Sample preparation

Jumbo squid mantles were obtained from a local fish market in Hermosillo, Mexico. The extraction of low-molecular-mass compounds (LMMC; fraction <1 kDa) from mantle muscle of jumbo squid was carried out by homogenizing the skinned muscle in ultrapure water, obtained from a Milli-Q system (Merck Millipore Ltd., Burlington, MA, USA). Briefly, 100 g of sample were homogenized (8000 rpm, 90 s) in Milli-Q water (300 mL) with the help of a tissue homogenizer (Tekmar Tissumizer SDT 1810; IKA-Werke, Breisgau, Germany). Then, the sample was centrifuged (12 000×g, 20 min) at 4 °C in a refrigerated centrifuge (Avanti J-26S XPI; Beckman Coulter Inc., Palo Alto, CA, USA). The supernatant was ultrafiltered using a 1 kDa (nominal molecular mass limit) Amicon ultrafiltration stirred cell membrane (Merck Millipore Ltd). Finally, the permeate was transferred into a clear tube and freeze-dried using a lyophilization system (FreeZone 12; Labconco™, Kansas City, MO, USA). Thus obtained sample is considered LMMC fraction.

Fourier transform infrared spectroscopy

In order to elucidate the composition of the lyophilized sample (LMMC, <1 kDa fraction) a FTIR spectrometer (IFS 66/S; Bruker Optics Ltd., Coventry, UK) was used. Sample was placed on a single-reflection diamond attenuated total reflectance (ATR) unit (Specac Limited, Orpington, UK) and carefully pressed down to ensure a good contact. Scans were obtained in the 4000–400 cm⁻¹ range. Analysis of the spectra was performed using the Omnic™ v. 6.1A spectra software (6). Three replicates were performed.

Free amino acid composition

Free amino acid analysis followed the procedures described by Vázquez-Ortíz et al. (7) with modifications. Briefly, 95 mg of freeze-dried LMMC fraction were dissolved in 200 µL of 5% trichloroacetic acid and centrifuged (1485×g, 15 min, 4 °C; Eppendorf® 5417R centrifuge; Eppendorf, Hamburg, Germany). Then, the supernatant was filtered using a 0.22-µm filter (Merck Milipore Ltd., Cork, Ireland) and the obtained filtrate was mixed with 200 µL of internal standard (10 μg/mL, l-α-aminobutyric acid) and 700 µL of ultrapure water. Sample derivatization was carried out by mixing 10 µL of sample with 10 µL of o-phthalaldehyde (OPA) solution: 10 mg of OPA dissolved in 250 µL of methanol with 37.5 µL of 30% Brij™ 35 (Sigma Diagnostics, St. Louis, MO, USA), 25 µL of 2-mercaptoethanol and gauged to 10 mL with 0.5 M borate buffer, pH=10.4. Derivatized sample was injected into a high-performance liquid chromatographer (Agilent® 1100; Agilent Technologies, Palo Alto, CA, USA) with a ZORBAX 300Extend-C18 (150 mm×4.6 mm, 3.5 µm) column. Free amino acids were eluted from the column using an elution gradient consisting of acetate buffer (pH=7.2, with 1% tetrahydrofuran) and methanol at a flow rate of 1.2 mL/min. Amino acid detection and quantification were conducted with a fluorescence detector at 350 and 450 nm for emission and excitation, respectively. The identification of free amino acid was conducted using amino acid analytical standards (purchased from Sigma-Aldrich, Merck) as reference. The obtained chromatograms were analyzed with the ChemStation^TM Rev. A.10.01 software (8).

Sarcosine determination

Sarcosine determination in LMMC fraction was carried out using the methodology described by Bank et al. (9) with modifications. The sarcosine derivatization was made by mixing 200 µL of fraction with 200 µL of 9-fluorenylmethoxycarbonyl chloride (FMOC-Cl) (6 mg/mL acetone). Then 600 µL of pentane was used to remove excess FMOC-Cl and hydrolysis products. Subsequently, the upper phase was removed after 30 s. The extraction process finished with the addition of 400 µL 25% (V/V) acetonitrile in 0.25 M boric acid. Finally, 50 µL of extract were injected in the HPLC system (Agilent Technologies), equipped with a ZORBAX 300Extend-C18 (150 mm×4.6 mm, 3.5 µm) column and a fluorescence detector used at 254 and 630 nm for excitation and emission, respectively. The used mobile phase, under isocratic conditions, was as follows: solvent A: 20 mM citric acid, pH=2.85, 20 mM sodium acetate, 5 mM tetramethylammonium chloride and 0.01% sodium azide; and solvent B: 80% 20 mM sodium acetate, 5 mM tetramethylammonium chloride and 0.01% sodium azide (in 20% methanol). The obtained chromatograms were analyzed with the ChemStation^TM Rev. A.10.01 software (8).

Total carbohydrate analysis

Total carbohydrates were determined by the phenol-sulfuric acid method (10). Briefly, 65 mg of LMMC fraction were dissolved in 1 mL of ultrapure water. Then, 200 µL of this solution were taken and mixed with 200 µL of phenol reagent (5% m/V) followed by a rapid addition of 1 mL of concentrated sulfuric acid. Colour development and cooling of the sample lasted 40 min. Finally, absorbance was measured at 490 and 480 nm for hexose and pentose quantification, respectively, using an UV-visible spectrophotometrer (Cary® 50; Varian, Inc., Santa Clara, CA, USA). d-glucose and d-xylose were used as standards.

Free monosaccharide composition

Free monosaccharides present in the LMMC fraction were analyzed by gas chromatography (GC) using the methodology described by Rouau and Surget (11) with a few modifications. The freeze-dried sample was re-suspended in Milli-Q water at a concentration of 150 mg/mL. Then, 100 µL of inositol (5 mg/mL H₂O) as internal standard, 400 µL of 2 M H₂SO₄, 500 µL of 25% ammonia and 1 mL of sodium borohydride (20 mg in 1 mL of dimethyl sulfoxide, DMSO) were added to 1 mL of sample solution and heated (35 °C, 90 min). After cooling of the sample, 100 µL of glacial acetic acid, 2 mL of acetic anhydride and 150 µL of methylimidazole were added and left to rest at room temperature (25 °C) for 20 min. The free monosaccharide extraction was performed by mixing 6 mL of Milli-Q water and 2 mL of chloroform, left to rest for 30 s and then the water phase was removed carefully. This procedure was conducted twice by adding 8 mL of ultrapure water for each washing. Finally, the non-aqueous phase was transferred into injection vials and maintained at –80 °C for future analysis.

For monosaccharide determination, 5 μL of extract were injected into a gas chromatograph system (HP 6890 GC series; Agilent, Santa Clara, CA, USA) equipped with a 225 DB (50% cyanopropylphenyl-dimethylpolysiloxane, 30 m×0.32 mm i.d., 0.15 μm) column. The chromatographic conditions were as follows: injection temperature 220 °C, detector temperature 260 °C and furnace temperature 205 °C, using nitrogen as the carrier gas at 1 mL/min. The obtained chromatograms were analyzed with the GC ChemStation^TM Rev. A. 09.01 software (12).

Ammonium chloride determination

The concentration of ammonium chloride was determined by flux injection analysis (FIA)/gas diffusion technique adapted from Clinch et al. (13). This method determines total ammonia nitrogen content (sum of NH₃ and NH₄⁺). However, the ammonium chloride in LMMC fraction solution was in the form of NH₄⁺. The LMMC fraction solution (500 µL) was injected directly into the mobile phase (0.1 M NaOH) using the low pressure injection valve (Rheodyne 5020; Anachem, Luton, UK). The mobile phase was carried out by a Gilson MINIPLUS 2 peristaltic pump (Gilson Incorporated, Middleton, WI, USA) to a gas diffusion cell (240 mm×1.5 mm×0.2 mm). Finally, the produced gas (NH₃), diffused through the membrane, was detected at 625 nm (PYE Unicam SP6-550 UV/Vis spectrophotometer; Phillips, Eindhoven, The Netherlands) by using a bromophenol blue (0.4 g/L in 0.1 M NaOH) indicator solution. A fresh standard curve using ammonium chloride in the range of 20–400 µmol/L was prepared daily.

Antifreeze activity

The antifreeze activity of LMMC fraction present in the jumbo squid muscle was evaluated by differential scanning calorimetry (DSC). The antifreeze activity is determined by measuring the temperature displacement of ice recrystallization conferred by solutes present in the solution in comparison with pure water, behaviour known as thermal hysteresis. This analysis was conducted employing a calorimeter (DSC7; Perkin Elmer de México, S.A., Monterrey, Mexico) following the method described by Lu et al. (14). Briefly, aliquots (approx. 40 mg) of lyophilized LMMC fraction solution (at 32.2 mg/g H₂O, solute mass fraction of LMMC present in the muscle) were transferred to stainless steel capsules and hermetically sealed. Then, capsules were gradually frozen from 20 to –30 °C at 1 °C/min. After reaching –30 °C, the temperature was maintained for 15 min and finally capsules were heated to 20 °C at 1 °C/min. Parameters such as crystallization point (onset temperature of crystallization transition), freezing point (onset of melting transition) and enthalpy of melting transition (ΔH_m) were determined from thermograms. The thermograms were analysed with the Pyris™ v. 4.02 sofware (15). The study was designed in order to know the interval of temperatures (t_h=holding temperatures) to ensure the presence of frozen and thawed water (partial thawing) during the unfreezing transition.

Once this thermal characterization was obtained, the sample was gradually frozen to –30 °C at 1 °C/min, maintained for 15 min, and then heated to a partially unfreezing temperature (t_h) at 1 °C/min, maintained for 3 min; and finally, the sample was again frozen to –30 °C. The enthalpy of recrystallization (ΔH_f) and recrystallization point (t₀=onset of recrystallization transition) were calculated with the help of the above mentioned software. This process was repeated at different t_h. The amount of ice fraction (x/%) (16) and thermal hysteresis were calculated using the following equations, respectively:
and

Thermal hysteresis = t_h – t₀ /2/

Protein cryostability

The protein cryostability conferred by osmolytes present in the LMMC fraction was expressed as the stability of myofibrillar proteins (mainly myosin and actin) to high temperatures after being subjected to a freeze-thawing process. For this objective, a DSC study was designed using a calorimeter (DSC7; Perkin Elmer de México, S.A.). First, a myofibrillar protein concentrate (MPC) from jumbo squid muscle was obtained by homogenizing 20 g of muscle in ultrapure water (at 1:3 ratio, mantle/water) at 8000 rpm for 90 s with the help of a tissue homogenizer (Tekmar Tissumizer SDT 1810; IKA-Werke). Then, homogenate was centrifuged (12 000×g, 20 min) at 4 °C, using a refrigerated centrifuge (Avanti J-26S XPI; Beckman Coulter Inc., Palo Alto, CA, USA). This process was repeated twice on the precipitate to improve the extraction of water-soluble compounds (mainly sarcoplasmic proteins and LMMC). The final precipitate was lyophilized using a freeze-dryer system (FreeZone 12; Labconco™). Then, lyophilized protein concentrate was re-suspended in a solution containing the LMMC fraction (at 32.2 mg/g H₂O, solute mass fraction of LMMC present in the muscle) to a final 86% (m/V) water content. MPC and ultrapure water (instead of LMMC fraction) were prepared and used as control. Aliquots (approx. 40 mg) were transferred to stainless steel capsules and hermetically sealed. Capsules were analyzed by calorimetry as follows: first, the capsules were gradually frozen to –30 °C at 1 °C/min, held at this temperature for 15 min, and then heated up to 90 °C at 5 °C/min. The temperatures of endothermic transition such as onset temperature (t_onset) and maximum temperature of denaturation (t_max) as well as their enthalpy of protein denaturation were observed.

Statistical analysis

The compositional characterization is shown as a mean value±standard deviation (S.D.). Protein cryostability mean values were analysed with a one-way analysis of variance (ANOVA) with 5% significance level. Tukey-Kramer test was used for multiple comparisons of mean values using the NCSS 97 software (17). All experiments were conducted in triplicate.

Sample compositional characterization

The LMMC fraction (<1 kDa fraction) from jumbo squid muscle characterized by FTIR spectroscopy had 32.2 mg of compounds per g of muscle. Lyophilized LMMC fraction absorbed the light within the mid infrared spectrum (4000–500 cm^–1) corresponding to X-H stretching (4000–2500 cm^–1), double bonds (2000–1500 cm^–1) and fingerprint (1500–600 cm^–1) regions (Fig. 1) (18). Thus, the 3500–2500 cm^–1 spectrum indicated the presence of hydrogen bonds mainly from carbohydrates (19). However, under this same spectrum, three characteristic peaks at 3359, 3024 and 2960 cm^–1 were observed, which correspond to N-H (from amine groups), C-H and -CH₃ stretching bonds, respectively.

Fig. 1

FTIR spectra of low-molecular-mass (LMMC; <1 kDa) fraction from jumbo squid muscle

Absorbance values associated with double bonds (2000–1500 cm^–1), in particular carboxylic groups (C=O) and amine groups (N-H) of peptide bonds, were detected at the 1620 and 1522 cm^–1 regions, corresponding to amine I and amine II groups, respectively (20). Besides, amine I group showed a more intense absorbance, indicating the presence of β-sheet peptides of low molecular mass (21) in the sample. Furthermore, there were peaks in the 1500–600 cm^–1 region related to molecular deformation (19). In this region, a peak at 1450 cm^–1 related to symmetric deformation of ammonium ion (NH₄⁺) was identified (21). This ion can be a product of ammonium chloride (NH₄Cl) ionization, compound resulting from the catabolism of proteins and amino acids. It has been identified as the responsible for the organism’s buoyancy and the sour-bitter taste of jumbo squid muscle (22, 23).

Besides, inside the 1200–1000 cm^–1 region, corresponding to the deformation of monosaccharides, peaks at 1042 and 1035 cm^–1 indicated the stretching of the C-OH and CH₂OH bonds, respectively, both structural components of monosaccharides (24). A similar pattern was reported by Dauphin (25), who found absorbances at 1150–1000 cm^–1 in a soluble fraction of Nautilus macromphalus L., associated with carbohydrates. Also, absorption peaks were found within the 1500–1300 cm^–1 region, which is correlated with the presence of organic acids (24). It is important to notice that there is no research on the elucidation of the chemical composition of water-soluble extracts from jumbo squid muscle by infrared spectrophotometry. This FTIR spectroscopy analysis indicated which type of compounds to search for in the LMMC fraction in the present study.

Total carbohydrate content

Carbohydrates found in the LMMC fraction totalled 67.01 µg/mg of sample (Table 1). These were divided in hexoses (35.2±5.2) µg/mg) and pentoses (31.9±4.6) µg/mg). These results are similar to the values reported by Lee (26), who mentioned that carbohydrate mass fraction in cephalopods is less than 1%. Given the presence of carbohydrates in jumbo squid muscle, it can be inferred that these compounds may influence the stability of the protein under freezing condition, as it has been observed that some carbohydrates, such as glucose and trehalose, can inhibit water crystallization (27). Besides, Baruch et al. (28) found that carbohydrates have potential use as cryoprotectants in freeze-thawing processes, since they observed carbohydrate antifreeze activity related to its structure and spatial disposition of the molecule hydroxyls groups in order to interact with the ice crystals.

Free monosaccharides

Glucose, arabinose and fucose were the main free monosaccharides in the LMMC fraction from jumbo squid mantle (Table 1). The profile of free monosaccharides from jumbo squid muscle, as well as for others cephalopods species, has never been reported. Only Kani et al. (29) have reported the mass fraction of glucose in the muscle of four squid species (Sepioteuthis lessoniana, Loligo bleekeri, Loligo edulis and Todarodes pacificus), ranging from 9–60 mg per 100 g of muscle against 100.2 mg per 100 g of jumbo squid muscle (data not shown). This variation is probably due to different metabolic characteristics of each species.

Among the identified free monosaccharides, glucose was found at a mass fraction of (32.7±2.1) µg/mg, representing 64% of total free monosaccharides in the LMMC fraction (Table 1). Total free monosaccharides (51.1 µg/mg) represented the major constituents (76%) of total carbohydrates (67.1 µg/mg) (Table 1) present in the fraction; the rest of carbohydrates were possibly other oligosaccharides or glycated biomolecules. Major free monosaccharides present in the fraction, glucose and arabinose, have been reported as cryoprotectants, stabilizing protein against its denaturation (28). Besides, glucose has been reported to confer protein stabilization and ice recrystallization inhibition during temperature fluctuations in frozen conditions (28). However, the identification of other carbohydrates such as di-, tri- and oligosaccharides, which could be present in the fraction, is needed in order to make a stronger inference about the role that carbohydrates play as protein cryostabilizers since it has been reported that di- and trisaccharides act more efficiently as osmolytes (30).

Free amino acid and sarcosine content

Total mass fraction of free amino acids (FAA) in LMMC was (in mg/g sample): 18.84, with arginine (8.82±0.15), sarcosine (6.95±0.76) and taurine (1.05±0.02) being the dominant FAA in the sample (Table 2). Sarcosine and taurine are characteristic amino acids of jumbo squid muscle (31), although these are not proper amino acids since both derive from them; sarcosine is an intermediary and byproduct in the synthesis and degradation of glycine, while taurine derives from cysteine (32). Urich (22) reported similar taurine mass fraction (10 µmol/g) in other cephalopods. Based on the study by Yancey (33), it can be suggested that the predominance of these amino acids can also contribute to the cryostability of jumbo squid proteins.

On the other hand, the protein stabilization mechanism varies among amino acids. Shukla and Trout (34) proposed that arginine interacts (cation-π interaction) with tryptophan residues of proteins, contributing to the protein stability; however, they also proposed that arginine accumulates in the hydrophobic regions of proteins, thus inhibiting their aggregation, effect that can be replicated in jumbo squid muscle under freezing conditions.

Sarcosine has been reported as an important osmolyte stabilizer of proteins that can counteract the destabilizing effect of urea by increasing its solvation and also by its direct interaction with urea, which in turn becomes less available to interact with proteins (35). Thus, sarcosine can show a similar protecting effect on jumbo squid proteins.

Fish protein tends to denature/aggregate under freezing conditions reducing their technological functionality; however, jumbo squid protein functionality remains stable even during freeze-thaw cycles (5), most probably due to the presence of these FAA, among other compounds.

Ammonium chloride content

Jumbo squid (Dosidicus gigas) species has an inherently high level of ammonium chloride in its muscle (23). This is an important compound that has been related to the organism´s buoyancy (22) and, thus, survival. In the present study, the ammonium chloride mass fraction of the LMMC from jumbo squid muscle was (220.4±1.6) µg/mg. The presence of ammonium chloride in the LMMC can also promote jumbo squid protein muscle stability since it has been reported that some cations (including NH₄⁺) can confer stability to the native structure of proteins (36).

Thermal characterization

The thermal characterization of jumbo squid muscle and its LMMC fraction (32.2 mg/g) during a freeze-thawing cycle is shown in Table 3. It can be observed that both samples displaced the freezing (FP) and crystallization (CP) points towards lower temperatures than of pure water (with FP=0 °C and CP=–15 °C). LMMC fraction, at the mass fraction of 32.2 mg/g, displaced the FP of water down to –1.2 °C. The lowest displacement (down to –3.5 °C) was observed in the muscle, which was to be expected due to the presence, besides the LMMC fraction, of other higher-molecular-mass compounds (i.e. peptides, proteins, etc.). Interestingly, the FP shown by the jumbo squid muscle was lower than the ones reported for other commercial fish species from the North Atlantic, which ranged from –1.1 to –2.0 °C (37). It is pertinent to mention that the thermal characterization (in freezing conditions) of jumbo squid muscle by DSC has not been reported elsewhere; however, Rahman and Driscoll (38) reported FP of different cephalopod species (Sepia officinalis, Octopus cyanea and Spirula spirula), ranging between –0.5 and –1.7 °C.

On the other hand, it could be observed (Table 3) that the LMMC fraction apparently did not displace the CP of water (–15 °C) towards lower temperatures, indicating that the LMMC at mass fraction of 32.2 mg/g did not confer more complexity to the aqueous medium (39). However, it is observed that higher extraction of heat is needed in that fraction (–119.2 J/g vs –110.2 J/g from that of muscle) for water molecules to reach a crystal structure (40). This behaviour must be mainly due to the high quantity of water available in the LMMC fraction for this process.

LMMC fraction antifreeze activity

The antifreeze activity analysis of the LMMC fraction is shown in Table 4. Based on a study by Mao et al. (41) on an insect antifreeze protein, our system can be considered to have an antifreeze capacity, showing that a delay in the onset temperature of refreezing increased from 0.39 to 0.66 °C, thus indicating that the LMMC exhibited a thermal hysteresis effect. Our result showed that the compounds present in the LMMC fraction inhibit recrystallization of water at 0.66 °C, contributing to the stabilization of muscle structure during temperature fluctuations under freezing conditions. This result is comparable with the report by Zhang et al. (42) about an antifreeze protein from Avena sativa L., whose antifreeze activity initiated at a thermal hysteresis of 0.75 °C.

The thermal behaviour of these LMMC present in the <1 kDa fraction should be due to their effect on the colligative water properties; however, Yeh and Feeney (43) mentioned that if a system shows thermal hysteresis during a recrystallization process, it is not due to colligative mechanisms but to compounds interacting with the ice crystal surface, thus avoiding the unfrozen water to reincorporate into their surface (evading recrystallization).

It is true that the antifreeze activity, expressed as thermal hysteresis, cannot only be ascribed to monosaccharides and other low-molecular-mass compounds in the fraction because there are reports of other oligosaccharides (i.e. lactose, trehalose and stachyose, among others) (28) and low-molecular-mass proteins/peptides (not quantified in the present study) conferring antifreeze activity; however, results shown in the present study could elucidate this behaviour. Besides, further characterization of the LMMC fraction is necessary to clarify which other low-molecular-mass compounds, besides the ones shown in the present study, confer cryoprotection to jumbo squid muscle.

The antifreeze activity conferred by the LMMC fraction is important for the jumbo squid processing and distribution industry, since its compounds can avoid the recrystallization/crystal growing of ice during temperature fluctuations at frozen storage, thus evading the rupture of cell membranes and muscular structure changes responsible for the syneresis during the thawing process and therefore, the loss of nutrients, mass and technological properties.

Myofibrillar protein cryostability

The jumbo squid myofibrillar protein cryostability (mainly myosin and actin proteins) was related to the thermal changes in their denaturation transitions (Table 5). The obtained thermal behaviour was similar to the one reported by García--Sánchez et al. (5). The thermograms of samples showed two main endothermic transition peaks, one around 40–50 °C and the other between 70 and 80 °C, corresponding to myosin and actin proteins, respectively (data not shown).

Table 5 shows that control treatment (myofibrillar concentrate without the LMMC fraction) reduced t_onset and t_max significantly (p<0.05) when comparing them with the muscle (unwashed) and the MPC with the LMMC fraction. On the other hand, these treatments (unwashed muscle and MPC with the LMMC fraction) showed similar (p≥0.05) t_onset and t_max, indicating the effect of the LMMC (among other compounds in that fraction and in the muscle) on the myosin stability of jumbo squid muscle during the frozen storage.

With respect to the enthalpy of denaturation (∆H) of myosin protein, all samples needed similar amount of heat (p≥0.05) to denature this protein. However, MPC without the LMMC fraction showed the lowest value (Table 5), indicating the relationship of LMMC fraction and the freezing stability of protein.

On the other hand, jumbo squid actin in MPC without the LMMC fraction had the lowest (p<0.05) values of t_onset and t_max (69.6 and 72.8 °C, respectively) of all samples (Table 5), showing the effect of that fraction on the protein stability of jumbo squid muscle during frozen storage. Actin showed a similar ∆H tendency as myosin, decreasing its value (p≥0.05) as the treatments with/without the LMMC fraction were applied. From these last results, it is evident that the MPC system, in the absence of the LMMC fraction, had lower stability (at frozen storage) than the systems with that fraction.

These results indicate that myofibrillar proteins (at frozen storage) are stabilized by osmolytes present in the LMMC fraction, such as arginine and sarcosine, both of which have been reported to show a high protein-stabilizing effect (44). However, more research is necessary to completely elucidate the stabilizing effect during freezing of these compounds on myofibrillar proteins and conduct a thorough compositional characterization of the LMMC fraction since the effect of other compounds (i.e. trimethylamine oxide and betaine) on protein stability has also been reported (45).

To summarise, results of the present study confirm the relevance of not washing the jumbo squid muscle, for example during the surimi production, since it only takes away compounds responsible for the muscle´s cryostability. Besides, after extensively washing the muscle, cryoprotectans (mainly sucrose and sorbitol) need to be added. Therefore, applying washing and adding cryoprotectans only increases the cost of surimi production. Instead, it is recommended to freeze the jumbo squid mantle muscle as such for further direct use as a raw material for fish analogues. However, if removal of the common sour-acidic squid flavour is required, washing the defrosted mantles is an alternative.