Carbohydrates are the primary energy source required for all living organisms' life activities. They are not only nutritional substances, but some also possess special physiological activities. Excessive intake of sugars such as sucrose and starch in daily diets can lead to increased blood lipids and blood sugar levels, which may result in obesity, diabetes, arteriosclerosis, and myocardial infarction. Therefore, converting absorbable carbohydrates into non-digestible sugars with specific physiological activities can alleviate the high sugar burden in diets and promote human health. It is a research hotspot in food nutrition and health.

Based on food enzyme engineering strategies, we utilized dextransucrase (EC 2.4.1.5) and inulosucrase (EC 2.4.1.9). We obtained two recombinant proteins through heterologous recombinant expression technology and explored their catalytic performance in converting sucrose into dextran and oligofructose, which are dietary fibers. Combining these two recombinant enzymes, we created a composite enzyme capsule, achieving a "sugar control-fiber enhancement" composite enzyme capsule that reduces carbohydrate absorption while promoting bowel movements and regulating gut microbiota.

To further investigate the optimal amounts and timing for combining the two recombinant enzymes, we used commercially available cane juice as the sample. We utilized triangular flasks as containers and incrementally added specific amounts of the two enzyme solutions (1 mg, 5 mg, 10 mg, 15 mg, and 20 mg). The reactions were conducted at 37°C, and samples were taken at set intervals (5 min, 10 min, 15 min, 30 min, and 60 min). After the reaction, samples were heated to 65°C to inactivate the enzymes. Subsequently, the changes in the contents of sucrose and D-glucose were measured using high-performance liquid chromatography (HPLC).

 

Table 1: Sucrose (g) vs. Time (min) and Composite Enzyme (mg) Experimental Results

 

Table 2: Glucose(g) vs. Time (min) and Composite Enzyme(mg) Experimental Results

3.1 Sucrose Data - Normality Test

First, we conducted normality tests on the measured sucrose variation data using the Shapiro-Wilk test (for small sample sizes, generally less than 5000) or the Kolmogorov-Smirnov test (for large sample sizes, usually greater than 5000) to assess significance. If the results do not show significance (P > 0.05), it indicates that the data follow a normal distribution; otherwise, it suggests a non-normal distribution. In practical research, meeting the normality assumption is often challenging. However, suppose the absolute kurtosis values are less than 10 and skewness is less than 3, along with normal distribution histograms, P-P plots, or Q-Q plots. In that case, the data can be described as approximately usual.

Table 3 presents descriptive statistics and results from the normality tests for the sucrose data collected at 5 min, 10 min, 15 min, 30 min, and 60 min. This includes median, mean, and other statistics used to assess normality.

For the samples at 5 min, 15 min, 30 min, and 60 min (N < 5000), the Shapiro-Wilk test was applied, yielding significance P-values of 0.244, 0.391, 0.186, and 0.140, respectively. None of these values show significance, allowing us to fail to reject the null hypothesis; therefore, the data satisfy the normal distribution assumption.

For the 10-minute sample (N < 5000), the Shapiro-Wilk test resulted in a significant P-value of 0.005***, leading us to reject the null hypothesis; hence, this data does not meet the average distribution criterion. Its kurtosis (4.687) has an absolute value of less than 10, and its skewness (2.146) has an absolute value of less than 3, which allows for further analysis using normal distribution histograms.

Table 3: Descriptive Statistics and Normality Test Results for Sucrose

 

Figure 1 displays the histogram of sucrose data variation at the 5 min mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can be generally accepted as approximately usual.

Figure 1. Normality Test Histogram of Sucrose Data at 5 Minutes

 

Figure 2 displays the histogram of sucrose data variation at the 10 min mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can be generally accepted as approximately usual.

Figure 2. Normality Test Histogram of Sucrose Data at 10 Minutes

 

Figure 3 displays the histogram of sucrose data variation at the 15 min mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can be generally accepted as approximately usual.

Figure 3. Normality Test Histogram of Sucrose Data at 15 Minutes

 

Figure 4 displays the histogram of sucrose data variation at the 30-minute mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can be generally accepted as approximately usual.

Figure 4. Normality Test Histogram of Sucrose Data at 30 Minutes

 

Figure 5 displays the histogram of sucrose data variation at the 60-minute mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can be generally accepted as approximately usual.

Figure 5. Normality Test Histogram of Sucrose Data at 60 Minutes

 

3.2 Glucose Data - Normality Test

We first performed normality tests on the measured glucose variation data using the Shapiro-Wilk test (for small sample sizes, generally less than 5000) or the Kolmogorov-Smirnov test (for large sample sizes, usually greater than 5000) to assess significance. If the results do not show significance (P > 0.05), it indicates that the data follow a normal distribution; otherwise, it suggests a non-normal distribution. In practical research, meeting the normality assumption is often challenging. However, if the absolute kurtosis values are less than 10 and skewness is less than 3, along with average distribution histograms, the data can be described as approximately usual.

Table 4 presents descriptive statistics and results from the normality tests for the glucose data collected at 5 min, 10 min, 15 min, 30 min, and 60 min. This includes median, mean, and other statistics used to assess normality.

For the samples at 5 min, 10 min, 15 min, 30 min, and 60 min (N < 5000), the Shapiro-Wilk test was applied, yielding significance P-values of 0.978, 0.440, 0.587, 0.859, and 0.373, respectively. None of these values show significance, allowing us to fail to reject the null hypothesis; therefore, the data satisfy the standard distribution assumption.

Table 4: Descriptive Statistics and Normality Test Results for Glucose

 

Figure 6 displays the histogram of glucose data variation at the 5 min mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can generally be accepted as approximately usual.

Figure 6. Normality Test Histogram of Glucose Data at 5 Minutes

Figure 7 displays the histogram of glucose data variation at the 10 min mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can generally be accepted as approximately usual.

Figure 7. Normality Test Histogram of Glucose Data at 10 Minutes

 

Figure 8 displays the histogram of glucose data variation at the 15 min mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can generally be accepted as approximately usual.

Figure 8. Normality Test Histogram of Glucose Data at 15 Minutes

Figure 9 displays the histogram of glucose data variation at the 30-minute mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can generally be accepted as approximately usual.

Figure 9. Normality Test Histogram of Glucose Data at 30 Minutes

Figure 10 displays the histogram of glucose data variation at the 60-minute mark. Suppose the histogram resembles a bell shape (high in the middle and low at both ends). In that case, it indicates that while the data may not be perfectly normal, it can generally be accepted as approximately usual.

Figure 10. Normality Test Histogram of Glucose Data at 60 Minutes

We used MATLAB software to perform three-dimensional visualization of the processed data and established functions to plot three-dimensional graphs. We also solved for the optimal solutions of the model and determined the extrema of the functions.

 

4.1 Sucrose Modeling Analysis

4.1.1 Fitting the Raw Sucrose Data

In this section, we fitted the raw sucrose data to determine the relationship between the variables. This process included selecting appropriate mathematical models, applying regression techniques, and analyzing the goodness of fit to ensure that the model accurately represents the observed data.

clear;clc;

x=[1 5 10 15 20];

y=[5 10 15 30 60];

z=[0.76 0.74 0.72 0.71 0.24;

    0.53 0.43 0.41 0.42 0.41;

    0.5 0.42 0.089 0.473333333 0.436666667;

    0.39 0.40 0.42 0.44 0.39;

    0.383333333 0.446666667 0.403333333 0.35 0.003333333];

%function [fitresult, gof] = createFit(x, y, z)

%CREATEFIT(X,Y,Z)

%  Create a fit.

%  Data for 'untitled fit 1' fit:

%      X Input : x

%      Y Input : y

%      Z Output: z

%  Output:

%      fitresult : a fit object representing the fit.

%      gof : structure with goodness-of fit info.

% Fit: 'untitled fit 1'.

[xData, yData, zData] = prepareSurfaceData( x, y, z );

 % Set up fittype and options.

ft = 'thinplateinterp';

 % Fit model to data.

[fitresult, gof] = fit( [xData, yData], zData, ft, 'Normalize', 'on' );

% Plot fit with data.

figure( 'Name', 'untitled fit 1' );

h = plot( fitresult, [xData, yData], zData );

legend( h, 'untitled fit 1', 'z vs. x, y', 'Location', 'NorthEast' );

% Label axes

xlabel x

ylabel y

zlabel z

grid on

view( -41.1, 28.8 );

 

Figure 11. Three-Dimensional Plot of Sucrose Content Variation

 

4.1.2 Solving for the Optimal Value of the Model

Based on the calculations, when the amount of composite enzyme is set to 10 mg and the digestion time is 48 minutes, the model computes a minimum negative value. Considering the practical context, we take this value as 0 g. Therefore, the sucrose is completely digested when the composite enzyme amount is 10 mg and the digestion time is 48 minutes.

The code used for this optimization is as follows:

clear;clc;

x=[1 5 10 15 20];

y=[5 10 15 30 60];

z=[0.76 0.74 0.72 0.71 0.24;

    0.53 0.43 0.41 0.42 0.41;

    0.5 0.42 0.089 0.473333333 0.436666667;

    0.39 0.40 0.42 0.44 0.39;

    0.383333333 0.446666667 0.403333333 0.35 0.003333333];

xx=[1:0.5:20];

yy=[5:1:60];

[xi,yi]=meshgrid(xx,yy);

zi=interp2(x,y,z,xi,yi,'spline');

surf(xi,yi,zi)

[imin,jmin]=find(zi==min(min(zi)))

xmin=xx(jmin),ymin=yy(imin),zmin=zi(imin,jmin)

[imax,jmax]=find(zi==max(max(zi)));

xmax=xx(jmax),ymax=yy(imax),zmax=zi(imax,jmax)

 

4.2 Glucose Modeling Analysis

4.2.1 Fitting the Raw Glucose Data

This section will fit the raw glucose data to identify the underlying relationships between the variables involved. This process will involve selecting suitable mathematical models, applying regression techniques, and evaluating the goodness of fit to ensure the model accurately represents the observed glucose data.

clear;clc;

x=[1 5 10 15 20];

y=[5 10 15 30 60];

z=[2.4 2.5 2.8 2.8 3.366666667;

   2.5 2.566666667 2.4 2.866666667 3.9;

   2.666666667 2.633333333 2.633333333 2.666666667 2.733333333;

   2.266666667 2.5 2.466666667 2.7 2.666666667;

   2.566666667 2.466666667 2.5 2.5 2.933333333];

%function [fitresult, gof] = createFit(x, y, z)

%CREATEFIT(X,Y,Z)

%  Create a fit.

%  Data for 'untitled fit 1' fit:

%      X Input : x

%      Y Input : y

%      Z Output: z

%  Output:

%      fitresult : a fit object representing the fit.

%      gof : structure with goodness-of fit info.

%% Fit: 'untitled fit 1'.

[xData, yData, zData] = prepareSurfaceData( x, y, z );

 % Set up fittype and options.

ft = 'thinplateinterp';

 % Fit model to data.

[fitresult, gof] = fit( [xData, yData], zData, ft, 'Normalize', 'on' );

 % Plot fit with data.

figure( 'Name', 'untitled fit 1' );

h = plot( fitresult, [xData, yData], zData );

legend( h, 'untitled fit 1', 'z vs. x, y', 'Location', 'NorthEast' );

% Label axes

xlabel x

ylabel y

zlabel z

grid on

view( -37.3, 8.5 );

 

Figure 12. Three-Dimensional Plot of Glucose Content Variation

 

4.2.2 Solving for the Optimal Solution of the Model

Based on the calculations, when the amount of composite enzyme is set to 1 mg and the digestion time is 46 minutes, the model computes a minimum value of 1.6318 g. This indicates that the addition of the composite enzyme indeed promotes glucose production.

Furthermore, when the enzyme amount is increased to 10 mg and the digestion time is set to 49 minutes, the model calculates a maximum value of 5.0486 g. Consequently, when the composite enzyme amount is set to 20 mg with the same digestion time of 49 minutes, the maximum glucose produced from degradation reaches 5.0486 g, while the remaining sucrose is converted into dietary fiber.

We utilized sugarcane juice as a model food suspension to test the functionality of composite enzymes under real conditions. By analyzing the raw data and three-dimensional plots, we can conclude that, generally, a longer reaction time leads to more complete catalysis, resulting in higher glucose concentrations and lower sucrose concentrations.

Specifically, with a composite enzyme amount of 10-20 mg and a digestion time of approximately 50 minutes, the residual sucrose concentration in the solution approaches 0 g, while the resulting glucose concentration is notably high, around 5.0486 g. The calculated results from the model align well with the actual measured data, confirming that the dextransucrase and inulosucrase break down sucrose into dextran and oligofructose (i.e., glucose).

This demonstrates the potential of these enzymes in enhancing sugar processing and improving nutritional outcomes in food applications.

1.      Scientific Platform Serving for Statistics Professional 2021. SPSSPRO. (Version 1.0.11) [Online Application Software]. Retrieved from https://www.spsspro.com.
2.      Zong Xuping, Yao Yulan. Rapidly Testing Statistical Distributions of Data Using Q-Q and P-P Plots. Statistics and Decision, 2010(20):2.