The second major component of sorghum and millet grains is protein. Both genetic and environmental factors affect the protein content of sorghum and millets. In sorghum the variability is large, probably because the crop is grown under diverse agroclimatic conditions which affect the grain composition (Burleson, Cowley and Otey, 1956; Waggle, Deyoe and Smith 1967; Deosthale, Nagarajan and Visweswar Rao, 1972). Fluctuations in the protein content of the grain are generally accompanied by changes in the amino acid composition of the protein (Waggle and Deyoe, 1966).
The quality of a protein is primarily a function of its essential amino acid composition. To assess the protein quality, Block and Mitchell (1946) introduced the concept of an amino acid or chemical score, in which the amount of the essential amino acid that is in greatest deficit is expressed as a percentage of the amount present in a standard or reference protein. Egg and human milk proteins, for their very high biological value, have been considered as reference standards. Sorghum and millet proteins differed in their essential amino acid profile (Table 21). However, the most common feature was that Iysine was always found to be the most limiting amino acid. The highest deficit of Iysine was in the protein of barnyard millet (chemical score 31), closely followed by little millet (chemical score 33). Sorghum protein, with a chemical score of 37, did not differ very much in quality from the proteins of barnyard and little millet.
The primary function of dietary protein is to satisfy the body's needs for nitrogen and essential amino acids. According to the World Health Organization (1985), the chemical score of a protein if calculated in relation to the essential amino acid requirement pattern as reference would be more realistic and indicative of the capacity of the protein to meet human requirements. Such data on chemical score relative to amino acid requirement for different age groups (Table 22) suggested that the inherent capacity of the existing varieties commonly consumed was not adequate to meet the growth requirements of infants and young children, though all of them except sorghum may be able to meet the maintenance requirements in adults.
Grain proteins are broadly classified into four fractions according to their solubility characteristics: albumin (water soluble), globulin (soluble in dilute salt solution), prolamin (soluble in alcohol) and glutelin (extractable in dilute alkali or acid solutions). In solubility fractionation studies with sorghum and pearl, finger and foxtail millets, five protein fractions were obtained (Table 23). The levels of albumin plus globulin were higher in pearl millet varieties than in sorghum, while amounts of the cross-linked prolamin, ß-prolamin, were higher in sorghum than in pearl millet.
TABLE 21: Essential amino acid composition (mg/g) and chemical score of sorghum and millet proteins
Grain | Isoleucine | Leucine | Lysine | Methi- onine |
Cystine | Pheny- lalanina | Tyrosine | Threonine | Tryptophan | Valine | Chemical score |
Sorghum | 245 | 832 | 126 | 87 | 94 | 306 | 167 | 189 | 63 | 313 | 37 |
Pearl millet | 256 | 598 | 214 | 154 | 148 | 301 | 203 | 241 | 122 | 345 | 63 |
Finger millet | 275 | 594 | 181 | 194 | 163 | 325 | - | 263 | 191 | 413 | 52 |
Foxtail millet | 475 | 1 044 | 138 | 175 | - | 419 | - | 194 | 61 | 431 | 41 |
Common millet | 405 | 762 | 189 | 160 | - | 307 | - | 147 | 49 | 407 | 56 |
Little millet | 416 | 679 | 114 | 142 | - | 297 | - | 212 | 35 | 379 | 33 |
Barnyard millet | 288 | 725 | 106 | 133 | 175 | 362 | 150 | 231 | 63 | 388 | 31 |
Kodo millet | 188 | 419 | 188 | 94 | - | 375 | 213 | 194 | 38 | 238 | 55 |
Sources: FAO. 1970a; Indira and Naik. 1971.
TABLE 22: Lysine amino acid scores for different age groups based on the 1985 WHO index
Grain | Infant
( <1 year) |
Preschool child (2 - 5 years) | Schoolchile
(10 - 12 years) |
Adult |
Wheat | 43 | 46 | 62 | 100+ |
Rice (hushed) | 57 | 61 | 82 | 100+ |
Maize | 41 | 43 | 58 | 100+ |
Sorghum | 17-51 | 18-55 | 25-74 | 71-100+ |
Pearl millet | 26-69 | 28-74 | 38-100+ | 100+ |
Foxtail millet | 28-38 | 30-40 | 40-55 | 100+ |
Finger millet | 39-63 | 41-68 | 56-91 | 100+ |
Kodo millet | 46-52 | 48-55 | 65-74 | 100+ |
Barnyard millet | 26 | 27 | 37 | 100+ |
Proso millet | 23-72 | 24-74 | 32-98 | 93-100+ |
Sources: WHO, 1985; Hulse, Laing and Pearson, 1980.
Apart from a favourable essential amino acid profile, easy digestibility is an important attribute of a good-quality protein. Chemical score does not take into account the digestibility of protein or availability of amino acids. Biological methods based on measurement of growth and nitrogen retention assess the overall nutritional quality of the protein. These methods include determination of protein efficiency ratio (PER), net protein utilization (NPU), biological value (BV) and true protein digestibility (TDP).
Sorghum
Wide variability has been observed in the essential amino acid composition of sorghum protein (Hulse, Laing and Pearson, 1980; Jambunathan, Singh and Subramanian, 1984). Lysine content was reported to vary from 71 to 212 mg per gram of nitrogen and the corresponding chemical score varied from 21 to 62.
Singh and Axtell ( I 973a) identified two high-lysine Ethiopian sorghum varieties, IS 11758 and IS 11167. The average lysine content of the whole kernel of IS 11758 was 3.13 g per 100 g protein and the total protein content of the kernel was 17.2 percent. IS 11167 contained 3.33 g lysine per 100 g protein and 15.7 percent protein. Normal sorghum grown under similar conditions contained 12 percent protein and 2.1 g lysine per 100 g protein. Feeding trials in rats have shown higher PER values for high-lysine varieties (1.78 and 2.05 for IS 11758 and IS 11167, respectively) than for normal sorghum (PER 0.74 and 1.24).
TABLE 23: Distribution of protein fractions in sorghum and millet grains (percentage of total protein)
Fraction | Sorghum |
Pearl millet |
Finger millet |
Foxtail millet |
||||
Range | Mean | Range | Mean | Range | Mean | Range | Mean | |
Albumin+globulin | 17.1-17.8 | 17.4 | 22.6-26.6 | 25.0 | 17.3-27.6 | 22.4 | 11.6-29.6 | 17.1 |
Prolamin | 5.2-8.4 | 6.4 | 22.8-31.7 | 28.4 | 24.6-36.2 | 32.3 | 47.6-63.4 | 56.1 |
Cross-linked prolamin | 18.2-19 5 | 18.8 | 1.8-3.4 | 2.7 | 2.5-3.3 | 2.78 | 6.4-17.6 | 8.9 |
Glutelin-like | 3.4-4.4 | 4.0 | 4.7-7.2 | 5.5 | - | - | 5.2-11.9 | 9.2 |
Glutelin | 33.7-38.3 | 35.7 | 16.4-19.2 | 18.4 | 12.4-28.2 | 21.2 | - | 6.7 |
Residue | 10.4-10.7 | 10.6 | 3.3-5.1 | 3.9 | 16.1-25.3 | 21.3 | - | 2.0 |
Total | 91.2-94.0 | 92.9 | 78.6-87.5 | 83.9 | 74.7-83.9 | 78.7 | - | 98.0 |
Sources: Jambunathan,Singh and Subramanian. 1984 ( sorghum and pearl millet); Virupaksha. Ramachandra and Nagarajut. 1975 ( finger millet ): Monteiro. Virupaksha and Rajagopol Rao. 1982 (foxtail millet).
Another high-lysine mutant, P721, was reported to have 60 percent more lysine than normal sorghum. Van Scoyoc, Ejeta and Axtell (1988) have demonstrated that the high lysine of P721 resulted primarily from unusually high amounts of Iysine-rich glutelin and low Iysine-poor prolamin.
Ejeta and Axtell (1987) observed that in all three of these high-lysine sorghum varieties the Iysine content of the germ was normal but the Iysine content of the endosperm was higher than in normal sorghum.
Naik (1968), using a modified extraction procedure, observed wide variations in the distribution pattern of protein fractions in the sorghum varieties. Albumin ranged from 2 to 9 percent of total protein, while globulin ranged from 12.9 to 16 percent, prolamin from 27 to 43.1 percent and glutelin from 26.1 to 39.6 percent. Seasonal differences in the distribution pattern of protein fractions were reported (Virupaksha and Sastry, 1968): sorghum varieties grown in the Rabi (dry) season had less prolamin than when grown in other seasons.
Studies on amino acid composition of the protein tractions (Ahuja, Singh and Naik, 1970) showed that the albumin and globulin fractions contained high amounts of Iysine and tryptophan and in general were well balanced in their essential amino acid composition. On the other hand, the prolamin fraction was extremely poor in Iysine, arginine, histidine and tryptophan and contained high amounts of proline, glutamic acid and leucine. Present in the form of protein bodies, prolamin was found to be a predominant protein fraction directly associated with the protein content of the grain. Glutelin, the second major protein traction, is a structural component, the protein matrix in the peripheral and inner endosperm of the sorghum kernel.
Both in vitro and in vivo studies have demonstrated wide variability in protein digestibility of sorghum varieties (Axtell et al., 1981). Values ranging from 49.5 to 70 percent (Nawar et al., 1970) and from 30 to 70 percent (Silano, 1977) have been reported. Elmalik et al. (1986) observed that in rats digestibility of protein of sorghum varieties with intermediate and corneous endosperm texture was 70.3 and 74.5 percent, respectively. These values were lower than that observed for corn protein (78.5 percent). In certain sorghum varieties the presence of condensed polyphenols or tannins in the grains is another factor that adversely affects protein digestibility and amino acid availability (Bach Knudsen et al., 1988; Bach Knudsen, Munck and Eggum, 1988; Whitaker and Tanner, 1989).
In tannin-free sorghum varieties, Sikabbubba (1989) observed that the protein digestibility was inversely correlated with total protein in the grain (r = -0.548, p < 0.1), total prolamin (r = -0.627, p < 0.25), cross-linked or ß-prolamin (r = 0.647, p < 0.05) and digestibility of ß-prolamin (r = -0.727, p < 0.01). In studies in boys aged 10 to 11 years (Kurien et al., 1960), progressive substitution of sorghum for rice in a predominantly vegetarian diet resulted in progressive decrease in protein digestibility from 75 to 55 percent and in apparent nitrogen retention from 4.5 to 2.1 percent. Similar observations were also made in 10- to 11-year-old girls fed sorghum proteins. In nitrogen balance studies conducted with 6- to 30-month-old children recovering from protein energy malnutrition, MacLean et al. (1981) observed that for whole-grain gruels prepared from four sorghum varieties including two high-lysine varieties, P721 opaque and IS 11758, the average protein digestibility was 46 percent. The protein digestibility of sorghum grain was thus found to be extremely poor as compared to that previously observed for wheat (81 percent), maize (73 percent) and rice (66 percent). However, in a study with decorticated and extruded sorghum product fed to young children (Maclean et al., 1983), the protein digestibility, 81 percent, was much higher than for the whole grain (46 percent). Nitrogen retention, which had been 14 percent in the whole-grain study, was also enhanced, to 21 percent. In vitro studies conducted on extruded sorghum (Mertz et al., 1984) also showed that extrusion processing of sorghum grain improved the protein digestibility and hence the nutritive value. Digestibility of sorghum protein was also improved after processing of the grain into nasha, a thin fermented porridge used as baby food in the Sudan (Graham et al., 1986). Nitrogen retention was improved in normal Nigerian men fed home-pounded and winnowed sorghum with reduced fibre content (Nicol and Phillips, 1978). These observations emphasize the importance of grain processing to improve the nutritive value of sorghum. A decrease in the protein digestibility of sorghum on cooking was attributed to reduced solubility of prolamin and its reduced digestibility by pepsin (Hamaker et al., 1986).
Pearl millet
Pearl millet, like sorghum, is generally 9 to 13 percent protein, but large variations in protein content, from 6 to 21 percent, have been observed (SernaSaldivar, McDonough and Rooney, 1991). Lysine is the first limiting amino acid of pearl millet protein. A significant inverse correlation has been reported between the level of protein in the grain and the Iysine content of the protein (Deosthale et al.,1971). In high-protein varieties of pearl millet with protein content ranging from 14.4 to 27.1 percent, significant inverse correlations have also been observed between protein and threonine, methionine and tryptophan. The essential amino acid profile shows more Iysine, threonine, methionine and cystine in pearl millet protein than in proteins of sorghum and other millets. Its tryptophan content is also higher (Table 21).
Wide variation is observed in the lysine content of pearl millet protein, with values ranging from 1.59 to 3.8 g per 100 g protein. From chemical scores calculated in relation to amino acid requirements for different age groups it was apparent that pearl millet has greater potential to meet the lysine requirements of growing children than most other cereals (Table 22). Pushpamma, Parrish and Deyoe (1972) observed in rat feeding trials a PER of 1.84 for pearl millet as against 1.74 for finger millet, 1.46 for sorghum and 1.36 for maize. This has supported the view that the protein quality of pearl millet ranks quite high in comparison with that of other cereals. On fortification of a pearl millet diet with 0.3 percent lysine hydrochloride, the growth response of rats was enhanced and nearly equalled that of controls fed a casein diet (Howe and Gilfillan, 1970).
Protein quality is associated with the distribution pattern of protein fractions in the grain. Sawhney and Naik (1969) observed large variability in the protein fractions of pearl millet varieties. Albumin ranged from 6.1 to 26.5 percent (mean 15.1 percent), globulin from 3.5 to 14.7 percent (mean 8.7 percent), prolamin from 21.3 to 38.0 percent (mean 30.2 percent) and glutelin from 23.8 to 37.7 percent (mean 30.3 percent). As in other cereals, albumin and globulin are rich in lysine as well as the other basic amino acids arginine and histidine. The globulin fraction appeared to be very rich in sulphur amino acids. The prolamin fraction is characterized by high glutamic acid proline and leucine and was also shown to be rich in tryptophan, whereas glutelin was found to contain more lysine and less tryptophan.
True protein digestibility in rats fed pearl millet varied little, from 94 to 97 percent ( Singh et al., 1987), and it was not affected by the protein content of the grain (Table 24). The digestible energy content was lower in high-protein types because of their high prolamin content. In high-protein genotypes, the lysine content of the protein was low and this was reflected in low biological value and low net protein utilization. But the net utilizable protein (percent protein x NPU) from the high-protein genotypes was two to three times higher than that from normal millets. Rats fed raw pearl millet flour exhibited higher digestibility of protein and energy than rats fed raw wheat flour (Dassenko, 1980). However, the digestibility and PER were lower when the millet was fed as chapatti, probably because the longer cooking time required for millet chapatti resulted in heat damage to the protein. In nitrogen balance studies in 11 - to 12-year-old boys, the apparent protein digestibility of a pearl millet-based diet was 52.9 percent and the nitrogen balance was positive (Kurien, Swaminathan and Subrahmanyan, 1961).
TABLE 24: Protein quality and digestible energy in dehulled millets (%)
Grain | True digestibility | Biological value | Net protein utilization | Digestible energy |
Pearl millet (low protein) | 95.9 | 65.6 | 62.9 | 89.9 |
Pearl millet (high protein) | 94.6 | 58.8 | 55.7 | 85.3 |
Foxtail millet | 95.0 | 48.4 | 46.3 | 96.1 |
Common millet | 99.3 | 52.4 | 52.0 | 96.6 |
Little millet | 97.7 | 53.0 | 51.8 | 96.1 |
Barnyard millet | 95.3 | 54.8 | 52.2 | 95.6 |
Kodo millet | 96.6 | 56.5 | 54 5 | 95.7 |
Sources:: Singh et al . 1987 (pearl millet); Geervani and Eggum,1989 (other millets).
Finger millet
Finger millet is poor in protein content compared with other common cereals (Table 17). Wide variability in the composition of the grain, including its protein content, was reported (Hulse, Laing and Pearson, 1980). Both genetic and environmental factors appear to have an important role in determining the protein content of finger millet (Pore and Magar, 1977; Virupaksha, Ramachandra and Nagaraju, 1975). Prolamin is the major protein fraction in finger millet (Table 23). The high protein of white-grain varieties was attributed to the higher prolamin content of the grain, while the lysine content and hence the protein quality of these varieties are low (Virupaksha, Ramachandra and Nagaraju, 1975). Differences in amino acid composition in different varieties of finger millet are large, and as in other cereals both the lysine content and the methionine content of the protein are inversely correlated with the protein content of the grain. The protein tractions also showed wide variation in their amino acid composition. While the albumin and globulin fraction was found to contain a good complement of essential amino acids, the prolamin fraction contained higher proportions of glutamic acid, proline, valine, isoleucine, leucine and phenylalanine but low lysine arginine and glycine. The amino acid composition of prolamin was almost the same as that of endosperm protein.
In vitro studies showed that proteins of finger millet and kodo millet were resistant to pepsin digestion unless the millet was first cooked in an autoclave for 15 minutes or boiled for at least two hours in water. Digestibility of the protein was found to be adversely affected by tannin in the grain, which was as high as 3.42 percent in some of the finger millet varieties studied (Ramachandra, Virupaksha and Shadaksharaswamy, 1977). A finger millet diet was found to be adequate to maintain a positive nitrogen balance in adults (Subrahmanyan et al., 1955). The subjects also showed positive calcium and phosphorus balances, and the digestibility of finger millet protein was found to be 50 percent. Supplementation of a finger millet protein diet with lysine or with leaf protein in addition to lysine significantly improved nitrogen retention in young boys. They also showed greater increase in height and weight (Doraiswamy, Singh and Daniel, 1969). The use of finger millet in child and infant feeding, however, appeared to be limited because of its poor digestibility and the large quantity required to meet energy requirements. The growth performance of growing rats fed sprouted finger millet was better than that of animals fed raw grain. However, the protein quality as judged by PER was unaltered by sprouting (Hemanalini et al., 1980). Further processing of germinated finger millet grains by drying, roasting and filtering through a cloth gave a product low in fibre. Animals fed this product as a source of protein showed improved calcium retention, probably because of the low fibre content of the flour.
Malted grains of finger millet have a significantly higher saccharifying enzyme activity useful in brewing. This activity is higger than that of malted sorghum, pearl millet or maize (Rao and Mushonga, 1985). A weaning food with low hot paste viscosity and high energy density was developed in which malted finger millet from which the vegetative portion had been removed was combined with green gram. Amix of 70 parts malted finger millet grains and 30 parts green gram, combined with 10 percent skimmed milk powder, had a PER of 2.7 and NPU of 63 percent (Malleshi and Desikachar, 1982). An extruded cooked product prepared from a blend of rice (42.5 parts), finger millet (42.5 parts) and defatted soy flour (15 parts) exhibited a significant improvement in protein quality over the unprocessed blend (Dublish, Chauhan and Bains, 1988). The PER values after extrusion had increased from 1.92 to 2.41, while the trypsin inhibitory activity was reduced by about 70 to 100 percent, the tannin content was below measurable amounts and the phytin phosphorus as percentage of total phosphorus had decreased by about 4 to 13 percent. These changes obviously might have contributed to the improved protein quality. A blend of finger millet and defatted soy flour (85:15) on extrusion had a PER of 1.81 before processing and 2.23 after extrusion.
Foxtail millet
The protein in foxtail millet is also deficient in lysine Its amino acid score (Table 22) is comparable to that of maize (Baghel, Netke and Bajpai, 1985). Monteiro, Virupaksha and Rajagopol Rao ( 1982) observed large variation in the grain protein content and in the distribution pattern of different solubility fractions. Prolamin constituted the major storage protein (Table 23) and showed a positive correlation with total protein in the grain. Evaluations of the amino acid composition of the protein fractions and of total protein in different varieties have confirmed that lysine is the first limiting amino acid, followed by tryptophan and sulphur amino acids.
With increase in grain protein, the lysine content of the protein decreased. The protein was found to be high in leucine. Naren and Virupaksha ( 1990) observed that the prolamin was relatively rich in the sulphur amino acid methionine and that the sulphur status of the soil affected the synthesis of prolamin in the grain. In studies on in vitro protein digestibility, 90.5 to 96.9 percent of the protein in foxtail millet was digestible by pepsin, and 89.7 to 95.6 percent by papain (Monteiro et al., 1988). Poor digestibility with trypsin (21.6 to 36.9 percent) was improved by prior treatment with acid. The protein quality of dehusked grain (Table 24) was the lowest among the minor millets tested (Geervani and Eggum, 1989). Heat treatment or lysine supplementation improved the protein quality (Geervani and Eggum, 1989). In growing rats fed 10 percent protein, the nitrogen balance improved from 19 to 31 percent when the foxtail millet diet was fortified with lysine protein digestibility and biological value were also enhanced (Ganapathy, Chitra and Gokhale, 1957). Supplementation of dehusked millet with chickpea raised the PER from 0.5 to 2.2.
Common millet
Though the range of protein content in common millet can be very wide, the values appear to lie most frequently in a narrow range of 11.3 to 12.7 percent, with a mean of 11.6 percent, on a dry-matter basis (Serna-Saldivar, McDonough and Rooney, 1991). The protein of common millet is deficient in lysine as well as threonine, and its tryptophan content is also marginal (Chung and Pomeranz, 1985). Studies on the protein solubility fractions of common millet showed that more than 50 percent of the grain protein was prolamin and the next predominant fraction was glutelin, about 28 percent. The prolamin fraction was very poor in lysine arginine and glycine compared to the albumin and globulin fraction and had more alanine, methionine and leucine (Jones et al., 1970). When ted as the sole source of protein (8.4 percent) in the diet, common millet had a PER of 0.95. According to data presented by Kuppuswamy, Srinivasan and Subramanian (1985), a common millet diet containing 9 to 11 percent protein had a PER of 1.2 and biological value of 56. A protein isolated (84.8 percent) by alkali extraction of common millet (Tashiro end Maki, 1977) was compared for its protein quality with casein and gluten. In a 21 -day feeding trial in young rats with 10 percent protein in the diet, the PER of the protein isolate of common millet was 3.1 while that of casein was 2.8. Animals fed whole millet flour as a source of protein failed to grow. In adult rats the biological value of millet flour was higher than that of the other protein sources. In in vitro studies, the isolated protein was digestible by pepsin and by pepsin-pancreatin but not by trypsin.
Other millets
Kodo, barnyard and little millets have been investigated less from the nutritional point of view. Kodo millet grains are enclosed in a hard, corneous husk which is difficult to remove. The fibre content of the whole grain is very high. Kodo millet has around 11 percent protein, and the nutritional value of the protein has been found to be slightly better than that of foxtail millet but comparable to that of other minor millets (Table 24). Apart from lysine the protein of kodo millet is deficient in tryptophan (Chung and Pomeranz, 1985). As with other foodgrains, the nutritive value of kodo millet protein was improved by supplementation with legume protein (Rajalakshmi and Mujumdar, 1966). The PER of kodo millet on supplementation with chickpea and amaranth leaves was increased from 0.9 to 1.9 (Patwardhan, 1961b).
Barnyard and little millets are comparable to common millet in their protein and fat content (Geervani and Eggum, 1989), and both are very high in fibre. With lysine amino acid scores of 31 and 33, little millet and barnyard millet have the poorest quality proteins among the millets. Barnyard and little millets are comparable in their protein digestibility, biological value, net protein utilization and digestible energy content (Table 24) and therefore in their overall nutritive value.