Difference between revisions of "Team:SCUT-China/Proof Of Concept"

The concept we are going to prove is that, the experience gained and concluded from this project, and the methodology explored. After our analysis and neaten, it finally condensed into essence as a handbook called ‘UP strategy’(UAS Promoter strategy). In order to verify the UP strategy we offer and advocate is realistically significant, we conduct the following proof of concept.

It is known from existing studies that promoters can bind to different transcription factors through UAS to achieve response to different environments (PH, carbon source, osmotic pressure, etc.). Our wet experiment was designed according to the secondary growth characteristics of Saccharomyces yeast, hoping to obtain a promoter that can maintain high expression activity even in the late stage of fermentation.

We therefore used the YeTFaSCo database to find CAT8 and ADR1, the UASs associated with ethanol induction or glucose repression. CAT8 is a zinc cluster transcriptional activator that regulates the expression of most enzymes in the gluconeogenic genes and the glyoxalate cycle. The CAT8 protein binds to a carbon-responsive element in the upstream region of the Saccharomyces cerevisiae gluconeogenic gene and activates the target gene in response to glucose depletion transcription of the ADH2 (alcohol dehydrogenase) gene upon ethanol induction. In addition, ADR1 may interact with CAT8 to activate gene transcription.

Based on the above considerations, in the first round of engineering, we selected the CAT8 binding sequence of FBP1p and the predicted ADR1 binding sequence in ALD4; in the second round of engineering, we replaced the CAT8 binding sequence with the predicted CAT8 binding sequence from SED1Lp, ALD4p. In the third round of engineering, we made multiple copies of the important UAS. In the third round of engineering, we made multiple copies of the important UAS to ensure that the nucleosome affinity was not lifted too much, and the yield of Valencienes in M2, M5, and M8 was increased by 18.9%, 29.5%, and 56.6%, respectively. Compared with pPDC1, the yield of Valencienes in the late fermentation stage was significantly higher in M8. It fully illustrates that, through the rational and eligible UAS insert strategy(make sure the insert region won't have too large nucleosome affinity to in order that the added UAS can be bound by transcription factors with high probability), it is possible to create a promoter that meets the expectations for production purposes.

Firstly, comparing promoters M1 and M2, the common design key UAS is in different positions, when UAS is placed in the higher affinity region, its promoter strength is lower. When the UAS is in a region with lower affinity for nucleosomes, the promoter has a higher intensity. In the first set of experiments with controlled variables, the UAS was fixed and the position or affinity was varied, and the promoter with lower affinity obtained higher scores, indicating that the position with lower affinity helped to enhance the function of UAS and thus influenced the promoter strength, and it is necessary to consider the choice of higher or lower affinity in the design of UAS embedding.

Figure 1. Nucleosome affinity curve and UAS arrangement of M1, M2

Secondly, comparing promoters M8 with M9, when the composition of UAS is the same but the arrangement is different, the yield is different. According to the nucleosome affinity distribution map, the M8 sequence contains a CAT8 TFBS from SED1Lp-1, which is located upstream of the promoter from 207 to 187.

The M9 sequence was obtained by moving the middle UAS to the left of the Adr1 UAS from the M8 promoter sequence. At this point, the moved CAT8 UAS is 234 to 254 bp upstream and the nucleosome affinity value in this region is at a relatively high position. The same conclusion as the previous set of experiments(M1 and M2) was obtained, saying the higher the nucleosome affinity the lower the local UAS effect. However, the difference between the affinities in this group is much smaller and only rises by a certain small margin leading to a decrease in yield as expected.

Figure 2. Nucleosome affinity curve and UAS arrangement of M1, M2

A similar experiment result could be seen in M8 and M11, there are less bp amount difference but still reflect on promoter strength.

Figure 3. Nucleosome affinity curve and UAS arrangement of M8,M11

Thirdly, another very representative case is the comparison between M8 and M12, where we tried to place the UAS in the region that was originally the "peak" as promoter M12 to see whether the yield would be reduced. Compared to M8, M12 moved the Adr1 UAS on the left side from about 226bp upstream to 581bp upstream. According to the nucleosome affinity distribution map, we can also easily find that the position around 581bp has a very high nucleosome affinity. This means that the improvement from M8 to M12 will bring about a sudden decrease in yield. And realistic experimental data are perfectly validating this argument.

The difference between this set of comparisons and the first set of comparisons is that the UAS setup of M1 is placed in a position shaped like a valley in the affinity plot of PDC1, and M12 is placed in a position shaped like a mountain peak. Case is that the UAS of M1 undergoes an abrupt change in affinity after insertion to the point of inappropriateness.

Figure 4. Nucleosome affinity curve and UAS arrangement of M8, M12

Since the TFBS sequence is actually a segment of motif containing a simple base obtained by experiment or prediction, if, in the experiment we tried several times to replace the UAS motif with a shorter or longer combination, in order to try to increase the strength of the promoter variants (M5, M6, M7 ), indeed we gain some meaningful result. If you are in a position to attempt base simplification, it is recommended to use the sequence lengths provided in the database and add 2 bases at a time in steps of left and right.

Table 2-1 UAS introduction

Table 2-1 shows our selected interested UAS, and most of these are associated with carbon source response and glucose repression according to the TF database description of these TFBS, in order to achieve the high expression activity maintained in the from early to late stages of our design. We also selected 14 natural strong promoters commonly used in brewer's yeast at the beginning. To select the strongest promoters in each metabolic pathway type, we constructed expression vectors to perform characterization experiments for these 14 promoters. Therefore the strength of these 14 promoters is also known to us.

Two important instrument used in UP strategy are NuPoP for nucleosome affinity prediction and TF database for TFBS finding. Prediction in silico always suffered users’ query, as same as these two tools. In this time, we verify them by a mutual confirmation during our usage.

On the one hand, all UASs’ location were labeled on our promoters by TF database, focusing on those key UAS. Also we added features by snapgene for the better visualization. On the other hand, the HMM integrated software NuPoP was performed on each promoters, creating all the nucleosome affinity curve prediction result.

Stacking the two graphs together, you would find an interesting pattern. The two prediction match surprisingly well. It is common for two people to come up with different wrong answers at the same time, but the fact that two people come up with the same answer at the same time indicates that this answer has significance to an extend. First, most of the carbon source responsive UAS, such as Gcr1, Rgt1, Azf1, which play a positive role, always appear in the nucleosome with low affinity, which we call 'valley', seen in the stacking graphs. It not only shows that the TF database and the affinity value distribution curve prediction have a certain degree of accuracy, but also shows that the UAS does need low nucleosome affinity as a condition to function. This implies that the region with low affinity values is more worthy of researchers' attention when designing promoter optimization pathways. The UAS located in that region may play a more important role and deserves more attention from researchers.

The software we created, NuPGO, performed very satisfactorily in the reduction of the nucleosome affinity value. Including the validation on the experimental side, a very significant enhancement has been achieved with a simple optimizing program different from ours according to the literature [3]. Implying that better results can be obtained using the algorithm that has been algorithmically optimized by us.

In order to make NuPGO easy to use for all researchers, we insisted on making it a user-friendly software when designing it. After downloading the installation package, you only need to enter one command to execute the optimization program. In addition, NuPGO is completely open source code, which can be made available to researchers for deeper optimization or for research modifications. These two points greatly ensure the availability of NuPGO.

View our GitHub！