Optimization of fortified sponge cake by nettle leaves and milk thistle seed powder using mixture design approach

Abstract Powdered nettle leaf and milk thistle (MT) seed were added to the cake batter with certain percentages selected by the Design‐Expert v. 10 software (0–25, 25–0, 18.75–6.25, 6.25–18.75, and 12.5–12.5). Addition of nettle and MT seeds to the cake reduced the moisture content, volume, and springiness and increased hardness of the samples. 12.5% nettle‐12.5% MT seed had the least hardness and the highest amount of springiness and cohesiveness. The highest BI, WI, SI, L*, a*, and b* and the lowest ΔE were observed in 12.5% nettle‐12.5% MT seed and 25% MT seed samples, respectively. Antioxidant activity and antimicrobial properties were increased in all samples compared to the control sample, so that 6.25% nettle‐18.75% MT exhibited the highest antioxidant activity and antimicrobial properties. The highest levels of quercetin and silymarin were observed in 25% nettle and 25% MT seeds, respectively. In the sensory evaluation, 12.5% nettle‐12.5% MT seed took the best scores regarding flavor, texture, color, and overall acceptance. Based on the lowest hardness, 13.65% nettle‐11.34% MT seed was determined as optimized points by the software, which was equivalent to desirability of 0.72. The optimum sample contained 62.90 mg quercetin and 886.70 mg silymarin. According to the HPLC analysis results, consumption of 10 optimal cakes daily could theoretically decrease the blood sugar level, which requires further studies. The remaining amount of quercetin and silymarin in 2.5 g of nettle leaves and 2.5 g of MT seeds after heating was 11 and 19 mg, respectively. In other words, heat did not have much effect on the destruction of quercetin and silymarin.

Milk thistle (MT) is an annual or biennial flowering plant growing all over the world. MT is recommended for the treatment of varicose veins, menstrual disorders, splenic congestion, kidney, and liver in the early 19th century (Abenavoli et al., 2018).
Silymarin, extracted from MT seeds and then dried in standard conditions, contains mostly flavonolignans (up to 80% w/w) in addition to polymeric and oxidized polyphenolic compounds. Silymarin is also reported to have positive effects on the blood sugar control and reduction (Luminita et al., 2016). Silymarin is composed mainly of silybinin, silychristin, silydianin, and isosilybinin with approximate percentages of 60, 20, 10 and 5%, respectively. Furthermore, diastereoisomers of the mentioned compounds, including silybinin A, silybinin B, isosilybin A, isosilybin B are found in addition to taxifolin. The use of nettle and MT in cereal products such as cakes and the resistance of silymarin and quercetin to the heat process can produce functional foods that have medicinal properties, texture quality and consumer acceptance, and lead to an increase in human health (Tayoub et al., 2018). Ataei and Hojjatoleslamy (2017) evaluated the amount of oleuropein remaining in the cake.
They reported that oleuropein remained more in the olive leaf powder than its extract did. Therefore, in this study, nettle leaves and MT seeds were added to the cake in powder form. Moreover, the amount of silymarin and quercetin, which is effective in reducing blood sugar, was evaluated (Ataei & Hojjatoleslamy, 2017).
In addition, the mixture design method was used, which contributes to the reduction of the number of treatments and repeat tests, so that results are obtained faster.

| Sponge cake preparation
We prepared the sponge cake according to the recipe introduced by the authors in an earlier study with some modifications. Table 1 shows the ingredients based on 100 g of flour. Nettle leaf and MT seed powder was added to the formulation in the last step of batter preparation. After cooling, the baked cakes were packed in poly propylene bags at room temperature and kept for future physicochemical and sensory evaluation analyses (Ataei & Hojjatoleslamy, 2017).

| Optimization of mixing ratio
The mixture Design of Experiment method was used to find the optimum mixture ratio of the nettle-MT powder (p < .05). The Design-Expert (V. 10) software, Stat-Ease Inc., USA, was used in this research.
Nettle leaf and MT seed were assumed to be independent variables with a total proportion of 25% in a 100-g sponge cake.
The software then suggested the amount of nettle and MT seed for different treatments (Table 1). The best model expressing the behavior of the treatments was chosen according to the obtained p-value. Moreover, more parameters, including R 2 and Adjusted R 2 , were studied where the lack of fit test was not significant.
Finally, the response and the dependent variable were considered desirable based on the independent variables as: (Mohammadi et al., 2019).
where; D, is total desirability, d, is desirability of each response, and n, is response number.

| Chemical properties of sponge cake
Chemical properties of the cake samples were measured using AACC approved methods as: moisture content, crude protein, ash content and crude fat by AACC 46-40, AACC 46-11A, AACC 08-01, and AACC 30-10, respectively. According to the method AACC 46-11A, the crude protein content was assessed using the nitrogen conversion factor of 6.25 (Ataei & Hojjatoleslamy, 2017).

| Physical properties of sponge cake
The cake volume and density were measured according to Prokopov et al. using rapeseed displacement (Prokopov et al., 2015). Samples with dimensions of 20 × 20 × 20 mm were taken from the midsection of the cakes on the first and after thirty days of storage for the texture profile analysis (TPA) using a texture analyzer (Brookfield Springiness quantifies the elasticity of the cake by measuring the distance recovered between the first and second compressions. Cohesiveness indicates the extent at which the product texture resists to the second deformation relative to the first one (Salehi et al., 2016). The Hunter Lab system was used to determine the color characteristics of the crumb of cake samples using a ColorFlex Ez, USA colorimeter. The L*, a* and b* values were determined. Where the L* value indicates the lightness from dark (0) to light (100), the a* value indicates the degree of the green to red color, the higher positive the value the more reddish. The value b* represents the yellow to blue color, the higher positive the value the more yellow. ΔE, which shows total color difference from the reference color, was obtained from Equation 2 (Hafez, 2012).

| Quercetin and Silymarin measurement
To evaluate the quercetin and silymarin retention in baked cakes, the amount of quercetin and silymarin was measured. The measurement was performed according to the method proposed by Haghi and Hatami (2010) and Bourgeois et al., 2016), using the HPLC technique.
The HPLC system (Azura, Knauer Co) with a k-1001 pump and a UV-Visible detector k-260 was used for this purpose. The Teknokrom C18 column was chosen for the analysis, and the UVvisible detector was set at 370 nm and 150 nm for quercetin and silymarin, respectively. The mobile phase to measure quercetin was acetonitrile (7%) (A) and phosphoric acid (B). Injection volume was 20 μl, and the column temperature was ambient temperature. The elution protocols applied for quercetin were: 0-9 min, 75% B; 9-19 min, 75%-25% B; and 19-24 min, 25% B (Bourgeois et al., 2016).

| Microbial load determination
The microbial load of cake samples was determined according to the International Commission on Microbiological Specification of Foods (ICMSF, 2009). For this purpose, a unit weight of a cake sample was taken and mixed with 9 ml of sterile distilled water in a test tube.
It was then serially diluted until the desirable dilutions were obtained. A volume of 0.01 ml aliquot from each diluent was aseptically transferred into different previously autoclaved plates poured with already sterilized molten media. The plates were then incubated at 28°C for 72 hr.
The plates in the incubator were daily checked for growth. The number of grown colonies on each plate was counted using a colony counter. The number of colonies was reported in term of the number of the colony forming unit (CFU).

| Determination of antioxidant activity
Free radical scavenging activity, DPPH radical elimination, was de- where absorbance control is the absorbance of DPPH solution without extract.

| Sensory evaluation
The hedonic scale as a unique and commonly used scale to measure food product liking and preference yields reliable results.
Hence, a 5-point hedonic scale, according to the method employed by the authors in a previous study was conducted to determine the degree of overall liking of the sponge cakes (Ataei & Hojjatoleslamy, 2017). A total of 20 semi-trained panelists were selected, each panelist received five cake slices, cut from the midsection of the cakes maintained at ambient temperature, and was asked to score each sample based on the degree of liking on a fivepoint hedonic scale (one: dislike very much, two: dislike, three: neither dislike nor like, four: like, five: like very much). The panelists received samples.

| Physicochemical characteristics of sponge cake
reported that addition of green tea to the sponge cake had no significant effect on the moisture content of the cake.
% inhibition of DPPH = 100 × (Absorbance control − Absorbance sample) Absorbance control  Table 2, addition of the nettle leaf and MT seed powders caused to decrease the cake volume and to increase its density, compared to the control sample. The cake volume is generally influenced by the characteristics of the fibers added, via both the nature and amount.
Fiber addition can disturb the gluten network, and eventually the gluten protein is diluted. A weak gluten network might result in letting gases such as carbon dioxide and water vapor to escape from the cake pores and consequently decrease the cakes volume (Kim et al., 2012;Tsong-Ming lu et al., 2010). It appears that it occurred more considerably in the sample containing the 25% MT, so that it had the lowest volume, while the 12.5% nettle-12.5% MT seed had the highest volume among the cake samples. These findings were consistent with those reported by Aydogdu et al. (2018) in one study of the rheological properties of cake batter and quality of the final product as affected by addition of different fibers. It was stated that a significant correlation was found between the consistency index and the specific volume. As the batter consistency index was increased, cakes with lower specific volume were obtained. The type of fiber is highly important. The consistency index of the batter containing lemon and apple was higher than that with oat and pea. Particularly, addition of lemon fiber resulted in the highest consistency index of batter and the lowest specific volume of cakes. Gómez et al. (2003) reported that the lowest volume of bread was obtained, when the coffee fiber with higher soluble fiber content was added to bread.
Furthermore, higher elastic and viscous modulus values inhibited cake development, leading to lower volume.  and their texture hardness. The 25% MT seed sample had the lowest moisture content and the highest hardness, while the 12.5% nettle-12.5% MT seed sample revealed the highest moisture content and the lowest hardness. Furthermore, a predictable reduction of the moisture content of the samples occurred during storage time, being concurrent with the hardness increase of the samples. Figure 3 These results correspond to those obtained by Seo et al. (2010) in the study of the sponge cake containing turmeric. Addition of turmeric to the cake increased the cake hardness, gumminess, and chewiness, and thus the softness of the cake was decreased due to its effect on the gluten network. By addition of 129.5% sugar, 0.5% turmeric powder, and 10.0% oil, linear models were selected for the hardness, gumminess, and chewiness (Seo et al., 2010). the storage period of the cake, this network gradually is weakened, the internal strength of the cake structure is decreased, and consequently the cohesiveness and springiness of the cake are decreased (Kim et al., 2012).
In this study, the results revealed that the weakening of the gluten network and eventually reduction of cohesiveness and springiness occurred largely when nettle and MT seed were simultaneously added to the cake compared to the control. However, addition of either nettle or MT seed alone led to further disruption of the gluten network and further reduction of the cake cohesiveness and springiness. The results were in agreement with the findings obtained by Tsong-Ming lu et al., 2010) for the cakes added with the green tea powder. They replaced different percentages (0%, 10%, 20%, and 30%) of the cake flour by the green tea powder and reported that the viscosity of the cake batter was higher than that of the control sample, but the cake volumes exhibited a reverse trend. It was attributed to the increased replacement of flour with cellulose and the weakened gluten matrix. Thus, the cohesiveness and springiness were decreased with addition of the green tea powder. Kim et al. (2012) reported that hardness and gumminess were reduced, and cohesiveness and springiness were increased with addition of the cactus Opuntia humifusa powder. This was due to the mucilaginous properties of the cactus pectin. Gums displaying viscous properties in the solution state were added to cake batters in order to increase the moisture retention during baking. Pectin preserves the moisture, increases the volume, springiness and cohesiveness of cake, and decreases its hardness.

| Antioxidant activity
Antioxidants serve as preventatives from the destructive and detrimental influence of free radicals, also known as oxidants, on the cells and the resulting diseases. Therefore, assessment of antioxidant and free radical scavenging characteristics of different natural products and additives has been the aim of many research studies.
Polyphenols, including phenolic acids, and flavonoids are the major and most popular compounds with antioxidant properties (Kamkar & Khodabakhshiyan, 2017;Viktorova et al., 2019). According to the literature, antiradical properties and inhibitory potential of nettle are greater than those of many synthetic antioxidants (Zeipina et al., 2015). Antioxidant activity of nettle leaves was reported to range from 17.31% to 80.77%. Zeipina et al. attributed this variation to the fertility of soil, the clone and the plant age. (Zeipina et al., 2015). Saa (2017) studied the antioxidant properties of silymarin and reported that silymarin had higher antioxidant properties than BHT had. As different concentrations of silymarin were added to the sunflower oil, the antioxidant activity of the sunflower oil was increased compared to BHT. Table 2 presents the results of the antioxidant activity measurement. The quartic model was selected by the software to fit the data. According to our findings, the highest and lowest DPPH scavenging elimination was observed in the 6.25% nettle-18.75% MT seed sample and the 25% nettle cake.
It appears that the combination of nettle and MT seeds used in the cake could exhibit higher antioxidant effect than when used alone.  so that as the amount of antioxidants in food increases so do antimicrobial properties (Fazelinasab et al., 2017). Both nettle and

| Antimicrobial activity
MT have antimicrobial and antioxidant properties owing to their phenolic and flavonoids content (Hasanloo et al., 2014;Moradi & Amini, 2017). The results of this study showed that the growth of mold in all treatments was lower than that of the control sample. It was due to the presence of nettle and MT seed in the formulations. spectively. The quercetin content ranged from 0 to 116mg, and the silymarin content ranged from 0 to 2,025 mg in the cake samples.

| Quercetin and Silymarin content
The linear model was selected as the best model fitting to experimentally measured quercetin and silymarin values. As nettle and MT seed were increased, the quercetin and silymarin content was increased in the samples.    Table 4 presents the results of the sensory evaluation performed based on the 5-point hedonic scale. Flavor, texture, and overall acceptance values were fitted by a quadratic model, while odor and color were predicted by linear and cubic models in the best way, respectively. Among all treatments, the cake with 12.5% nettle leaf and 12.5% MT seed gained the highest overall acceptance score.

| Optimization
The optimized combination of the nettle leaf powder and the MT seed powder to be used in the cake was determined by the Design-Expert software. Independent variables were assumed to be different amounts of nettle and MT seed, up to 25%, according to the minimum hardness value and the highest sensory test scores, and then optimization was performed by the software. According to  The results demonstrated that approximately 10 optimal cakes could theoretically decrease the blood sugar level, which needs further studies. The amount of quercetin and silymarin in 2.5 g of nettle leaves and 2.5 g of MT seed was 0.44% w/w equal to 11 mg and 0.87% w/w equal to 11 mg, respectively. After heating, their amount was 11 mg and 19 mg, respectively. In other words, heat did not have considerable effect on the destruction of quercetin and silymarin, and the substances could resist the thermal processes of the food industry. It is recommended that nettle and MT seeds be used in various food industries, particularly in the cereal industry in order to have acceptable physical and chemical properties and to use its medicinal properties like lowering the blood sugar. In this respect, further studies are required.

ACK N OWLED G M ENTS
This work was supported by Azad University, Shahrekord branch

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the Shahrekord Azad University upon reasonable request.