Phages in medicine and biotechnology

In our latest blog Adam Ostrowski discusses #Phages in #medicine and #biotechnology

Bacteria-killing viruses

Bacteriophages, or “Phages” for short, were first described in 1915 by William Twort and in 1917 a French microbiologist, Felix d’Herelle, discovered their ability to kill bacteria. He named his discovery by blending Greek words for “bacteria” and “eater” arriving at “Bacteriophages”.

Phages were discovered over a decade before penicillin – the first known antibiotic – which made them the first dedicated antimicrobial weapon in the medicinal arsenal. Later, d’Herelle begun to collaborate with a Georgian scientist George Eliava and moved to Georgia. When the World War II divided Europe with the Iron Curtain, the West have focused its attention at the newly discovered antibiotics, whereas the Soviet Union, not having access to substantial parts of this research, heavily invested into the work on phages. This resulted with phages being mostly forgotten until fairly recent times, and the key centres for use of phages in medicine still remain the Eliava Institute for Bacteriphage Therapy in Tibilisi, Georgia, as well as the Hirszfeld Institute in Wroclaw, Poland, also one of the pioneering sites of phage research. Nowadays, multiple hospitals around the world are investing in expanding their phage research.

In fact, bacteriophages have been gathering increasing attention over the last few years, mostly in the reports of successful treatments of antibiotic resistant infections. Indeed, the use of phages to tackle life-threatening bacterial infections is a subject of intense study and interest. Especially now, in the times of increasing antibiotic resistance of pathogens, we need new tools to keep these diseases at bay and phages are an ideal alternative to the traditional antibiotics. They are naturally occurring antagonists of bacteria and are characterised with very high specificity, which means they can be used like homing missiles against the pathogenic microbes.

Due to their incredible specificity, phages are being used not only in medicine to cure human conditions, but also in veterinary medicine and animal husbandry, where abuse of antibiotics has already led to widely spread resistances, requiring alternative approaches. Looking at wider biotechnology, phages are also used to extend the shelf life of fresh produce by eradicating spoilage bacteria, preventing crop blights in the fields and keeping food industry free from bacteria, which would interfere in food production, especially in the dairy industry.

The incredibly broad applicability of the bacteriophage technology to a variety of aspects of biotechnology has made them one of the fastest developing tools in applied biological sciences of the last decade. The Phage Therapy market alone was estimated in 2021 for $38M and projected to grow to $100M by 2030. The whole bacteriophage technology market, including agricultural and food production uses among others, was estimated for $1.1B globally in 2023. However, selecting, optimising and producing phages at an industrial scale, or in a manner compatible with the requirements of medical application, comes with many unique requirements. These include the type of data generated during the development phase, requirements for production processes and traceability.

How are bacteriophages turned into a product?

Bacteriophages are found everywhere in the natural environment, and the location where they were found usually indicates the presence of the type of bacteria the isolated phage is active against. These bacteria are referred to as “Phage hosts”. When looking for new phages and selecting ones for further study, it is important to keep the track of their origin, as well as the that of their host. A phage and its host are an unbreakable pair, you cannot make more phages without a suitable host!

Once a phage is isolated, it’s a common practice to check which known types of bacteria related to its original host the phage is active against. Microbiological labs have huge archives of thousands of bacterial strains, so obviously keeping these records for every phage is a no small feat! It is, however, essential for selecting phages that work against many types of bacteria of interest and are efficient in killing them. Surprisingly, this process is usually completely manual with the scientist filling out laboratory records and conventional spreadsheets. This is obviously not sustainable with the rapid growth of the demand for the alternatives to the conventional antimicrobials, and the typical lab information management systems don’t provide the flexible solutions suitable for the type of data phage hunting generates. It is up to the scientists to maintain the knowledge and control over their stocks and choose which phages to select for further work.

Often a decision has to be made quickly and having the information on phage activity to hand helps to quickly choose the best phages in an emergency, for example when a patient is admitted to a clinic with an infection not responding to treatment and their health, or life, is at stake. This is especially important, as usually more than one phage is administered for treatment. Therefore, phage researchers construct complex cross-reactivity matrixes to aid in the phage mix formulation process.

Finally, once an optimal mix of phages is known, these phages have to be produced at a suitably large scale, purified and stored. This again generates a lot of information on the optimal conditions for phage fermentation and instructions how to purify a particular phage while minimising loss of its bactericidal activity and how long any given phage can be stored for and under all conditions.

To create a viable product, or a medical treatment, all of the above steps have to be strictly documented, processes tracked and data stored securely and in a traceable manner. It is a complex and mundane task, which is all too often performed manually and relays on disconnected spreadsheets and paper-based notes. But is there another way…?

Data, data, data

Looking at the bacteriophage production from the data perspective, a fairly straightforward structure emerges. Every phage can be treated as a record in a database with a tree of metadata associated with each record. The same can be said for each of the bacterial isolates in the collection. This metadata includes essential information on each stocked microbe and the best conditions for its maintenance and amplification. In many cases, these data will be stored separately, with no option to search them quickly. Sometimes a complex spreadsheet will host all the data. In each case these data will be prone to errors, changes and accidental deletion. Of course there is a number of LIMS systems that could be used, but the niche nature of the bacteriophage field means none of the existing solutions have been made to its specific needs and peculiar nature. This means a lot of important data still might get lost as improperly tagged notes or comments on the side of larger data entries. Furthermore, the requirements of each phage lab will be different and will strongly depend on the field, which field does the lab operate in: biotechnology, agritech, medtech, etc. It certainly is a situation where one size database does not suit everyone’s needs!

In addition to the above, the cross-reactivity matrix of phages against the bacterial isolates can grow into hundreds or thousands of combinations. It is not possible to maintain such a data structure manually and not only build upon it, but also search, filter and call up information in an efficient fashion.

How correct data handling can speed up decision making

Finally, once an optimal mix of phages is known, these phages have to be produced at a suitably large scale, purified and stored. This again generates a lot of information on the optimal conditions for phage fermentation and instructions how to purify a particular phage while minimising loss of its bactericidal activity and how long any given phage can be stored for and under all conditions.

To create a viable product, or a medical treatment, all of the above steps have to be strictly documented, processes tracked and data stored securely and in a traceable manner. It is a complex and mundane task, which is all too often performed manually and relays on disconnected spreadsheets and paper-based notes. But is there another way…?

Only by being able to extract all relevant information quickly, a phage producer can decide which phages to use in a mix and create the most efficient treatment. An approach consisting of a complex, purpose-built database, where the cross-reactivity matrix can be queried and filtered easily, will return the best candidates quickly. An increasing number of biotech businesses involved development of bacteriophage-based products engage machine learning algorithms to help them segregate the vast amount of data and select the optimal phage mixes for any given set of bacterial isolates found in the area they are focusing on.

Imagine a situation where upon collecting all phage reactivity data, you can request a small pool of hits of phages with strong reaction against a maximum number of selected isolates to design the most efficient phage mix. Indeed you need more than one or two phages in the mix, but a mix with more than six phages might be economically not viable to produce. Algorithms are much better in filtering data like that than humans are. Now that you have your list of candidates, the system instantly produces all relevant information: which hosts to use, how to amplify the phages, as well as all associated information required to fill out regulatory paperwork for the new product.

Perhaps there is an automated screening system, which automatically captured the reactivity data or an automated process for controlling bioreactors for each phage, which can pull the required protocol directly from the database. Altogether, the phage product development process is a data-intensive one requiring specific computational solution to achieve the global success the phage community is pushing for. This fascinating area of biotechnology is still in early stages of development, which allowed it to thrive without more sophisticated solutions, however, it is now booming. In the near future bespoke data handling solutions will be needed to ensure continuous field expansion.

Why choose Firefinch

Firefinch Software is a company dedicated to creating bespoke software for the needs of the Biotechnology and Life Sciences sectors. Many of our software developers come from a STEM background and include former research scientists, product managers and facility managers. We have a documented history of delivering projects involving complex databases, data capture and management from hardware and hardware interfacing. We treat each of our Clients individually and build their unique story to capture all the needs and provide a solution that perfectly fits their requirements.

All of the projects we undertake are expertly managed by a team of Project Managers with significant scientific experience, including in the bacteriophage research field. They will ensure the software we deliver adheres to our customers’ requirements and regulatory needs. Our advanced understanding of both biological and technological problems surrounding software development for biotechnology places us in a unique position to deliver a tailored solution to both small and medium scale enterprises as well as to multinational corporations.

Find out more

If managing phage research and production data is something you are struggling with and you require new tools, or you would like to brain storm some ideas around this subject, the Firefinch team is always happy to help. Reach out to us for a non-obligations consultation via our website www.firefinch.io, or talk to Adam on LinkedIn https://www.linkedin.com/in/adam-ostrowski-phd/

References

1. https://microbiologysociety.org/publication/past-issues/future-tech/article/phage-therapy-future-tech.html
2. https://www.frontiersin.org/journals/microbiology/articles/10.3389/fmicb.2021.793135/full
3. https://www.sciencedirect.com/science/article/pii/S2214799320300849
4. https://www.mdpi.com/2076-2607/12/2/253
5. https://publications.parliament.uk/pa/cm5804/cmselect/cmsctech/328/report.htm