D-psicose 3-epimerase plays a crucial role in the biotransformation of D-psicose. Under the catalysis of cobalt chloride, D-psicose-3-epimerase(DAE) can convert fructose into D-psicose ( Figure.1 ). However, the reported D-psicose 3-epimerase has poor thermal stability and low expression level, which cannot meet the requirements of industrial production. In this study, we found the coding genes of D-alloxone-3-epimerase from different sources, in order to screen D-alloxone-3-epimerase with strong thermal stability (Table 1).
Table1 : DAE from different sources
| Name | Sources |
| TcDAE | Thermoclostridium caenicola |
| TtDAE | Thermogutta terrifontis |
| NtDAE | Novibacillus thermophilus |
| DsDAE | Dorea sp. CAG 317 |
Therefore, using the yield of D-psicose to characterize the enzyme activity of DAE, we set different temperatures, but the artificially set temperature gradient is limited, and the optimal temperature of D-psicose 3-epimerase can be accurately predicted by modeling.. In the subsequent production process, we can try to ensure the most suitable temperature for conversion, and ultimately achieve the purpose of increasing conversion rate or yield.
Figure 1. DAE converts fructose into D-alloulose
1.Analysis process
2.Analysis results
The Friedman test was used to analyze the significant differences of multiple data. If the significant P value was 0.011 **, the statistical results were significant, indicating that there were significant differences between TcDAE, TtDAE, NtDAE and DsDAE. The Cohen 's f value is 1.073, which is a large difference.
3.Analysis steps
a. the normality test of variables, whether the overall distribution of the data presents a normal distribution. If the test passes, analysis of variance can be used.
b.Look at the Friedman test table. If significant, you can look at the median to analyze the difference.
c.If the Friedman test is significant, the difference can be quantitatively analyzed by effect quantitative analysis.
4. Detailed conclusions
Output Result 1 : Normality Test Results
Table 3.Normality test index results
| Variable name | Sample size | Mean value | Standard deviation | Bias angle | Kurtosis | S-W test | K-S test |
| TcDAE | 5 | 1.391 | 0.941 | -0.999 | -0.928 | 0.834(0.149) | 0.318(0.592) |
| TtDAE | 5 | 1.752 | 0.436 | 0.003 | -0.379 | 0.984(0.953) | 0.178(0.989) |
| NtDAE | 5 | 2.83 | 0.376 | 1.769 | 3.07 | 0.789(0.066*) | 0.279(0.746) |
| DsDAE | 5 | 1.52 | 0.425 | 0.438 | -3.002 | 0.839(0.162) | 0.243(0.867) |
| Note : * * *, * *, * represents the significance level of 1 %, 5 %, 10 % respectively | |||||||
Table 3 shows the results of descriptive statistics and normality test of quantitative variables TcDAE, TtDAE, NtDAE, DsDAE, including mean, standard deviation, etc. Normal distribution is usually applied to small sample data ( sample size ≤ 5000 test method is Shapiro-Wilk test, ) ;
The sample N of TcDAE was less than 5000, and the S-W test was used. The significant P value was 0.149, which was not significant at the level, and the original hypothesis could not be rejected. Therefore, the data met the normal distribution, and variance analysis was recommended.
The sample N of TtDAE was less than 5000, and the S-W test was used. The significant P value was 0.953, which was not significant at the level, and the original hypothesis could not be rejected. Therefore, the data met the normal distribution, and variance analysis was recommended.
The sample N of NtDAE is less than < 5000, and the S-W test is used. The significant P value is 0.066 *, which is not significant at the level, and the original hypothesis cannot be rejected. Therefore, the data meet the normal distribution, and variance analysis is recommended.
The sample N of DsDAE was less than 5000, and the S-W test was used. The significant P value was 0.162, which was not significant at the level, and the original hypothesis could not be rejected. Therefore, the data met the normal distribution, and variance analysis was recommended.
Output result 2 : Normality test histogram
Figure 2.Results of normality test. A.represents the results of the data positive test of TcDAE. B.represents the results of the data positive test of TtDAE. C represents the results of the data positive test of NtDAE. D represents the results of the data positive test of DsDAE.
Figure 2 shows the results of the normality test of data TcDAE, TtDAE, NtDAE and DsDAE. The normal graph basically shows a bell shape ( high in the middle and low at both ends ), which is basically acceptable as a normal distribution. Table 1 and Figure 2 prove the credibility of the data.
1.Modeling process
1.1 The model fitting process of TcDAE
After simulation, TcDAE conforms to the Gaussian model ; the functional relationship is as follows :
The model is characterized by a single peak ( maximum ) during the change process, and we used it to predict the optimum temperature of the enzyme TcDAE.
1.2 The model fitting process of TtDAE
After simulation, TtDAE conforms to the Gaussian 1 model, and its functional relationship is as follows:
The model is characterized by a single peak ( maximum ) during the change process, and we used it to predict the optimum temperature of the enzyme TtDAE.
1.3 The model fitting process of NtDAE
After simulation, NtDAE conforms to the Gaussian 1 model, and its functional relationship is as follows:
The model is characterized by a single peak ( maximum ) during the change process, and we used it to predict the optimum temperature of the enzyme NtDAE.
1.4 The model fitting process of DsDAE
After simulation, DsDAE conforms to the Gaussian 1 model, and its functional relationship is as follows:
The model is characterized by a single peak ( maximum ) during the change process, and we used it to predict the optimum temperature of the enzyme DsDAE.
All the above model codes are as follows :
TcDAE=[1.935 2.0465 2.145 0.83 0];
TtDAE=[1.173562673 1.666112152 2.316611215 2.016611215 1.585];
NtDAE=[2.557384824 2.65597281 2.860396544 3.47 2.607850652];
DsDAE=[1.12336839 1.36465 2.017 1.930 1.165];
[xData1, yData1] = prepareCurveData( temperature, TcDAE );
ft1 = fittype( 'gauss1' );
opts1 = fitoptions( 'Method', 'NonlinearLeastSquares' );
opts1.Display = 'Off';
opts1.Lower = [-Inf -Inf 0];
opts1.StartPoint = [2.145 60 19.7821621234724];
[fitresult1, gof] = fit( xData1, yData1, ft1, opts1 );
x=[40:0.5:80];
y1=fitresult1(x);
i1=find(y1==max(y1));
x1max=x(i1)
y1max=max(y1)
figure(1)
h1=plot(fitresult1);
h1.LineWidth = 1.2;
hold on
plot(xData1, yData1,'*','LineWidth',1.2)
plot(x1max, y1max,'*','LineWidth',1.2)
plot([x1max,x1max],[0,y1max],'b','LineWidth',1.2)
grid on
hold off
[xData2, yData2] = prepareCurveData( temperature, TtDAE );
ft2 = fittype( 'gauss1' );
opts2 = fitoptions( 'Method', 'NonlinearLeastSquares' );
opts2.Display = 'Off';
opts2.Lower = [-Inf -Inf 0];
opts2.StartPoint = [2.316611215 60 12.6234446945792];
[fitresult2, gof] = fit( xData2, yData2, ft2, opts2 );
y2=fitresult2(x);
i2=find(y2==max(y2));
x2max=x(i2)
y2max=max(y2)
figure(2)
h2=plot(fitresult2);
h2.LineWidth = 1.2;
hold on
plot(xData2, yData2,'*','LineWidth',1.2)
plot(x2max, y2max,'*','LineWidth',1.2)
plot([x2max,x2max],[0,y2max],'b','LineWidth',1.2)
grid on
hold off
[xData3, yData3] = prepareCurveData( temperature, NtDAE );
ft3 = fittype( 'exp2' );
opts3 = fitoptions( 'Method', 'NonlinearLeastSquares' );
opts3.Display = 'Off';
opts3.StartPoint = [2.52433868727362 0.0024161436616496 -6.29250334111916e-06 0.13530715255296];
[fitresult3, gof] = fit( xData3, yData3, ft3, opts3 );
y3=fitresult3(x);
i3=find(y3==max(y3));
x3max=x(i3)
y3max=max(y3)
figure(3)
h3=plot(fitresult3);
h3.LineWidth = 1.2;
hold on
plot(xData3, yData3,'*','LineWidth',1.2)
plot(x3max, y3max,'*','LineWidth',1.2)
plot([x3max,x3max],[0,y3max],'b','LineWidth',1.2)
grid on
hold off
[xData4, yData4] = prepareCurveData( temperature, DsDAE );
ft4 = fittype( 'exp2' );
opts4 = fitoptions( 'Method', 'NonlinearLeastSquares' );
opts4.Display = 'Off';
opts4.StartPoint = [-0.000745137762826976 0.100929373825454 0.451811175587761 0.0258951876842494];
[fitresult4, gof] = fit( xData4, yData4, ft4, opts4 );
y4=fitresult4(x);
i4=find(y4==max(y4));
x4max=x(i4)
y4max=max(y4)
figure(4)
h4=plot(fitresult4);
h4.LineWidth = 1.2;
hold on
plot(xData4, yData4,'*','LineWidth',1.2)
plot(x4max, y4max,'*','LineWidth',1.2)
plot([x4max,x4max],[0,y4max],'b','LineWidth',1.2)
grid on
hold off
figure(5)
h1=plot( fitresult1,'r');
h1.LineWidth = 1.5;
hold on
h2=plot( fitresult2,'b');
h2.LineWidth = 1.5;
h3=plot( fitresult3,'g');
h3.LineWidth = 1.5;
h4=plot( fitresult4,'y');
h4.LineWidth = 1.5;
plot(56.76,2.105,'*','LineWidth',1.2)
plot(59.72,1.873,'*','LineWidth',1.2)
grid on
hold off
2.Modeling results
2.1. Model fitting diagram of TcDAE
Figure 3. Gaussian model of TcDAE
The yield of D-psicose can be used to characterize the enzyme activity of TcDAE.The higher the yield of D-psicose, the higher the activity of TcDAE. Figure 3 shows that the yield of D-psicose increases first and then decreases with the increase of temperature. It was proved that the enzyme activity of TcDAE increased first and then decreased with the increase of temperature. At 50.5 degrees, the enzyme activity TcDAE was the highest.
2.2. Model fitting diagram of TtDAE
Figure 4. Gaussian model of TtDAE
Figure. 4 shows that the yield of D-psicose increases first and then decreases with the increase of temperature. It was proved that the enzyme activity of TtDAE increased first and then decreased with the increase of temperature. At 63.5 degrees, the enzyme activityTtDAE was the highest.
2.3. Model fitting diagram of NtDAE
Figure 5. The model fitting of NtDAE
Figure. 5 shows that the yield of D-psicose increased first and then decreased with the increase of temperature. It was proved that the enzyme activity of NtDAE increased first and then decreased with the increase of temperature. At 74 degrees, the enzyme activity NtDAE was the highest.
2.4. Model fitting diagram of DsDAE
Figure 6. The model fitting of DsDAE
Figure.6 shows that the yield of D-psicose increased first and then decreased with the increase of temperature. It was proved that the enzyme activity of DsDAE increased first and then decreased with the increase of temperature. At 67.5 degrees, the enzyme activity was the highest.
2.4. Comparison of TcDAE, TtDAE, NtDAE and DsDAE model
Figure. 7.Model fitting graphs of TcDAE, TtDAE, NtDAE and DsDAE
Compared with TcDAE, TtDAE, NtDAE and DsDAE, the yield of D-psicose increased first and then decreased(Figure 7). The higher the yield of NtDAE was significantly higher than that of other groups.