Preventing lipid oxidation during recovery of functional proteins from herring (Clupea harengus) fillets by an acid solubilization process

It has previously been found that a process based on solubilization at pH 2.7 gives high yields of herring muscle proteins with good functionality. In this study, the development of lipid oxidation during acid processing of herring mince was studied. It was tested how modifications of the process conditions and/or additions of antioxidants could prevent lipid oxidation during the actual process and then during ice storage of the protein isolates. Processing parameters evaluated were prewash of the mince, exposure time to pH 2.7, inclusion or exclusion of a high-speed centrifugation, and addition of antioxidants. Antioxidants tested were erythorbate (0.2%, 9.3 mM), sodium tripolyphosphate (STPP; 0.2%, 5.4 mM), ethylenediaminetetraacetic acid (EDTA; 0.044%, 1.5 mM), and milk proteins (4%). The first three antioxidants were added in the prewash or during the homogenization step, whereas milk proteins were added to the final precipitate. At time 0, all isolates were analyzed for pH, moisture content, and thiobarbituric reactive substances (TBARS). Selected isolates were also analyzed for lipid and protein content. Stability during ice storage was followed in terms of odor, TBARS, and color (a/b values). Extensive lipid oxidation took place using the "control" process without high-speed centrifugation. This was not significantly (p < or = 0.05) affected by a prewash or varied exposure time to pH 2.7. Including high-speed centrifugation (20 min, 10,000g) significantly (p < or = 0.05) reduced TBARS values, total lipids, a values and b values. Erythorbate alone, or in combination with STPP/EDTA, significantly (p < or = 0.05) reduced lipid oxidation during processing if added in the prewash or homogenization step. During ice storage, better stability was gained when antioxidants were added in both of these steps and when EDTA was used instead of STPP.     

Dynamic headspace gas chromatography/mass spectrometry characterization of volatiles produced in fish oil enriched mayonnaise during storage

Protection against lipid oxidation and formation of unpleasant fishy and rancid off-flavors in oil-in-water food emulsions, such as fish oil enriched mayonnaise, is difficult to achieve. Volatile profiles from stored mayonnaises with different oil phase compositions were collected using a developed dynamic headspace sampling technique, in which interfering acetic acid was removed in situ with potassium hydroxide, and subsequently 148 volatiles were characterized and monitored by gas chromatography/mass spectrometry. Multivariate statistics showed correlation between the concentration of 62 volatiles and the fish oil and storage parameters, indicating the formation of lipid oxidation products, which impose fishy off-flavors. Further verification was obtained by gas chromatography/olfactometry, by which, among 78 odors, cis-4-heptenal and trans,cis-2,4-heptadienal were detected as distinct fishy notes. In total, 27 volatiles, including 1-penten-3-one, cis-2-penten-1-ol, cis-3-hexenal, cis-4-heptenal, 1-octen-3-one, 1,cis-5-octadien-3-one, 1-octen-3-ol, trans,cis-2, 4-heptadienal, and trans,cis-2,6-nonadienal, were suggested to contribute to the developed unpleasant fishy and rancid off-flavors.     

Inhibition of hemoglobin-mediated oxidation of regular and lipid-fortified washed cod mince by a white grape dietary fiber

The effectiveness of a white grape dietary fiber concentrate (WGDF) against hemoglobin-mediated oxidation of washed cod mince, with and without 10% added herring oil, was evaluated during ice storage. WGDF was added at two different levels: 2 and 4% based on final weight. An ethanol extract with the ethanol extractable polyphenols (EPP) and the ethanol-extracted grape dietary fiber residue were also tested as antioxidants in the washed cod mince. The addition of WGDF to the model system completely and significantly (p 相似文献   

Freshness assessments of Moroccan sardine (Sardina pilchardus): comparison of overall sensory changes to instrumentally determined volatiles

Freshness of ice-stored sardine was assessed by two sensory methods, the quality index method (QIM) and the European Union freshness grading system, and by instrumental means using the method of aroma extract dilution analysis. Screening of sardine potent volatiles was carried out at three freshness stages. In the very fresh state, the plant-like fresh volatiles dominated the odor pattern, with the exception of methional. Overall odor changes in sardine throughout storage correlated with changes in the concentration of some potent volatiles: after 2 days of ice storage, (Z)-4-heptenal, (Z)-1,5-octadien-3-one, and methional imparted an overall "fishy" odor character to sardine, whereas at a lower sensory grade (B), the compounds (E)-2-nonenal and (E,Z)-2,6-nonadienal could be, in part, associated with the slightly rancid aroma top notes. Trimethylamine was detected as a highly volatile odorant using solid-phase microextraction (SPME) headspace analysis of refrigerator-stored sardine. Intensity and sensory characteristics of some SPME determined volatiles, for example, 3-methylnonane-2,4-dione, were closely related to overall odor changes. SPME headspace analysis may be useful in the characterization of off-flavors in fish.     

Hemoglobin-mediated lipid oxidation and compositional characteristics of washed fish mince model systems made from cod (Gadus morhua), herring (Clupea harengus), and salmon (Salmo salar) muscle

The use of washed cod light muscle minces in mechanistic studies of hemoglobin (Hb)-mediated fish lipid oxidation has largely increased in the past 5 years. Although cod light muscle has a low level of intrinsic lipid oxidation catalysts, a prerequisite for a good oxidation model system, we believe it cannot fully mimic the oxidation kinetics taking place in other fish species being more susceptible to lipid oxidation. The aim of this study was to systematically investigate whether washed mince model systems useful in Hb-mediated oxidation studies could be prepared also from herring (Clupea harengus) and salmon (Salmo salar) light muscles. The kinetics of oxidation in the washed models was measured during ice storage (+/-Hb), and the results were related to compositional differences. Minces from cod, herring, and salmon light muscles were washed 3 times with 3 volumes of water and buffer. A 20 microM portion of Hb and 200 ppm streptomycin was then added, followed by adjustment of pH and moisture to 6.3 and 86%, respectively. Samples with or without Hb were then stored on ice, and oxidation was followed as peroxide value (PV), rancid odor, redness (a*) loss and yellowness (b*). Prior to storage, all minces and models were also analyzed for total lipids, fatty acids, alpha-tocopherol, proteins, Hb, Fe, Cu, and Zn. Hb-mediated lipid oxidation appeared within 2 days on ice in all models. Small differences in the oxidation rates ranked the models as herring > cod > salmon. These differences were ascribed to more preformed peroxides and trace elements in the herring model, and more antioxidants in the salmon model. Controls, without Hb, stayed stable in all cases except herring, where a very slight oxidation appeared, especially if the herring raw material had been prefrozen. In conclusion, fattier fish like dark muscle species and salmonoids are useful for making washed mince model systems and would be a better choice than cod if there is an interest in the oxidation kinetics of such species.     

Loss of redness (a*) as a tool to follow hemoglobin-mediated lipid oxidation in washed cod mince

Instrumental measurement of redness loss (decrease in a* value) was evaluated as a tool to follow hemoglobin (Hb)-mediated lipid oxidation in fish muscle. Two washed cod mince model systems were used (prepared at pH 6.5 and 5.5), both fortified with 15 micromol/kg of trout Hb and adjusted to pH 6.5 and 81% moisture. The rate of oxidation was varied through pH alterations (pH 6.1 and 6.9) and addition of an antioxidative cod muscle press juice. During ice storage, TBARS, painty odor, and a* values were followed. In all "oxidizing" samples, a* values correlated well with TBARS and painty odor development; r = -0.95 and -0.77, respectively. In press juice containing samples, the correlation was lower (0.55 for a vs TBARS) because there was a slight a* value decrease even in the absence of measurable lipid oxidation. a* values distinguished between "oxidizing" and stable samples within 1 day, before any lipid oxidation products could be chemically detected. It was confirmed in an aqueous phosphate buffer model system that the redness loss corresponded to a buildup of brownish met-Hb at the expense of oxy- and deoxy-Hb. The a* value data were best used as a lipid oxidation index by calculating the rate of decrease (k value) in the "initial phase" of the redness loss (before accumulation of lipid oxidation products) or in the "differentiation phase" (during the exponential raise in TBARS/painty odor). Calibration to lipid oxidation products must, however, be made for each specific sample type. Washing method, pH, Hb-type, etc., all affected both k values and absolute a* readings. Small yellowness (b*) increases also occurred along with a value losses, possibly the result of polymerized Schiff bases.     

Added triacylglycerols do not hasten hemoglobin-mediated lipid oxidation in washed minced cod muscle

Hemoglobin-mediated lipid oxidation in washed, minced cod muscle was related to the triacylglycerol to membrane lipid ratio. The same rapid development of thiobarbituric acid reactive substances (TBARS) and painty odor occurred with and without the presence of up to 15% menhaden oil. Without hemoglobin, development of TBARS and painty odor was slow, despite a high amount of hydroperoxides in samples with oil added (1135 micromol/kg muscle). This suggested that hemoglobin reacted by cleaving preformed hydroperoxides into secondary oxidation products. Nearly doubling the hemoglobin concentration approximately doubled the extent of lipid oxidation with and without added oil. This indicated that hemoglobin was limiting for the oxidation reaction. The noneffect of added oil suggests that membrane lipids and/or preformed membrane lipid hydroperoxides provided sufficient substrate in hemoglobin-catalyzed oxidation of washed minced cod muscle. Fe(2+-)ADP did not induce any oxidation of washed minced cod with/without added oil. Results suggest that lipid oxidation in fatty fish may be more related to the quantity and type of the aqueous pro-oxidant and the membrane lipids than to variations in total fat contents.     

Aqueous extracts from some muscles inhibit hemoglobin-mediated oxidation of cod muscle membrane lipids

It was evaluated whether trout hemoglobin (Hb)-mediated oxidation of minced washed cod muscle lipids could be prevented by an aqueous isolate from cod and some other muscle sources. Lipid hydroperoxides and painty odor developed approximately 4 days faster in washed than unwashed cod mince. When adding back an aqueous fraction (press juice) isolated from unwashed mince to washed mince at 2-6-fold dilutions, development of hydroperoxides and painty odor was either delayed or completely prevented. The inhibitory substances were heat stable, and their effect was slightly reduced at reduced pH. The <1 kDa fractions of whole and heated press juices were as inhibitory as the unfractionated press juices. Inhibition by the unheated, heated, and ultrafiltered (30 kDa) press juices was lost after dialysis. These findings implied the presence of one or more highly effective aqueous low molecular weight antioxidants in cod muscle press juice. The same antioxidative properties were found in heated haddock, dab, and winter flounder muscle press juices but not in heated herring and chicken muscle press juices. Unheated chicken press juice was however highly inhibitory.     

Chemical and olfactometric characterization of volatile flavor compounds in a fish oil enriched milk emulsion

Development of objectionable fishy off-flavors is an obstacle in the development of fish oil enriched foods. Only little is known about the sensory impact of specific volatile fish oil oxidation products in food emulsions. This study examined the volatiles profiles of fish oil enriched milk during cold storage (2 degrees C) for 14 days by dynamic headspace sampling followed by gas chromatography-mass spectrometry analyses. Different volatiles (n = 60) comprising alkenals, alkadienals, alkatrienals, and vinyl ketones were identified in the fish oil enriched milk. The potent odorants identified by gas chromatography-olfactometry were 1-penten-3-one, (Z)-4-heptenal, 1-octen-3-one, (Z)-1,5-octadien-3-one, (E,E)-2,4-heptadienal, and (E,Z)-2,6-nonadienal, but despite their potency, none of the separated volatiles imparted a fishy or metallic odor. Two isomers, (E,Z,Z) and (E,E,Z) of 2,4,7-decatrienal were identified in fish oil enriched milk emulsions with peroxide values 0.8 and 3.4 meq/kg, respectively. To our knowledge, this is the first report on appearance of these decatrienals in food emulsions having a relatively low peroxide value.     

Volatile composition of sunflower oil-in-water emulsions during initial lipid oxidation: influence of pH.

The formation of odor active compounds resulting from initial lipid oxidation in sunflower oil-in-water emulsions was examined during storage at 60 degrees C. The emulsions differed in initial pH, that is, pH 3 and 6. The volatile compounds were isolated under mouth conditions and were analyzed by gas chromatography/sniffing port analysis. The lipid oxidation rate was followed by the formation of conjugated hydroperoxide dienes and headspace hexanal. The initial pH affected the lipid oxidation rate in the emulsions: the formation of conjugated diene hydroperoxides and the hexanal concentration in the static headspace were increased at pH 6. Pentanal, hexanal, 3-pentanol, and 1-octen-3-one showed odor activity in the emulsions after 6 days of storage, for both pH 3 and 6. Larger amounts of odor active compounds were released from the pH 6 emulsion with extended storage. It was shown that this increased release at pH 6 was not due to increased volatility because an increase in pH diminished the static headspace concentrations of added compounds in emulsions.     

Chemical and sensory evaluation of crude oil extracted from herring byproducts from different processing operations

Fish oils extracted from marinated herring (frozen and unfrozen) byproducts and maatjes herring byproducts were evaluated on their chemical and sensory properties. The obtained crude oils had very low content of copper (<0.1 mg/kg oil), and iron values were 0.8, 0.1, and 0.03 mg/kg oil, respectively, for oil from maatjes and frozen and fresh byproducts. For the maatjes oil, a much lower value was found for alpha-tocopherol compared to the other oils. Storage stability results showed that the oils behave differently. Secondary oxidation products were measured for fresh oil, while for the maatjes and frozen byproducts' oil, tertiary oxidation products were detected. Over storage time, the maatjes and frozen byproducts' oils became more intense in odor, correlating positively at the end with sensory attributes of train-oil, acidic, marine and fishy. The best correlation between sensory and chemical analyses was found for FFA and fishy off-odor (r = 0.781).     

Lipid oxidation in fillets of herring (Clupea harengus) during ice storage

The influence of ice storage on lipid oxidation, odor, antioxidants, water-soluble catalysts, and microorganisms was investigated in fillets of herring (Clupea harengus) during 15 days. Based on linear regression analyses of the data, significant rises (p ascorbic acid > glutathione peroxidase (GSH-px); however, GSH-px correlated best to the development of lipid oxidation products (r(mean) = -0.96). The activity of aqueous pro-oxidants, which were enzymatic in nature to a great extent, had decreased by 75% at day 15. No significant increase in total bacteria was seen until after 7 days. There were major local differences in both composition and stability throughout the fillet. Oxidation proceeded most rapidly in the tissue right under the skin, probably explained by its high initial pro-oxidative activity.     

Antioxidative properties of press juice from herring (Clupea harengus) against hemoglobin (Hb) mediated oxidation of washed cod mince

The antioxidative effect of herring (Clupea harengus) light muscle press juice (PJ) against hemoglobin-(Hb-) mediated oxidation of washed cod mince during ice storage was tested. The PJ was fractionated into low-molecular-weight (LMW; <1 [corrected] kDa) and high-molecular-weight (HMW; >1, >3.5, and > 50 kDa) fractions; it was preheated (10 min, 100 degrees C) and tested with or without removing heat coagulated proteins. Its antioxidative effect was compared with that given by endogenous levels of two tentative antioxidant candidates: ascorbic acid and uric acid. Oxidation was followed by determining rancid odor, peroxide value, and redness. Whole herring PJ and the LMW-PJ fraction significantly (p < 0.001) extended the oxidation lag phase of controls, from 2 up to 8 and 7 days, respectively. The HWM-PJ fractions were significantly (p < 0.05) less efficient than the whole and LMW-PJ samples, giving only 3.5-4.5 days of lag phase. Heat-treated PJ, with and without the heat-coagulated proteins, gave 7 and 5 days of oxidation lag phase, respectively. Heating different batches of the LMW-PJ fraction grouped the results into two categories: one where heating almost fully destroyed the antioxidative activity (fractions prepared from spring-caught herring) and another where heating had no or a minor effect (fractions prepared from fall-caught herring). The spring LMW-PJ had low ascorbic acid levels (18-42.6 microM), and 50-100% were destroyed by the heating. In fall LMW-PJ, the levels were 76.2-137.6 microM, and only 43-51% were destroyed. Ascorbic acid fortification of heated spring LMW-PJ to reach the levels found in the corresponding unheated spring LMW-PJ sample and the heated fall LMW-PJ gave back most of the antioxidative activity, which proved an important role of ascorbic acid for the antioxidative activity of LMW-herring PJ. This conclusion is drawn despite the fact that pure solutions with endogenous levels of ascorbic acid (giving 8.4-19.6 microM in final model) only very slightly delayed Hb-mediated oxidation of the washed cod mince.     

Lipid oxidation in fish oil enriched mayonnaise: calcium disodium ethylenediaminetetraacetate, but not gallic acid, strongly inhibited oxidative deterioration

The antioxidative effects of gallic acid, EDTA, and extra emulsifier Panodan DATEM TR in mayonnaise enriched with 16% fish oil were investigated. EDTA reduced the formation of free radicals, lipid hydroperoxides, volatiles, and fishy and rancid off-flavors. The antioxidative effect of EDTA was attributed to its ability to chelate free metal ions and iron from egg yolk located at the oil-water interface. Gallic acid reduced the levels of both free radicals and lipid hydroperoxides but promoted slightly the oxidative flavor deterioration in mayonnaise and influenced the profile of volatiles. Gallic acid may therefore promote the decomposition of lipid hydroperoxides to volatile oxidation products. Addition of extra emulsifier reduced the lipid hydroperoxide levels but did not influence the level of free radicals or the oxidative flavor deterioration in mayonnaisse; however, it appeared to alter the profile of volatiles. The effect of the emulsifier on the physical structure and rheological properties depended on the presence of antioxidants.     

Modified atmosphere packaged cheddar cheese shreds: influence of fluorescent light exposure and gas type on color and production of volatile compounds

The influences of fluorescent light exposure and packaging atmosphere on the headspace volatiles and color of Cheddar cheese shreds were evaluated using gas chromatography-mass spectrometry and spectrocolorimetry, respectively. Cheddar cheeses were packaged under atmospheres of 100% carbon dioxide or 100% nitrogen and stored at 4 degrees C under fluorescent light for 6 weeks. Cheeses stored under carbon dioxide contained higher concentrations of aldehydes and fatty acids and lower concentrations of alcohols and esters than cheeses stored under nitrogen. Carbon dioxide atmospheres potentiated light-induced oxidation in shredded Cheddar cheeses, as evidenced by aldehyde and fatty acid headspace volatiles measured following storage. Color bleaching occurred only in cheeses packaged under carbon dioxide and exposed to light. The shift in color is proposed to be due to an interaction between carbon dioxide and high-intensity light, leading to the oxidation of the pigment molecule, bixin. The results have significant implications for procedures used to handle and store pigmented cheeses to ensure desirable flavor and consumer acceptability.     

Solid phase microextraction-gas chromatography for quantifying headspace hexanal above freeze-dried chicken myofibrils

A method using solid phase microextraction (SPME) combined with gas chromatography/mass spectrometry (GC/MS) was developed and used to determine the oxidation of freeze-dried chicken myofibrils spiked with methyl linoleate. Freeze-dried chicken myofibrils were found to act as a significant reservoir for hexanal. Recovery of hexanal emissions from the headspace above spiked myofibrils was 95% using a 5 min sampling time, with a total analysis time of approximately 12 min/sample. The SPME-GC/MS working linear response was from 0.01 to 10 mg hexanal/L (r( 2) = 0.995). Freeze-dried chicken myofibrils with added methyl linoleate (0.6 mmol/g of protein) were stored at 50 degrees C at water activities of 0.30 and 0.75 for 0, 12, 27, and 50 h. Lipid oxidation was determined using SPME-GC/MS to measure headspace hexanal concentration, the thiobarbituric acid reactive substances assay (TBARS) to quantify malonaldehyde, and a conjugated diene assay. Lipid oxidation was influenced by storage time and water activity. A strong correlation (r = 0.938) existed between SPME-GC/MS and TBARS. The use of SPME-GC/MS was a sensitive and rapid method for detecting hexanal as an indicator of lipid oxidation in chicken myofibrils.     

Chitosan as an edible invisible film for quality preservation of herring and atlantic cod

The effect of chitosan with different molecular weights as coatings for shelf-life extension of fresh fillets of Atlantic cod (Gadus morhua) and herring (Clupea harengus) was evaluated over a 12-day storage at refrigerated temperature (4 +/- 1 degrees C). Three chitosan preparations from snow crab (Chinoecetes opilio) processing wastes, differing in viscosities and molecular weights, were prepared; their apparent viscosities (360, 57, and 14 cP) depended on the deacetylation time (4, 10, and 20 h, respectively) of the chitin precursor. Upon coating with chitosans, a significant (p < or = 0.05) reduction in relative moisture losses of 37, 29, 29, 40, and 32% was observed for cod samples coated with 360 cP chitosan after 4, 6, 8, 10, and 12 days of storage, respectively. Chitosan coating significantly (p < or = 0.05) reduced lipid oxidation as displayed in peroxide value, conjugated dienes, 2-thiobarbituric acid reactive substances and headspace volatiles, chemical spoilage as reflected in total volatile basic nitrogen, trimethylamine, and hypoxanthine, and growth of microorganisms as reflected in total plate count in both fish model systems compared to uncoated samples. The preservative efficacy and the viscosity of chitosan were inter-related; the efficacy of chitosans with viscosities of 57 and 360 cP was superior to that of chitosan with a 14 cP viscosity. Thus, chitosan as edible coating would enhance the quality of seafoods during storage.     

Hemoglobin-mediated oxidation of washed minced cod muscle phospholipids: effect of pH and hemoglobin source

Lipid pro-oxidative properties and deoxygenation/autoxidation patterns of hemoglobins from nonmigratory white-fleshed fish (winter flounder and Atlantic pollock) and migratory dark-fleshed fish (Atlantic mackerel and menhaden) were compared during ice storage at pH 7.2 and 6. A washed cod mince model system and a buffer model system were used for studying lipid changes and hemoglobin changes, respectively. TBARS and painty odor were followed as markers for lipid oxidation. At pH 6, all four hemoglobins were highly and equally active as pro-oxidants. At pH 7.2, pro-oxidation by all hemoglobins except that from pollock was slowed down, and activity ranked as pollock > mackerel > menhaden > flounder. The higher catalytic activities of the hemoglobins at pH 6 than at pH 7.2 corresponded with higher formation of deoxyhemoglobin and methemoglobin. Pollock had the most extensive formation of deoxy- and methemoglobin at both pH values, which could explain its high catalytic activity. The pro-oxidative differences among the other hemoglobins at pH 7.2 did not correlate with deoxygenation and autoxidation reactions. This indicates involvement of other structural differences between the hemoglobins such as differences in the heme-crevice volume. It is suggested that a biological reason for the species differences was their adaptations to different depths/water temperatures.     

Identification and sensory evaluation of volatile compounds in oxidized porcine liver

Headspace solid phase microextraction (HS-SPME) was used to isolate the off-flavor volatile compounds, which are formed during the oxidation of porcine liver induced by iron. Poly(dimethylsiloxane)/divinylbenzene fiber was used in the HS-SPME. Changes in the volatile compounds of oxidized porcine liver and unsaturated fatty acids induced by iron were examined. Results showed that 1-octen-3-one (metallic), hexanol (weak metallic), 1-octen-3-ol (mushroomlike), (E)-2-nonenal (cardboardlike), and (E,E)-2,4-decadienal (fatty, oily) were the main contributors to the overall off-flavor of porcine liver. The results of the sensory evaluation revealed that oxidized arachidonic acid has a major impact on metallic and liverlike off-flavor and that when liverlike off-flavor is perceived, metallic is also included. Oxidized linolenic acid was the most important contributor to the objectionable fishy off-flavor. Oxidized porcine liver exhibited distinct metallic, liverlike, and weak fishy background notes. Liverlike flavor had a high correlation coefficient with odor characteristics such as metallic (0.839) and fishy (0.777). In this study, it was clearly observed that the stronger the metallic and fishy off-flavor the higher the perception of liverlike off-flavor.     

Protein and lipid oxidation during frozen storage of rainbow trout (Oncorhynchus mykiss)

This study aimed at investigating protein and lipid oxidation during frozen storage of rainbow trout. Rainbow trout fillets were stored for 13 months at -20, -30, or -80 degrees C, and samples were analyzed at regular intervals for lipid and protein oxidation markers. Lipid oxidation was followed by measuring lipid hydroperoxides (PV), as well as secondary oxidation products (volatiles) using dynamic headspace GC-MS. Free fatty acids (FFA) were measured as an estimation of lipolysis. Protein oxidation was followed using the spectrophotometric determination of protein carbonyls and immunoblotting. Significant oxidation was observed in samples stored at -20 degrees C, and at this temperature lipid and protein oxidation seemed to develop simultaneously. FFA, PV, and carbonyls increased significantly for the fish stored at -20 degrees C, whereas the fish stored at -30 and -80 degrees C did not show any increase in oxidation during the entire storage period when these methods were used. In contrast, the more sensitive GC-MS method used for measurement of the volatiles showed that the fish stored at -30 degrees C oxidized more quickly than those stored at -80 degrees C. Detection of protein oxidation using immunoblotting revealed that high molecular weight proteins were oxidized already at t = 0 and that no new protein oxidized during storage irrespective of the storage time and temperature. The results emphasize the need for the development of more sensitive and reliable methods to study protein oxidation in order to gain more explicit knowledge about the significance of protein oxidation for food quality and, especially, to correlate protein oxidation with physical and functional properties of foods.