Table 1 and 2 show the point mutation sites provided by Hotspot Wizard. In the search for point mutations, reference was made to the original analysis of residue 291. All conserved residues in enzymes are composed of polar, nonpolar, acidic and basic groups. Most of the conserved residues, such as Tyr, Asn, Gly, Thr, Gln and Cys, belong to polar groups, followed by non-polar groups (Phe, Trp, Ala, Val and Pro), acidic groups (Glu and Asp) and basic groups (Arg and His). So we evaluated some functional residues as well as other residues.

Rosseta's Relax module was used to score these point mutation models. The initial model energy was calculated as -607.691. The energy of No. 18 Arg mutation was -332.407 and the structure was disorganized（Figure 1）, so the mutation scheme has been abandoned. The possible reason was that site 18 was mutated into Arg, and the surrounding area was positively charged, while Arg was negatively charged in the past, and the energy was too high, so the structure appeared to fall apart.

Figure 1

The energy of the Gln mutation 18 was -605.077. Although its structure did not fall apart, its energy increased and thus it was abandoned. The structure of Arg 20 was disintegrated after mutation, so it was abandoned too. (Figure 2)

Figure 2

However, the energy of Gly 16 was -610.561 after mutation. Meanwhile, compared with the structure before mutation, its RMSD value was 0.683, so it was selected for further mutation.

Figure 2

Finally, we screened the model with Leu mutation at site 67, and its energy was -614.826. Compared with the original model, the RMSD value was 0.97. Its reduced energy and relatively stable structure suggested that it might be a better variant of bromelain, as the experiments would prove.

The definition of upper relationship means that some single mutations of A are unfavorable, but the addition of B mutation will have better performance than single B mutation, which means that B has upper relationship with A. After calculating by Relax, we found that the Gly single mutation at site 16 had an upper relationship with the Leu single mutation at site 67: when Leu single mutation was detected, the structure was disordered, RM value was 1.4, and energy was -348. But with the addition of Gly, the structure stabilizes and the energy drops to -614.82. At the same time, the single mutation of Gly at site 16 has an upper relationship with the single mutation of Val at site 64: the energy of single mutation of Val is -599, while the energy of mutation of both Gly and Val decreases to -608.

After literature review and network search, BAEE was finally determined as the reaction substrate of bromelain, and its docking and energy calculation were performed with bromelain and mutated variants. The calculations were done online by Hdock.

Finally, the binding energy of bromelain in wild-type was -125.09, while that in mutant was -126.26, which decreased. This also proves that the stem bromelain variant we developed is more easily bound to the substrate and its reactivity is enhanced（Figure 4, 5, 6 and 7）

Figure 4

Figure 5

Figure 6

Figure 7

A molecular dynamics simulation was conducted to analyze the binding stability of Stem Bromelain(212aa) and BAEE complexes, where multiple descriptors were analyzed to understand the flexible and stable nature of the complexes. The system has been balanced in advance (the result of temperature and pressure balance is shown in below(Figure 8,9).

Figure 8: the result of temperature balance

Figure 9: the result of pressure balance

After the whole system was balanced, the molecular simulation started to run, which took 10h (1ns) in total.

RMSD and RMSF analysis were performed on the simulated results, and the results were shown in the figure below. It can be seen that Stem bromelain(212) has good reactivity with BAEE in neutral environment. The RMSD value was less than 0.4, and the RMSD value structure did not change much before and after simulation, indicating that the reaction between the complex was very stable. At the same time, the RMSF value changed greatly, which reflected that the atomic motion of Stem bromelain (212) was relatively free when it reacted with BAEE.

Figure 10: the RMSD of Stem Bromelain(212) and BAEE complex

Figure 11: the comparing RMSD of Stem Bromelain(212) and BAEE complex

Figure 12: the RMSF of Stem Bromelain(212) and BAEE complex

As for the mutant, since it can also dock with BAEE and its binding energy is better than that of Stem Bromelain (212), we predict that its RMSD map fluctuation should be smaller, and the data will be made up in the later molecular simulation.

Bromelain wild type, bromelain variant 1 and bromelain variant 2 were synthesized at 37°C, pH=7 and induced by iptg (50mg/mL, Sangon Biotech). After 3-4 hours all sluices containing the three proteins were collected and sonicated to lyse the bacteria. Correctly transformed and induced E. coli expressed bromelain wild type, bromelain variant 1 and bromelain variant 2 (Fig 13). We can see the bromelain (variant) bands around 25-35 kDa. By isolating and purifying pineapple protease variant 1 and variant 2 by Ni columns, we obtained a protein solution containing all three of these proteins.

Fig 13: Electrophoresis of proteins before and after transformation

In the experiment of formulating the working curve, we configured several standard L-tyrosine with gradient concentrations of 0, 10, 20, 30 and 40 respectively μg/ml; After a thorough reaction with Folin reagent for 20min, microplate reader was applied to measure the absorbance of the system after reaction at the wavelenth680nm. The A values obtained are 0.001, 0.436, 0.801, 1.096 and 1.293 respectively. From this series of data, the following working curve can be obtained:

The formula we applied to is: A=0.03246c+0.076, and the reaction is carried out at 40℃ then the absorbance of the system is measured:

Taking the wild stem bromelain for an example. The average value of the data is 0.552, and the concentration of L-tyrosine can be calculated as 14.66μg/ml according to the function. The amount of substance consumed in the reaction time is 0.08091μmol per minute. According to the formula provided in GB/T23527-2009:

Where X is the enzyme activity of the sample, A₁ is the enzyme activity of the reaction diluent obtained from the working curve (0.08091μmol here), V₁ is the volume of reaction vessel (15ml here), V₀ is the volume of the reaction system (10ml here), n is the dilution multiple of the sample (2 here), m₀ is the mass of the sample taken (0.5g here), and T is the time consumed (10min here). It can be calculated that the activity of our p.(S16G), p.(S16G,W67L) and wild type are respectively 6900U, 7344U, 4854U per gram of coarse cell extracts in the buffer.

——Only p.(S16G,W67L)

The experimental process is basically consistent with the contents above. The reaction temperature is raised to 50, 55 and 60℃, and three parallel groups are still set at each temperature. After reacting with Folin reagent, the absorbance of the system is measured at 680nm:

The calculation process will not be repeated. Therefore, the average absorbance values at three temperatures are 0.510, 0.266 and 0.087, and the enzyme activities at these three temperatures are 4412U, 1915U and 1541U per gram of coarse cell extracts in the buffer respectively. The enzyme activity-temperature diagram can be plotted from the above data:

Overall, our enzyme activity data confirms the success of the initial mutant as well as the model. We produced a bromelain variant with a 1.5-fold increase in activity in our final design.

In order to determine the optimal induction concentration of iptg, we prepare to add iptg to induce expression when the od values of E. coli are 0.6, 0.8, 1.0, 1.2, and study the expression results. However, after many times of measurement with the microplate reader, we found that the od value of the bacterial fluid always fluctuate around 0.3, and no matter how we adjusted the culture time, this value would not increase too much. We speculate that this is because the T7 expression system which we selected has leaked expression. However, the leaked expression of bromelain decomposes the endogenous protein of E. coli, causing it to be unable to continue to increase in value ---the bands showed at 25-35KD in the sample which was not induced can prove this hypothesis to a certain extent.

We have transformed the bromelain variant into Bacillus subtilis, but due to the epidemic, we have not fully completed our experiments with Bacillus subtilis as the chassis for bromelain production and believe that our mutant will perform better after undergoing glycosylation in a eukaryotic chassis.