A New, Simple Method for the Production of Meat-Curing Pigment Under Optimised Conditions Using Response Surface Methodology

Meat Science 92 (2012) 538–547

Contents lists available at SciVerse ScienceDirect

Meat Science

journal homepage: www.elsevier.com/locate/meatsci

A new, simple method for the production of meat-curing pigment under optimised conditions using response surface methodology

Naﬁseh Soltanizadeh ⁎, Mahdi Kadivar

Department of Food Science and Technology, College of Agriculture, Isfahan University of Technology, Isfahan 84156, Iran article info abstract

Article history: The production of cured meat pigment using nitrite and ascorbate in acidic conditions was evaluated. HCl, Received 22 August 2011 ascorbate and nitrite concentrations were optimised at three levels using the response surface method Received in revised form 15 March 2012 (RSM). The effects of process variables on the nitrosoheme yield, the wavelength of maximum absorbance Accepted 22 May 2012 (λmax), and L*, a* and b* values were evaluated. The response surface equations indicate that variables exerted a significant effect on all dependent factors. The optimum combinations for the reaction were Keywords: HCl=−0.8, ascorbate=0.46 and nitrite=1.00 as coded values for conversion of 1 mM hemin to Nitrosoheme Yield nitrosoheme, by which a pigment yield of 100%, which was similar to the predicted value of 99.5%, was fi λ Absorbance obtained. Likewise, the other parameters were not signi cantly different from predicted values as the max, Optimisation L*, a* and b* values were 558 nm, 47.03, 45.17 and 17.20, respectively. The structure of the pigment was iden- FTIR tified using FTIR and ESI/MS. ESI/MS © 2012 Elsevier Ltd. All rights reserved.

1. Introduction 2007). The N-Nitrosamines in meat products, are compounds that could cause carcinogenicity. These compounds (e.g., N-nitrosopyrrolidine and Current meat-curing practice, which is founded upon the ancient N-nitrosodimethylamine) are formed, albeit in the parts-per-billion art of preserving meat with salt, employs the addition of nitrite range, by the reaction of nitrite with the amines or amino acids that are (and in certain products, nitrate) along with salt, sugar, reducing present in foods. In addition, the existence of residual nitrite in cured agents, and phosphates to meat (Rubin, Diosady, O'Boyle, Kassam, & meat increases the body's total nitrite load, which in turn may lead to Shahidi, 1992). Nitrite has beneficial effects on meat products and po- an increased likelihood of nitrosamine formation within the human di- tentially detrimental effects on human health. The role of nitrite in gestive tract (Rubin et al., 1992). However all health implications being cured meat is four-fold: i) it provides the characteristic pink-red associated with nitrite/nitrate consumption are tenuous and suggestive cured-meat colour to the lean tissue; ii) it inhibits the growth of a at present and no known case of human cancer has been shown to result number of bacteria that cause food poisoning or spoilage; iii) it con- from exposure to N-nitroso compound and all datas are according to indi- tributes to the distinctive flavour of cured meats; and iv) it retards rect observations. It should be mentioned that inspite of concerns about oxidative rancidity in processed meat products, principally through consumption of cured meat products, recent studies demonstrate that ni- a process of metal chelation. tritehavebenefitialeffectsonhealthanduponitsingestionandmixture Despite these beneficial effects on cured-meatproducts,therewere with gastric acid, is a potent bacteriostatic and/or bactericidal agent for deep concerns about the use of nitrite and nitrate as food additives be- gastrointestinal, oral, and skin pathogenic bacteria (Archer, 2002). The cause both are potentially toxic for humans. The lethal oral doses for potential role in hypoxic vasodilation and protective action against ische- humans have been established in the range of 80–800 mg of nitrate and mia are other physiological and pharmacological properties of nitrite that 33–250 mg of nitrite/kg body weight (Honikel, 2008). Over time, nitrite has been recently considered (Butler & Feelisch, 2008; Parthasarathy & has been suspected of playing a role in the development of cancer, methe- Bryan, 2012) but due to the potential health hazard associated with the moglobinemia in infants, and even reproductive toxicities such as birth use of sodium nitrite, extensive studies have been conducted to find defects (Archer, 2002). Recently, epidemiological relationships were methods to reduce nitrite in cured meat products. The chance of finding shown between cancer incidence and intake of processed meats a single compound which duplicates all functions of nitrite is very slight. (Demeyer, Honikel, & De Smet, 2008; Faramawi, Johnson, Fry, Sall, & Yi, Therefore, the development of a multifunctional system, including the synthetic cooked cured meat pigment have been considered (Shahidi, Rubin, Diosady, & Wood, 1985). Cured meat pigment is ordinarily developed by the reaction of nitrite with the natural meat pigment myoglobin ⁎ Corresponding author. Tel.: +98 311 3913382; fax: +98 311 3912254. to form dinitrosyl ferrochrome (DNFH). The pigment, which gives meat E-mail address: [email protected] (N. Soltanizadeh). its characteristic cured-meat colour, is formed from the meat pigment

0309-1740/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.meatsci.2012.05.024 N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547 539 myoglobin, which consists of an iron porphyrin complex, the heme group, Table 1 attached to the protein globin. In the presence of nitrite, the bright red Coded settings for the process parameters for nitrosoheme synthesis, according to a central composite design. nitrosomyoglobin is formed, in which the H2Ointheaxialpositionon the heme iron is replaced by nitric oxide (NO). The NO is formed from ni- Independnt variables Symbols Range and levels trite by the natural reducing activity of the muscle tissue, which is accel- −1.68 −1 0 1 1.68 erated by the addition of reductants such as ascorbic acid. In heat- HCl concentration (%) X 0.32 1 2 3 3.68 processed cured meat, the globin has been split off to a heat-stable pink 1 Ascorbic acid concentration (mM) X2 65.91 100 150 200 234.08 pigment, nitrosyl hemochromogen (Pegg & Shahidi, 2000). The synthetic Nitrite concentration (mM) X3 15.91 50 100 150 184.08 production of this pigment has been the subject of some researches (Shahidi & Pegg, 1991, 1995a, 1995b; Shahidi & Pegg, 1992; Shahidi et al., 1985). In a series of studies, curing colour was prepared from red blood cells and nitric oxide (Shahidi & Pegg, 1991, 1995a, 1995b; Shahidi & Pegg, 1992; Shahidi et al., 1985). The pigment 2.4. Yield could be prepared in a direct, one-step process or by an indirect method through a hemin intermediate. Hemin was dissolved in a di- The yield of nitrosoheme pigment was determined according to lute sodium carbonate solution and added to a mixture containing Hornsey (1956). Brieﬂy, the absorption of preformed cooked cured sodium tripolyphosphate, sodium ascorbate, and sodium acetate. pigment in 80% acetone solution was read in 540 nm. With multiply- Nitric oxide was then introduced into the solution to produce the ing the absorbance in 652 (Molecular weight of hemin) and dividing mM nitrosoheme, which then was precipitated out by lowering the pH to 11.3 (E 540), the concentration of nitrosoheme was obtained as (Pegg, Shahidi, & Fox, 1997). ppm. Yield was calculated by division of nitrosoheme concentration A new, rapid and easy procedure for the production of cooked to initial concentration of hemin. cured-meat colour was developed in this study. The inﬂuence of nitrite, ascorbic acid levels, and HCl concentration on pigment formation was evaluated in the model system of hemin using response 2.5. Colour of preformed pigments surface methodology, and after comparison of results with those of others under actual curing conditions and model systems, the The lightness/darkness (L* value), red/green (+/−a* value), and optimum conditions for in vitro production of nitrosoheme were yellow/blue (+/−b* value) of the nitrosoheme pigments were deter- established. mined according to Yam and Papadakis (2004). In this procedure, 2 ml of diluted pigment was poured into a glass container with an in- ternal diameter of 2 mm. The colour of the samples was measured 2. Material and methods using an apparatus constructed in the Department of Food Science, Isfahan University of Technology, Iran. It consists of a chamber with 2.1. Material a trapezoidal cross section that is equipped with two D65 (daylight) lamps for illumination of the samples. An 8.1-mega-pixel camera All chemicals and solvents used in this study were analytical grade (Samsung L830, South Korea) was used to record the images. Samples commercial products. Sodium nitrite, chloridric acid (37%) and ace- were placed in the centre of chamber. The images were captured after tone (pro analysis grade, 99.8%) were purchased from the Merck adjustment of the lens and focusing of the camera. Colour measure- Chemical Company, Germany. L-ascorbic acid sodium salt was ments for each sample were made at 5 different locations using obtained from Alfa Aesar Chemical Company, Germany. Hemin adobe Photoshop CS version 8.0 to determine colour coordinates, (98% pure, HPLC grade) was purchased from the Fluka Company, i.e., the L*, a* and b* values. For standardization of colour, the images Switzerland. were also recorded from the Ral colour standard.

2.2. Preparation of nitrosoheme Table 2 The nitrosoheme pigment was prepared from hemin and nitrite. Experimental design used in the RSM studies and the responses. Bovine hemin (6.52 mg) was dissolved in 1.8 ml of a 0.1 N NaOH so- Exp. Coded level of variables Yield λmax L* a* b* lution. This solution was diluted with 8 ml of acetone, and then no. (%) (nm) value value value 0.2 ml of concentrated hydrochloric acid was added, giving a solu- X1 X2 X3 tion of acid hematin in 80% acetone. The level of HCl changed 1 0 0 1.68 97.01 537 45.74 42.35 18.80 according to Tables 1 and 2, and the amount of NaOH determined 2 −11−1 100.00 541 50.74 43.35 21.60 3 1.68 0 0 89.40 537 49.54 34.86 23.40 as the ﬁnal volume was 10 ml. The nitrite and ascorbic acid were 4 0 0 0 95.20 522 48.54 41.17 19.20 then added in proper concentrations (Tables 1 and 2). The container 51−1 1 80.74 511 56.38 18.83 23.80 was gently shaken until nitric oxide slowly bubbled into the mixture 6 0 0 0 95.13 522 48.54 39.47 19.80 and was emitted into the air from the top of container. The container 7 −1.68 0 0 85.67 510 57.15 21.74 23.60 was then capped and shaken vigorously for 30 s. The preformed, 8 0 0 0 94.51 522 49.14 38.50 20.00 9 −1 1 1 100.00 540 46.74 47.24 18.21 cooked cured-meat pigment was stored in the dark until further 10 0 0 0 94.97 522 47.74 40.20 19.20 application. 11 1 1 1 85.02 511 48.54 29.51 19.80 12 −1 −1 1 98.25 538 46.34 43.35 21.00 13 1 −1 −1 76.20 508 59.95 14.22 24.40 14 1.68 0 0 74.09 510 56.15 13.00 26.00 2.3. Absorption spectra of nitrosoheme 15 0 0 0 96.59 522 47.54 40.93 18.80 16 −1 −1 −1 95.25 512 50.34 35.34 22.20 The pigments from hemin-nitrite synthesis were diluted 12.5 17 0 1.68 0 93.58 537 49.14 41.90 19.00 times with a 4:1 (v/v) acetone/water solution, and their absorption 18 0 0 0 94.63 522 47.54 40.68 19.40 − spectra were recorded using a Camspect spectrophotometer (M 350, 19 1 1 1 79.31 511 56.75 21.14 23.80 20 0 −1.68 0 83.40 510 53.54 19.07 24.40 UK). All absorption spectra in the visible range were recorded. 540 N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547

2.6. Experimental design (4) a* value model:

¼ : − : þ : Minitab software version 14.0 was employed for experimental de- y4 40 045 8 958X1 4 970X2 þ : − : 2− : 2− : 2 sign, data analysis and model building. The Central Composite design 4 360X3 5 000X1 2 638X2 2 130X3 with 3 variables was used to determine the response pattern and then to establish a model. The 3 variables used in the study were HCl concentration (X1), ascorbic acid concentration (X2), and nitrite concen- (5) b* value model: tration (X3), with 3 levels for each variable, and the dependent ¼ : þ : − : − : þ : 2 variables were yield (y1), λmax (y2), lightness (y3), redness (y4), and y5 19 433 0 964X1 1 249X2 1 263X3 1 688X1 þ : 2 þ : 2− : yellowness (y5) of the produced pigments. The symbols and levels 0 628X2 0 452X3 0 699X2X3 are shown in Table 1. The actual set of experiments performed (experimental runs 1–20) and the responses are shown in Table 2. Six replicates at the centre of the design were used to allow for estima- The experimental and predicted values from these models are tion of a pure error of the sum of the squares. Experiments were presented in Tables 2 and 3, respectively. The predicted response randomised to minimize the effects of extraneous factors. A full qua- values are slightly different from the experimental data. The analysis dratic equation, or the diminished form of this equation (shown of variance and error for the response surface model are given in below) was used: Table 4. The p-values for the 5 models are all b0.001, which indicates that all models are significant. Xk Xk ¼ β þ β þ β 2 þ ∑∑β ; Y 0 jXj jjXj ijXiXj 3.2. Effects of variables on yield b j¼1 j¼1 i j According to the results, the conversion of hemin to nitrosoheme where Y is the estimated response and β0, βj, βjj, and βij are the re- pigment ranged from 76 to 100%, depending on the treatment. gression coefficients for the intercept, linearity, square, and interac- Shahidi, Rubin, Diosady, and Chew (1984) used nitrite and ascorbic tion terms, respectively, and Xi and Xj are the coded independent acid to synthesize cooked cured-meat pigments in an aqueous solution variables. of hemin. They found that the purity of this pigment using spectropho- tometric analysis in 80% acetone solution was 65–72%, but it did not im- 2.7. FTIR analysis part the characteristic cured-meat colour when applied to meat. In another study, Shahidi et al. (1985) prepared cooked cured-meat pig- The FTIR from 4000 to 600 cm− 1 were recorded on a Bruker TEN- ment in pure form using hemin and nitric oxide, the latter acting as SOR27 IR spectrometer (Bruker Instruments, Germany). the nitrosating agent. The yield increased to 96% in this procedure, but it was still less than the yield of nitrosoheme production in our experi- 2.8. Mass spectroscopy ment. In fact, by replacing the water with acetone, the solubility of nitrite and ascorbic acid was reduced. Therefore, nitric oxide, which is a The mass spectral data were obtained in the positive ion mode on colourless and poisonous gas and over-exposure to which may cause an LCQ Advantage (Thermo Finnigan Co., USA), which was equipped methemoglobinemia, cyanosis, delayed pulmonary oedema, mental with an electrospray ionisation source. The electrospray ionisation confusion, unconsciousness and death, is produced under a controlled source and capillary were operated at 3.06 kV and 5.41 V, respective- manner while over-processing is prevented. Considering these factors, ly. The capillary temperature was set to 200 °C. First-order ESI mass the use of the gas on an industrial scale is almost impossible. spectra were recorded in the mass range m/z 100–1000. The heme As shown in Fig. 1A and B, at different concentrations of HCl, the and nitrosoheme solutions were introduced into the mass spectrom- yield changes very slightly with the increase in ascorbic acid and ni- eter at a constant flow rate of 5 μL/min by a syringe pump employing trite concentrations. The concentration of HCl influences the yield a 100 μL syringe.

3. Results and discussion

Table 3 3.1. Model ﬁtting and analysis Predicted value obtained from estimated quadratic model.

Exp. no. Yield (%) λmax(nm) L* value a* value b* value The mathematical models representing the yield, λmax, L*, a* and b* values of nitrosoheme as a function of the independent variables 1 99.07 531.11 46.56 41.35 18.59 within the region of investigation are expressed by the following 2 96.17 538.92 52.24 39.84 21.95 3 95.30 538.74 48.02 40.97 22.59 equations: 4 95.09 522.07 48.17 40.04 19.43 5 82.01 518.21 54.75 20.71 23.85 (1) Yield model: 6 95.09 522.07 48.18 40.04 19.43 7 91.11 513.03 56.17 26.69 22.83 ¼ : − : þ : y1 95 088 7 174X1 2 273X2 8 95.09 522.07 48.18 40.04 19.43 þ 2:367X −4:188X2−1:806X2 9 100.90 542.17 46.53 48.56 18.03 3 1 2 10 95.09 522.07 48.18 40.04 19.43 (2) λmax model: 11 86.56 516.84 49.22 30.65 19.95 12 96.36 532.54 46.16 38.62 21.92 ¼ : − : þ : 13 77.28 499.96 60.46 11.99 24.98 y2 522 068 9 915X1 5 815X2 þ : − : − : 14 71.17 505.39 57.51 10.84 25.83 5 375X3 2 750X1X2 3 750X2X3 15 95.09 522.07 48.18 40.04 19.43 (3) L* value model: 16 91.93 514.29 51.86 29.90 23.05 17 93.80 531.85 49.10 40.94 19.11 18 95.09 522.07 48.18 40.45 19.43 ¼ : þ : − : − : þ : 2 y3 48 176 2 822X1 1 290X2 2 856X3 1 624X1 19 81.82 513.59 54.93 21.93 23.88 þ : 2 þ : 2− : 20 86.16 512.29 53.44 24.22 23.31 1 093X2 1 128X3 1 476X1X2 N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547 541

Table 4 A Analysis of the variance (ANOVA) for the ﬁt of the experimental data to response sur- 90 80 face model. 2 Source Degree of Sum of Mean F P freedom squares square value value

Yield 1 Regression 9 1139.08 126.564 11.63 0.000a Linear 3 849.95 283.315 26.03 0.000a Square 3 282.05 94.016 8.64 0.004a 0 Interaction 3 7.08 2.361 0.22 0.882ns Residual error 10 108.85 10.885 –– Total 19 1247.92 - –– Ascorbic acid -1 λ 95 max 70 Regression 9 2470.18 274.464 11.65 0.000a Linear 3 2198.89 732.963 31.11 0.000a -2 Square 3 0.29 0.095 0.00 1.000ns 75 85 Interaction 3 271.00 90.333 3.83 0.046b Residual error 10 235.57 23.557 –– -2 -1 0 1 2 Total 19 2705.75 ––– HCl L* value Regression 9 326.824 36.314 24.55 0.000a Linear 3 242.882 80.960 54.74 0.000a B a Square 3 62.052 20.684 13.99 0.001 1.5 90 80 Interaction 3 21.891 7.297 4.93 0.023b Residual error 10 14.789 1.479 –– Total 19 341.613 ––– 1.0 a* value Regression 9 2157.45 239.716 13.92 0.000a 0.5 Linear 3 1692.98 564.328 32.76 0.000a Square 3 460.23 153.409 8.91 0.004b Interaction 3 4.23 1.411 0.08 0.968ns 0.0

Residual error 10 172.25 17.225 –– Nitrite Total 19 2329.70 ––– -0.5 b* value Regression 9 104.902 11.656 21.63 0.000a 70 Linear 3 55.805 18.602 34.52 0.000a -1.0 Square 3 45.006 15.002 27.84 0.000a 95 Interaction 3 4.092 1.364 2.53 0.116ns -1.5 85 75 Residual error 10 5.389 0.539 –– Total 19 110.291 ––– -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

ns Not significant at P>0.5. HCl a Significant at Pb0.001. b Significant at Pb0.05. C

1.5 92.5

1.0 more than two other variables. The contour plots (Fig. 1A and B) indicate that at HCl concentrations of around 1%, the maximum pref- 0.5 ormed nitrosoheme yield is obtained. It seems that HCl has two effects on nitrosoheme synthesis: it decreases pH and facilitates reac- 0.0 87.5 tions that transform nitrite to nitric oxide. Nitrite is the conjugate Nitrite 95.0 base of a weak acid, nitrous acid (HNO2). In low concentrations of -0.5 HCl, because the pH of solution is approximately the pKa of HNO2 (3.36), the majority of the nitrite will be present in the form of -1.0 HNO2. In the presence of reductants (ascorbic acid), nitric oxide is produced from HNO . Because the reactivity of the nitrite/nitrous 2 -1.5 82.5 acid system increases with decreasing pH, the concentration of nitric 85.0 90.0 oxide is maximized, and the yield increases to 100%. The influence of -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 pH on the curing process has been well established, and it provides Ascorbic acid the basis for formulating acidulants to accelerate this process. Fox, Fig. 1. The contour plots (A) mutual effect of HCl and ascorbic acid, (B) mutual effect of Townsend, Ackerman, and Swift (1967) indicated that a pH decrease HCl and nitrite, (C) mutual effect of nitrite and ascorbic acid on the yield of of 0.2 units doubles the rate of colour formation due to the nitrite- nitrosoheme production. heme interaction. Also, in low concentrations of HCl, chloride ions play a role in nitrosating reactions. Sebranek (1979), examining the importance of sodium chloride in meat curing, concluded that chloride ions could actually help catalyse nitrosation reactions. He went on to state that the significance of these reactions in practical meat oxide from hemin, as its absorption spectra is similar to hemin at processing may have little influence on nitrite behaviour because HCl concentrations of more than 3% (Fig. 2). highly acidic conditions would be required for the reaction. HCl pro- With simultaneous increases in nitrite and ascorbic acid concentra- vides chloride ion and acidic pH that is needed in this regard. Increas- tions, the yield of nitrosoheme production is enhanced (Fig. 1C). When ing the HCl concentration to more than 1% causes the nitrosoheme nitrite and ascorbic acid are present at high levels, nitric oxide is pro- pigment to form, but the acidic pH catalyses the separation of nitric duced in adequate quantities to transform all hemin to nitrosoheme. 542 N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547

of the corresponding ﬁve-coordinate gas complexes. It seems that in samples containing 2% HCl, one nitric oxide molecule is bound to iron and forms ﬁve-coordination, but in other concentrations of HCl, dinitrosylheme is obtained. As shown in Fig. 4A and B, when the concentration of HCl increases, maximum absorbance shifts to shorter wavelengths, and with increases in the concentrations of nitrite and ascorbic acid, maximum absorbance is seen at longer wavelengths. Fig. 4C indicates that A

1.5 540 520

550 1.0

0.5

0.0

-0.5 Ascorbic acid 530 Fig. 2. Absorption spectra of hemin and nitorosoheme produced using 3% HCl. -1.0

3.3. Effects of variables on λmax -1.5 510 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 The most important property of nitrosoheme pigment for judging HCl its quality is its ability to reproduce the typical nitrite-cured colour in meat. Fig. 3 shows the absorption patterns, of the visible region of hemin and two synthesized nitrosoheme pigments. All pigments B 1.5 showed the characteristic absorption pattern of the iron-porphyrin 550 540 520 compound with a red colour and had maxima at 454–481, 508–540 and 537–561 nm. Shahidi and Pegg (1991) reported maximum absor- 1.0 bance for synthesized nitrosoheme at 540 and 563 nm. However, Tarladgis (1962) showed maxima at 482, 552 and 572 nm for nitric 0.5 oxide, haemoglobin and freshly cured ham, respectively, and 482, 548 and 576 nm for cooked ham. Fox and Thomson (1963) followed 0.0

the conversion of metmyoglobin to nitrosylmyoglobin by evaluating Nitrite the absorption spectra and reported that the maximum absorption -0.5 of nitrosylmetmyoglobin is at 535 nm and that a slight shift toward 530 longer wavelengths would be expected if nitrosylmetmyoglobin -1.0 converted to nitrosylmyoglobin. The same trend was observed in this study after nitrosoheme production from hemin. -1.5 510 The Soret band was observed at 415 nm, except in samples prepared with 2% HCl, which had a Soret band at 395 nm (Fig. 3). Miller, Pedraza -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 and Chance (1997) observed a Soret band for nitrosylmyoglobin at HCl 422 nm before it shifted to 393 nm, which is indicative of the formation C 1.5 530

1.0

0.5

0.0 Nitrite 510 -0.5

-1.0

520 -1.5 500

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Ascorbic acid

Fig. 3. Absorption spectra of (A) hemin, (B) nitrosoheme synthesized in 2% HCl, (C) Fig. 4. The contour plots (A) mutual effect of HCl and ascorbic acid, (B) mutual effect of nitrosoheme synthesized in other HCl concentrations. HCl and nitrite, (C) mutual effect of nitrite and ascorbic acid on the λmax. N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547 543 A A

1.5 1.5 30

50 1.0 1.0

0.5 55 0.5

0.0 0.0

-0.5 Ascorbic acid -0.5 Ascorbic acid 40 10 -1.0 -1.0

60 -1.5 50 -1.5 20 0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 HCl HCl B

B 1.5 30 1.5 1.0 1.0 0.5 0.5 55 0.0

0.0 Nitrite

Nitrite -0.5 -0.5 40 10 -1.0 50 -1.0 -1.5 20 -1.5 60 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 HCl HCl C 1.5 C 30 40 1.5 525 4 48 1.0

1.0 0.5 46

0.5 0.0 Nitrite

0.0 -0.5 Nitrite -0.5 50 -1.0 20 35

-1.0 56 -1.5 15 25 54 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -1.5 58 Ascorbic acid -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Ascorbic acid Fig. 6. The contour plots (A) mutual effect of HCl and ascorbic acid, (B) mutual effect of HCl and nitrite, (C) mutual effect of nitrite and ascorbic acid on the a* value. Fig. 5. The contour plots (A) mutual effect of HCl and ascorbic acid, (B) mutual effect of HCl and nitrite, (C) mutual effect of Nitrite and ascorbic acid on the L* value. from iron, and as shown in Fig. 2, their absorption spectra have a pattern similar to the spectrum of hemin. the simultaneous increase of nitrite and ascorbic acid concentrations shifts λmax to 522–540 nm. As mentioned before, with increases in 3.4. Effects of variables on L* value the HCl concentration to 1%, the yield of nitrosoheme production increases, which is conﬁrmed by the longer wavelengths observed in The contour plots indicate that with increases in HCl concentra- these samples. At higher concentrations of nitrite, the nitrosating tion and reductions in nitrite and ascorbic acid concentrations, light- agent will be available in sufﬁcient amounts. Ascorbic acid also catal- ness is enhanced (Fig. 5A and B). At low concentrations of HCl, the yses the conversion of nitrosylmetheme to nitrosoheme, which yield is higher and the L* value does not change with increases in produces pigments with a λmax of 540 nm. In extremely acidic condi- ascorbic acid concentrations, but at higher concentrations of HCl, tions, however, nitrosating agents are produced, but they dissociate the lightness variation is greater. As mentioned before, HCl has a 544 N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547 A A Wavelength 1.5 22 530 2 545 1.0 Yield 96 0.5 1 100 b value 17 0.0 0 20 a value -0.5 20 Ascorbic acid

Ascorbic acid 43 26 -1 48 -1.0 L value -2 45 -1.5 28 50 24 24 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -2 -1 0 1 2 HCI HCl B 1.5 B Wavelength 22 530 1.0 2 545 Yield 0.5 96 1 100 0.0 b value

Nitrite 17 0 20 -0.5 20 Nitrite 26 a value -1 43 -1.0 48 L value 24 -1.5 24 -2 45 50 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 HCI -2 -1 0 1 2 C HCl

1.5 20 C Wavelength 1.0 530 2 545 0.5 18 Yield 96 0.0 1 100

Nitrite b value -0.5 17 0 20 Nitrite a value -1.0 43 -1 48 24 22 -1.5 L value 45 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -2 50 Ascorbic acid

Fig. 7. The contour plots (A) mutual effect of HCl and ascorbic acid, (B) mutual effect of -2 -1 0 1 2 HCl and nitrite, (C) mutual effect of nitrite and ascorbic acid on b* value. Ascorbic acid

Fig. 8. Optimum region identified by overlaying plots of the five responses (yield, λmax, L* value, a* value and b* value) as functions of (A) HCl and ascorbic acid concentration, dual role in nitrosoheme production by lowering the pH and provid- (B) HCl and nitrite concentration, (C) nitrite and ascorbic acid concentration. ing the chloride ion. Gimeno, Astiasarán, and Bello (2001) evaluated calcium ascorbate at different concentrations as a partial substitute The role of chloride ions in nitrosoheme production and the light- for NaCl in dry-fermented sausages. After a simultaneous reduction ness of pigments, at higher concentrations of HCl, is insignificant. It in NaCl and increase in ascorbic acid content, the L* value decreased. seems that under those conditions, iron dissociates from the heme Gimeno, Astiasaran, and Bello (1999) reported a higher L* value in ring, and the decrease in nitrosoheme formation enhances the light- fermented sausage in which the NaCl was replaced with CaCl2. ness. Legge and Lemberg (1941) studied the role of reductants and These studies confirm the role of chloride ions in the colour of acidification on choleglobin formation and the detachment of iron nitrosoheme pigments. from heme, by introducing ascorbic acid to decrease the dissociation N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547 545

determined as follows, according to coded values: HCl percentage= −0.80 (1.19%), Ascorbic acid concentration=0.46 (123.08 mM) and Nitrite concentration=1.00 (200 mM). After production of nitrosoheme

under optimal conditions, the yield, λmax, L*value,a* value and b*values were 100±0.0%, 540±1 nm, 47.03±0.7, 45.17±0.85 and 17.2±0.28, respectively. These values are not signiﬁcantly (p>0.05) different from the predicted values of 99.5%, 540 nm, 46.32, 44.12, and 18.17.

Fig. 9. The FTIR spectrum of Heme and nitrosoheme. of iron by acids. Ascorbic acid is also a reducing agent at low concentrations; it evacuates oxygen from the reaction solution and inhibits both the formation of NO2 and the reaction with hemin that produces a pigment with darker colour. When the proportion of ascorbic acid increases, other pigments such as choleglobin will be formed, which have higher L* values. As shown in Fig. 5C, simultaneous increases in nitrite and ascorbic acid concentrations cause the L* value to decrease. Heaton, Cornforth, Moiseev, Egbert, and Carpenter (2000) and Deniz and Serdarolu (2003), using different amounts of nitrite for production of meat products, found that lightness is reduced by increasing the level of nitrite.

3.5. Effects of variables on a* value

In a comparison of Fig. 6AandBwithFig. 1A and B, the same pattern can be seen. At lower concentrations of HCl and higher concentrations of ascorbic acid, nitrosoheme production increases, accompanied by a higher a* value. With increasing HCl concentrations, the produced B pigment decomposes, leading to a decrease in redness. The greater redness of pigments with sodium nitrite is due to the fact that it is reduced to nitric oxide by added reductant (i.e., ascorbate) and reacts with hemin to form the red nitrosoheme. Although the a*valueshavenot been evaluated in synthesized pigments, the effect of higher levels of nitrite on increasing the redness of meat products has been conﬁrmed (Deniz & Serdarolu, 2003; Heaton et al., 2000; Pegg, 1993). The chloride ion of HCl at low concentrations inﬂuences the production of pigments with higher a* values, and increasing the concentration of HCl or in other strongly acidic conditions, nitric oxide dissociates after production of nitrosoheme, a phenomenon like light fading that decreases the a* value and produces a pigment with green colour.

3.6. Effects of variables on b* value

As shown in Fig. 7, variation of the b* value is not extensive with the changes of variables. When the concentration of ascorbic acid or nitrite decreases, the b* value increases, and HCl at low or high concentrations enhances this value. In these conditions, cooked cured- meat pigments do not form or dissociate after production and convert to oxidized heme.

3.7. Optimisation of nitrosoheme production

3.7.1. Numerical optimisation From the computed predictions, the optimal conditions to obtain the highest yield and λmax in the range of 530–545 nm were Fig. 10. The ESI mass spectrum of (A) heme and (B) nitrosoheme. 546 N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547

3.7.2. Graphical optimisation of 658.5 is produced. Other abundant peaks have m/z of 646 and With the aim of determining the optimal conditions for nitrosoheme 736.4; this is related to mononitrosylheme and non-covalent bound be- production, a graphical optimisation was conducted using Minitab soft- tween the compound with m/z 658.8 and acetone–H2O, respectively. ware. Such a methodology essentially consists of overlaying the curves of the five models, obtained from the Central Composite design, 4. Conclusion according to the specificcriteriaimposed. The optimal working conditions were defined so as to obtain a The suggested procedure could produce cooked cured-meat pig- high yield of nitrosoheme production with the absorbance and Hunt- ment simply and in high yield, and the pigment generates the appropri- er Lab in optimum range. Thus, the following constraints, including ate colour in cured meat. Results show that all three reaction variables Yield>96%, 530 nmbabsorbanceb545 nm, 45bL* valueb50, 43ba* have effects on the characteristics of the pigment, both individually valueb48 and 17bb* valueb20, were imposed. Fig. 8 shows the over- and in association with other reaction parameters. A yield of 100% lay plot in which the non-shaded area represents the factors satisfy- was achieved after optimisation using RSM. HCl, ascorbic acid and ni- ing the imposed criteria. trite concentrations of 1.19%, 123.08 mM and 200 mM, respectively, were found to be the best conditions for conversion of 1 mM hemin to 3.8. FTIR of the synthesized pigment nitrosoheme. In the evaluation of cured pigment structure, it was found that in addition to Fe, the carboxylic group of heme reacts with The FTIR spectra of heme and the synthesized pigment are shown NO during the production of nitrosoheme. At the end, although we − 1 in Fig. 9. The hemin showed a significant band at ν=1703 cm , could produce the cooked cured meat pigment in vitro, the other char- which is the dominant signal in the middle of the infrared (IR) acteristics of nitrite must be considered in further studies. range; this can be clearly attributed to the carbonyl stretching mode of the protonated heme propionates (Dörr, Schade, & Hellwig, Acknowledgments 2008). After the reaction of heme with nitric oxide, this band dis- appeared and strong signals appeared at 1031, 1066, 1111, 1141, The authors gratefully acknowledge the financial support of Iran − 1 1261, 1591, 1720 and 1790 cm . It seems that the reaction of the National Science Foundation. We also express our sincere gratitude carboxylic groups of hemin propionate with nitric oxide is the main to Dr. Niazi, Director of Diagnostic Center, Drugs and Biological Mate- reason for the disappearance of the carbonyl band in the IR spectra, rials Control of Iran Vetrinary Organization for his enthusiastic help − 1 whereas the two bands obtained at 1720 and 1790 cm may be re- and Mr. Tavassoli for his technical assistance. lated to the interaction of nitric oxide with the carboxylic acid group of propionate. The peak at 1591 cm− 1 corresponds to the presence of References a bent Fe–NO moiety and a pentacoordinate complex. Miller, Pedraza and Chance (1997) indicated that the primary ligand-bound state Archer, D. L. (2002). Evidence that ingested nitrate and nitrite are beneficial to health. falls at 1613 cm− 1 for Mb14NO and shifts to 1587 cm− 1 for Journal of Food Protection, 65(5), 872–875. 15 Butler, A. R., & Feelisch, M. (2008). Therapeutic uses of inorganic nitrite and nitrate: Mb NO. Sun, Zhou, Xu, & Peng (2009) studied the IR spectra of From the past to the future. Circulation, 117(16), 2151–2159. − 1 extracted cured pigment and related the band at 1653 cm to the Demeyer, D. l., Honikel, K., & De Smet, S. (2008). The World Cancer Research Fund Fe–NO complex. Maxwell and Caughey (1976) reported that the report 2007: A challenge for the meat processing industry. Meat Science, 80(4), – – 953 959. stretching frequency of the bent Fe NO moiety of preformed cured Deniz, E. E., & Serdarolu, M. (2003). Effects of nitrite levels, endpoint temperature and − 1 pigment was in the range of 1600–1700 cm . Killday, Tempesta, storage on pink color development in turkey rolls. European Food Research and Bailey and Metral (1988) reported that the infrared spectrum of Technology, 217(6), 471–474. Dörr, S., Schade, U., & Hellwig, P. (2008). Far infrared spectroscopy on hemoproteins: A extracted cooked cured meat pigment had a nitrosyl stretch at model compound study from 1800–100 cm-1. Vibrational Spectroscopy, 47(1), − 1 1656 cm . Pegg and Shahidi (1996) and Jankiewicz, Kwany, 59–65. Wasylik, and Graczyk (1994) reported that the stretching frequency Faramawi, M. F., Johnson, E., Fry, M. W., Sall, M., & Yi, Z. (2007). Consumption of differ- (m=1659 cm− 1) of preformed cured pigment was consistent with ent types of meat and the risk of renal cancer: Meta-analysis of case control studies. Cancer Causes & Control, 18(2), 125–133. the first nitrosyl group bound with the ferrous atom of the heme Fox, J. B., & Thomson, J. S. (1963). Formation of bovine nitrosylmyoglobin. I. pH 4.5–6.5. molecule. According to Killday et al. (1988), the second nitrosyl group Biochemistry, 2(3), 465–470. in cured pigment has the stretching band at m=ca. 1900 cm−1, Fox, J. B., Townsend, W. E., Ackerman, S. A., & Swift, C. D. (1967). Cured color development during frankfurter processing. Food Technology, 21, 386–388. which was not observed in this study; however, it seems that two Gimeno, O., Astiasaran, I., & Bello, J. (1999). Influence of partial replacement of NaCl nitrosyl groups are bound to hemin, one to propionate and the other with KCl and CaCl2 on texture and colour of dry fermented sausages. Journal of Ag- to Fe. ricultural and Food Chemistry, 47, 873–877. Gimeno, O., Astiasarán, I., & Bello, J. (2001). Calcium ascorbate as a potential partial substitute for NaCl in dry fermented sausages: Effect on colour, texture and hy- 3.9. ESI/MS of synthesized nitrosoheme gienic quality at different concentrations. Meat Science, 57(1), 23–29. Heaton, K. M., Cornforth, D. P., Moiseev, I. V., Egbert, W. R., & Carpenter, C. E. (2000). Minimum sodium nitrite levels for pinking of various cooked meats as related to The ESI mass spectra obtained for heme and nitrosoheme are shown use of direct or indirect-dried soy isolates in poultry rolls. Meat Science, 55(3), in Fig. 10. In the spectrum of heme, the following ions are abundant: 321–329. Honikel, K. O. (2008). The use and control of nitrate and nitrite for the processing of Heme (m/z 616.3), [Heme: Acetone: H2O] (m/z 693.8). After conversion meat products. Meat Science, 78(1–2), 68–76. of heme to nitrosoheme, the most intense peak is m/z=658.5; that is Hornsey, H. C. (1956). The colour of cooked cured pork. I.-Estimation of the nitric not consistent with the m/z of mononitrosylheme (m/z~646) or oxide-haem pigment. Journal of the Science of Food and Agriculture, 7, 534–540. dinitrosylheme (m/z~676). Sun et al. (2009) also obtained a peak at Jankiewicz, L., Kwany, M., Wasylik, K., & Graczyk, A. (1994). Structure studies on the – m/z 657.3 for a cooked cured pigment that had been extracted from nitrosyl derivative of heme. Journal of Food Science, 59(1), 57 59. Killday, K. B., Tempesta, M. S., Bailey, M. E., & Metral, C. J. (1988). Structural character- cured meat. However, they used HPLC/ESI-HR-MS with acetonitrile as ization of nitrosylhemochromogen of cooked cured meat: Implications in the a mobile phase for analysis of extracted pigment and related the molec- meat-curing reaction. Journal of Agricultural and Food Chemistry, 36(5), 909–914. ular ion peak to coordination of iron porphyrin with an acetonitrile. As Legge, J. W., & Lemberg, R. (1941). Coupled oxidation of ascorbic acid and haemoglobin: The labile iron in blood and its increase in choleglobin formation. mentioned in discussion of the FTIR spectrum of nitrosoheme, we pro- Biochemical Journal, 35(3), 353–379. pose that two nitric oxide molecules have been bound to heme. One of Maxwell, J. C., & Caughey, W. S. (1976). An infrared study of NO bonding to heme B and them coordinates with iron at the centre of heme, and the other reacts hemoglobin A. Evidence for inositol hexaphosphate induced cleavage of proximal histidine to iron bonds. Biochemistry, 15(2), 388–396. with the propionate side chain of heme that is accompanied by loss of Miller, L. M., Pedraza, A. J., & Chance, M. R. (1997). Identification of conformational sub- onemoleculeofH2O. Therefore, a compound with a molecular weight states involved in nitric oxide binding to ferric and ferrous myoglobin through N. Soltanizadeh, M. Kadivar / Meat Science 92 (2012) 538–547 547

difference Fourier transform infrared spectroscopy (FTIR). Biochemistry, 36(40), Shahidi, F., & Pegg, R. (1992). Powdered Cooked Cured-Meat Pigment Which is a 12199–12207. Non-Nitrite Meat Preservative: WO Patent WO/1992/007,476. Parthasarathy, D. K., & Bryan, N. S. (2012). Sodium Nitrite: The "Cure" for Nitric Oxide Shahidi, F., & Pegg, R. B. (1995a). Proteinaceous food product containing stabilized cooked Insufﬁciency. Meat Science, 92(3), 274–279. cured-meat pigment: US Patent 5443852. Pegg, R. B. (1993). Development of nitrite-free meat curing systems. Unpublished Ph.D., Shahidi, F., & Pegg, R. B. (1995b). Stabilized cooked cured-meat pigment: US Patent 5425956. Memorial University of Newfoundland Canada. Shahidi, F., Rubin, L. J., Diosady, L. L., & Chew, V. (1984). Preparation of dinitrosyl fer- Pegg, R. B., & Shahidi, F. (1996). A novel titration methodology for elucidation of the rohemochrome from hemin and sodium nitrite. Canadian Institute of Food Science structure of preformed cooked cured-meat pigment by visible spectroscopy. Food and Technology Journal, 17(1), 33–37. Chemistry, 56(2), 105–110. Shahidi, F., Rubin, L. J., Diosady, L. L., & Wood, D. F. (1985). Preparation of the cooked Pegg, R. B., & Shahidi, F. (2000). Nitrite curing of meat: The N-nitrosamine problem and cured-meat pigment, dinitrosyl ferrohemochrome, from hemin and nitric oxide. nitrite alternatives. Connecticut: Food & Nutrition Press, INC. Journal of Food Science, 50(1), 272–273. Pegg, R. B., Shahidi, F., & Fox, J. B. (1997). Unraveling the chemical identity of meat pig- Sun, W. Q., Zhou, G. H., Xu, X. L., & Peng, Z. Q. (2009). Studies on the structure and ox- ments. Critical Reviews in Food Science and Nutrition, 37(6), 561–589. idation properties of extracted cooked cured meat pigment by four spectra. Food Rubin, L., Diosady, L., O'Boyle, A., Kassam, N., & Shahidi, F. (1992). Nitrite-free meat cu- Chemistry, 115(2), 596–601. rative and process of curing meat: WO Patent WO/1992/003,223. Tarladgis, B. G. (1962). Interpretation of the spectra of meat pigments. II.—Cured meats. Sebranek, J. G. (1979). Advances in the technology of nitrite use and consideration of The mechanism of colour fading. Journal of the Science of Food and Agriculture, alternatives. Food Technology, 33(7), 58–62,93. 13(9), 485–491. Shahidi, F., & Pegg, R. (1991). Encapsulation of the preformed cooked cured meat pig- Yam, K. L., & Papadakis, S. E. (2004). A simple digital imaging method for measuring ment. Journal of Food Science, 56(6), 1500–1504. and analyzing color of food surfaces. Journal of Food Engineering, 61(1), 137–142.