A quick recap of our processes
We used nanopore technology to sequence (‘read’) the DNA of the microbes in a variety of the fermented foods. To do this we extracted the DNA from the various ferments and performed an amplification step that simultaneously selects a bacterial gene of interest and amplifies it. This bacterial gene acts as a marker that we can use to census the entire community of bacteria present. We then feed this DNA into the MinIon nanopore sequencer. The DNA is read by passing through a protein pore (which regulates speed) and then across an electrically charged membrane. The different bases (or ‘letters’) of the genetic code alter the charge across the membrane and these changes are recorded. Once we have collected this raw data we must de-code it back into DNA bases using a machine-learning algorithm, which we call ‘basecalling’. Once the raw data has been basecalled we have usable strings of DNA sequences. We can then use a computer program to match these strings against a known database of bacterial genes. Finally, this gives us an output that can be tallied up and visualised.
The graphs below are from our first batch of samples. Most of the samples are dominated by lactobacillus, which is reassuring as this is what has been described in previous studies. Interestingly, even the old fish sauce (made in 2014) and the old kimchi (~ 1 year old) have a variety of lactic acid producing bacteria. Also, our two replicates from the same kimchi (old kimchi 1 and old kimchi2) found high agreement between them. This suggests that the technology and protocols we were using are fairly robust. One other point to note is that the younger ferments (~ 1 week old or less) have a markedly different microbiome. Many of these are likely environmental bacteria found on the raw ingredients, chef’s hands and the kitchen. It’s also important to note the ‘high’ value of Escherichia shouldn’t necessarily be cause for concern. Firstly, from the gene we used (16S), it’s impossible to say anything about the exact Escherichia species/strain present. Many are harmless, commensal bacteria that live in our bodies (although, some strains do cause severe food poisoning). Secondly, the value itself doesn’t tell us anything about total numbers. This value is the relative abundance, i.e. out of all the DNA in the sample, what proportion is Escherichia. This means that whilst it is the most dominant genus (or group), it could be present in the food at a very low level if there are aren’t many bacteria in total.
In the next batch of samples, we wanted to explore where the lactic acid bacteria were coming from. To get an idea of this we took samples from a recently made kimchi at different times, as well as the ingredients. The recipe uses mooli radish- which also grows at the Eden project. We took samples from a radish leaf, the radish pre and post salting and he kimchi sauce.
Whilst we found lactobacillus and other lactic acid bacteria (Leuconostoc, Weissella) after 6 hours in the kimchi, we couldn’t detect it on the leaf, mooli or kimchi sauce. This means it is either coming from an alternative source (the chef’s hands / the fermentation vessel / other), or present at levels so low it is undetectable.
These results tie back to a theme we were keen to explore: “Everything is everywhere, but the environment selects”. This means that all types of bacteria are able to disperse everywhere, but they only thrive when the environmental conditions are right This, in effect, is fermentation in a nutshell. By creating the right conditions (high salt, low oxygen), we are enabling lactobacillus and other lactic acid bacteria to thrive. In turn, this prevents the growth of other harmful bacteria and fungi as the pH decreases, ultimately preserving the food.
The ease of modern sequencing technologies allows us to look under the hood of ancient fermentation recipes. Hopefully, this project was useful to visitors at the Eden project to think about fermentation at home and demystify the process. Many visitors were interested in the probiotic health effects of fermented foods, which we perhaps a little underprepared for. There is a lot of information online about this topic, and filtering the solid research from over hyped fiction is not easy- particularly when much of the science is locked behind paywalls. In the future, hopefully there will sufficient research to reveal or dispel these effects. In the meantime, hopefully we can enjoy the food and enable better preservation (and less waste) of seasonal vegetables.
All data and code for he analysis can be found at https://github.com/s-meaden/ferment.