We demonstrated that our insulin analogues Cytoplasmic Proinsulin, Cytoplasmic Winsulin, Ecotin Proinsulin and YncM Winsulin all bound insulin antibodies. Therefore they all have a structure very similar to that of native human insulin, which was an important precursor result to validate going on to test for bioactivity in a glucose uptake assay. Furthermore, we confirmed that insulin constructs were being successfully expressed in the cell lysates. The ELISA assay also allowed us to demonstrate that YncM-Winsulin was working as expected. Since the YncM tag, when expressed as a fusion protein in Bacillius, is a secretory tag, we confirmed its functionality by detecting its presence in the surrounding media, but not in the cell lysate.

While we used well characterised commercial strains of E.coli, BL21 and SHUFFLE, to express the insulin in the periplasm and cytoplasm respectively, the B.subtilis strain that we were using for secretion, WB800, is a non-comercially used strain and was donated to us from the Professor Sui-Lam Wong’s lab at the University of Calgary, Canada. This strain was appealing to use in our project due to the 8 proteases which it has knocked out. In order to demonstrate that it was in fact an 8 protease knockout, we plated this strain, along side wild-type B.subtilis 168 on a milk plate. If the WB800 strain was in fact a protease knockout, a clear zone around the colony on the milk plate, which results from the proteolysis of the milk proteins, would be very minimal or non-visible compared to the WT168 strain. Our results showed no signs clearing resulting from protease activity in the WB800 strains and therefore we demonstrated that this strain was in fact deficient in proteases compared to the WT168 strain, making it ideal to use in protein expression.

Again, our E.coli expression vector, pET15b, is a well-known and characterised system for inducible expression that has been demonstrated to work. The expression vector we chose for expressing in our Bacillus secretion system, pUS258 was constructed in our lab by a PhD student. This vector is an E.coli/Bacillus shuttle vector that is also integrative in B.subtilis. Integration of the gene inserted into this vector occurs in the region of the B.subtilis genome which results in the knockout of 2 amylase genes, AmyE. It was important that we did not use an integrative vector which integrates into a protease gene that was knocked out in our WB800 strain as this would not be possible. In order to demonstrate whether this vector was in fact integrative and did in fact knockout these two genes, our transformants were plated on LB+starch plates, and if integration did occur, starch would not be broken down and when flooded with iodine, the plate would not show any clearing around the colonies. Our results showed many colonies which did not result in starch breakdown, as well as very few colonies which were able to break down starch, and hence we were able to demonstrate that this vector was in fact integrative and functional. Therefore, we were able to effectively demonstrate that our expression systems were favourable for protein expression and functional, in particular, our B.subtilis expression system which was not well characterised.

The next step was to demonstrate that through the use of this expression system, that we were able to induce expression of both our Winsulin and human proinsulin. We were able to demonstrate this in our ELISA assay through the detection of binding to anti-insulin antibodies. In this assay we used whole cell lysates which were induced to express each of our 6 insulin constructs as we were assaying whether any insulin was produced, regardless of where it was being targeted. Along side this, we also used the his-tag purified media from the B.subtilis cells. The ELISA demonstrated that in our whole cell lysates for E.coli cells induced to express Cytoplasmic proinsulin, Ecotin proinsulin, Cytoplasmic winsulin, as well as the purified medium from B.subtilis expressing YncM Winsulin, which were all treated with proteases to cleave off secretion tags, and for proinsulin, the c-peptide to produce functional insulin, insulin which was able to bind to anti-insulin antibodies was produced. This was particularly exciting for the single chain insulin we designed, as this is novel and not previously tested and means that we designed an insulin that is similar enough to native human insulin and binds to anti-insulin antibodies and does not need an extra cleavage step of the c-peptide. The binding to anti-insulin antibodies demonstrates that the protein is not only expressed but also correctly folded, as antibodies are extremely specific to the protein in which they bind.

Furthermore, we also set out to demonstrate that our expressed insulin would be targeted to the cytoplasm and periplasm of E.coli, and secreted into the medium by B.subtilis through the use of sequence specific tags. The cytoplasmic constructs were untagged and hence, would remain where they were expressed, the cytoplasm. The positive result for the cytoplasmic constructs in the ELISA was enough to demonstrate that correctly folded protein was being produced in the cytoplasm. The fact that the ELISA detected neither proinsulin nor winsulin in the B.subtilis whole cell lysates, however, did detect Winsulin in the medium, is proof that the YncM tag is targeting the protein to be secreted into the medium. We were not able to prove that the Ecotin tag used to target the protein to the periplasm of E.coli was functional as when we ran the periplasmic extract from these cells induced to express ecotin-tagged insulin on an SDS-page gel, we were not able to see any indication of induced protein when compared to the negative control.

To validate physiological activity of Cytoplasmic proinsulin, Ecotin proinsulin, and secreted YncM winsulin which were detected in the highest concentrations in the ELISA, We performed glucose uptake assays using both mice and human cell lines. Ecotin and Cytoplasmic proinsulin which had the c-peptide cleaved to produce functional human insulin, proved to have significant activity compared to basal levels. This was seen in the glucose uptake assay where we were measuring the rate of glycogen synthesis. There is also a strong indication that secreted YNCM Winsulin is biologically functional, however the results from the glucose oxidation assay were not significant and hence could not be definitively proven. This could be due to the fact that we did not have enough (maybe the His-tag purification step was inefficient or not enough protein was secreted – we will further testing for this.

Overall, we believe that this project was a success!