Retos del uso de las TIC en la educación

6.1. Milk, Ice Cream, and Dairy Products

Udy (9) analyzed whole milk and spray-dried milk samples by Orange-G binding. The milk samples and Kjeldahl protein values were supplied by Ashworth and co-workers at the Department of Dairy Science, Washington State University (Pullman, WA). Dye-binding studies at Ashworth's

laboratory led to one of the ®rst Ph.D. dissertations in this area (68). Literature covering milk or dairy protein analysis using Orange-G and Acid Orange 12 is summarized in Table 6.

Ashworth et al. (24) analyzed 354 milk samples from six breeds of cows. Milk powders were also analyzed. The average protein content for milks was 3.49 (+ 0.273)%. Some 95% of protein determinations were within + 0.67% of the crude protein content. NPN, proteose peptone, milk fat, and lactose caused little or no interference. Sample preservatives (hydrogen peroxide, formaldehyde, or mercuric hydrochloride) also did not affect the results. Adding mercuric chloride (1.35 mg %) to milk samples allowed room temperature storage before analysis. Antibiotics were not effective preservatives.

TABLE6 Dye-Binding Assay of Milk and Dairy Protein

Dye, application Reference

Orange-G

Milk (fresh, powder) Udy (9), Ashworth et al. (24), Dolby

(25), Ashworth and Chaudry (26), and Conetta et al. (69), Park and King (70) Milk (fresh, evaporated, powdered),

buttermilk, cheese, sherbet, cream, ice cream

Ashworth (20)

Ice cream, frozen desert Kroger et al. (71)

Acid Orange 12

Milk (fresh, evaporated, powdered), buttermilk, cheese, sherbet, cream, ice cream

Ashworth (20)

Milk Sherbon (21)

Chocolate milk drink, buttermilk Sherbon and Luke (72)

Nonfat dry milk powder, ice cream,

half-and-half Sherbon and Luke (73)

Various dairy products Sherbon (74)

Milk Conetta et al. (69), Lakin (75,76),

Wilkinson and Richardson (77),

Cheese Kristoffersen (78)

Various dairy products, NFDM Sherbon and Fleming (79),

Ice cream, ice milk, diet ice cream,

dietetic ice cream Bruhn et al. (80)

Other dyes

Milk powder (delactosed)ÐRamazol

The DBC for milk protein fractions was assessed by Ashworth and Chaudry (26). Milk protein fractions should have similar DBC values; otherwise, assays may be affected by variations in milk composition. Compared with whey proteins, caseins had a lower DBC (Table 7). Presumably the proportions of casein and whey protein remain fairly constant in different milk samples. The quantity of protein in several brands of milk drink (chocolate milk, two-ten, half-and-half, vitamin D milk, etc.) were determined by Ashworth (20).

The Orange-G binding capacity for milk proteins was remarkably constant, notwithstanding processing into products such as ice cream and evaporated milk (Table 8). Fresh milk had a protein content of 3.5%, whereas evaporated milk contained 7% protein. Cheese manufacture had a signi®cant lowering effect on DBC, probably because of proteolysis. Notice that the values for the DBC are 50% lower than those reported in Table 3 for reasons discussed earlier.

6.2. Of®cial Approval of Dye-Binding Assays

Collaborative studies led to dye-binding assays being granted AOAC approval for milk protein determination (21,72,83). The studies involved laboratories attached to the of®ces of the U.S. Federal Milk Market

TABLE7 Apparent DBC for Different Milk Protein Fractionsa

Milk protein (fraction) AB 10B (mmoles g 1_{cP) Orange G (mmoles g} 1_cP)

Whole milk 566.6 389.4 Skim 561.7 396.0 NFDP 595.8 407.1 Whole casein 589.3 446.9 Paracasein 556.8 415.9 a-Casein 577.9 398.2 k-casein 592.5 420.4 b-casein 504.9 376.1 Whey protein 730.5 539.8 b-Lactoglobulin 763.0 557.5 Proteose peptone 584.4 278.8 Average 602.1 420.6 SD 76.1 76.2

a_{DBC was calculated from Ref. 26 using dye molecular weights from Table 2. Values are}

between 50 and 100% lower than expected from the basic amino acid content in milk proteins (see Table 3).

Administrators.* The extent of interlaboratory differences was considered practically insigni®cant. The agreement between dye-binding results and Kjeldahl results was excellent. Dye binding was granted of®cial ®rst action status for protein quantitation in fresh milk, dairy chocolate milk drink, cultured buttermilk, and half-and-half cream.

A collaborative study to examine dye-binding assays for ice cream mix and nonfat dry milk powder is described by Sherbon and Luke (73). Five commercial samples of vanilla ice cream mix and 10 samples of nonfat dry milk were analyzed in powder form or after reconstitution. Seven laboratories performed dye-binding assays while three did Kjeldahl analysis. Tests using nonfat dry milk powder gave the same results as for the reconstituted material. The protein content for nonfat dry milk was 34.946% as determined by Kjeldahl analysis and 35.141% by dye binding. The average difference between Kjeldahl and dye-binding results was ‡ 0.195%.

TABLE8 Dye-Binding Capacity and Protein Content for Different Milk Productsa

Product OG (mmoles g 1_{) AO12 (mmoles g} 1_{) Protein (%)}

Milk Fresh 393.5 1000.0 3.5 Evaporated 386.8 932.8 7.0 Powdered, nonfat 394.6 1000.0 36.0 Powdered, whole 366.9 963.6 26.0 Buttermilk 381.1 902.0 4.0 Cottage cheeses Creamed 394.8 966.4 15.0 Dry curd 394.8 966.4 18.0 Cheddar cheese Mild 306.8 801.1 24.0 Sharp 245.4 700.3 24.0 Cream 20±50% fat 394.8 1000.06 2.5 Ice cream Vanilla 394.8 1000.0 4.0 Chocolate 362.3 963.6 4.0

a_{A comparison with data from Table 3 shows that the values for OG are lower than expected.}

Source: DBC values calucalated from Ref. 21.

* One lot of sterile, canned milk was analyzed by ®ve laboratories. Three laboratories carried out crude protein determination by the Kjeldahl method. A further 25 fresh milk samples were analyzed via 140 replicate determinations.

Ice cream mix had 3.852% protein or 3.968% protein by the Kjeldhal and dye-binding analysis, respectively. Consequently, dye binding received AOAC approval as a method for protein determination in ice cream and nonfat dry milk Analysis of ice cream mix and ice cream proteins using dye binding is further described by Kleyn (84) and by Bruhn (85). Further examples of these studies are listed in Table 9.

Sherbon (32) compared the Pro-Milk and Udy methods for the analysis of milk proteins. The two techniques gave comparable results. Greater care was needed in calibrating the Pro-Milk instrument. A recommendation to grant of®cial status to the Pro-Milk method was deferred pending further work. A year later, a six-laboratory study of the Pro-Milk method led to AOAC approval (33). In addition to reports cited in Table 9 the Pro-Milk Analyzer is discussed in Refs. 86±91.

6.3. Wheat and Other Cereal Proteins

Udy (8) found that wheat albumin, gluten, albumin ‡ gluten, and residue proteins bound constant amounts of Orange-G regardless of the variety of seed. These observations paved the way for a quantitative analysis of wheat

TABLE9 Application of Amido Black 10B for Milk and Dairy Protein Analysisa

Amido Black 10B Reference

Cheese Kroger and Weaver (92)

Condensed milk Lueck (93)

Ice cream, frozen dessert Kroger et al. (71)

Milk Dolby (25), Ashworth and Chaudry

(26), Radcliffe (94), O'Connell (86), Conetta et al. (69), Sherbon (95), Kroger (96), Uzonyi (97), Patel et al. (98), Ng-Kwai-Hang and Hayes (99), van Reusel and Klijn (100)

Whey protein, casein Roper and Dolby (101), McGann et al.

(102), Renner and Ando (103), Reimerdes and Flegel (104)

Milk (goat) Grappin et al. (105), Mabon and

Brechany (106)

Milk (mastitis) Waite and Smith (107)

Skimmed milk powder Sanderson (108), O'Connell and

McGann (109)

protein using dye binding (9). In all, 128 samples of whole wheat ¯our and 218 samples of re®ned wheat ¯our (from 58 known and 34 unknown wheat varieties) were examined for Orange-G binding capacity at pH 2.2. The samples were also analyzed for crude protein content by the Kjeldahl method. The correlation between amounts of dye bound and crude protein (%N 6 6.25) is described by Equation (30) (re®ned wheat ¯our) or Equation (31) (whole wheat ¯our).

cP ˆ 1:092Db 4:62 …30†

cP ˆ 1:000Db 5:53 …31†

The nonzero intercept indicates dye binding to nonprotein components, probably starch and/or wheat bran. Greenaway (110) at the U.S. Department of Agriculture (USDA;Beltsville, MD) reported a positive correlation between dye binding and Kjeldahl results for soft wheat (< 10% protein), hard winter wheat, hard spring wheat, and durum wheat;

cP ˆ 0:8842X ‡ 1:7938 …R ˆ 0:988† …32†

where X (% protein) is wheat protein content determined from dye binding and cP ˆ Kjeldahl protein (%N 6 6.25). Over 220 protein assays were performed. Methods were compared with respect to average protein content, correlation coef®cients, and standard error of estimates. The dye- binding assay gave reliable estimates for protein content for hard red spring wheat. For other classes of wheat, protein results were slightly lower than expected from Kjeldahl analysis. The mean difference between Kjeldahl and dye-binding tests was  0.5%. For wheat samples having less than 10% protein, dye-binding and Kjeldahl results differed by *1%.

The reliability of the Udy method was compared with that of ®ve other techniques (Kjeldahl, alkaline distillation, biuret, Dumas, and infrared re¯ectance) for wheat protein determination by Pomeranz and More (111). Reliability encompasses speci®city, accuracy, precision, sensitivity, and the LLD (Chapter 1). Dye binding was the least precise method tested. Interestingly, no strong case is made for preferring any one method. Forty- ®ve varieties of rice produced in the 1966 season in the Philippines by the International Rice Research Institute (IRRI) were analyzed for protein (112). A typical set of results highlighting the performance of the dye- binding assay for rice protein analysis is given in Table 10.

Barley and malt proteins were analyzed by Pomeranz et al. (113) using ®ve methods including the Udy method. About 120 samples each of barley and malt from all over the United States were analyzed. No details of the dye-binding assay were given other than a reference to Udy's 1971 paper

(22). The correlation coef®cient for dye-binding and Kjeldhal results was 0.974 (barley) or 0.984 (malt). The average mean squared error for analysis (with the Kjeldahl method as reference) was 0.897%. Using commercial apparatus, 200 protein determinations were completed daily.

Baker and Hunt (114) evaluated the Pro-meter instrument (Foss America Inc) for dye-binding analysis of wheat protein. About 107 wheat samples were ground to pass 20-mesh screen and then analyzed according the instrument manufacturer's instructions. The graph of instrument response versus protein content was curvilinear for 50 samples of red wheat (hard red spring, hard red winter, durum wheat). By contrast, a linear calibration graph was obtained for 57 white wheat samples. The Pro-meter instrument was judged satisfactory despite some mechanical dif®culties. The ®lter system was periodically clogged, necessitating dismantling and cleaning of the measuring unit.

6.4. Legumes and Other Seed Proteins

The Udy method is applicable to range of legumes including, chickpeas, cowpeas, gram, mungbeans, peas, and soybean (Table 1). Pomeranz (115) analyzed 24 soy ¯our samples using Orange-G and commercial apparatus from the Udy Corporation. The results were compared with the biuret and Kjeldahl methods. A highly signi®cant correlation was found between protein content assessed by dye binding (X, %) and crude protein (N 6 6.25%). For samples containing up to 80% protein, the regression equation was

cP ˆ 1:003X 4:559 …R ˆ 0:989† …33†

The standard error of analysis was 1.8213%. Flour particle size had negligible effects on protein results. Mild heat did not affect dye-binding

TABLE10 Determination of Rice Protein Using Acid Orange 12 Dye-Binding

Assay

Parameter Milled rice Brown rice

Regression linea _{Y ˆ 14.67 13.60A}

485 Y ˆ 14.78 14.12A485

Protein (%) 5.55±11.65 6.00±11.95

SY.X(%) + 0.48 + 0.26

R 0.961 0.984

a_{Y ˆ crude protein content (%). Protein (%) is range for 45 samples. S}

Y.X(%) ˆ standard

results although other studies show that severe heating reduces the DBC of soy proteins (116±120).

Romo et al. (121) assessed seed protein extractability using the Udy method. The following DA482 changes were noted for the different seed

protein solutions (10 mg mL 1_{): 1.363 (®eld bean), 1.197 (cowpea), 2.976}

(rapeseed), 2.2454 (sesame seed), and 1.203 (cotton seed). Clearly, the assay sensitivity is different for different seed proteins. Medina et al. (122) proposed that a single calibration graph might be used for cereal, legume, and oilseed protein analysis. A composite graph would save time. Sesame ¯our, rapeseed meal, and rapeseed ¯our were analyzed by the standard Udy (shaker mixing) method. Fig. 8 shows a composite calibration graph for Acid Orange 12 binding to cereal, legume, and oilseed proteins. The least- squares equation for the composite graph is*

cP ˆ 0:2152Db‡ 4:7333 …R ˆ 0:981† …34†

FIGURE8 A composite calibration graph relating dye binding (X-axis) and crude protein content for *28 samples of legumes, cereals, and oilseeds. (Drawn from Table IV in Ref. 123.)

* Actually, the regression equation reported in the literature was Y ˆ 0.245X ‡ 2.532 (R ˆ 0.995). In contrast, Fig. 1 was drawn using only 50% of the experimental data.

Apparently the average DBC for Acid Orange 12 is 465 (+ 17.11) mg g 1

(cP) or 1328.6 mmole g 1_.

The Udy method was further optimized for seed protein determination (122). Vacuum drying (558C) or atmospheric drying (1008C) had no effect on dye binding. Improving the degree of mixing, extending the shaking time from 30 to 150 minutes, and/or reducing the particle size to 40 mesh increased the perceived sample protein content. Table 11 summarizes results for sesame ¯our, rapeseed ¯our, and rapeseed meal. For all cases, a good correlation was obtained between dye binding and Kjeldahl results.

Rapeseed protein was also determined by Goh and Clandinin (123). Twelve commercial meals and two laboratory samples were analyzed using the Udy method with Orange G dye reagent. The investigators also examined Acid Orange 12 as a dye reagent. Rapeseed meal had 30.1±44.8% (w/w) protein. For Orange-G the least-squares line relating dye binding and Kjeldahl results was

cP ˆ 0:49Db‡ 3:91 …R ˆ 0:93† …35†

For Acid Orange 12 dyes the corresponding equation is

cP ˆ 0:28Db‡ 0:36 …R ˆ 0:98† …36†

From such data it may be shown that the apparent DBC for Orange-G is 204 mg (dye) g 1_{(cP) or 490 mmoles (dye) g} 1_{(cP). With Acid Orange 12 the}

DBC is 357 mg g 1_{(cP) or 580 mmoles (dye) g} 1_{(cP). The calibration data}

were not affected by ¯our particle size (40 or 60 mesh). However, DBC was higher for a protein/dye ratio of 2:1 as compared with a ratio of 4:1. Values for the DBC were proportional to the net concentration of arginine, histidine, and lysine. The standard deviation for analysis was 1.3% (Orange G) or 0.80% (Acid Orange). From the higher DBC (per weight), precision,

TABLE11 Protein Content in Sesame and Rapeseed Products

Determined from Acid Orange 12 Dye Binding and Kjeldahl Analysis

Sample Protein (% w/w)a

Dye binding Kjeldahl

Sesame ¯our 59.4 (+ 0.524) 58.9 (+ 1.093)

Rapeseed ¯our 59.4 (+ 1.743) 60 (+ 2.91)

Rapeseed meal 36.1 (+ 0.595) 36 (+ 0.338)

a_{Values are mean (+ SD).}

and sensitivity of analysis, Goh and Clandinin (123) concluded that Acid Orange 12 was a more suitable dye reagent for rapeseed protein determination.

6.5. Fish, Meat, and Egg Products A. Animal Feedstuffs

Bunyan (124) determined the protein content of feedstuffs containing animal protein. The procedure using Orange-G was essentially as given in Method 2. The dye-binding response was dependent on sample particle size. The extent of dye binding also increased with mixing time. With care, values of Db could be correlated with the protein content (Table 12). However,

animal feeds were found to be highly heterogeneous owing to their different manufacturing and thermal histories. A number of meat meal samples had an unusually high content of gelatin. In one case, meat meal was positively identi®ed as feather meal (techniques for establishing protein authenticity are described in Chapters 9±11). It was concluded that dye-binding assays were not suited for animal feedstuffs. Differences in processing history, protein quality, and possible adulteration led to large variations in results.

TABLE12 Analysis of Protein Content of Animal Feeds Using Orange G Dye

Binding

Sample (na_{) Regression line (mmole g}DBC1_{cP) % Error (CV)}

Meat meal (21) cP ˆ 0.278Db‡ 30 796±925 6.4 Whale meat meal (12) cP ˆ 0.216Db‡ 30 842±770 2.0 Fish meal (8) cP ˆ 0.325Db‡ 24 675 2.3 Soy bean (8) or Groundnut meals (6) cP ˆ 0.217Db‡ 28 1020 2.0 Miscellaneous foodsb cP ˆ 0.414Db‡ 12 Ð

a_{n ˆ number of different feed samples.}

b_{Including casein, dried blood protein, egg, brewer's yeast, roller dried milk, and grass meal.}

B. Meat Proteins

Raw beef, chicken, pork (loin), and cod ®llets were analyzed using Orange- G or Amido Black 10B* dye binding by Torten and Whitaker (125). Their procedure was described in Chapter 4. A signi®cant correlation was observed between crude protein values (Kjeldhal-N 6 6.25) and Db (Table

13).

The DBC for raw meat proteins decreased linearly with increasing sample protein (see last column of Table 13). The amount of dye bound depended on the dye/protein ratio. In general, inadequate amounts of Orange-G were used in many early studies. Dye limitations and inadvertent changes in protein/dye ratio for different assays reduced the reliability of dye-binding assays. The regression equation (cP ˆ 0.301Db‡ 8.18) for beef

applies over a restricted range of protein content. The effect of a changing DBC is shown in the simulations reported in Fig. 9. One set of results are computed on the basis that the DBC is ®xed. Where DBC varies the dye/ protein ratio the simulated calibration graphs were nonlinear (Fig. 9). The curves are remarkably like actual calibration curves reported for ground pork and chicken (125). These samples showed a high dependence of DBC on protein content and large deviations from linearity. A linear equation did ®t the data but only over a highly restricted range of protein content. Ground chicken, pork loin, or cod ®llet having greater than 50% crude protein content should probably not be analyzed by Orange-G dye binding. It was on account of the dependence of DBC on protein content that Amido Black 10B was judged unsuitable for meat protein analysis.

TABLE13 Protein Determination in Raw Meat Using Orange G Dye Binding

Sample Regression linea _{R DBC (mg g} 1_cP)

Beef cP ˆ 0.301Db‡ 8.18 0.90 DBC ˆ 209.2 1.135cP

Chicken breast cP ˆ 0.602Db 2.50 0.94 DBC ˆ 265.5 3.721cP

Pork loin cP ˆ 0.367Db‡ 5.45 0.80 DBC ˆ 271.2 4.040cP

Cod ®llet cP ˆ 0.632Db‡ 3.00 0.95 DBC ˆ 246.9 3.397cP

a_{Symbols cP and D}

bare as de®ned previously.

Source: Summarized from Ref. 125.

* As Amido Black 10B was found to be unsuitable for raw meat analysis, the following discussion focuses on results obtained with Orange-G.

FIGURE9 The effect of a changing dye-binding capacity on calibration graphs for

protein analysis in raw meat samples using Orange G binding. Shaded squares show normal response according to regression equations in Table 13. The open circles show simulated graphs for the assay response when DBC changes with sample protein content.

C. Egg, Chicken, and Meat Protein

Egg, chicken, and meat products were analyzed by Ashworth (126) (Table 14). As he was one of the ®rst investigators to apply dye-binding assays to foods, his approach merits attention. Reliable results were obtained provided that the free Acid Orange 12 concentration (after shaking with the protein sample) was kept within the range of 0.4±0.6 mg mL 1_{. To achieve this, the initial the}

dye/protein ratio was kept within a range of 0.64±0.92. Pork had the same DBC as beef, which was lower than the value of chicken. The DBC for mixtures of meat could be deduced from values for individual components. Dye binding was not affected by the presence of fat or by normal cooking (1608C, 40 minutes). It was concluded that dye binding is useful method for composition control in ground meats, eggs, and prepared mixes.

D. Sausage Protein

Seperich and Price (127) determined protein in model sausage emulsions and muscle components (myo®brillar protein, sarcoplasmic protein, and stroma) from which they were produced. The approach was modi®ed from Ref. 128.* These studies con®rmed that protein dye binding was not affected by sausage emulsion fat content from 20 to 40%. The DBC was a function of

TABLE14 Acid Orange 12 DBC of Egg and Meat Products

Meat product DBC (mg g 1_cP)a

Egg (whole) 410±440

Egg albumin (egg white) 390±410

Chicken meat 460±480

Chicken liver 360±390

Beef or pork (ground) 430±440

Beef liver 420±440

Proteose peptone 90±145

Gelatin 310±350

a_{Ranges of values are given for analysis performed in the presence of}

excess of dye concentration of 0.4±0.6 mg mL 1_.

Source: Ref. 126.

* Sausage emulsion samples (3.5 g) were homogenized with 51 mL of citrate (0.2 M)±phosphate (0.1 M) buffer (pH 5.5). Ten milliliters of the resulting homogenate was retained for Kjeldahl analysis. The remainder was shaken with 80 mL of Acid Orange 12 (0.56±3.64 mM;0.2±

1.27 mg mL 1_{) in a 250-mL centrifuge tube for 30 minutes and then centrifuged (5.680g; 5}

dye/protein ratio. At the highest dye concentration examined the DBC was of the order of 400 mg g 1 _{(cP), in line with values reported by other}

investigators. However, DBC decreased to about 33±34 mg g 1_{(cP) at a dye}

concentration of 0.2 mg mL 1_.

6.6. Mushrooms

Nine strains of Agaricus bisporus (Lange) Imbach were analyzed by Weaver et al. (128) using dye binding, Kjeldahl, and quantitative amino acid analysis.*. The average protein content was 29.4 (+ 6.2)% by Kjeldahl analysis, 22.4 (+ 2.4)% by dye binding, and 28% (+ 3.4)% by amino acid analysis. Per wet weight basis, Agaricus had 2.6±2.8% protein. Quantitative amino acid analysis was more correlated with dye-binding analysis (R ˆ 0.74) than Kjeldahl analysis (R ˆ 0.4). Mushrooms are thought to contain high amounts of NPN, which could lead to errors in Kjeldahl analysis. Braaksma and Schaap (129) reported the protein content for Agaricus as 0.5% fresh weight or 7% per dry weight basis (Chapter 7). REFERENCES

1. A Grollman. The combination of phenol red and proteins. J Biol Chem 64:141±160, 1925.

2. WW Smith, HW Smith. Protein binding of phenol red, Diodrast, and other substances in plasma. J Biol Chem 124:107±113, 1938.

3. HW Robinson, CG Hogden. The in¯uence of serum protein on the spectrophotometric absorption curve of phenol red in a phosphate buffer mixture. J Biol Chem 137:239±254, 1941.

4. LM Chapman, DM Greenberg, CLA Schmidt. Studies of the combination between certain acid dyes and proteins. J Biol Chem 72:707±729, 1927. 5. LMC Rawlings, CLA Schmidt. Studies on the combination of certain basic

dyes and proteins. J Biol Chem 82:709±716, 1929.

6. LMC Rawlings, CLA Schmidt. The mode of combination between certain dyes and gelatin granules. J Biol Chem 83:271±284, 1930.

7. H Fraenkel-Conrat, M Cooper. The use of dyes for the determination of acid and basic groups in proteins. J Biol Chem 154:239±340, 1944.

In document ISBN: La Editorial Tecnocientífica Americana se encuentra indizada o referenciada en las siguientes bases de datos: (página 17-23)