. / Marine Environmental Research 62 (2006) S101–S104
proteomes (literature cited in McDonagh et al., 2005). However, our experiments were performed against a largely stable proteomic background where little appreciable difference exists in the intensity of specific protein spots. We previously identified actin as a target both for carbonylation and glutathionylation (McDonagh et al., 2005). Use of diagonal gels shown here extends these observations by revealing that a range of disulphide-bridged variants of actin exists. This may provide a subtle means for the gill cytoskeleton to ‘‘sense’’ change in redox status. It is known that endoplasmic reticulum proteins such as protein disulphide isomerase (which forms disulphide bridges in proteins) are targets for OS as this compartment is necessarily partly oxidizing. References Brennan, J.P., Wait, R., Begum, S., Bell, J.R., Dunn, J., Eaton, P., 2004. Journal of Biological Chemistry 279, 41352–41360. Costa, V.M.V., Amorim, M.A., Quintanilha, A., Moradas-Ferreira, P., 2002. Free Radical Biology and Medicine 33, 1507–1515. England, K., Cotter, T., 2004. Biochemical and Biophysical Research Communications 320, 123–130. Fitzpatrick, P.J., Sheehan, D., Livingstone, D.R., 1995. Marine Environmental Research 39, 241–244. Fratelli, M., Demol, H., Puype, M., Casagrande, S., Eberini, I., Salmona, M., Bonetto, V., Mengozzi, M., Duffieux, F., Miclet, E., Bachi, A., Vandekerckhove, J., Gianazza, E., Ghezzi, P., 2002. Proceedings of the Naional Academy of Sciences USA 99, 3505–3510. Livingstone, D.R., Lemarie, P., Matthews, A., Peters, L., Bucke, D., Law, R.J., 1993. Marine Pollution Bulletin 26, 602–606. Lyons, C., Dowling, V., Tedengren, M., Hartl, M.G.J., O’Brien, N.M., van Pelt, F.N.A.M., O’Halloran, J., Sheehan, D., 2003. Marine Environmental Research 56, 585–597. McDonagh, B., Tyther, R., Sheehan, D., 2005. Aquatic Toxicology 73, 315–326. O’Sullivan, D., Neale, P.J., Coffin, R.B., Boyd, T.J., Osburn, C.L., 2005. Marine Chemistry 97, 14–33. Stadtman, E.R., Levine, R.L., 2000. Annual of the New York Academy of Sciences 899, 191–208. Valavanidis, A., Vlahogianni, T., Dassenakis, M., Scoullos, M., in press. Ecotoxicology and Environmental Safety.