Effect of Genetic Elimination of Kunitz Trypsin Inhibitor on Agronomic and Quality Traits in Soybean [Glycine max (L.) Merril]

Nutritional status of soybean makes it suitable for human and animal diet owing to presence of 40% protein, 20% oil and nutraceutical components (lecithins, tocopherols, saponins, isoflavones). Despite all the benefits, the presence of anti-nutritional factors in soybean seed is major constraint limiting its use as human food. Being paramount element among other antinutritional factors present in soybean, trypsin inhibitor serve as an enzyme which retards the activity of trypsin, a core protein for proper digestion of food in humans. There are two types of trypsin inhibitors present in soybean i.e. kunitz trypsin inhibitor (20 KDa) and bowman birk inhibitor (8KDa) as major antinutritional factors (Clemente et al., 2013). Soybean kunitz trypsin inhibitor (Kti) was first isolated and studied by Kunitz (1945). Kunitz trypsin inhibitor inhibits activity of trypsin protein by forming of stoichiometric complex with it thereby blocking its active site (Kunitz, 1945). Since its discovery, numerous functions of Kti have been reported. Kti is known as serine protease inhibitor and have been associated with adverse effects like growth inhibition, pancreatic hypertrophy and hyperplasia in mammals (Liener, 1995). Other than ill effects of Kti, there are some benefits like its ability to provide defense against insect pests in plants. The well-known feeding deterent Kti confers resistance to brown plant hopper (Nilaparvta lugens) in transgenic rice (Lee et al., 1999) and different levels of insect resistance to tobacco and potato transgenic plants (Maechetti et al., 2000). Though, it serves an important tool in insect resistance but the kunitz trypsin inhibitor, which constitutes 80% of the trypsin inhibitor activity is the main culprit of ill effects of soy foods in human beings. The activity of kunitz trypsin inhibitor is retained only in native state. Increase in temperature leads to its denaturation, ultimately causing loss of its inhibiting power (Kunitz, 1945). But heat treatment adds to cost of soy processing and it denatures desirable food proteins as well. Therefore development of kunitz trypsin inhibitor free genotypes is the viable alternative to render the soy foods healthy.
       
Mainly, Kti is translated from m-RNA of Kti3 gene although there are 10 minor genes governing the trypsin inhibitor enzyme production. As a result of frame shift mutation in Kti3 gene, there is premature termination during its translation leading to production of truncated protein with lost inhibiting power in genotypes carrying it. The mutated gene was named as null allele (null Kti) which is recessive counterpart of Kti3 gene (Walling et al., 1986, Jofuku et al., 1989). Null Kti allele has been used to develop genotypes with reduced trypsin inhibitor activity. In Serbia, two Kti-freesoybean genotypes viz. ‘Laura’ and ‘Launa’ (Peric et al., 2014) and in India, Kti-free soybean genotypes, namely, ‘NRC101’ and ‘NRC102’ have been developed (Rani et al., 2010). But till date, no work has been reported to assess the effect of elimination Kti allele from soybean on other traits of interest. Therefore, the objective of the present study was to determine the effect of introgression of null Kti allele on characters of economic importance in soybean.

The research material used in this study was 125 F₅ genotypes segregating for Kti gene obtained by crossing SL525 × NRC101. SL525 is well adapted and high yielding variety recommended for cultivation in North-West part of India. It has normal content (21 KDa) of Kunitz trypsin inhibitor (Kti) in its seeds. The other parent of the cross i.e. NRC101 has been developed by Indian Institute of Soybean Research, Indore and it lacks kunitz trypsin inhibitor (null Kti). Gene specific marker and SSR marker linked the kti locus were used to identify null Kti and Kti positive plants (Moraes et al., 2006). A total of 50 null Kti (homozygous) and 65 Kti positive (homozygous) were used in the study and 10 heterozygous plants were eliminated from analysis. The experimental lines (125) along with both the parents of the cross viz. SL525 and NRC101 were grown in randomized complete block design with three replications. Each entry had single row of three meter length with line spacing of 45 cm and plant spacing of 5 cm within rows. The overall mean of Kti and null Kti genotypes for different characters was compared using Minitab 18 software. Among all the lines in F₄ generation, 5 lines segregated for Kti gene in F₅ generation. The F₅ plants originating from same F₄ lines were expected to be genetically more similar to each other but only differing for Kti gene. Therefore, these plants were compared separately for their performance to have better understanding of affect of null Kti gene introgression on characters studied.
       
In each genotype and replication, 10 random plants were selected and observations on agronomic parameters such as germination (%), days to flowering, days to maturity, plant height (cm), 100-seed weight (g), yield per plant (g), oil content (%), protein content (%), trypsin inhibitor activity (TIU/g) and fatty acids composition (%) was recorded. Per cent germination after harvest was recorded in replicates of three by paper towel method. A pre-treatment of HgCl₂ (0.1%) was given to seeds to prevent pathogenic infection prior to germination test. Oil content (%) was measured by using 5 g sample of each genotype by Nuclear Magnetic Resonance (NMR). Protein content was determined by N concentration using Kjeldahl analysis (AOAC, 2000) and the concentration of N was converted to crude protein by the factor 6.25 (Khalil and Manan, 1990). Fatty acids in oil were analyzed and trans-esterified by using gas liquid chromatography (GLC) by following protocol developed by Appleqvist (1967).

Comparison of null Kti and Kti positive plants
 
To understand the underlying effect of Kti gene, genotypes lacking Kti allele were compared with Kti +ve genotypes. The foremost comparison was done for various parameters using unpaired t-test (Table1) between Kti +ve (65 plants) and null Kti group (50 plants) of F₅ generation derived from cross of SL525 × NRC101. Removal of  the Kti allele resulted in decrease in overall mean grain yield of null Kti group (27.1 g/plant) in comparison to Kti +ve group (30.3 g/plant) although the difference was not statistically significant and some null Kti lines performed well in terms of yield (Table 3 and Fig 2B). Majority of null Kti lines were early maturing in comparison to kti genotypes with range of 119-137 days and statistically difference between maturity of the two groups was significant, although it is not much important agronomically (<2 days). However, overall comparison of means of two groups indicated that elimination of Kti allele had no effect on days to flowering, 100-seed weight and plant height. Trypsin inhibitor activity declined sharply in null Kti genotypes as shown in Fig 2D and difference of means of the two groups was statistically highly significant. The null Kti plants had trypsin inhibitor activity of 7.4 TIU/g in comparison to 88.8 TIU/g of Kti +ve plants. The germination after harvest was found to be adversely affected with elimination of Kti allele in genotypes but range (36-100%) showed that there were some null Kti lines with 100 per cent germination. The overall mean of germination for null Kti plants was 81.2 per cent whereas plants having Kti allele had 90.4 per cent germination. Trypsin inhibitors compromise at least 6 per cent of the total protein present in soybean seeds (Ryan, 1973). However, in our study the protein content was not affected with removal of kunitz trypsin inhibitor gene. This is because with elimination of Kti allele, trypsin inhibitor activity is lost but truncated protein is still produced in genotypes carrying null allele. The protein content of null Kti  and Kti +ve plants was 40.6 and 40.7 per cent, respectively. Oil content remained same irrespective of the allele present in genotypes i.e. 20.6 per cent for null Kti plants and 20.7 per cent for Kti +ve plants. Furthermore, the fatty acids composition viz. palmitic acid, stearic acid, oleic acid, linoleic acid and linolenic acid also remained unchanged in both the groups of plants. Thus lines carrying null allele were similar as Kti lines for protein, oil and fatty acid composition. The differences in two groups of plants for maturity and grain yield may be attributed to linkage drag of null Kti allele and plants carrying this allele are expected to have more of NRC101 genome which is better suited to climatic conditions of central India (Rani et al., 2010). Furthermore, SL525 has contributed more to genome of Kti plants which is well adapted variety of Punjab (Singh et al., 2006). NRC101 is early maturing, lower in grain yield than SL525 in Punjab.
 

 

       
To prevent bias due to genomic differences between two groups i.e. null Kti and Kti +ve plants, second comparisonwas made between the F₅ lines (5 pairs of lines) originating from same F₄ line showing segregation for Kti gene (Table 2, Fig 1). This comparison is more reliable as F₅ plants from same F₄ line would be genetically more similar to each other but only differing for Kti gene and furthermore in F₅ generation, homozygosity reaches 93.7%. The results of this comparison showed that germination was adversely affected by removing Kti gene. These results coincide with overall comparison of kti positive and kti negative plants (Table 1). But the other parameters such as days to maturity, oil content, 100-seed weight, grain yield per plant, protein content and plant height showed no significant differencesbetween the two groups (Table 2). However, trypsin inhibitor activity was significantly lower in null Kti lines (9.8 TIU/g) in comparison to Kti plants (88.5 TIU/g). Till date, no work on effect of null Kti allele on agronomic performance has been reported yet, although a comparison of Kti and null Kti plants for nutritional quality is available (Friedman et al., 1991). They compared two isolines differing for Kti allele for amino acid composition, fatty acid content, trypsin inhibitor activity. The amino acid composition and fatty acid content were same in both isolines whereas trypsin inhibitor content analyzed by ELISA showed that the translated protein’s functional and structural properties were less than 0.2% in null kti line in comparison to Williams 82 (kti +ve). Almost similar results have been obtained in the present study. The kunitz trypsin inhibitor follows similar pattern as other storage seed proteins during germination and degraded upon imbibition by cysteine protease (Wilson et al., 1988). Germination is governed by several environmental and genetic factors (Panfield et al., 2015, Chen et al., 2008), therefore the interaction of null Kti with germination needs further studies on this aspect.




       
The Kti gene is responsible for trypsin inhibition activity, therefore, genotypic and phenotypic correlation analysis of the all the F₅ genotypes was performed (Table 4). Specifically focusing on correlation of trypsin inhibition activity with other parameters, this analysis would be helpful to understand association Kti gene with other traits. The results showed that trypsin inhibitor activity was positively correlated with germination (r=0.38), days to maturity (r=0.18), grain yield (r=0.15) and plant height (r=0.14) at 1% level of significance and with seed weight (r=0.11) at 5% level of significance. This indicates that with increase in trypsin inhibitor activity, the germination is enhanced, maturity is delayed, higher the grain yield, plant height and seed weight in Kti positive genotypes. These results coincide with results discussed earlier i.e. comparison of means of plants (Kti +ve and null Kti). Though differences between mean values of two groups for days to maturity plant height and seed weight were statistically not significant but values were numerically higher in kti +ve lines as compared to null kti lines. The cause of positive correlation between two characters may be due to linkage, environmental effect or pleiotropy. Here, linkage seems to be genetic basis of correlation of trypsin inhibitor activity with other characters. Any loose linkage can be broken down during segregation and recombination which occurred in the present study over the generations. The results showed non-significant correlation between Kti activity and protein content which match with a previous study (Arefrad et al., 2014).
 

 
Comparison of null Kti plants with parent, SL525
 
In order to develop agronomically desirable genotypes, the null Kti plants were compared with Kti parent, SL525 (Fig 1) for those characters which were found to be significantly different in null Kti and Kti +ve plants. Fig 2(A) shows that although some of the null Kti genotypes show less germination than kti positive plants but most of them exceeded the Kti parent, SL525 (70%) for germination. Forgrain yield (Fig 2 B), a few of null Kti genotypes touched the Kti parent mark and some of them surpassed SL525 in yield (Table 3) indicating that these can serve as potential null Kti varieties. Fig 2(C) indicates that null Kti genotypes are early maturing in comparison to SL525, only one of high yielding genotypes was late maturing. The early maturing trait is desirable agronomically. The comparison of lines lacking trypsin inhibitor allele with kti positive parent, SL525 revealed that out of 50 null kti lines, 14 lines had significant higher grain yield than SL525. The SLNR08-02 genotype produced highest grain yield (52.3 g/plant) followed by SLNR08-02 (48.7 g/plant). SL525 recorded grain yield of 27.3 g/plant. The other characters of identified lines are comparable to SL525. Some plants lacking trypsin inhibitor activity have been identified which are equal to or better than Kti parent, SL525 for investigated traits.
 

The research offers comparison of Kti and null Kti genotypes for parameters of economic importance in order to understand the effect of removal of Kti allele which has not been reported yet. Based on comparison of kti and null-kti pants, our study suggested that null Kti gene adversely affected germination and grain yield which opens path for further studies. Other characters like protein, oil and fatty acid composition remained unaltered. Moreover, null Kti genotypes performing better than Kti parent have been identified. The lines lacking protease inhibitors which have good germination, high yield and early maturity can serve as potential genotypes after adaptive and multi-location trials. The null Kti lines are expected to be preferred by processors as well as consumers. The cost of soy processing will reduce by using Kti free genotypes because it lacks protease inhibitors for which  heat treatment prior to processing is required to eliminate them. Furthermore, the nutritional status of crop will be uplifted with removal of main anti-nutritional factor i.e. Kunitz trypsin inhibitor and the highly nutritious crop popularly known as miracle bean is expected to gain attention for food purposes particularly at household level by fortifying soybean with wheat to prepare chapattis.

Conceptualization of research (Dr (Mrs.) S.K. Dhillon, Dr. B.S. Gill); Designing of the experiments (Dr. B.S. Gill); Execution of field/lab experiments and data collection (Sandeep Kaur Dhaliwal, Dr. Sunita Sharma and Dr. Gurpreet Kaur); Analysis of data and interpretation (Sandeep Kaur Dhaliwal); Preparation of the manuscript (Sandeep Kaur Dhaliwal and Dr. (Mrs.) Asmita Sirari).

The authors should declare that they do not have any conflict of interest.

We are thankful to Indian Council of Agriculture Research for the financial support to conduct experiments under scheme ICAR-1 and we are also indebted to Indian Institute of Soybean Research, Indore and Dr. Anita Rani for providing kind assistance and experimental material.