Discussion and Perspectives

We report the use of science crowdfunding and citizen science to characterize and compare the ITS profile of 39 beers. Specifically, we focussed on the fungal (i.e. yeast) diversity and mapped the representation of a gene (ITS1) that is highly variable among fungi. As expected, DNA sequences from S. cerevisiae and  S. pastorianus were largely represented in our data. More interestingly, we measured DNA traces from 88 different species (i.e. wild yeast) with 52 species unique to different beers. Our analysis suggests that wild yeasts may be present with high diversity in beer. DNA is readable in both live and dead organisms, therefore we cannot infer if the presence of one species (i.e. Brettanomyces) has a direct impact on the fermentation process. Rather, with controlled experiments it could be possible to vary the fermentation process and investigate if the representation of specific microorganisms is affected \cite{Hong_2016} . The source/origin of each species could also be investigated by directly sequencing raw materials (i.e. malts). Some yeast species are present in the air, and techniques to analyse airborne DNA are available \cite{Jiang_2015}. Plant DNA may also present in beer \cite{NAKAMURA_2013} and our approach could be potentially extended to map the diversity of bacteria, as well as the malt and hops varieties.
We believe that small, independent “craft” science can branch out from mainstream scientific research. Our choice to run BeerDeCoded independently and inside a citizen-science lab parallels what is happening in the brewing establishment. In 1978 there were 89 big, mainstream breweries in the US. In 2016, there are 5,301. Among them, 3,132 small, independent microbreweries represents 60% of the registered businesses (source: American Brewers Association LINK). The phenomena of “craft” beer is expanding also in Europe and Asia. Supported by the favorable demand, one driver of this expansion is the awareness that brewing is attainable with limited resources. Individuals and small teams can brew good beer independently from the endorsement of important stakeholders. Our first dataset demonstrates that it is possible to execute complex molecular analyses on beer using limited resources, and with negligible support from traditional funding agencies and research institutions. Sustainable funding may be achieved by offering “craft science as service” to micro- and homebrewers.
Will molecular data be useful for the brewing industry? Microbreweries and homebrewers face sanitisation challenges. Information encoded in the beer DNA could help brewers to control the quality and the reproducibility of their processes, or to decide whether to use the final product for short term or long term storage based on an index of spoilage microbes. Other targeted analyses could help to validate procurement and to insure compliance if specific ingredients are declared on the label (i.e. is this yeast/hops/barley really what is declared?). Relevant data are accumulating. For instance, the full genome of 157 brewing yeast strains has been sequenced \cite{Gallone_2016}.
With untargeted molecular analyses, brewers could benchmark their production batches against known commercial beers and differentiate better in the brewing landscape. Major brands face losses due to counterfeiting and adulteration. In this context, molecular fingerprints could provide proof of authenticity for final products and for proprietary yeast strains.
With DNA sequencing costs dropping dramatically, and with the emergence of portable and low cost instrumentation, we believe that it is a favorable time to expand DNA analysis to novel fields, beer included. Still, BeerDeCoded remains a small “craft science” hub dedicated to Open Science. We invite researchers from other laboratories, microbreweries and citizen laboratories to build on our open data set, or to contribute additional data at their will.

Methods

Detailed methods are available on Github https://github.com/beerdecoded/Beer_ITS_analysis.

Beer sample preparation

The content in the beer bottle was mixed by inversion. 50 mL were transferred into a conical tube and centrifuged (5000 rpm, 20 min, 4 °C) to collect cells and precipitable material. Low fermentation beers (i.e. lager, pilsner) had a more unstable pellet compared to high fermentation beers (i.e. ales). The pellets were resuspended with 1 mL TE buffer (Tris 10 mM, EDTA 1 mM, pH 8.0) and transferred into 1.5 mL tubes. The samples were centrifuged (10000 rpm, 10 min, 4 °C), the supernatant removed and the pellet were stored at -20°C for future analyses. For DNA extraction, the ZR Fecal DNA MiniPrep Kit from Zymo Research (Catalog No. D6010) was used with minor modifications.

Quatity control for DNA extraction

The absorbance of the extracted DNA was measured at 230, 260 and 280 nm using a nanodrop instrument to ensure the DNA was free from protein and chemical contaminants.

PCR1: ITS amplification

To amplify yeast genomic DNA, we used the fungal hypervariable region ITS1 (internal transcribed spacer 1) as described in \cite{Bokulich_2015} using the following primers:  BITS (5′–CTACCTGCGGARGGATCA–3′) and B58S3 (5′–GAGATCCRTTGYTRAAAGTT–3′). Typical PCR reactions contained 5–100 ng DNA template. Amplicon size (500nt) was verified with gel electrophoresis and with a fragment analyser. ITS amplicon were purified from free primers and primer dimer species using AMPure XP beads following the manufacturer’s instructions. Dual indices and Illumina sequencing adapters were attached using the Nextera XT Index Kit following manufacturer’s instructions.

MiSeq Sequencing

MiSeq sequencing was performed using the MiSeq v3 reagent kit protocol. Briefly, we quantified the amplified DNA with a fluorimetric method that uses dsDNA binding dyes (Qubit). We diluted each DNA sample to 4 nM using 10 mM Tris pH 8.5 and pooled 5 uL of diluted DNA from each library. In preparation for cluster generation and sequencing, 5 uL of the pooled final library was denatured with 5 uL of freshly diluted 0.2 N NaOH, and combined with 30% PhiX control library to serve as an internal control for low-diversity libraries. After loading the samples on the MiSeq, we generated paired 2x 300bp reads that were exported as FASTq files.

Bioinformatics Analysis

We downloaded the curated set of ITS sequences from the  Refseq database (https://www.ncbi.nlm.nih.gov/refseq/targetedloci/). We used these sequences to build an ITS index for the Burrows-Wheeler Aligner \cite{Li_2009}. We used BWA to map the reads of each beer from the fastq files  to our ITS index. Then we sorted and indexed the bam files using samtools. Subsequently, we counted the number of ITS per beer and per species and we kept only the species where we found more than 10 reads.

Dataset content

The dataset contains the metagenetic profiles for 39 beers. The data was obtained using a targeted approach based on the phylogenetic typing with internal transcribed spacers (ITS) of ribosomal sequences.

Data and code statement

The raw data are stored in the SRA database in the bio project PRJNA388541
All methods, quality control, processed tables, metadata and code are accessible in the following git repository: https://github.com/beerdecoded/Beer_ITS_analysis

Author contributions

JS, LH and GR conceived the projects and communicated with the public. JS, NR and GR designed the experiments and carried out the research. JS, LH and GR wrote the manuscript.

Aknowledgements

The authors would like to thank Gabrielle Salanon for her help with sample extraction and analysis. Keith Harshman (Lausanne Genomic Technology Facility) for the access to the sequencing platform. Onecodex for the access to their metagenomic analysis tool. UniverCité and InArTiS for their support of the Hackuarium association and Rachel Aronoff for her critical review of the manuscript

Grant information

This project was funded by the 124 backers of the BeerDeCoded crowdfunding campaign that happened in June 2015. For a full list of backers, see [LINK]. These funders played a significant role in the data collection as they sent their beer samples for analysis. All backers from the crowdfunding campaign were given a chance to participate in the study design and analysis, decision to publish, or the preparation of the manuscript.
Copyright:  © 2017 Sobel J et al. This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Data associated with the article are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).
Competing interests: GR is shareholder of SwissDeCode, a company selling to food manufacturers point-of-need DNA tests for food safety and compliance.