Effect of Previous Frozen Storage and Coating Medium on the Essential Macroelement and Trace Element Content of Canned Mackerel

The effect of previous frozen storage (−18 °C for 6 months) and different coating media (aqueous: water and brine; oily: sunflower, refined olive, and extra-virgin olive oils) on the essential macroelement and trace element content of canned Atlantic mackerel (Scomber scombrus) was studied. Previous frozen storage led to an increased (p < 0.05) content of canned samples of K (oil-coated samples) and Ca (all coating conditions) and to a decreased (p < 0.05) content of P (aqueous-coating samples) and S (water- and oil-coated samples). For trace elements, a content increase (p < 0.05) in Cu and Se (brine-canned samples) and Mn (water- and refined-olive-oil-coated samples) was detected in canned fish muscle with frozen storage. Concerning the coating effect, aqueous-coating samples showed lower (p < 0.05) Mg, P, S, K, and Ca contents than their corresponding oil-coated samples. For trace elements, lower average contents were found for Co, Cu, Mn, Se, and Fe in aqueous-coating fish muscle when compared to their counterparts coated in oily media. Content changes in the different elements in canned fish muscle are discussed based on interactions with other tissue constituents and modifications that such constituents undergo during processing (i.e., protein denaturation, liquor losses from the muscle, lipid changes).


Introduction
The majority of the macroelements and trace elements considered essential for human biological processes, including growth, antioxidant defence, reproduction, and hormone metabolism, are present in marine fish and invertebrates [1][2][3]. Marine species can accumulate minerals from the diet and deposit them in their skeletal tissues and organs. Remarkably, the concentration of elements in seafood may be influenced by different factors such as nourishment source, environment, species, season, tissue, maturation degree, and processing [4][5][6].
Canning is a traditional process widely employed for marine species preservation [7,8]. The nature of the raw material can be altered substantially by the heat treatment involved to produce a food product with different characteristics [9,10]. Both bacteria and endogenous enzymes should be permanently inactivated by heat treatment and lead to a safe and durable food, provided that further contamination does not occur. Different kinds of coating media (namely, aqueous and oily) can be employed for canned fish commercialisation, all of them receiving great acceptance from consumers [11]. However, marine species are reported to be highly sensitive to heat processing, which leads to heat degradation, toughening and drying of fish muscle, oxidation of constituents, and leaching of watersoluble constituents [12][13][14]. With regards to the mineral content of canned fish muscle, The fish white muscle of two cans with the same coating medium was pooled together, minced, and employed for the different analyses. Cans corresponding to each coating medium were analysed by means of five replicates (n = 5).

Determination of Moisture Content
Moisture was determined as the weight difference (1-2 g) before and after 4 h at 105 • C in agreement with the official method, 950.46B [25]. Results were calculated as g·kg −1 .

Determination of Element Presence in Initial Fish and Canned Samples
Contents of essential macroelements (Ca, K, Mg, Na, P, and S) and trace elements (Co, Cu, Fe, Mn, and Se) were determined in initial and canned fish. About 300 mg of ground sample was put into a digestion flask with 9 mL of 69% nitric acid Hiperpur, 3 mL of H 2 O 2 (for ultratrace analysis), and 3 mL of Milli-Q water and digested according to the procedure based on EPA 3050B (US-EPA 1996). Quintuplicate samples, plus five blanks and five samples of certified reference material, were digested in an advanced microwave digestion system (ETHOS TM EASY, Milestone, Sorisole, Italy). Sample solutions were transferred to 50 mL flasks. A clean ISO 5 laminar flow cabinet (Cruma 670 FL, Barcelona, Spain) was employed for sample handling.
ICP-MS analysis using an Agilent 7900 instrument (Agilent Technologies, Inc., Santa Clara, CA, USA) was employed for determination of the eleven aforementioned elements. External calibration with element standards traceable to NIST standards was employed. The limits of detection (LD) were calculated by comparison to the standard deviation of the blanks (LD = 3·SD blanks). Procedural blanks always accounted for <1% of element concentrations in the samples. Accuracy of the analytical procedures was ensured using certified reference material DORM-2, prepared by the National Research Council of Canada (Table S1). The reference values of six macroelements in DORM-2 were reported by Engstrom et al. [26] and Millos et al. [27]. The results of the DORM-2 analysis were always within the ranges of certified reference material, except for Na, whose contents were slightly lower than certified values (Table S1).
Results were calculated as mg·kg −1 dry muscle.

Statistical Analysis
Data (n = 5) corresponding to the content of the different macroelements and trace elements were subjected to one-way ANOVA (p < 0.05) to analyse the effect of previous frozen storage and coating condition (Statistica version 6.0, 2001; Statsoft Inc., Tulsa, OK, USA) on the content of the different essential elements in the canned product. Differences between initial fish and canned samples were also analysed. In all cases, comparison of means was performed using a least-squares difference (LSD) method.

Determination of Essential Macroelements
Concerning the Na content, comparison between initial fish and canned fish revealed a marked decrease (p < 0.05) in water-coated samples and a strong increase (p < 0.05) in brine-canned ones (Table 1); no effect (p > 0.05) of the canning process was detected in the case of oil-coated samples.
Concerning the coating medium effect, canned samples corresponding to watercoating condition showed the lowest (p < 0.05) contents, while those coated in brine showed the highest (p < 0.05) contents. Additionally, comparison of the three oil-coated samples did not lead to observing differences (p > 0.05). The use of the 6-month frozen storage (Group II samples) led to higher average Na contents in water-canned and oil-canned samples but lower contents in brine-canned ones; however, differences were not significant (p > 0.05).
Compared to the initial fish, a marked decrease (p < 0.05) in the K content was detected in canned fish muscle that was coated in both aqueous media ( Figure 1). Fish muscle subjected to oil coating showed lower average contents than the initial fish; however, Foods 2023, 12, 2289 4 of 12 differences were not found significant (p > 0.05). Aqueous-coating samples showed lower (p < 0.05) K contents than their corresponding oil-coated samples. A K content increase (p < 0.05) was observed in oil-coated fish muscle as a result of the previous frozen storage; in the case of aqueous-coating samples, no effect (p > 0.05) was determined for the previous storage period.   Table S1.
Different tendencies were observed for the Ca content when considering fish with or without previous frozen storage ( Figure 2). Compared to the initial fish muscle, canned muscle without previous frozen storage showed lower average contents in samples coated in aqueous-coating systems and higher average values in oil-coated samples; differences from initial fish were found significant (p < 0.05) in the cases of brine-, ROO-, and EVOO-coated conditions. For canned fish muscle of Group II, higher average Ca contents were obtained in all kinds of canned samples when compared to initial fish; differences were only found significant (p < 0.05) in oil-coated samples. Higher Ca average contents  Table S1.
Compared to the initial fish, a remarkable decrease (p < 0.05) in Mg content was observed in aqueous-coating samples (Table 1); on the contrary, oil-coated ones did not show differences (p > 0.05) from initial fish. Among coating conditions, canned fish corresponding to aqueous media showed lower (p < 0.05) contents than their corresponding oil-coated samples. Additionally, brine-coated mackerel corresponding to Group II had lower (p < 0.05) Mg contents than samples corresponding to the water-coated condition.
Different tendencies were observed for the Ca content when considering fish with or without previous frozen storage ( Figure 2). Compared to the initial fish muscle, canned muscle without previous frozen storage showed lower average contents in samples coated in aqueous-coating systems and higher average values in oil-coated samples; differences from initial fish were found significant (p < 0.05) in the cases of brine-, ROO-, and EVOOcoated conditions. For canned fish muscle of Group II, higher average Ca contents were obtained in all kinds of canned samples when compared to initial fish; differences were only found significant (p < 0.05) in oil-coated samples. Higher Ca average contents were found in aqueous-coating fish muscle than in their corresponding samples that were coated in oil; differences were found significant (p < 0.05) for samples of Group II. A remarkable (p < 0.05) Ca content increase was detected in all kinds of canned samples as a result of the previous frozen period. fish; however, no differences (p > 0.05) were proved for canned ones corresponding to the remaining coating conditions tested. Water-coated samples revealed lower (p < 0.05) S contents than their corresponding oil-coated samples. Related to the effect of the previous frozen period, average S contents were found lower in canned fish muscle corresponding to Group II; differences were found significant (p < 0.05) in all cases except for the brine-canned condition.  Table S1.

Determination of Trace Elements
A marked increase (p < 0.05) in Co content was detected in all kinds of canned samples when compared to initial fish ( Table 2). In canned fish corresponding to Group I, higher Co contents (p < 0.05) were detected in fish coated in SFO and EVOO than in samples corresponding to both aqueous media. For canned fish corresponding to Group II, SFO-coated muscle showed higher (p < 0.05) contents than samples corresponding to brine-and ROO-coating conditions. For all kinds of canned fish, no effect (p > 0.05) of the previous frozen period could be determined for the Co presence.   Table S1.
Compared to the initial fish, a marked P content decrease (p < 0.05) was observed in canned fish coated in both aqueous media (Table 1). In contrast, no effect (p > 0.05) was proved when taking into account the oil-canned samples. Oil-coated muscle showed higher (p < 0.05) P levels than their corresponding aqueous-coating samples. A 6-month previous storage led to a marked decrease (p < 0.05) in P content in aqueous-coating samples but did not provoke differences (p > 0.05) when employing an oil-coated system.
Comparison to initial fish revealed different tendencies for the S content in canned fish depending on whether the fish muscle considered was from Group I or II (Table 1). In the case of Group I, no differences (p > 0.05) could be outlined by comparison of initial fish and fish muscle canned in both aqueous coating systems; on the contrary, oil-coated fish revealed a remarkable increase (p < 0.05). For fish samples of Group II, water-coated samples indicated a relevant decrease (p < 0.05) in S content when compared to initial fish; however, no differences (p > 0.05) were proved for canned ones corresponding to the remaining coating conditions tested. Water-coated samples revealed lower (p < 0.05) S contents than their corresponding oil-coated samples. Related to the effect of the previous frozen period, average S contents were found lower in canned fish muscle corresponding to Group II; differences were found significant (p < 0.05) in all cases except for the brinecanned condition.

Determination of Trace Elements
A marked increase (p < 0.05) in Co content was detected in all kinds of canned samples when compared to initial fish ( Table 2). In canned fish corresponding to Group I, higher Co contents (p < 0.05) were detected in fish coated in SFO and EVOO than in samples corresponding to both aqueous media. For canned fish corresponding to Group II, SFOcoated muscle showed higher (p < 0.05) contents than samples corresponding to brine-and ROO-coating conditions. For all kinds of canned fish, no effect (p > 0.05) of the previous frozen period could be determined for the Co presence.   Table S1.
Average Cu content showed a general increase in canned samples (Table 2). Differences from initial fish were found significant (p < 0.05) for oil-coated fish (Groups I and II) and for Group II samples coated in both aqueous media. Higher average Cu contents were detected in oil-coated samples than in samples corresponding to aqueous-coating conditions; differences were found significant (p < 0.05) in Group II samples. For all coating systems, higher average Cu contents were detected in Group II samples when compared to their counterparts from Group I; however, differences were only found significant (p < 0.05) for brine-canned samples.
Comparison with initial fish revealed an average increase in Fe content in all canned samples ( Figure 3); differences were found significant (p < 0.05) for oil-coated samples for Groups I and II. Comparison among coating systems showed that both olive oil systems led to the highest average Fe contents. For samples of Group I, differences between ROO-coated fish and aqueous-and SFO-coating systems were significant (p < 0.05). No significant differences (p > 0.05) were found as a result of the previous frozen storage; however, oil-coated samples corresponding to Group II led to higher average Fe values.
Comparison with initial fish revealed an average increase in Fe content in all canned samples ( Figure 3); differences were found significant (p < 0.05) for oil-coated samples for Groups I and II. Comparison among coating systems showed that both olive oil systems led to the highest average Fe contents. For samples of Group I, differences between ROO-coated fish and aqueous-and SFO-coating systems were significant (p < 0.05). No significant differences (p > 0.05) were found as a result of the previous frozen storage; however, oil-coated samples corresponding to Group II led to higher average Fe values.  Table S1.  Table S1.
A general average decrease in the Mn content was obtained by comparison of initial fish and any kind of canned sample ( Table 2). This decrease was found significant (p < 0.05) in all kinds of canned samples of Group I. For canned samples corresponding to this group, water-coated fish revealed lower (p < 0.05) Mn contents than all other canned samples; in the case of canned fish of Group II, ROO-coated fish showed a higher (p < 0.05) Mn content than aqueous-coating samples. An Mn content increase (p < 0.05) was observed in waterand ROO-coated samples as a result of the frozen storage period.
The Se content showed scarce differences between initial fish and canned samples (Table 2). Thus, only ROO-coated fish muscle (Groups I and II) and brine-coated fish (Group II) led to higher (p < 0.05) contents than the initial fish. For samples of Group I, ROO-coated fish muscle showed a higher (p < 0.05) Se content than their corresponding samples coated in aqueous conditions. Previous frozen storage time led to a Se content increase (p < 0.05) in the case of brine-coated samples, but no differences (p > 0.05) were detected in those corresponding to the remaining coating systems tested.

Discussion
Previous research provides extensive information on chemical changes related to protein and lipid fractions during seafood canning [16][17][18][19][20][21][22]. On the contrary, data concerning the changes in mineral content related to this technological process can be considered relatively scarce. The present study has shown remarkable changes in the content of most elements in canned fish as a result of the different processing conditions considered. Thus, differences between initial and canned fish corresponding to the different kinds of canned samples have shown different trends (increase, decrease, or no variation) according to the processing condition and the element considered. In order to discuss and justify such variations, a review of the most important events produced by each single processing step or condition applied ought to be carried out.
Concerning the effect of canning (i.e., sterilisation step), constituents from marine species have been shown to be highly sensitive [7,9]. In this sense, oxidation of protein and lipid fractions, heat breakdown, and leaching of constituents into the coating medium have been mentioned [13,28,29]. Remarkably, a detrimental effect on mineral content may be produced as a result of liquor losses from canned fish muscle. Therefore, reduction in mineral contents in fish muscle during the heating process may be related to the protein denaturation and release of elements with the loss of water as free salts, possibly associated with soluble free amino acids and hydrophobic vitamins [12,14,30,31]. This loss ought to be more remarkable when an aqueous medium is employed as coating medium than in the case of using an oil-coating system. On the other hand, denatured proteins are reported to become more reactive and be damaged easily by interacting with other constituents, especially if a heating step such as sterilisation is included. Furthermore, release of prooxidant elements such as non-heme Fe from heme-Fe complexes may have detrimental consequences for the quality of canned muscle [32,33]. As a result, canned fish may undergo a notable lipid and protein oxidation, leading to breakdown and production of low-molecular-weight molecules susceptible to being lost from the fish tissue [12,13]. Accordingly, a content decrease in constituents such as proteins and lipids in the canned fish muscle would lead to a relative increase in other constituents such as essential minerals.
According to such considerations, relevant differences between initial fish and canned samples were proved for macroelements and trace elements in the current study. In our study, we observed a decreased content of Na and Mg (water-coated samples), P (water-and brine-coated samples), S (water-coated samples of Group II), K (water-and brine-coated samples), and Ca (brine-coated samples of Group I). On the contrary, a content increase was detected for Na (brine-coated samples), S (oil-coated samples), K (olive-oil-coated samples of Group II), and Ca (ROO-and EVOO-coated samples of Group I and oil-coated samples of Group II). As a predominant trend, it can be concluded that a mineral content decrease was observed for aqueous-coating samples, while oil-coated ones showed higher contents than the initial fish.
Previous research accounts for the effect of canning on mineral content in fish muscle. Thus, Seet and Brown [15] proved some loss of minerals (Ca, Na, K, Mg, P, Cu, and Fe) from the canned muscle into the water-coating medium in canned tuna (Thunnus alalunga). Later on, a loss of protein and water and an increase in lipid content were detected by Castrillón et al. [16] after tuna (T. alalunga) canning (sterilisation at 115 • C for 55 or 90 min) by using soy bean oil as coating medium; additionally, a content decrease in some elements (Mg, K, and P) and no effect on others (Na, Ca, Cu, Fe, and Zn) were detected after steaming. Recently, the canning treatment of brine-canned chub mackerel (Scomber colias) led to increased contents of Na, Ca, Mn, Fe, Cu, Se, and S but to lower contents of K, Mg, Co, and P in canned fish muscle [34].
Concerning the frozen storage of seafood, different damage pathways have been described as being responsible for sensory and nutritional losses during such technological processing [17,20,21]. Among them, microstructural changes, protein denaturation, and lipid hydrolysis and oxidation development have proved to be important events for quality loss. As a result of freezing and frozen storage, the fish tissue is reported to become less elastic, more fibrous, harder, and lacking in water-holding capacity [19,35]. This last effect can be especially important for mineral content in frozen seafood and, therefore, in the corresponding canned product. Thus, liquor produced from fish tissue, especially during the needed thawing step, can lead to an important decrease in the mineral content. This element release ought to increase with protein damage and, accordingly, with time and temperature of frozen storage [13,23].
In our study, a strong effect of previous frozen storage on the content of the macroelements tested was observed. Thus, the inclusion of the storage period led to an increased effect in canned fish on K (oil-coated samples) and Ca (all coating conditions) content and to a decreased content of P (aqueous-coating samples) and S (water-and oil-coated samples). In the case of trace elements, a content increase in Cu and Se in brine-canned samples and of Mn (water-and ROO-coated samples) was detected as a result of the previous frozen storage period.
Previous research accounting for the effect of frozen storage on the element content in seafood can be considered scarce. Thus, Karl et al. [36] detected a notable reduction in I content in different kinds of fish after deep-freezing (−40 • C) and thawing. Pourashouri et al. [37] showed an increase in the non-heme Fe content as a result of the Fe release from heme-Fe complexes during the frozen storage (6 months at −18 • C) of Wels catfish (Silurus glanis); this change was explained according to the oxidative cleavage of the porphyrin ring [32,33]. Recently, a previous 6-month frozen storage (−18 • C) led to a general decrease in essential element (K, Mg, Ca, Mn, Fe, Se, P, and S) content in brine-canned mackerel (S. colias) [34]. In a subsequent study, Prego et al. [23] analysed the effect of the previous frozen storage time (-18 • C, up to 15 months) on the content of essential elements in the corresponding brine-canned mackerel (S. colias); as a result, an increased frozen storage time led to an increase in Ca and Mn contents but produced a decrease in K content.
Another effect taken into account in the current study is that of the coating medium. The coating medium acts as an extracting medium susceptible to provoking the loss of certain elements in canned muscle. In this sense, the more or less lipophilic or hydrophilic behaviour of molecules in which minerals are integrated would be mandatory in order to foresee important content decreases. It is worth pointing out that interaction of mineral elements with other fish constituents can vary in a wide range and show a great dependence on their chemical characteristics [38,39]. Thus, transition metals (Fe, Cu, etc.) and nonpositive elements (S, P, etc.) have been reported to be strongly bound to other tissue constituents and lead to a wide number of functional molecules. Contrary, alkali earth (Mg and Ca) and alkali (Na and K) elements are known to be present in cellular media as sulphates, organic salts (pyruvates, lactates, citrates), or chlorides. Therefore, a higher loss of alkali and alkali earth elements would be expected to occur when employing an aqueous-coating medium than in the case of using an oil-coated one.
In our study, we observed a marked effect of the coating medium on the content of the macroelements analysed. Thus, aqueous-coating samples showed lower Mg, P, K, and Ca contents than their corresponding oil-coated samples. Additionally, a lower Na and S content was observed in water-coated samples than in those coated in any of the three oils tested. In the case of trace elements, a lower Co content was observed in brine-coated samples than in SFO-coated ones; additionally, lower Cu contents in brine-coated samples than in their corresponding coated ones in SFO and EVOO were obtained. Water-coated samples showed lower Mn contents than fish muscle canned in ROO or EVOO media. Finally, aqueous coating led to lower Se (Group I) and Fe contents than ROO-coated samples. Because no differences in macroelement and trace element contents have been detected between ROO-and EVOO-coated samples, it is concluded that the presence of natural antioxidant compounds in EVOO (i.e., polyphenol compounds) [17,40] has not led to a different trend in element composition.
Considering the different effects that each condition or single processing step may have on the mineral composition of the current canned mackerel, two opposite effects can be signalled related to changes in the canned seafood. On one hand, each of the different processing steps or conditions applied (freezing, frozen storage, thawing, coating, and sterilisation), especially the last one, would lead to oxidation breakdown, denaturalisation, and partial loss of the main constituents (proteins and lipids, especially) [9,13,19]. As a result, a relative increase in other constituent content such as macroelements and trace elements would be expected to be produced. On the other hand, modifications of constituents, breakdown of binding of minerals to other tissue constituents, and liquor losses from the muscle would lead to a partial loss of minerals into the coating medium [16,30,31]. The significance of this effect would depend on the kind of binding to other tissue constituents and the more or less lipophilic or hydrophilic nature of molecules in which minerals are integrated [39,40]. Elements whose linkage to other constituents is easily lost during any of the processing steps would lead to a more relevant decrease content. On the contrary, those whose linkage to other tissue constituents is not modified during processing would not be likely to be lost and lead to a relative content increase according to the content decrease in other constituents.
Great attention is being accorded currently to the Na/K ratio value in human diet. Yang et al. [41] found a reduced risk of cardiovascular diseases if a Na/K ratio lower than 1.0 value was present. Remarkably, Na/K ratios in the current study were found below 0.7 value in all kinds of canned samples except for those that included brine as coating medium. In that case, higher ratio values than 2 were found, this value being explained on the basis of the great presence of NaCl in the coating medium.

Conclusions
The results show that previous frozen storage led to an increased (p < 0.05) content in canned fish of K (oil-coated samples) and Ca (all coating conditions) and to a decreased (p < 0.05) content of P (aqueous-coating samples) and S (water-and oil-coated samples). For trace elements, a content increase (p < 0.05) in Cu and Se (brine-canned samples) and Mn (water-and refined-olive-oil-coated samples) was detected in canned fish of Group II. Concerning the coating effect, aqueous-coating samples showed lower (p < 0.05) Mg, P, S, K, and Ca contents than their corresponding oil-coated samples. For trace elements, lower average contents were detected for Co, Cu, Mn, Se, and Fe in aqueous-coating samples when compared to those coated in oily media.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/foods12122289/s1, Table S1: Accuracy control of the analytical procedure for the determination of macroelements and trace elements in initial and canned samples corresponding to Groups I and II*. Certified reference material DORM-2 (National Research Council of Canada, NRCC) was employed for trace elements. In the case of macroelements, contents come from Engstrom et al. [26] and Millos et al. [27].  Data Availability Statement: All related data and methods are presented in this paper. Additional inquiries should be addressed to the corresponding author.