1. Comparison between 2A system and CRISPR-Csy4 system
We selected two fluorescent proteins (yeGFP and DsRed) as reporter genes to compare and analyze 2A system and CRISPR-Csy4 system on four aspects, which include plasmid construction methods, fluorescence observation results, protein expression and mRNA expression (through RT-qPCR).
1.1 Plasmid selection
In our project, pY26TEF-GDP was selected as the plasmid vector. It was designed as shuttle plasmid between Escherichia coli and Saccharomyces cerevisiae, which has ampicillin resistance gene and URA3 selective marker. In addition, the plasmid also has two constitutive promoters, pGAP and pTEF1, which reduces the difficulty of constructing CRISPR-Csy4 system.
Fig 1. The pY26TEF-GDP on snapgene
1.1.1 Plasmid construction of 2A system
We added BamH I, Hind III sites and Not I, Bgl II sites to NLS-yeGFP-T2A-DsRed by PCR (Fig. 2a) and assembled them to the downstream of pGAP (BBa_K3544206) and the downstream of pTEF1(BBa_K3544504) (Fig. 3).
The two plasmids were constructed to verify that 2A system is effective in S. cerevisiae, and to verify that these two promoters can express the target protein in S. cerevisiae strain EGY48 with similar intensity.
Fig 2. (a)PCR M: Takara DL5,000 DNA Marker; 1-3: NLS-yeGFP-T2A-DsRed with BamH I, Hind III; 4-6: NLS-yeGFP-T2A-DsRed with Not I, Bgl II
(b) Enzyme digestion M1: Takara DL5,000 DNA Marker; M2: Takara DL10,000 DNA Marker; 1-2: NLS-yeGFP-T2A-DsRed (digested by BamH I, Hind III); 3-4: NLS-yeGFP-T2A-DsRed (digested by Not I, Bgl II); 5: pY26TEF-GDP ; 6: pY26TEF-GDP (digested by BamH I, Hind III)
C) Enzyme digestion M: Takara DL10,000 DNA Marker; 1: pY26TEF-GDP ; 2: pY26TEF-GDP (digested by Not I, Bgl II)
The constructed plasmids were transferred to DH5α, spreaded on LB solid medium (with ampicillin, following as A+), and cultured overnight at 37℃. The next day, the single-colony was selected for expansion culture, and the plasmid was extracted after 16h. Four kinds of restriction enzymes were used for digestion verification (Fig. 3).
Fig 3. Enzyme digestion M1: Takara DL10,000 DNA Marker; M2: Takara DL2,000 DNA Marker; 1,3: pY26TEF-GDP with BBa_K3544206; 2,4: pY26TEF-GDP with BBa_K3544206 (digested by BamH I, Hind III); 5,7: pY26TEF-GDP with BBa_K3544504; 6,8: pY26TEF-GDP with BBa_K3544504 (digested by Not I, Bgl II)
Plasmids verified by enzyme digestion were sequenced and checked correctly (Fig. 4)，which were stored at -20℃ for the yeast transformation.
Fig 4. The sequencing results of BBa_K3544206 and BBa_K3544504 sequencing
1.1.2 Plasmid construction of CRISPR-Csy4 System
In order to change T2A in NLS-yeGFP-T2A-DsRed into CRISPR(Csy4) sequence, we designed two pairs of primers refer to the MCS behind pGAP. One pair of primers was used to add BamH I and EcoR I sites on the end of NLS-yeGFP (Fig. 6a), and the other pair used DsRed as a template, adding EcoR I sites and CRISPR(Csy4) to the upstream of DsRed, and Hind III sites to the downstream of DsRed (Fig. 6b).
Fig 5. The primer design of CRISPR-Csy4 System
Fig 6. PCR M:Takara DL2,000 DNA Marker
(a)1-5: NLS-yeGFP with BamH I and EcoR I
(b)1-4: DsRed with EcoR I, CRISPR, Hind III
We hope to link the digested NLS-yeGFP and CRISPR-DsRed with the carrier vector in multiple fragments, but after two attempts, we find that the ratio of the two fragments is difficult to control.
Due to the short experimental period caused by the COVID-19, we stopped trying this method. Instead, we assembled the CRISPR-DsRed (BBa_K3544104) into the plasmid firstly, and assembled the NLS-yeGFP(BBa_K3544103) to the upstream of CRISPR-DsRed (Fig. 7, Fig. 8). (The intermediate product (CRISPR-DsRed) of this step played a positive control role in the 2A system verification.)
Fig 7. Enzyme digestion (a) M1: Takara DL15,000 DNA Marker; 1-2: pY26TEF-GDP (digested by BamH I, EcoR I); 3-5: CRISPR-DsRed (digested by BamH I, EcoR I)
(b) M:Takara DL10,000 DNA Marker; 1: pY26TEF-GDP; 2-4: pY26TEF-GDP with BBa_K3544104 (digested by EcoR I, Hind III); 5-7: NLS-yeGFP (digested by EcoR I, Hind III)
Fig 8. Enzyme digestion (a) M1: Takara DL10,000 DNA Marker; 1: pY26TEF-GDP; 2: pY26TEF-GDP (digested by BamH I, EcoR I); 5: pY26TEF-GDP with CRISPR-DsRed; 6: pY26TEF-GDP with CRISPR-DsRed (digested by BamH I, EcoR I)
(b) M:Takara DL10,000 DNA Marker; 1: pY26TEF-GDP ; 2: pY26TEF-GDP with BBa_K3544105 (digested by EcoR I, Hind III)
Fig 9. The sequencing results of BBa_K3544105
After the successful sequencing of NLS-yeGFP-CRISPR-DsRed (BBa_K3544105), we added Not I and Bgl II sites on both ends of csy4 gene by PCR and assembled them downstream of pTEF1 (Fig. 10). The downstream primer of Csy4 and the downstream primer of DsRed were used to sequence the plasmid (with BBa_K3544503 & BBa_K3544888). After the whole system was constructed, the plasmid was stored at -20℃ to prepare for yeast transformation.
Fig 10. Enzyme digestion (a)M: Takara DL10,000 DNA Marker; 1: pY26TEF-GDP with BBa_K3544888; 2: pY26TEF-GDP with BBa_K3544888 (digested by Not I, Bgl II); 3-4: NLS-Csy4 (digested by Not I, Bgl II)
b）M: Takara DL10,000 DNA Marker; 3,5,7,9,11: pY26TEF-GDP with BBa_K3544888 & BBa_K3544503; 4,6,8,10,12: pY26TEF-GDP with BBa_K3544888 & BBa_K3544503 (digested by Not I, Bgl II)
C) The sequencing results of BBa_K3544503 and BBa_K3544888
1.2 Fluorescence observation
In this part, we used stereo fluorescence microscope and inverted fluorescence microscope to observe the yeast colonies and single-cell fluorescence after the plasmid was transformed. This can not only help to select yeast single-colonies that successfully transformed plasmid, but also preliminarily verify the feasibility of 2A system and CRISPR-Csy4 system.
1.2.1 Yeast selection
Last year, 2019 SCU-China encountered a problem when we want to transfer plasmids into yeast strains BY4741 and YM4271. The success rate of plasmid transformation decreased significantly when the size of plasmid was greater than 10 kbp. Unfortunately, in this year, the final plasmid size for ABA producing is 13 kbp. After consulting the technical staff of Shanghai Weidi Biotechnology company, he recommended the EGY48 strain to us, which has higher plasmid transformation efficiency than BY4741 and uses URA3 as the main selective marker. All of these cases of EGY48 is very consistent with our plasmid vector pY26TEF-GDP.
1.2.2 Fluorescence observation of 2A system
The plasmids (1.1.1) constructed with the 2A system were transferred to EGY48. They were spread on a SD-Ura plate (with Ampicillin) and cultured at 29℃ for 48 hours. After the white single-colonies grow out, observe the fluorescence of the two groups of yeasts under the stereo fluorescence microscope (Fig. 11, shooting environment is set to 12 x magnification,500ms, 1 times the brightness gain).
Fig 11. The result of fluorescence observation under stereo fluorescence microscope
The results indicate that pTEF1 and pGAP can effectively turn on protein expression in EGY48, and their intensities are similar (details are compared in fluorescence of microplate reader). However, we also observed that some tiny colonies appeared on the plate under the bright field of the microscope. At first, we thought they were yeast colonies that failed to grow up successfully, while the subsequent selecting experiment of miscellaneous bacteria showed that these may be miscellaneous bacteria brought into the system during plasmid transformation (see 3.2 laboratory miscellaneous bacteria for details).
To observe the nuclear entry of NLS-yeGFP-T2A-DsRed, we selected single-colonies on two groups of plates and cultured them in SD-Ura liquid medium containing chloramphenicol (as C+ in the following parts) for two days, and then took 5μL bacterial solution to observe it under inverted fluorescence microscope (Fig. 12, shooting environment is set to 1000 x magnification, 100 ms, 64 times the brightness gain).
Fig 12. The result of fluorescence observation under inverted fluorescence microscope
As shown in the above figure, yeGFP of both groups have obvious nuclear entry phenomenon and DsRed has shown brightness in the whole cell, which indicates that NLS sequence is effective in EGY48, and 2A system can separate NLS-yeGFP-T2A-DsRed and leave DsRed in cytoplasm. However, we observed that there was a certain red fluorescence in the nucleus. We speculated the reason was some fusion proteins brought into the nucleus for NLS, which was related to the efficiency of T2A cleavage (see 1.3 Protein Verification for details).
1.2.3 Fluorescence observation of CRISPR-Csy4 system
The plasmid NLS-Csy4/NLS-yeGFP-CRISPR-DsRed(pY26TEF-GDP) and intermediate product CRISPR-DsRed(pY26TEF-GDP) were transferred into EGY48. They were spread on the SD-Ura plate (C+) and cultured at 29℃ for 48 hours, and observed under stereo fluorescence microscope(Fig 13, shooting environment is set to 12 x magnification,500ms, 1 times the brightness gain).
Fig 13. The result of fluorescence observation under stereo fluorescence microscope
The fluorescence of CRISPR-Csy4 system in the above figure is consistent with the expectation in Design, that yeGFP can be observed obvious fluorescence while DsRed is not expressed. To further verify that the red fluorescence in the nucleus is caused by fusion proteins entering the nucleus, we observed the expression of DsRed on CRISPR-DsRed(pY26TEF-GDP). Unexpectedly, the yeast that only express DsRed has weak green fluorescence under the excitation light of 488nm, which is presumed to be caused by the longer excitation spectrum (400-600nm) of DsRed[1](see 1.2.4 fluorescence comparison of microplate reader for details).
After the expanding cultured under the same conditions of 2A system group, 5μl of bacterial liquid was taken for observation(Fig. 14, shooting environment is set to 1000 x magnification, 100 ms, 64 times the brightness gain).
Fig 14. The result of fluorescence observation under inverted fluorescence microscope
The result of CRISPR-DsRed is as expected, the red fluorescence is evenly distributed in the cell without entering the nucleus. However, we found that under the same observation conditions, yeGFP, which should have entered the nucleus in the CRISPR-Csy4 system, has a certain accumulation in the cytoplasm, and this is not present in the comparison 2A system. In terms of the size of the proteins carried by NLS, the 2A system group contains NLS-yeGFP and NLS-yeGFP-T2A-DsRed, while the CRISPR-Csy4 system group only expressed NLS-yeGFP, which might be easier to enter the nucleus. Therefore, we presume that the reason for this result is the high expression of NLS-yeGFP, which failed to enter the nucleus in time (see 1.3 protein expression verification).
1.2.4 Relative fluorescence intensity of yeGFP and DsRed
To visualize the expression level of pGAP and choose the appropriate promoter for NLS-Csy4 and function genes, we measure the fluorescence intensity, and the data was divided by OD.
Fig 15. Relative fluorescence intensity of two fluorescence proteins
The abnormal GFP signal of pGAP-R could be attributed to immaturity of DsRed[1].
The signal of DsRed in pGAP-R is low, this is related to our plasmid design. We use double enzyme digestion and T4 ligation to construct pGAP-NLS-yeGFP-CRISPR-DsRed, the first step is construction of pGAP-CRISPR-DsRed, which is called pGAP-R here. The secondary structure of CRISPR，which is same like an attenuator， may influences the expression level of DsRed.
In short, pTEF1 is stronger than pGAP (judged by the fluorescence of GFP and DsRed), so we use pTEF1 to express NLS-Csy4 for high cutting efficiency, and to design the construction of function genes additionally according to our modeling result.
There is also an interesting phenomenon that the GFP fluorescence of pTEF1-NYR is weaker than pTEF1-NC/PGAP-NYR. The difference between them is the additional polypeptide -RAEGRGSLLTCGDVEENPG formed after 2A self-cutting, which means 2A may influence function of gene upstream.
We applied modeling to verify our suppose.
See more in Modeling.
1.3 Verification of protein expression
To observe whether the 2A system cleaves DsRed from NLS-yeGFP more intuitively, we extracted yeast proteins from two groups of 2A systems and two groups of CRISPR-Csy4 systems for SDS-PAGE experiment and Western Blot. experiment.
1.3.1 Protein extraction
In order to ensure that the total amount of yeast protein extracted is similar, the above four groups of yeasts were added into 15ml SD-Ura liquid medium (C+). At same time, we added EGY48 to YPDA medium (C+) as control, and cultured them in 50ml centrifuge tube at 29℃ until the OD 600 was about 1.00.
We used Sangon Biotechnology Yeast Total Protein Extraction Kit to extract yeast total protein (see protocol for details). About 500μL of total protein with a concentration of 2.50 mg / ml could be extracted from 10 ml yeast culture media. The following results can be obtained by observing the protein solution under a stereo fluorescence microscope(Fig. 16, shooting environment is set to 1 x magnification,100ms, 1 times the brightness gain).
Fig 16. The result of fluorescence observation under stereo fluorescence microscope
The results indicated that although the yeast protein extraction kit could extract yeGFP and DsRed protein, the fluorescence intensity of DsRed was significantly lower than yeGFP. Because of time limitation, we didn't go deep into this problem, instead we focused on yeGFP for the verification of 2A system.
Fig 17. Protein extraction NC: EGY48; 1: EGY48 with pY26TEF-GDP(BBa_K3544206);
2: EGY48 with pY26TEF-GDP(BBa_K3544504); 3: EGY48 with pY26TEF-GDP(BBa_K3544888 & BBa_K3544503);
4: EGY48 with pY26TEF-GDP(BBa_K3544204)
As shown in the above figures, the total protein concentration of four tubes protein solution were similar, while the fluorescence intensity of yeGFP in CRISPR-Csy4 system is higher than that in 2A system of two groups. On the one hand, it maybe occur for that the cleaved short-chain mRNA in CRISPR-Csy4 system is more suitable for protein expression, on the other hand, the residual 2A sequence in the tail of yeGFP may affects its fluorescence.
See Modeling for details.
1.3.2 Verification of coomassie brilliant blue staining by SDS-PAGE
To show that the red fluorescence in the nucleus is caused by incomplete separation of 2A sequence, we hoped to verify by SDS-PAGE and obtained the following figure.
Fig 18. M: TransGen protein ruler IV; NC: EGY48; pGAP-NYD: EGY48 with pY26TEF-GDP(BBa_K3544206); pTEF1-NYD: EGY48 with pY26TEF-GDP(BBa_K3544504); pTEF1-DsRed: EGY48 with pY26TEF-GDP(BBa_K3544888 & BBa_K3544503); pGAP-NC/pTEF1-NYD: EGY48 with pY26TEF-GDP(BBa_K3544204)
According to the prediction of Snapgene, NLS-yeGFP-T2A-DsRed is 57kDa, NLS-yeGFP is 31kDa, DsRed is 26kDa, and NLS-yeGFP in CRISPR-Csy4 is 28kDa. Unfortunately, these molecular weights can find dense bands in yeast negative control group, which makes us unable to distinguish the target bands, so we chose Western Blot instead to verify our conjecture.
1.3.3 Protein verification by Western Blot
To show that the red fluorescence in the nucleus is caused by incomplete separation of 2A sequence, we hoped to verify by SDS-PAGE and obtained the following figure.
After 5 groups samples finished SDS-Page electrophoresis, we transferred the protein to PVDF membrane by wet transfer method at 100V for 95min. Then, the membrane was sealed with sealing buffer (2g skimmed milk powder +50ml PBS) for 1h. After sealing, the primary antibody was incubated overnight at 4℃ and the secondary antibody was incubated at room temperature for 1h. Finally, ECL developer was used for chemiluminescence and photography(Fig. 19).
Fig 19. Western Blot result of empty vector, yeGFP alone, DsRed alone, 2A system. （From left to right) lane 1: empty carrier, lane 2: pGAP-yeGFP-T2A-DsRed, lane 3: pTEF1-NLS-yeGFP-T2A-DsRed, lane 4: pTEF1-DsRed, lane 5: pTEF1-NLS-yeGFP
The results showed that the correct size bands appeared at 57kDa and 31kDa in Lane 2 and Lane 3, corresponding to the size of NLS-yeGFP-T2A-DsRed and NLS-yeGFP, which indicated that our 2A system could achieve the effect of protein separation to a certain extent, but this was still limited by the efficiency of 2A sequences. This can also explain what we saw before under the fluorescence microscope: under the excitation light of 488nm, our yeast produced red fluorescence both inside and outside the nucleus.
At the same time, we also noticed that the expression of yeGFP in Lane 5 was significantly higher than the total amount of the upper and lower bands in Lane 2 and Lane 3 when the concentration of proteins were similar. This is consistent with the results of fluorescence intensity control (1.2.4) of microplate reader, indicating that the protein expression in CRISPR-Csy4 system was higher than 2A system.
See Measurement for details.
1.4 RT-qPCR reveals the cleveage of CRISPR
To investigate the cutting efficiency of Csy4 in EGY48, we design three pairs of primers showed as follow:
Fig 20. Primers for RT-qPCR
F1 and R1 match with NLS-yeGFP, F3 and R3 match with DsRed. Meanwhile, F2 matches with the downstream of NLS-yeGFP and R2 matches with the upstream sequence of DsRed. As the result, the relatively normalized expression of the yeGFP-CRISPR spacer-DsRed(credible,Cq<30) is far less than yeGFP and DsRed gene. Considering yeGFP-CRISPR spacer-DsRed has no value in three duplications of negative control, we suggested the yeGFP-CRISPR spacer-DsRed transcribed successfully in S.cerevisiae and the CRISPR spacer showed a high efficiency of cleavage by Csy4.Our result shows the high cutting efficincy of Csy4.
See More in Measurement.
Fig 21. The Result of RT-qPCR
However, the RFU of DsRed has a higher level than yeGFP, this may attribute to these two reasons: 1. The decrease of stability of yeGFP RNA due to the lack of 3’ tail. 2. The increase of stability of DsRed RNA due to the 5’-OH formed by digestion of Csy4, which is consistant with our reference.
See More in Measurement.
1.5 Discussion
Through the above four experiments, we have summarized the following differences between two systems:
Fig 22. The differences between two systems
As shown in that above table, compared with CRISPR- Csy4 system, which require additional expression of the NLS-Csy4 protein, 2A system can express multiple proteins in the same ORF only by linking the target protein with 2A sequence, which is more convenient and faster in plasmid construction.
However, the fluorescence intensity of NLS-yeGFP expressed in CRISPR-Csy4 system using low-intensity pGAP promoter was higher than that expressed in 2A system using high-intensity pTEF1. The modeling results suggest that this may be due to the reduction of protein monomer fluorescence brightness caused by the residual 2A sequence in the tail of NLS-yeGFP. This is to say, only looking at the expression effect of yeGFP, CRISPR-Csy4 system is better.
Furthermore, the results of Western blot and RT-qPCR showed that the mRNA cleavage efficiency of CRISPR-Csy4 system was much higher than the protein separation efficiency of 2A system. Based on the above results, we believe that after solving the 5 '-end cap problem at the downstream of mRNA after Csy4 cleavage, CRISPR-Csy4 system can not only be applied in eukaryotic systems, but also surpass the existing 2A system in some aspects.
2. Expression of bcABA gene in yeast and production of ABA
2.1 Construction of pTEF1-BcABA3-2/pGAP-BcABA1-4
As our mathematic model revealed that expression of these four enzymes should be as BcABA3>BcABA1>BcABA2>
BcABA4, and the efficiency of 2A peptide could lead to a lower expression level to gene downstream, we choose to construct BcABA3 in the ORF of promoter pTEF1, which is stronger than pGAP(See our results of relative fluorescence intensity of yeGFP and DsRed), with BcABA2 linked by T2A. Not I and Bgl II were added to BcABA3-2A-2 by PCR to construct it into the ORF of promoter pTEF1. By double enzyme digestion and T4 ligation, pTEF1-BcABA3-2A-2 was constructed successfully and was examined by sequencing.
In all, the fluorescence intensity of GFP and DsRed demonstrates that pTEF1 is stronger than pGAP, and this is consistent with our reference.
According to this result, we use pTEF1 to express NLS-Csy4 and function genes according to our modeling.
Fig 23. The sequencing result ofpY26TEF-GPD-pTEF1-BcABA3-2A-2
Then we tried to ligate BcABA1-2A-4 into the ORF of pGAP. We attempted double enzyme for several times, but it did not pay off. This may due to the low efficiency of T4 ligation in long fragment—the linear pY26TEF-GPD-pTEF1-BcABA3-2 is about 10kb.
After failures, we adjusted our strategy from restriction cloning to Gibson assembly. The reaction is carried out under isothermal conditions using three enzymatic activities: a 5’ex-onuclease generates long overhangs, a polymerase fills in the gaps of the annealed single strand regions, and a DNA ligase seals the nicks of the annealed and filled-in gaps.
Sites of Hind III and BamH I are reminded in our primers for Gibson. The result was examined by double enzyme digestion.
Fig 24. Enzyme digestion M:Takara DL15,000 DNA Marker; 1: pY26TEF-GDP with BBa_K3544666 & BBa_K3544777; 2: digested by BamH I, Hind III, Not I, Bgl II; 3: digested by Not I, Bgl II; 4: digested by BamH I, Hind III
2.2 Transformation of function plasmid into Saccharomyces cerevisiae
Function plasmid pY26TEF-GPD-pTEF1-BcABA3-2 was transformed into strain EGY48 by chemical transformation, positive transformations were selected by URA3 nutritional deficiency screening and colony PCR.
The protocol of yeast colony PCR is as follow: 1. Add 10μL 10mM NaOH into a 200μL tube, pick a colony and dip it into this tube. 2.Boil this mix in 98℃ for 15 minutes. 3.Use 1μL solution and conventional PCR polymerase(NOT T5 MIX) to process colony PCR(in our study, we use KOD FX).
Pay attention that the difficulty of cell-wall broken has a positive relationship with the storage time of yeast component cells, and the concentration of NaOH and boil time chould be adjusted to break cell wall.
Fig 25. Colony PCR of transformats
2.3 Confirm the production of ABA by high-performance liquid chromatography
Before HPLC, we shake EGY48 transformed with function plasmid for 5 days, and detected the fermentation liquid.
Fig 26. Result of HPLC
The injection volume was 2uL. The gradient elution was achieved as follows: the percent-age of mobile phase (vol/vol)was 0.2 from 0 to 1 min, then increased linearly to 0.8 overthe next 8 min and maintained at 0.8 for the next 1 min; then it was decreased linearly to 0.2 in 10--10.1 min, and finally maintained at 0.2.
Fermentation yield could be calculated roughly according to the peak area of standard sample and product. Peak area of standard sample are 5456 and 7.2±1, and 110mg/L ABA solution was used, so the fermentation yield is 141±20µg/L, which is effective in preventing LSC.
See More in Measurement.
Fig 27. Miscellaneous bacteria culture
Fig 28. Laboratory EGY48 competent yeast suspension (left); Angel high activity dry yeast suspension (right); Outer packaging using yeast
Fig 29. Dough with laboratory EGY48 competent yeast (left); Dough with Angel high activity yeast (right); (a) Dough with laboratory EGY48 competent yeast (left); Dough with Angel high activity yeast (right);(b)After 3-4 hours of fermentation, there are obvious differences in dough size
Fig 30. Comparison table of noodle making efficiency between laboratory EGY48 competent yeast and angel high activity dry yeast