MATERIALS AND METHODS

Sudanese sorghum cultivar (feterita) was obtained from the local market, Khartoum north, Sudan while the Indian cultivar (CSH5) was provided by department of grain science and technology , CFTRI , Mysore, India. The seeds were cleaned and freed from foreign material and broken kernels. The clean seeds were milled in Barabeder Quadrumat Junior Mill (Regulation No 1) into flour to pass a 0.4 mm screen. The flour was stored in polyethylene bags at 4?C for further analysis. Unless otherwise stated, all reagents used in this study are of lab-grade

Cooking: Cooking of samples was done according to method followed by El Tinay et al. (1979).

Proximate analysis: The determination of moisture, crude fibre, crude fat and ash were carried out according to AOAC (1984) methods.

Determination of total minerals: Minerals were extracted from the samples by dry ashing method that was described by Chapman and Pratt (1982). The amount of iron, Ca and Cu were determined using atomic absorption spectroscopy (Perkin-Elmer 2380). Ammonium vandate was used to determine phosphorous along with ammonium molybdate method of Chapmann and Pratt (1982). Sodium and potassium contents were determined by flame photometer (CORNIGEEL) according to AOAC (1984).

Determination of tannin content: Quantitative estimation of tannins was carried out using the modified vanillin – HCl method according to Price et al. (1978).

Determination of in vitro protein digestibility: In vitro protein digestibility was carried out according to Saunder et al. (1973) method.

Isolation and characterization of phenolic acids
Isolation and characterization of free phenolic acids: Free phenolic acids were isolated according to the method of Ayumi et al. (1999). Two grams of flours were extracted with 70% ethanol (4 x 50 mL., 1 h each); the supernatants were obtained by centrifugation and concentrated, and the pH was adjusted in the range of 2 - 3 with 4 M HCl. Phenolic acids were separated by ethyl acetate phase separation (5 x 50 mL), and the pooled fractions were treated with anhydrous disodium sulfate to remove moisture, filtered and evaporated to dryness. Phenolic acids taken in methanol were estimated colorimetrically by using Folin-Cioclateu method with gallic acid as the reference standard (Kaluza et al. 1980) as well as by HPLC (model LC-10A, Shimadzu), on a reversed phase Shimpak C₁₈ column (4.6 x 250 mm), using a diode array detector (operating at 280 nm). A solvent system consisting of water/acetic acid/methanol (isocratic, 80:5:15) was used as mobile phase at a flow rate of 1 mL/min (Wulf and Nagel, 1967). Standfards such as caffeic, coumaric, ferulic, gallic, gentisic, protocatechuic, syringic, and vanillic acid were used for identification and quantification of phenolic acids present in the flour samples. Quantification of phenolic acids present in the sample was carried out by measuring the area under respective peaks and plotting against a standard graph prepared (2-10 μg) for each individual phenolic acid using the above-mentioned standards.

Isolation and characterization of bound phenolic acids: Bound phenolic acids were extracted according to the method of Erk-Nordkvist et al. (1984). Sorghum flour of two cultivar (2 g each) were extracted with 70% ethanol (4 x 50 mL) and hexane (4 x 50 mL) to remove free phenolic acids and fat, respectively. The dried samples were extracted with 1 M sodium hydroxide (2 x 100 mL, 2 h each) containing 0.5% sodium borohydride under nitrogen atmosphere, and the clear supernatants were collected followed by centrifugation. The combined supernatants were acidified with 4 M HCl to pH 1.5, and the phenolic acids were processed and analyzed by colorimetry as well as by HPLC as mentioned in the case of free phenolic acids.

Statistical analysis: Each determination was carried out on three separate samples and analyzed in triplicate and figures were then averaged. Data was assessed by the Analysis of Variance (ANOVA) (Snedecor and Cochran, 1987). Duncan Multiple Range Test (DMRT, 1955) was used to separate means. Significance was accepted at P<0.05.

RESULTS AND DISCUSSION

Chemical composition of sorghum: Table 1 shows the results of the proximate composition of Sudanese (feterita) and Indian (CSH5) sorghum cultivars. Data are expressed on dry matter basis (per 100 gm material). The moisture content of Sudanese and Indian cultivars was assessed as 7.49 and 6.77% respectively. These values are comparable to the range of 5.7 to 10% reported by Yousif and Magboul (1972), but significantly lower than the range of 8.89 to 9.88 stated by Arbab (1995) may be due to climatic or location differences. Results show that Sudanese and Indian sorghum cultivars contain ash 2.29 and 1.54% respectively. The value are within the range of 1.5 to 2.6%, 1.4-1.8%, 1.5 - 3.9% reported by Awad El Kareem (2002); Abdel Rahman (2002); Hassan (1995), respectively. The crude protein content of two sorghum cultivars Sudanese and Indian is given in Table 1. Results, however, showed values of 14.0 and 10.02% respectively. The protein content of Sudanese cultivar is significantly higher than Indian cultivars. The values are within the range of 8.61 to 18.21% reported by Sastry et al. (1968). The protein content of Sudanese cultivar (feterita) is higher than the value stated by Awad El Kareem (2002) who reported the protein content of feterita was 13.13.The crude fibre analysis for the two sorghum cultivars Sudanese (feterita) and Indian (CSH5) showed the values of 1.65 and 1.72% respectively. The fibre content of Indian variety was significantly higher than Sudanese cultivar. Results obtained were found to be within the range of 1.2 to 1.9% and 1.4 to 2% reported by El Tinay et al. (1979) and Abdel Rahman (2002) respectively. However, the crude fibre content of the two cultivars(1.65 and 1.72%) were significantly lower (P< 0.05) than those reported by Hassan (1995) who stated the fiber content of sorghum cultivars ranged between 1.8% and 2.17%. The fat content is 3.12% (for Sudanese) and 2.84% (for Indian). The fat content of Sudanese cultivar is significantly (P< 0.05) higher than Indian cultivar. The fat content of Sudanese cultivar was in the range reported by Awad El Kareem (2002) who stated the fat content of two Sudanese sorghum cultivar (Dabar and Fetarita) ranged between 3.1 and 3.8%, while the fat content of Indian cultivar within the range reported by Shepherd et al. (1970) who stated that the fat content of sorghum cultivars ranged from 1.5 to 2.5%.The results of carbohydrates content are viewed in Table 1 as 71.46 and 77.11% for Sudanese and Indian cultivars, respectively. The carbohydrates content of Indian cultivar was significantly (P<0.05) higher than Sudanese cultivar. The results obtained were in the range reported by Osman (2004) who recorded carbohydrates content of three Sudanese local cultivars (Tabat, mugud and feterita) to be ranging between 71.33 and 78.78%.

Mineral content: The mineral content of Sudanese and Indian sorghum cultivars are showen in Table 2. Total sodium content of Sudanese and Indian sorghum cultivars was 6.18 and 5.83 mg/100g respectively. Sodium content of Sudanese cultivar agrees with the result stated by Badi (2004) who reported that sodium content of two sorghum cultivars ranged from 6.3 to 7.0 mg/100g and both of the cultivars showed lower results than the results recorded by Khalil et al. (1984). Total potassium for both Sudanese and Indian were 225.23 and 367.51 mg/100g, respectively. Indian cultivar contains much amount of potassium compared to Sudanese. Potassium content of Sudanese and Indian is lower than 441.7 and 450 mg/100g reported by Badi (2004) and 430-458 mg/100g reported by Khalil et al. (1984), while potassium content of Indian cultivar within the range of 363-901 mg/100g stated by Deosthale and Belvady (1978). Total calcium content of Sudanese and Indian cultivars were 2.43 and 3.33 mg/100g respectively. The results obtained for both cultivars were less than results recorded by Badi (2004) and Khalil et al. (1984) who recorded 10.8 and 18 mg/100g, respectively. Total iron contents for Sudanese and Indian cultivars were 15.54 and 11.32 mg/100g respectively. Results obtained from two cultivars were higher than results reported by Badi (2004) who reported iron content of the Sudanese sorghum cultivars (Wad Ahmed and Tabat) as 3.8 and 4.5 mg/100g, respectively, while the results were in agreement with results stated by

Table 1:	Proximate composition of African and Asian sorghum cultivars
Values are means (±SD) of 3 replicates per treatment abMeans with different superscripts in the same row were significantly different (P<0.05)

Table 2:	Minerals content (mg/100g) of Sudanese and Indian sorghum cultivars
Values are means (±SD) of 3 replicates per treatment. ab Means with different superscripts in the same row were significantly different (P<0.05).

Table 3:	Tannins and Invitro protein digestibility (IVPD) of Sudanese and Indian sorghum cultivars
Values are means (±SD) of 3 replicates per treatment. ab Means with different superscripts in the same column were significantly different (P<0.05).

Deosthale and Belvady (1978) who reported the iron content of sorghum cultivars to be ranging from 4.70 to 14.05 mg/100g. Total phosphorus content of two sorghum cultivars (Sudanese and Indian ) were 263.30 and 314.15 mg/100g respectively, which are less than 407 and 396 mg/100g reported by Khalil et al. (1984) and 388 to 756 mg/100g stated by Deosthale and Belvady (1978). Phosphorus content of Indian cultivar is significantly (P< 0.05) higher than Sudanese cultivar. Total copper content for two sorghum cultivars Sudanese and Indian were 0.41 and 0.32 mg/100g, respectively. Results obtained were in agreement with Deosthale and Belvady (1978) who reported the copper content of sorghum cultivars to range from 0.39 to 1.58 mg/100g.

In vitro protein digestibility of sorghum: Table 3 showed the in vitro protein digestibility of Sudanese and Indian cultivars as 49.25 and 55.85% for uncooked sample, while in vitro protein digestibility was 26.11 and 33.11% for Sudanese and Indian cultivar, respectively. Indian cultivar had significantly (P< 0.05) higher in vitro protein digestibility for cooked and uncooked compared to Sudanese. These results obtained agree with Chibber et al. (1980) who reported that, uncooked and cooked high tannin sorghum varieties both shown to have low in vitro protein digestibility. The lowest in vitro protein digestibilities obtained in case of Sudanese cultivar positively correlated to its tannin content, this finding agrees with Chavan et al. (1979) who observed significant lowering in in vitro protein digestibility (IVPD) in a high tannin cultivar. The IVPD of high tannin grain was found to be improved to the level of low tannin grains upon dehulling the grain. The minimum level of tannin requires to show the growth depressing effects in rats was 0.64 to 0.84% sorghum tannin (Fuller et al., 1996). The tannin content of white and yellow grain cultivar grown in India for human consumption is below this level. Hence, the tannins present in Indian cultivars may not pose a significant problem of protein digestibility. Many efforts should be done to improve in vitro protein digestibility including sodium hydroxide, potassium hydroxide and sodium carbonate treatments which significantly increased in vitro protein digestibility from 48 to 69%, 69.1 and 72% respectively (Chavan et al., 1979). In vitro protein digestibility of both cooked cultivars (Sudanese and Indian) is significantly decreased, these findings agree with Duodo et al. (2002) who reported that cooking significantly decrease in vitro protein digestibility of sorghum and maize cultivar. Results obtained also agreed with Mertz et al. (1984) and Maclean et al. (1981) who reported effect of cooking on protein digestibility at the three level of organizations, cooking caused a significant reduction in protein digestibility for both sorghum varieties (high and low tannins). Results obtained from the two cultivars Sudanese (high tannin sorghum) and Indian (low tannin sorghum) indicate that tannin content is not the only responsible factor for lowering in vitro protein digestibility and may be many other factors had a role in this process. An old hypothesize that cooling cooked porridge leads to formation of resistant starch which may form complexes with kafirin (major fraction) proteins that are less susceptible to enzyme attack (Bach Knudsen and Munk, 1985).

Table 4:	Phenolic acid (%) in Indian and Sudanese sorghum cultivar (identified and quantified by HPLC)
FPA = Free phenolic acid. BPA = Bound phenolic acid. ND = Not detected

The results obtained for Sudanese (high tannin) and Indian cultivars (low tannin content) were significantly (P< 0.05) lower than the result stated by Ibrahim (2004) who analyzed two Sudanese sorghum cultivars [Wad Ahmed (high tannin) and Dabar (low tannin)] and recorded 47.9 and 53%, respectively. Fermentation of sorghum cultivar (high and low tannin) significantly decreased in vitro protein digestibility by 63% (low tannin) and increased (IVDP) by 17.5% of high tannin sorghum cultivar (Romo Parad et al., 1985). Ibrahim (2004) stated that supplementation of high tannin (Wad Ahmed) and low tannin (Dabar) sorghum cultivar with 5 and 10% whey protein significantly increased protein digestibility. For Sudanese cultivar (Feterita) may be supplementation with highly rich protein concentrate or isolate of legumes increased in vitro protein digestibility. In vitro protein digestibility of Sudanese cooked samples was higher than the value reported by Arbab and El Tinay (1997) who reported the in vitro protein digestibility of cooked sorghum (12 and 18%) will treatment with reducing agent. The results obtained for Sudanese and Indian were higher than the results stated by Arbab and El Tinay (1997) for uncooked samples. It is interesting that the use of a reducing agent during cooking doesn’t appear to completely reverse the effect of lowered sorghum protein digestibility on cooking. Cooking sorghum flour with a reducing agent did improve protein digestibility but not to the level of uncooked sorghum flour (Oria et al., 1995). Similar results have been reported from SDS-PAGE analysis of pepsin indigestible residues of sorghum protein body preparation (Duodo et al., 2002). During cooking, more of such disulphide cross-linked protein oligomers and polymers are formed. When sorghum is cooked, enzymatically resistant protein polymers are formed through disulphide bonding of the β-kafirin and δ-kafirin (Oria et al., 1995). Finally, the causes of the poor digestibility of sorghum proteins appear to be multi-factorial. Depending on the nature of the sorghum used (whole grain, endosperm protein body preparation, high tannin grain or condensed – tannin – free grain), different factors may contribute with some being more important than others. The probable causes of reduced protein digestibility indicate that a number of currently used processing technologies may be applied to improve sorghum protein digestibility. These processing technologies including dry cooking "popping”, extrusion, malting, fermentation and grain refinement (Duodu et al., 2001; Hamaker et al., 1994; Elkhalifa and Chandrashekar, 1999; Taylor and Taylor, 2002 and Duodo et al., 2002).

Tannin content of sorghum cultivars: The tannin content of Sudanese and Indian cultivar showed in Table 3, as 1.19 and 0.08% as catechin equivalent respectively. Sudanese cultivar had significantly (P< 0.0) higher tannin content compared to Indian cultivar. Results obtained for two cultivars agree with Jambunthan and Mertz (1973) who reported that tannin content of high tannin sorghum is 2.69% and for low tannin sorghum is 0.5%. Result findings agreed with Radhakrishnan and Sivaprasad (1980) who reported arrange of 0.01 to 2.056% tannin as catchin equivalents in the sorghum grown in India. The tannin content of cultivars commonly grown and consumed in India ranged from 0.43 to 0.64% tannic acid equivalents. For Sudanese sample (high tannin content) several investigators have implicated the high tannin character of certain sorghum genotypes to their ability to resist bird damage. It is observed that bird resistance is offered only during the milk stage of grain development, where the tannin content is higher. As the grain matures, the tannin content is reduced and the seeds become palatable (Price et al., 1967). The usefulness and practicability of high tannin cultivars to save the yield from bird damage in the field and subsequent processing of grain to remove tannins before consumption needs to be evaluated. Results obtained were comparable to result stated by Chavan et al. (1979) who analyzed two low tannin sorghum cultivars and two high tannin sorghum and recorded the tannin content for low tannin cultivars range from 0.40 to 0.46% (catechin equivalent) and for higher tannin sorghum ranged from 3.44 to 3.60% (as catechin equivalent).

Phenolic acid content: The phenolic acids (PAs) (free and bound) content separated and identified by high performance liquid chromatography (HPLC) of the two cultivars are shown in Table 4. Free phenolic acids of Indian cultivar detected were 13.40% syringic acid, 69.30% Coumaric and 16.0% ferulic acid while free phenolic acids of Sudanese cultivar were callic acid (30%), 6.8% protocatechuic acid, 17.30 gentisic acid, 38.5% caffeic acid, 3.90 p-comaric acid and 3.0% ferulic acid. Gallic, protocatechuic, gentisic, and p-coumaric were not detected as free phenolic acids for Indian cultivar while syringic acid was not detected in Sudanese cultivars. Indian cultivar contained high caffeic (69.30%) and ferulic (16.00%) compared to Sudanese which contained 38.80 and 3.00% respectively. Syringic acid content of Indian cultivar was 13.40% .The bound phenolic acid figures of Indian cultivar were 1.90, 1.70, 7.00 and 65.0% as syringic, caffeic, P-coumaric and ferulic acids, respectively. Gallic, protocatechuic and gentisic acids were not detected in free or bound form while p-coumaric was detected in bound form of Indian cultivar. Bound phenolic acids values of Sudanese cultivar were 8.5, 11.80, 1.40, 1.10, 6.70 and 45.0%, as gallic, protocatechuic, syringic, caffeic, p-coumaric and ferulic acids, respectively. Syringic, caffeic, p-coumaric and ferulic acids content in bound form of Indian cultivar were higher than Sudanese cultivar. The results obtained from two cultivars are comparable with Hahn et al. (1983) and Adam and Liu (2002) who reported that, the phenolic acids exist mostly in bound form (esterifies to cell wall polymers), with ferulic acid as being the most abundant bound phenolic acid in sorghum and other cereals. Findings obtained agreed with Hahn et al. (1984); Waniska et al. (1989) who reported that, several other PAs have been identified in sorghum including syringic (15C), protocatechuic (16C), caffeic (17C), p-coumaric (20C) as more abundant. The PAs like other phenols are though to help in plant defense against pests and pathogens. The PAs show good anti-oxidant activity (Subba Rao and Muralikrishna, 2002), and thus may contribute significantly to the health benefits associated with whole grain consumption. From these findings in spite of Sudanese cultivar demonstrated as high tannin sorghum cultivar and Indian as low tannin cultivar, phenolic acid content of syringic, cuffeic, p-coumanic and ferulic for Indian cultivar were higher than Sudanese. These observations agreed with Waniska et al. (1989) who reported that, in sorghum the level of PAs do not correlate with the presence of level of other phenol (anthocyanin or tannin). Waniska et al. (1989) observed increased level of free PAs in certain sorghums with pigmented testa (containing tannin) compared to the ones without pigmented testa. In general sorghums have level of PAs comparable to those of other cereals (Andersen et al., 2001; Adam and Liu, 2002; Hahn, 1983, 1984) significant varietals differences are observed in phenolic acid composition and ratios of bound and free forms of these compounds in sorghum. Adam and Liu (2002) found a strong correlation between anti-oxidant activity and levels of bound ferulic acid in wheat, corn, rice and oats.