Difference between revisions of "Part:BBa K4491007:Design"

Revision as of 16:46, 11 October 2022

Design rationales

Understanding the overall mechanism of araBAD promoter allowed us to pinpoint several aspects of the structure that can be rationally improved.

Rationale 1: Spacing between araI1 - araO2

To minimise the length of pBAD, we investigated the presence of araO2 site, as well as the 211 bp spacer between araI1 and araO2 (this region includes CAP binding sites and araO1). Previous literature suggested that complete or even partial deletions of araO2 would increase leaky expression of pBAD, emphasising that the existence of this region is essential for normal repression. Strangely, most commercially available sequences for pBAD (in CIDAR MoClo kit, for example) omitted the araO2 site, so its significance is still uncertain.

The spacing will dictate the size of the loop, which also controls the level of repression under the absence of arabinose and determines the promoter’s leakiness. Here, the spacing is defined to be between position -59 within araI1 and -270 within araO2 (underlined and asterisked). It was shown that there are no lower bounds for loop size - a functional araBAD promoter was designed with a 34-bp loop, after deleting a significant portion of the spacer []. While it maintained similar leakiness to that of the wild-type counterpart, the loss of CAP binding site drastically reduced the maximal strength. Still, the finding demonstrated the great flexibility of bulky araC proteins in mediating small loop formation.

Figure 1: Schematic representation of length modification within the araO2-araI1 spacing.

Another important observation was that as the spacing varied, the promoter activity oscillated with a 11.1 bp periodicity. Specifically, any insertion or deletion of integer multiples of 5bp noticeably increased leakiness, while integer multiples of 11.1 bp retained wild-type’s full repression [8]. This value was determined to be approximately equivalent to one helical repeat of DNA (10.5 bp). It was further explained that insertion of 5 bp between araO2 and araI1 would rotate one site halfway around the DNA double helix with respect to the other and impede repression. Despite araC’s flexibility, the torsional stress of DNA makes such looping much more energetically unfavourable.

Taking into account these results, our first design strategy is to reduce the spacer region down to 56 bp while still maintaining the CAP binding site downstream. We removed the araO1 site completely due to its minor role on P_BAD activity. The schematic of our rationale is depicted below.

Rationale 2: araI1 and araI2

We then investigated araI1 and araI2 17-bp regions, both containing two unique sites called the A- and B-box, which serve as specific binding sites for araC. In previous literature, Niland et al (1996) showed that any single base-pair substitution occurring in these two sites would drastically reduce binding of araC. Flanked between the two boxes are seven invariant nucleotides that, upon selected single substitution, demonstrated higher binding affinity to araC by somewhat 140% compared to that of wild-type araI1 []. We therefore modified the araI1 site to contain all the different substitutions which initially yielded tighter binding, while keeping the A- and B-boxes unchanged.

In another paper, Reeder (1993) found that the B-box of araI2 overlaps with four base pairs of the -35 consensus sequence [9]. Thus, any substitution in this box will negatively impact P_BAD’s activity, either resulting in very high leaky expression or lowered inducibility. However, the author remarked that araI1 has much higher affinity to araC than araI2 does, especially when no arabinose is present. This makes sense, as araC prefers binding to distal araO2 and araI1 than the nearby araI2. From this insight, we questioned whether a duplicate araI1-I1 may confer higher maximal activity than wild-type araI1-I2. We then sought to change the last nucleotide of the A-box and the interbox sequence of araI2 to that of wild-type araI1, but still leaving the araI2 B-box untouched. In a sense, we created a chimeric half-araI1-half-araI2 in place of wild-type araI2. We did not duplicate the entire araI1 as this would affect the overlapping -35 consensus sequence and make the promoter extremely leaky.

Figure 2: Schematic representation of nucleotide substitutions within the araI1-araI2 region.

We group the modifications for both araI1 and araI2 as our second strategy.

Rationale 3: -35 and -10

Our final design input comes from the work of the 2013 DTU iGEM team (see here). They managed to create a synthetic promoter library (SPL) for araBAD promoter by randomly mutating different base-pairs between the -35 and -10 boxes, right downstream of araI2. We decided to use the Col15 sequence, which showed promising low level of leakiness and high induced strength.

Combinatorial design

We finally opted for a combinatorial approach for the three strategies, and thus have initially designed 23 = 8 different pBADs, with or without the preceding optimizations. After some thoughts, we introduced another design with a slightly larger spacing between araO2 and araI1 (PB5). These designs were tested against the wild-type, full-length araBAD promoter, corresponding to part BBa_K2442101in the registry. We were aware that while an individual strategy may yield noticeable improvements, this might not be true when combining them together. In fact, some can result in antagonistic effects rather than the desired synergy. Still, we hoped that within the different permutations, some good designs may emerge. We also thought that, given this variety of promoter expression, we should not restrict our aim to only creating a single better pBAD, but also making a “family” of the promoter with different strengths, such as weak, medium and strong, for various purposes (in some cases, lower maximal activity might be necessary).