Our results

Unfortunately, due to unsuccessful cloning, no measurements were obtained. Despite the several protocols tested, no cloning have been confirmed.

We do not know why the cloning did not work.

With the AQUA cloning, we frequently obtained red colonies, which is a strong indication that pSB1C3 with the RFP was integrated in E. coli. The issue here is that we used the linear plasmid provided by iGEM. The latter did PCR on pSB1C3 with RPF to provide us the linear plasmid. Some template should have stay in the linear plasmid that they sent us. To counter this phenomenon we used Dpn1 to digest the circular plasmid following the right protocol but it still remains some template.

AQUA cloning and enzymatic cloning were equally concerned about problem on analysis. As soon as colonies were present, we performed colony PCR which gave nothing and then when we did enzymatic digestions the results were inexplicable. It was no contamination because the negative control was always empty, meaning that the issue was not with the water, the primers or the buffers. The stripes were never the right length and yet we tried numerous enzymes and enzymes blends. It could be non specific hybridization but we verified with the ApE software and did not find any hybridation area. Even when a single enzyme was used and should have given us the length of the plasmid with the insert, sometimes several stripes were present. It was like there was a contamination or that two empty plasmid bound. We do not understand how it could be something else than our plasmid because it should have possessed the chloramphenicol resistance.

In a nutshell, the profiles did not correspond to the expected clones and the analysis by sequencing did not help to understand what was going on with these clonings.

Regarding the chemistry part of the project, it is a real shame that so many problems with the HPLC occured because it does not allow any conclusion. The issue with the acid hydrolysis is that initially it is used to produce sugars and that HMF is just an unwanted co-product. The quantity of HMF is very low and is insufficient to be quantified. Originally, we aim to use ionic liquids which are present in a significant amount of scientific articles but we learn that they are very expensive and they need chrome salts which are banned at the industry level. Since our purpose was to adapt our method in an industrial way, there was not interest in using ionic liquid even if they worked well.

Toxicity results coming from IGEM Leiden

Hydroxymethylfurfural (HMF) is an organic compound of the furfural family formed by dehydration of sugars (1). It is a multifunctional molecule because of its structure (Fig 1),It has a furan ring with an aromatic alcohol and an aromatic aldehyde. HMF is a versatile intermediate allowing to obtain by a transformation some high value-added chemical. It is a valuable and interesting molecular building block which can be catabolized in some useful molecule such as 2,5-furandicarboxylic acid (FDCA) (2).

Our project is about producing a bacterial strain that transforms hydroxymethylfurfural (HMF) into furandicarboxylic acid (FDCA). HMF is a byproduct of the degradation of cellulose or other polysaccharides. FCDA is a platform molecule that can be polymerized into plastics or biofuels.

Our researches in the literature on the enzymes that can catalyse the reaction from HMF to FDCA, have shown that in some cases and for some organisms it may be toxic.

Several studies in different strains of Saccharomyces cerevisiae show that cell growth was significantly affected (3). Studies also show mutagenic effects on human cells (3). HMF is also known to cause specific damage on microorganisms with DNA mutations in Salmonella typhimurium (7) or Escherichia coli (8). But some studies also show that for some strains of E. coli or some others bacteria, there is not effect of toxicity (9).

The effects come from a reduction in available cellular energy caused by the inhibition of enzymes such as alcohol dehydrogenase, aldehyde dehydrogenase, pyruvate dehydrogenase (4), and two key glycolytic enzymes (hexokinase and glyceraldehyde-3-phosphate dehydrogenase) (5). The reactive aldehyde groups in furfural family increases accumulation of reactive oxygen species (ROS). Consequences of ROS are well known for causing DNA damage, membrane damage and protein misfolding (6).

We decided to investigate the toxicity of HMF. We found out through social media, that the project of iGEM Leiden involved the detection of bacterial cell stress. Because HMF is a known to be a toxin, we suggested that it would be interesting for us to test this on iGEM Leiden’s bacterial stress detection system, and they wholeheartedly agreed.

Following a skype session to discuss the details of our collaboration, we send them 2mL of solution containing 5 g/L HMF.

It allowed them to test the compound at various concentrations on four of their stress detection strains: pDnaK (10), pRpoE (11), pSoxS (12) and pCspA (13).

Results of exposure to HMF show no stress reaction for the pCspA-GFP (figure 2A) and pRpoE-GFP (figure 2C) strains. CspA is involved in metabolism related stress and RpoE is involved in protein synthesis stress. Because both strains did not invoke a stress reaction, there is no indication that HMF delivers stress in these pathways. However, to firmly conclude these stresses are not being experienced, more extensive testing is needed, since these are just two promoters involved in these stress pathways.

Both the pDnaK-GFP (figure 2B) and pSoxS-GFP strains (figure 2D) do show some responses when exposed to 0.5µg/mL HMF. DnaK is involved in the process of chromosomal DNA replication and SoxS is involved in DNA damage repair. Therefore, these results indicate that HMF delivers stress to bacteria by damaging their DNA and disturbing their replication process.

However, the stress being experienced at these HMF concentrations seems to be low. They shared these results with us which help us with our project, provide us with insight into the effect of HMF on bacterial cell processes. We are very thankful for this meaningful collaboration and look forward to meeting their team again in Boston.

From their stress detection strains, pCspA-GFP and pRpoE which are involved in stress metabolism did not show indication that HMF delivers stress. But, pDnaK-GFP and pSoxS-GFP (which are involved in DNA damage) show some responses.

It would be great to investigate more the toxicity of HMF with an other assay (like Minimal Inhibitory Concentration Assay) (14), we could do more tests and add some statistics tests for more relevance.