Development of Gluten-Free Rice Flour Noodles That Suit the Tastes of Japanese People

Gluten-free rice flour noodles with a flavor and texture profile preferred by the Japanese people were developed. The rice noodles contained potato starch (PS) as a binder. “Koshihikari” was selected from several candidate varieties based on its pasting properties. Since the Japanese people prefer the chewy texture of wheat flour “Udon” noodles, first, the stress−strain characteristics of “Udon” noodles in Japan were quantified, using a mechanical test. Next, different formulations of rice noodles were prepared by changing the amount of PS blended into the noodles. The mechanical tests on wheat and rice noodles show that rice noodles made from 85% rice flour and 15% PS have a texture similar to that of “Udon” noodles. Brown rice noodles containing roasted brown rice flour were also developed. Since brown rice flour hinders the binding of the dough, it was necessary to increase the amount of PS to increase the binding of roasted brown rice flour. Finally, noodles with 70% white rice flour, 10% brown rice flour, and 20% PS were produced. The gas chromatography−mass spectrometry analysis of the volatile compounds contained in white rice noodles and brown rice noodles identified the volatile compounds characteristic each of type.


Introduction
Rice is a staple food of the Japanese population. However, according to data released by the Japanese government, rice consumption per capita has fallen to less than half in the last 50 years [1]. This may be because the Japanese people have come to prefer wheatflour-containing foods, such as bread and noodles, to rice. This shift may hamper food self-sufficiency in Japan.
Increasing the intake of wheat flour foods by Japanese people may increase the number of people with wheat allergies and celiac disease (CD) due to gluten, which is a hydrate of the protein contained in wheat flour [2]. Celiac disease affects approximately 1% of the world's population [3]. In Japan, Fukunaga et al. reported that the presence of celiac disease in a non-clinical Japanese population was low, at 0.05% [4]. Although the number of patients suffering from this condition in Japan is smaller than that of Western countries at present, several researchers have pointed out the possibility of an increase in its prevalence in Japan in the near future [5,6]. The only effective and safe treatment of CD remains the strict exclusion of gluten for life under a so-called gluten-free diet [7]. Thus, we aimed to develop gluten-free rice noodles.
Many studies have been performed on the development of gluten-free noodles. Martil et al. extracted starch from parboiled rice flour and produced gluten-free pasta via an extrusion cooking process [8]. They reported that the starch was gelatinized by extrusion and then cooled in order to make the starch retrograde, and the pasta exhibited a hard In this study, with reference to the above works, we analyzed the volatile compounds of the white and brown rice noodles, using HS-SPE-GC-MS, and we identified the volatile compounds that were characteristic of each noodle.

Selection and Preparation of the Raw Rice Flour Variety
Based on the ease of procurement of rice, six varieties from the Fukushima Prefecture and one variety from Niigata Prefecture were used as candidates (Table 1). In the table below, the amylose content values from the literature are quoted [24][25][26][27][28]. Amylose content refers to the amylose ratio in the starch contained within the rice. Each type of raw rice was purchased in the form of brown rice, and it was milled and ground by consignment manufacturing (Maruko Foods Inc., Marumori, Japan) to obtain rice flour. For the development of brown rice noodles, brown rice was ground and roasted (140 • C, 10 min) by consignment manufacturing to obtain brown rice flour.

Preparation and Treatment of PS to Be Mixed as a Binder
The PS was purchased from a starch maker (Nishida Denpun Koujou, Ltd., Hokkaido, Japan). It was blended into rice flour (white or brown) in various ratios. Mixed flour was used for noodle making.

Noodle Making
The rice flour noodles (white noodles, brown noodles) used for the test were manufactured on consignment using a commercial scale process (Maruko Foods Inc., Miyagi, Japan). Rice flour (white or brown flour) and PS were mixed in various ratios. Tap water was added at 44% by mass to the flour. After mixing, the dough was passed into a roller 5 times until a smooth and elastic sheet was formed; this was aged in a refrigerator overnight, cut, steamed, cooled, and vacuum-packed.

Cooking Procedure of Rice Noodles Tested
Water (1 L) was boiled, and approximately 20 g of noodles was cooked for different lengths of time. The texture of noodles is affected not only by its composition but also by the cooking duration. Boiling the noodles for a short duration leads to a firm texture, whereas cooking for a longer time results in a softer texture. To ensure that the water content of the cooked noodle samples, which were boiled for different periods, was the same, the contents were measured to determine the relationship between the boiling time and water content. The water content shows an approximately linear relationship with the boiling time in a realistic boiling time range. The boiling time at which the water content of the rice noodles was 70% (wet basis) was determined from the obtained relationship. The water content was determined from the weight of the noodles before drying and that after drying at 130 • C for 2 h, having allowed them to cool in a desiccator for 30 min.

Quantification of Mechanical Properties
A blade (aluminum, thickness 3 mm, rounded tip) simulating a human tooth was attached to a material test device (EZ-TEST LX, Shimadzu Corporation, Kyoto, Japan) and perpendicularly lowered onto a noodle string placed on a horizontal base fixed to the device used to perform the shear test. The lowering rate of the blade was 10 mm/min. The shear test was carried out within 30 min of cooking. Each sample was analyzed six to eight times. Generally, udon noodles have a rectangular cross-section, and in the mechanical tests, the blade penetrated vertically toward the longer side. The length of this longer side was called the "width" of the noodles. The "width" of the cooked noodles was measured using a vernier caliper set on the base prior to each test. The product of the thickness of the blade and the width of the cooked noodles was designated as the apparent bottom area, and the stress value was obtained by dividing the generated force by the apparent bottom area. The distance from the contact point of the blade with the noodles to the base was defined as unity, and the change in this distance was referred to as the strain. The texture of the noodles was evaluated based on the stress-strain curve.

Pasting Properties
The candidate rice varieties were examined based on their pasting properties, which were quantified using the rapid visco analyzer (RVA: RVA4500, PerkinElmer, Waltham, MA, USA).
A sample of rice noodles (3.50 g) with 14% water content was weighed and suspended in 25 mL of water, and then, pasting properties were analyzed using the RVA. The measurement cell was held at 50 • C for 1 min, heated from 50 • C to 93 • C for 4 min, held at 93 • C for 7 min, and then cooled to 50 • C for 4 min, before finally being held at 50 • C for 3 min, based on the procedure of Nakamura et al. [29]. Each sample was analyzed three times.

Analysis of Volatile Compounds
A quadrupole GC−MS (GC: 7890A, Agilent Technologies, Santa Clara, CA, USA; MS: JMS-Q1000GC Mk-II, JEOL, Tokyo, Japan) was used for the analysis of the volatile compounds present in the noodles. A DB-WAX capillary column (Agilent Technologies, Santa Clara, CA, USA; inner diameter, 0.25 mm; length, 60 m; film thickness, 0.25 µm) was used for gas chromatography, and high-purity He (purity 99.9995%) was used as the carrier gas (flow rate: 1.5 mL/min).
An uncooked noodle sample (3.0 g) was placed in a 20 mL vial, and the headspace gas in the vial, which was thermally equilibrated at 80 • C for 20 min, was absorbed onto a solid phase (GL trap1; GL Science Inc., Tokyo, Japan) by pressurizing the vial at 140 kPa in an autosampler (S-trap; JEOL, Tokyo, Japan). After absorption, the solid phase was dried by flowing He for 1.5 min and then heated at 200 • C for 3 min to desorb and inject gas into the GC column. The GC oven temperature was maintained at 40 • C for 10 min, increased to 220 • C at a rate of 5 • C /min, and maintained at 220 • C for 10 min.
An n-alkane reagent (C8 to C20 in n-Hexane; GL Science Inc., Tokyo, Japan) was used to determine the Kovats' retention index (RI) [30], and 1 mL of 2-methyl-2-butanol aqueous solution (1 ppm) was used as the external standard sample.
The obtained chromatogram was deconvolved using the automated mass spectral deconvolution and identification system (AMDIS; National Institute of Standards and Technology (NIST), Gaithersburg, MD, USA), and candidate compounds were identified using MS-Search (NIST, Gaithersburg, MD, USA) [31]. For each candidate compound, the RI obtained from the analysis of the n-alkane reagent was calculated, and the compound was identified using the mass spectrum of the pure compound and the RI database provided in the NIST Chemistry WebBook [32]. The target ion was determined when the compound was identified, and the peak area of the target ion was used as a measure of the concentration of each compound in the noodle sample. Each sample was analyzed three times.

Statistical Analysis
Tukey's HSD test was applied for the comparison of yield stresses between cooked rice noodles using R version 3.1.3 (R Core Team, Vienna, Austria) [33]. Statistical significance was assessed with a confidence level of 0.99 (p = 0.01). A principal component analysis (PCA) of the volatile compound data was performed using the 64-bit statistical analysis tool and Excel 2010 (RIKEN Center for Sustainable Resource Science, Saitama, Japan) [34]. The PCA is a statistical method whereby a multidimensional phenomenon is reduced to a few dimensions. The principal components 1 (PC1) and 2 (PC2) emerge as variables when a multidimensional phenomenon is reduced to two dimensions, and information is lost as a result of this reduction [35]. The "proportion of variance" is a measure of how much information is retained.

Mechanical Properties of Cooked Udon Noodles
The primary aim of the study was to develop rice noodles that suit the taste preferences of the Japanese people, while trying to simultaneously achieve a texture similar to that of udon noodles. Therefore, in order to identify the numerical index of a chewy texture, we examined the texture of udon noodles.
The representative stress-strain characteristics of five kinds of cooked commercial udon noodles are shown in Figure 1. In all cases, the stress increased from the start of the penetration of the blade, before peaking. At this value, noodles break, causing stress relaxation. Consequently, the stress value drops, the blade contacts the base, and the test is completed. Since the peak stress value corresponds to the force exerted when a human tooth bites and breaks a noodle, it can be used as an index of the chewy texture. This stress value is called the yield stress, according to material mechanics. "Gotoh udon" and "Inaniwa udon" noodles show sharp peaks near the yield point; "Sanuki udon" and "Hakata udon" noodles show upwardly convex stress-strain curves, indicating that plastic deformation occurs during the shear test, which suggests a chewy texture that suits the tastes of the Japanese people; and "Ise udon" noodles show a broad peak. These "Ise udon" noodles have a texture that is too soft for Japanese people, and therefore the noodles are thickened. Furthermore, in the case of "Gotoh udon" and "Inaniwa udon" noodles, the noodles are thinned because the texture is too firm. Thus, we aimed for a texture in our rice noodles that is the same as that of "Sanuki udon" noodles. These discussions indicate that the texture and the width of the noodles are related to each other. Therefore, we also plotted the relationship between the width of the cooked noodles and the yield stress of the stress-strain curve. The results are shown in Figure 2. The yield stress values correspond to the peak stress values in Figure 1. We found that a specific relationship holds, as shown in Figure 2. It is important to note that this relationship is based on Japanese food preferences, and it does not have scientific criteria. When measuring the width and yield stress of cooked rice noodles, we tried to determine the formulation of rice noodles, aiming for the texture of "Sanuki udon".

Selection of Raw Rice Flour Variety
A representative RVA measurement result (Koshihikari flour) is shown in Figure 3. The viscosity in the cell increases sharply as the temperature rises before reaching its maximum value (peak viscosity). After that, the viscosity reaches its lowest value (trough viscosity). The viscosity begins to increase as cooling begins, and it finally approaches a constant value (final viscosity). The difference between the peak viscosity and trough viscosity values is called the breakdown value, and it is used as an index of gelatinization. The difference between the trough viscosity and final viscosity values is called the consistency value, which is a measure of how hard the gel becomes upon cooling [36]. In terms of noodle texture, the breakdown value is related to the chewiness of the noodles, and the consistency value is related to the tendency of the noodle to harden after cooking. In this study, we evaluated the difference in the pasting properties of samples using breakdown and consistency values.

Selection of Raw Rice Flour Variety
A representative RVA measurement result (Koshihikari flour) is shown in Fig  The viscosity in the cell increases sharply as the temperature rises, before reachi maximum value (peak viscosity). After that, the viscosity reaches its lowest value (t viscosity). The viscosity begins to increase as cooling begins, and finally approa constant value (final viscosity). The difference between the peak viscosity and troug cosity values is called the breakdown value, and is used as an index of gelatinizatio difference between the trough viscosity and final viscosity values is called the consis value, which is a measure of how hard the gel becomes upon cooling [36]. In ter noodle texture, the breakdown value is related to the chewiness of the noodles, an consistency value is related to the tendency of the noodle to harden after cooking. study, we evaluated the difference in the pasting properties of samples using break and consistency values.

Selection of Raw Rice Flour Variety
A representative RVA measurement result (Koshihikari flour) is shown in Figure 3. The viscosity in the cell increases sharply as the temperature rises before reaching its maximum value (peak viscosity). After that, the viscosity reaches its lowest value (trough viscosity). The viscosity begins to increase as cooling begins, and it finally approaches a constant value (final viscosity). The difference between the peak viscosity and trough viscosity values is called the breakdown value, and it is used as an index of gelatinization. The difference between the trough viscosity and final viscosity values is called the consistency value, which is a measure of how hard the gel becomes upon cooling [36]. In terms of noodle texture, the breakdown value is related to the chewiness of the noodles, and the consistency value is related to the tendency of the noodle to harden after cooking. In this study, we evaluated the difference in the pasting properties of samples using breakdown and consistency values. The relationship between the breakdown and consistency values of rice flour samples is shown in Figure 4. Although the "Koshinokaori (Niigata)" variety had the largest breakdown value, it also had a large consistency value. It would not be appropriate to select this variety, because noodles made from this variety tend to harden over time after cooking. The "Kamenoo", "Koshinokaori (Fukushima)", and "Tennotubu" varieties were excluded because their breakdown values were relatively small; consequently, their chewy texture would likely be weak. The "Sasashigure", "Hitomebore", and "Koshihikari" varieties from Fukushima Prefecture were selected as the candidate varieties (circled red dashed line) because of their relatively large breakdown values and low consistency values. We decided to use "Koshihikari", which has the lowest consistency value among the three candidate varieties; that is, it is resistant to retrogradation, according to various analyses and tests.
In addition, brown rice flour showed small breakdown (1646 mPa s, SD282) and consistency (1167 mPa s, SD30) values, and PS showed an excessive breakdown value (1.000  The relationship between the breakdown and consistency values of rice flour samples is shown in Figure 4. Although the "Koshinokaori (Niigata)" variety had the largest breakdown value, it also had a large consistency value. It would not be appropriate to select this variety, because noodles made from this variety tend to harden over time after cooking. The "Kamenoo", "Koshinokaori (Fukushima)", and "Tennotubu" varieties were excluded because their breakdown values were relatively small; consequently, their chewy texture would likely be weak.  The relationship between the breakdown and consistency values of rice flour samples is shown in Figure 4. Although the "Koshinokaori (Niigata)" variety had the largest breakdown value, it also had a large consistency value. It would not be appropriate to select this variety, because noodles made from this variety tend to harden over time after cooking. The "Kamenoo", "Koshinokaori (Fukushima)", and "Tennotubu" varieties were excluded because their breakdown values were relatively small; consequently, their chewy texture would likely be weak. The "Sasashigure", "Hitomebore", and "Koshihikari" varieties from Fukushima Prefecture were selected as the candidate varieties (circled red dashed line) because of their relatively large breakdown values and low consistency values. We decided to use "Koshihikari", which has the lowest consistency value among the three candidate varieties; that is, it is resistant to retrogradation, according to various analyses and tests.
In addition, brown rice flour showed small breakdown (1646 mPa s, SD282) and consistency (1167 mPa s, SD30) values, and PS showed an excessive breakdown value (1.000  The "Sasashigure", "Hitomebore", and "Koshihikari" varieties from Fukushima Prefecture were selected as the candidate varieties (circled red dashed line) because of their relatively large breakdown values and low consistency values. We decided to use "Koshihikari", which has the lowest consistency value among the three candidate varieties; that is, it is resistant to retrogradation, according to various analyses and tests.
In addition, brown rice flour showed small breakdown (1646 mPa s, SD282) and consistency (1167 mPa s, SD30) values, and PS showed an excessive breakdown value (1.000 × 10 4 mPa s, SD1448) and a medium consistency value (1777 mPa s, SD472). Here, SD means standard deviation. These data were not plotted in Figure 4 to ensure the readability of the other data. The low breakdown value of brown rice flour results in difficulty when making the dough. The large breakdown value of PS contributes to its chewy texture, but excessive blending induces hardness in the noodles.

Percentage of PS in the Rice Noodles
Gluten-free noodle dough cannot contain 100% rice flour because a binder to replace gluten is required. In this study, PS was chosen as the binder. The higher the amount of PS, the easier it is to form noodle dough; however, the texture becomes firm beyond chewy. On the contrary, if the content is too low, the dough cannot be appropriately prepared, and the noodles collapse during the noodle-making process. Therefore, the optimum amount of PS should be determined based on the mechanical properties of cooked noodles.
Different formulations of rice noodles were made by changing the amount of PS added to the noodle dough, i.e., from 5% to 20%. A PS addition rate of 5% indicates that the flour contains 5% PS on a mass basis. At 5% PS, the dough cannot be formed. We then decided to increase the PS addition rates to 10, 15 and 20%.
On the other hand, in the case of brown noodles, the addition of roasted brown flour impairs the ability to form a dough, and so the addition was therefore limited to 10%. Therefore, we set the blend amount of roasted brown rice powder to 10% and conducted a test in which the amount of PS added was changed to adjust the texture. When the PS was set to 15%, the noodles remained brittle, and when the amount of PS added was 20%, the brittleness was improved. Even then, the width of the noodles needed to be 7 mm rather than 4 mm. Table 2 shows the formulations and widths of the cooked rice noodles examined in this study. RUN1 to 4 show white rice noodles, and RUN5 shows a brown rice noodle. A representative stress-strain curve of each condition is shown in Figure 5. The yield stress value tended to decrease as the amount of PS added decreased. "≈" means "approximately equal". A bar graph plotted using all the yield stresses of these five formulations is shown in Figure 6. There was no significant difference at 20% PS, even if the width of the noodles changed. However, the yield stress decreased significantly with a decrease in the blending ratio of PS. This graph also shows the data for "Sanuki udon". It can be seen that the values  A bar graph plotted using all the yield stresses of these five formulations is shown in Figure 6. There was no significant difference at 20% PS, even if the width of the noodles changed. However, the yield stress decreased significantly with a decrease in the blending ratio of PS. This graph also shows the data for "Sanuki udon". It can be seen that the values of "Sanuki udon" are similar to those of the 15% PS-supplemented rice flour noodle. Figure 5. Stress-strain diagram of cooked rice noodles: a representative measurement of each condition. "≈" means "approximately equal".
A bar graph plotted using all the yield stresses of these five formulations is shown in Figure 6. There was no significant difference at 20% PS, even if the width of the noodles changed. However, the yield stress decreased significantly with a decrease in the blending ratio of PS. This graph also shows the data for "Sanuki udon". It can be seen that the values of "Sanuki udon" are similar to those of the 15% PS-supplemented rice flour noodle. The measurement results of the yield stress and widths of the rice flour noodles are superimposed on the relationship diagram of the yield stress and width of the udon noodles. The results are shown in Figure 7. The data points of udon noodles are indicated by black circle symbols, "•". The data on rice flour noodles intersect with those of udon noodles, and the data of the 15% PS-supplemented noodles just overlap those of udon noodles. The data of the udon noodles in the overlapping part correspond to those of "Sanuki udon". This means that the texture of rice flour noodles with the addition of 15% PS is similar to that of "Sanuki udon".

Aroma Characteristics
We analyzed the aroma characteristics of the white and brown rice noodles before cooking. The GC-MS analyses detected 39 compounds in common between white rice noodles and brown rice noodles. Table 3

Aroma Characteristics
We analyzed the aroma characteristics of the white and brown rice noodles before cooking. The GC-MS analyses detected 39 compounds in common between white rice noodles and brown rice noodles. Table 3 shows a list of the detected compounds, with the corresponding RI and target ion. The values obtained from the analysis are designated as RI_exp, and the values from the database are designated RI_ref. In Table 3, the unknown compounds refer to those for which there were no polar column data in the database. The peak areas of the target ions were determined for these components.  The score plot shows that both samples were mapped to different positions. The l ing plot shows that the positions of the white rice noodles were determined by esters, as ethyl acetate and ethyl 2-hydroxyhexanoate, alcohols, such as 2-ethyl-1-hexanol an octanol, and aldehydes, such as nonanal and (E)-2-nonenal. In particular, alcohols aldehydes have been shown to contribute to the aroma of rice [37]. On the other hand The score plot shows that both samples were mapped to different positions. The loading plot shows that the positions of the white rice noodles were determined by esters, such as ethyl acetate and ethyl 2-hydroxyhexanoate, alcohols, such as 2-ethyl-1-hexanol and 1-octanol, and aldehydes, such as nonanal and (E)-2-nonenal. In particular, alcohols and aldehydes have been shown to contribute to the aroma of rice [37]. On the other hand, the positions of brown rice noodles were determined by their low boiling point, as well as their content of sweet-flavored aldehydes such as 2-methyl-propanal, 2-methyl-butanal and 3-methyl-butanal, Maillard reaction products such as furfural, and nut-flavored compounds such as benzaldehyde and acetophenone [38]. The Japanese people have historically tried to incorporate roasted aromas into their traditional foods by applying a Maillard reaction, utilizing ingredients such as miso and soy sauce. The white rice noodles developed in this study have a rich rice flavor, whereas the brown rice noodles have a roasted aroma, which is preferred by the Japanese people.
In addition, the signals of the volatile compounds in the cooked rice noodles were weak, and it was hard to sufficiently compare the white rice noodles with the brown rice noodles. Therefore, in this study, we analyzed the aroma characteristics of the noodles before cooking. The analysis of the aroma of the noodles after cooking is an issue to be addressed in the future, with the examination of analysis conditions. Furthermore, the scientific sensory evaluation of aroma could not be carried out due to the limitations of human resources. Sensory tests will be performed in further studies.

Conclusions
In this study, we aimed to develop gluten-free rice noodles with a texture and flavor profile preferred by the Japanese people. The most appropriate raw rice flour for the noodles was selected based on its pasting properties. PS was used as the binder, and the blending ratio was set to 15% based on the stress-strain characteristics of the cooked noodles and upon comparison with those of udon noodles. In addition to the white rice noodles, as a product variation, brown rice noodles containing roasted brown rice flour were also developed. Both white and brown rice noodles were evaluated for their aroma, and the two types exhibited different aroma characteristics. We are currently offering prototypes of our developed product to the market, and they have been well-received. We believe that these products will enhance the gluten-free food industry.
Issues to be addressed in the future include scientific sensory evaluations of the texture and aroma characteristics, GC-MS analyses of the cooked rice noodles, and the development of brown rice noodles with better texture and flavor. These challenges will contribute to further improving the quality of gluten-free rice flour noodles.