Novel cascade technology for optimal utilization of animal and marine by-products

Cod heads and poultry carcasses are underutilized by-products with a complex composition consisting of proteins, fats, minerals, and other nutrients of high nutritional value that can be utilized for a globally growing population. In recent years, enzymatic protein hydrolysis (EPH) has gained significant attention as a versatile processing technology of animal and marine by-products. Currently, this technology cannot take full advantage of the multitude of components in the by-products. In Notably, we have developed novel multistep bioprocessing approaches for solving this challenge. The chemical similarity between cod heads and poultry carcasses is first and foremost related to protein composition and high contents of collagen. Thus, a main emphasis in the project has been related to developing bioprocesses to separate collagen from typical muscle proteins. For poultry carcasses, we have screened over 20 different commercial food grade enzymes to be able to choose enzymes with specific actions towards either collagen or muscle proteins, and we have been able to select two enzymes that show these characteristics. Poultry collagen has some unique features due to its high thermostability and high melting point. We have shown that this knowledge can be used together with the selected enzymes to separate the different protein classes in the poultry carcasses. By running a sequence of EPH reactions and adjusting enzymes used, temperatures and reaction times, we were able to produce separate fractions rich in muscle peptides and collagen peptides, respectively. This approach has been validated both in laboratory and in a pilot scale demonstration. In this way, we have developed a novel process where four fractions can be taken out from poultry carcasses: lipids, muscle peptides, collagen peptides, and a mineral rich sediment. For cod heads, a similar screening of enzymes as described for poultry carcasses has been performed, and other enzymes have been identified as promising targets for cascade processing. With the application of optimal control of EPH in mind, the simulation work of Notably has focused mainly on the forward model part of process optimization. Using classical models based on partial differential equations, we have developed a framework for establishing parameter-robust preconditioners for a broad class of coupled multiphysics systems. The solvers established within the proposed framework ensure that the solution time stays roughly constant regardless of the parameters of the model. Using a data-driven modelling approach we have obtained convolutional encoder-decoder networks which predict future FTIR spectra, continuously in time, based on initial conditions from the early stages of EPH. These network models present a first step towards purely data-driven forward models suitable for EPH optimization. By immobilizing enzymes on magnetic particles, the enzymes can be extracted and reused in the EPH process by magnetic separation of the particles. Different immobilization strategies were developed for covalent attachment of the protease Subtilisin A, and the enzyme-functionalized particles were shown to be reusable over more than six cycles, with more than 80 percent of the initial activity retained after more than two years' storage for the best systems. The concept of reusability was demonstrated for hydrolysis of chicken meat and turkey tendons. The best-case immobilization strategy was used to immobilize two other enzymes, demonstrating enzymatic activity of the resulting particles. Bromelain-functionalized magnetic particles were implemented in a multistep enzymatic protein hydrolysis of turkey residue, showing clear differences in the molecular weight distribution of the resulting hydrolysates compared to free enzyme. In parallel to immobilization, encapsulation of enzymes was investigated to potentially control the time of release, thereby being able to allow for sequential introduction of enzymes throughout the EPH process. Alginate was determined to be the most versatile and scalable material for enzyme encapsulation. Our results show that it is possible to produce a temperature-responsive lipid coating on the enzyme-loaded alginate capsules that can be used to release the enzyme into the reaction mixture upon a slight temperature increase. A prerequisite for process optimization and control of cascade bioprocesses is rapid analytical tools for characterisation of the unit operations. In Notably, the development of these tools has been performed in parallel to bioprocess development. We have developed new possibilities for raw material characterization based on Raman spectroscopy. In addition, Fourier transform infrared spectroscopy has been developed as a rapid technique for quantifying the degree of protein breakdown, collagen contents and thus essential product characteristics from the bioprocesses.

The Notably project has resulted in a sound scientific production, including 14 scientific papers (and more to come), 5 book chapters, and 5 master theses. Results have also been communicated in different media channels throughout the project. We have also arranged two international workshops where participants from academia and industry have met to share results and discuss challenges related to bioprocessing. EPH complies with the general biorefinery concept for total biomass utilization, and an upgrade of by-products as proposed in Notably can potentially have a huge economic impact for the food processing industries. Despite recent industrial efforts, poultry carcasses are still mainly used in feed applications, and cod heads are often exported for food purposes in Nigeria, which is regarded as an instable market. One main result of the project is a completely new bioprocessing approach for converting poultry carcasses into lipids, collagen peptides, muscle peptides and mineral-rich sediments. In Notably we have shown that such a process is possible, and current work is going in two directions: 1) optimization of each processing step to produce tailored peptides of specific properties; and 2) development of applications as ingredients in food, pharma and medical devices. We are among other things now working together with NMBU and the University of Oslo to study the collagen peptides produced and their potential use in e.g. bioinks and drug-delivery systems. Together with one of the industrial partners of Notably, we have applied for funding to further develop the process and the product applications. We are also working together with the industrial partners of the project to see if our results can be implemented in current industrial setups. Similarly, we are looking into possibilities for pursuing our initial results on processing of cod heads. The analytical work that has been started in Notably will continue with full speed in the RCN-funded projects SFI DigiFoods and TailoTides. In DigiFoods, one PhD-student is now working to implement inline Raman spectroscopy for food characterization, taking the results of Raman raw material characterization done in this project into industrial environments. Another PhD-student is working together with Nofima and Sintef Digital to develop and use an industrial prototype FTIR system for analysis of proteins and peptides. In this context, Nofima, NMBU and Simula also plan to apply the concept of data-driven forward models for EPH optimization. Finally, the Notably project initiated a fruitful collaboration between Sintef Industry and Nofima, and these partners have collaborated in several EPH-related projects in recent years. In 2021, Norilia, Nortura, Nofima and Sintef Industry was granted an IPN project (Hydrosens) regarding the production of taste neutral poultry protein ingredients. In this project, encapsulation of peptides will constitute an important part.

Annually, large quantities of by-products are produced from the animal and marine food processing industry. Cod heads and mechanically deboned poultry residues are two of the most underutilized by-products in Norway. These by-products are complex materials consisting of components such as proteins, fats, minerals, and other nutrients of high nutritional value for a globally growing population. If successfully processed into refined and tailor-made products, there is a great economic potential for high-paying markets. The proteins alone has an estimated 25 to 65% increase in sale price if processed optimally, as compared to the raw material. In recent years, enzymatic protein hydrolysis (EPH) have gained significant attention as a mild and versatile processing technology of animal and marine by-products. However, currently this technology cannot be used to take full advantage of the multitude of components in the by-products. In this project, we will develop a novel cascade bioprocessing technology that will solve this challenge. The cascade will constitute a combination of several different processing steps, each one aimed at giving separate products with the highest possible yield and quality. To be able to achieve this goal, we will combine research from several different scientific disciplines, namely bioprocessing and enzyme stabilization technologies, computer simulation and modelling, spectroscopy and analytical chemistry. By adopting the cascade bioprocessing technology, potential products from the cascade bioprocess can be formulated into a multitude of refined high-quality products for high-paying markets. To our knowledge, there are currently no companies in the world using cascade bioprocessing to facilitate full utilization of marine and animal-based by-products. In the future, our industrial partners and the rest of the by-product processing industry will have the desired technological for total utilisation and optimal value-creation of by-products.