Re: 발효과정에서 식물단백질은 어떻게 분해되는가?

[문형철]

Extracellular microbial proteases with specificity for plant proteins in food fermentation

Author links open overlay panelLise Friis Christensen a, Beatriz García-Béjar b, Claus Heiner Bang-Berthelsen a, Egon Bech Hansen a

Show more

Add to Mendeley

Share

Cite

https://doi.org/10.1016/j.ijfoodmicro.2022.109889Get rights and content

Under a Creative Commons license

Open access

Highlights

• •

  Microbial extracellular proteases can target plant proteins.

• •

  A knowledge gap exists for microbial assisted degradation of plant proteins.

• •

  Microbial assisted proteolysis displays strain specificity and substrate selective.

• •

  Plant protein hydrolysate can provide valuable food properties.

• •

  Proteolysis of plant proteins is affected by the synergies in the food matrix.

• 미생물 세포 외 프로테아제는 식물 단백질을 표적으로 삼을 수 있습니다.
• •식물 단백질의 미생물 보조 분해에 대한 지식 격차가 존재합니다.
• •미생물 보조 단백질 분해는 균주 특이성과 기질 선택성을 나타냅니다.
• •식물 단백질 가수분해물은 귀중한 식품 특성을 제공할 수 있습니다.
• •식물 단백질의 단백질 분해는 식품 매트릭스의 시너지 효과에 의해 영향을 받습니다.

Abstract

Plant-based food products are generating a growing interest as part of the ongoing transition to a primarily plant-based diet, which makes demands to the quality, functionality, and health properties of plant proteins. Microbes used for traditional food fermentations such as lactic acid bacteria (LAB) and fungi (yeasts and molds) carry out enzymatic changes on their protein substrates by which technological and sensorial characteristics can be improved. The literature on extracellular proteases targeting plant proteins, on the other hand, is scattered with only a narrow representation of plants even for traditionally plant-based products. Therefore, this review aims to explore the current state of knowledge regarding the application potential of microbial extracellular proteases targeting plant proteins, with a focus on traditional applied food microbes. Plant proteins are targeted by proteolytic microbes of both animal and plant origins, and their proteases show a wide range of activities. Extracellular microbial proteases can hydrolyze specific protein-based allergens and even reduce the toxicity of plant proteins. Additionally, microbial assisted proteolysis can improve plant protein digestibility by increasing availability of peptides and amino acids. This catabolic process will change the organoleptic characteristics of fermented plant proteins, and the release of bioactive peptides can provide additional functionalities to the plant matrix. The proteolytic activity is determined by the microbial strain, and it can be quite substrate selective, which is why proteases may be overlooked by the preval‎ent use of casein as substrate in proteolytic screenings. The synergetic effects of LAB and fungal species consortia can facilitate and steer plant protein hydrolysis by which co-fermentation may increase or change the properties of plant protein hydrolysates. Microbes do not necessarily require extracellular proteases because endogenous proteases in a plant-matrix may meet the microbial amino acid requirements. However, extracellular proteases have the potential to provide central properties to diverse food-matrixes by which the full proteolytic potential of food microbes needs to be explored in order to facilitate the development of high-quality plant-based food products.

요약

식물성 식품은

식물성 식단으로의 지속적인 전환의 일환으로 점점 더 많은 관심을 받고 있으며,

식물성 단백질의 품질, 기능성, 건강 특성에 대한 요구가 증가하고 있습니다.

젖산균(LAB)과 곰팡이(효모 및 곰팡이)와 같은 전통적인 식품 발효에 사용되는 미생물은

단백질 기질에 효소 변화를 일으켜

기술적, 감각적 특성을 향상시킬 수 있습니다.

한편,

식물 단백질을 표적으로 하는 세포 외 프로테아제에 관한 문헌은

전통적인 식물 기반 제품에 대해서도 식물에 대한 표현이 매우 제한적으로 흩어져 있습니다.

따라서

이 리뷰는 전통적인 응용 식품 미생물에 초점을 맞추어

식물 단백질을 표적으로 하는 미생물 세포 외 프로테아제의 응용 가능성에 관한

현재의 지식 상태를 탐구하는 것을 목표로 합니다.

식물 단백질은

동물과 식물 기원의 단백질 분해 미생물에 의해 표적이 되며,

이들의 프로테아제는 광범위한 활성을 나타냅니다.

세포 외 미생물 프로테아제는

특정 단백질 기반 알레르겐을 가수분해할 수 있으며,

식물 단백질의 독성을 감소시킬 수도 있습니다.

또한,

미생물의 도움을 받은 단백질 분해는

펩타이드와 아미노산의 가용성을 증가시킴으로써

식물 단백질의 소화율을 향상시킬 수 있습니다.

이 이화 작용은

발효된 식물 단백질의 관능적 특성을 변화시킬 것이며,

생체 활성 펩타이드의 방출은

식물 매트릭스에 추가적인 기능을 제공할 수 있습니다.

단백질 분해 활성은

미생물 균주에 의해 결정되며,

기질 선택성이 매우 높을 수 있습니다.

따라서

단백질 분해 스크리닝에서 카제인을 기질로 사용하는 것이 널리 사용되지만,

프로테아제는 간과될 수 있습니다.

LAB(latic acid bacteria)와 곰팡이 종의 공생 효과는

식물 단백질 가수분해를 촉진하고 유도할 수 있으며,

이때 공동 발효는 식물 단백질 가수분해물의 특성을 증가시키거나 변화시킬 수 있습니다.

미생물은

반드시 세포 외 프로테아제를 필요로 하지 않습니다.

식물 매트릭스에 존재하는 내인성 프로테아제가

미생물의 아미노산 요구량을 충족시킬 수 있기 때문입니다.

그러나

세포 외 프로테아제는

다양한 식품 매트릭스에 핵심적인 특성을 부여할 수 있는 잠재력을 가지고 있기 때문에,

고품질의 식물 기반 식품의 개발을 촉진하기 위해서는

식품 미생물의 완전한 단백질 분해 가능성을 탐구해야 합니다.

Questions answered in this articleBetaPowered by GenAI

This is generative AI content and the quality may vary. Learn more.

1. How can microbial proteolysis improve the digestibility of plant proteins?

2. What role does microbial facilitated proteolysis play in traditional fermented food products?

3. What are the main microbial species discussed in relation to extracellular proteolytic activity on plant proteins?

4. What role do endogenous proteases play in the development of fermented food products?

5. What benefits do the peptides and amino acids generated by microbial proteolysis offer in food processing?

• Previous article
• Next article

Keywords

Extracellular microbial proteases

Fermentation

Plant proteins

Lactic acid bacteria

Fungi

1. Introduction

Fermentation is an ancient method of preserving perishable food products. A wide range of species, including bacteria, yeast, and molds, have a long history of beneficial use in food fermentations, including the fermentation of plant raw materials (Bourdichon et al., 2012). This process has been accepted by humans owing to the positive sensorial and technological changes of foods, which have allowed for the development of a wide variety of food products thanks to the various fermentative pathways of microorganisms (Steinkraus, 2004). The periodic table of fermented foods by Gänzle, 2022 shows this global diversity of fermented food products where a considerable part of the fermented foods are plant-based with the use of diverse microorganisms. Lactic acid bacteria (LAB) appear to dominate the bacteria group in food fermentation, and LAB, yeast and filamentous fungi species appear to be the central microbiological workhorses of both plant- and animal-based food fermentation. The taxonomy of the LAB genus Lactobacillus has been continuously reorganized with the most recent update in 2020 (Zheng et al., 2020). This review makes use of the current taxonomy.

The fermentation of plants has traditionally been applied as a method of preservation but has also gained appreciation as a way to improve the nutritional value and organoleptic quality. The plant-based food products experience a growing interest. To meet the United Nations' Sustainable Development Goals, a dietary transition from primarily animal-based towards plant-based protein products is required (Aiking and de Boer, 2020). It is proposed that the current Western dietary patterns can be improved by reducing protein and calorie overconsumption, reducing food waste, and replacing animal-based products with plant-based products. The tendency of protein overconsumption is observed particularly in the Western diet, where a 10–15 % reduction in the protein intake per capita by 2050 has been stated as an aim (Fig. 1). At the same time, the plant-based protein content should increase from 40 % to 60 % of the total food proteins in the diet (Aiking and de Boer, 2020). These objectives impose requirements to the quality, functionality, nutritional value and toxicity of the plant proteins, thereby necessitating new food development strategies (Day, 2013). Hydrolysis of plant proteins is one approach for expanding the utilization catalog of plant proteins because proteolysis may release peptides with valuable bioactivities or cleave allergenic compounds (Day, 2013). The properties of the plant protein hydrolysate strongly depend on the selected protease and the choice of reaction condition. The emulsification, foaming and gel formation properties of the proteins can also be released through specific hydrolysis, whereas such functionalities may be lost through unspecific or heavy hydrolysis. Olsen et al., 2020 has described an algorithm to identify antioxidant peptides embedded within the sequences of native proteins. Bioinformatics and specific hydrolysis have been used to identify and release emulsifying and antioxidant peptides from potato, seaweed, and single cell proteins (García-Moreno et al., 2020a; Yesiltas et al., 2021).

1. 서론

발효는

부패하기 쉬운 식품을 보존하는 고대 방법입니다.

박테리아, 효모, 곰팡이 등

다양한 종들이 식물 원료의 발효를 포함하여

식품 발효에 유용하게 사용되어 왔습니다(Bourdichon et al., 2012).

이 과정은

미생물의 다양한 발효 경로 덕분에

다양한 식품의 개발이 가능해진 긍정적인 감각적, 기술적 변화로 인해

인간에 의해 받아들여졌습니다(Steinkraus, 2004).

Gänzle, 2022의 발효 식품 주기율표는 발효 식품의 세계적인 다양성을 보여줍니다.

https://localfoodconnect.org.au/wp-content/uploads/2022/05/fermented-foods.pdf

발효 식품의 상당 부분은

다양한 미생물을 사용하여 식물성입니다.

유산균(LAB)은

식품 발효에서 박테리아 그룹을 지배하는 것으로 보이며,

LAB, 효모 및 사상균 종은 식물성 및 동물성 식품 발효의 중심적인 미생물 작업용으로 사용되는 것으로 보입니다.

LAB 속 락토바실러스(Lactobacillus)의 분류는 2

020년 가장 최근의 업데이트를 통해 지속적으로 재구성되었습니다(Zheng et al., 2020).

이 리뷰에서는 현재의 분류를 사용합니다.

식물의 발효는

전통적으로 보존 방법으로 사용되어 왔지만,

영양가와 관능적 품질을 향상시키는 방법으로도 인정받고 있습니다.

식물성 식품에 대한 관심이 증가하고 있습니다.

유엔의 지속 가능한 개발 목표를 달성하기 위해서는

동물성 단백질 위주에서

식물성 단백질 위주로 식단을 전환해야 합니다(Aiking and de Boer, 2020).

현재 서구식 식습관은

단백질과 칼로리 과잉 섭취를 줄이고,

음식물 쓰레기를 줄이며,

동물성 제품을 식물성 제품으로 대체함으로써 개선될 수 있다고 제안됩니다.

단백질 과잉 섭취 경향은

특히 서구식 식습관에서 두드러지게 나타나며,

2050년까지 1인당 단백질 섭취량을 10-15% 줄이는 것이 목표라고 합니다(그림 1).

동시에 식물성 단백질 함량은

식이요법에서 총 식물성 단백질의 40%에서 60%로 증가해야 합니다(Aiking and de Boer, 2020).

이러한 목표는

식물성 단백질의 품질, 기능성, 영양가 및 독성에 대한 요구 사항을 부과하므로

새로운 식품 개발 전략이 필요합니다(Day, 2013).

식물 단백질의 가수분해는

식물 단백질의 활용 범위를 확장하는 한 가지 접근법입니다.

단백질 분해는

유용한 생체활성을 가진 펩타이드를 방출하거나 알레르기 유발 화합물을 분해할 수 있기 때문입니다(Day, 2013).

식물 단백질 가수분해물의 특성은

선택된 프로테아제와 반응 조건의 선택에 따라 크게 달라집니다.

단백질의 유화, 거품, 젤 형성 특성은

특정 가수분해를 통해 방출될 수 있지만,

비특이적 또는 과도한 가수분해를 통해 이러한 기능성이 손실될 수 있습니다.

Olsen et al., 2020은

천연 단백질 서열 내에 포함된 항산화 펩티드를 식별하는 알고리즘을 설명했습니다.

생물정보학 및 특정 가수분해는

감자, 해조류 및 단일 세포 단백질에서

유화 및 항산화 펩티드를 식별하고 방출하는 데 사용되었습니다(García-Moreno et al., 2020a; Yesiltas et al., 2021).

1. Download: Download high-res image (591KB)
2. Download: Download full-size image

Fig. 1. Green protein transition including plant protein and hydrolysis

A) The diagram illustrates how the Western world’s ongoing green protein transition should progress in order to reach the UN’s sustainable development goals. This includes a 10–15 % reduction in the total protein intake per capita over the time frame depicted. The figure numbers appear in Aiking and de Boer, 2020. B) Extracellular proteases (scissors) of three groups of food microbes, including lactic acid bacteria (LAB), yeast, and mold, have traditionally hydrolyzed animal proteins (orange chain of pearls). The bioprocessed protein hydrolysis can both induce (↑) or remove (↓) protein properties. Similar properties are desired for plant proteins, which is why similar bioprocessing of plant proteins using the same or other microbial extracellular proteases may promote the green protein transition. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

그림 1. 식물성 단백질과 가수분해를 포함한 녹색 단백질 전환

A) 이 도표는 서구 세계의 녹색 단백질 전환이 유엔의 지속 가능한 개발 목표에 도달하기 위해 어떻게 진행되어야 하는지를 보여줍니다. 여기에는 표시된 기간 동안 1인당 총 단백질 섭취량을 10-15% 줄이는 것이 포함됩니다. 그림의 숫자는 Aiking and de Boer, 2020에 나와 있습니다. B) 유산균(LAB), 효모, 곰팡이 등 세 가지 식품 미생물 그룹의 세포 외 프로테아제(가위)는 전통적으로 동물성 단백질(오렌지 체인 진주)을 가수분해했습니다. 생물학적 처리된 단백질 가수분해는 단백질 특성을 유도(↑)하거나 제거(↓)할 수 있습니다. 식물성 단백질에 대해서도 유사한 특성이 요구되는데, 이것이 바로 식물성 단백질에 대해 동일하거나 다른 미생물 세포 외 프로테아제를 사용하는 유사한 바이오 프로세싱이 녹색 단백질로의 전환을 촉진할 수 있는 이유입니다. (이 그림의 범례에 있는 색상 참조에 대한 해석은 이 기사의 웹 버전을 참조하시기 바랍니다.)

Microbes traditionally used in fermented food have shown valuable proteolytic activities as part of their fermentation processes. The fermentation involves multiple reactions to convert complex substrates into simple compounds, including microbiological, enzymatic, chemical, biochemical, and physical processes. These complex biosystems contain enzymes from both raw materials and microorganisms that are in charge of hydrolysis reactions such as proteolysis (Bourdichon et al., 2012; Joshi et al., 2018). Some microbes produce extracellular proteases, which are either bound to the microbial cell envelope or secreted. These proteases degrade environmental proteins to facilitate microbial growth, but the process may also support food developments as particularly described for dairy and meat products. However, the information on extracellular proteases of microbes targeting plant-based proteins is rare and scattered even for the traditional plant-based products as sourdough, wine, and sauerkraut. This is also illustrated in Fig. 2 in which the innovation curves for plant- and animal-based dairy products are represented as publications in PubMed (February 2022). This indicates that we are facing knowledge gaps in the scientific space where we are early on the innovation curve in the case of plant dairy alternatives. Microbial fermented dairy alternatives have emerged as prototypes (Madsen et al., 2021a, Madsen et al., 2021b) and further reviewed by Tangyu et al., 2019. However, there is no knowledge on the effect of extracellular proteolytic activity on dairy alternatives.

전통적으로 발효 식품에 사용된 미생물은

발효 과정의 일부로서 유용한 단백질 분해 활성을 보였습니다.

발효는

복잡한 기질을 단순한 화합물로 전환하는

여러 가지 반응(미생물학적, 효소적, 화학적, 생화학적, 물리적 과정 포함)을 포함합니다.

이러한 복잡한 생물학적 시스템에는

단백질 분해와 같은 가수분해 반응을 담당하는 원료와

미생물 모두에서 유래된 효소가 포함되어 있습니다(Bourdichon et al., 2012; Joshi et al., 2018).

일부 미생물은

세포 외 단백질 분해효소를 생산하는데,

이 단백질 분해효소는 미생물 세포막에 결합하거나 분비됩니다.

이 단백질 분해효소는

환경 단백질을 분해하여 미생물의 성장을 촉진하지만,

이 과정은 유제품과 육류 제품에 특히 설명된 것처럼 식품 개발을 지원할 수도 있습니다.

그러나

식물성 단백질을 표적으로 하는 미생물의 세포 외 단백질 분해효소에 대한 정보는 드물고,

전통적인 식물성 제품인 사워도우, 와인, 소금에 절인 양배추에 대해서도 흩어져 있습니다.

이는 그림 2에서도 잘 나타나 있는데,

식물성 및 동물성 낙농 제품에 대한 혁신 곡선이 PubMed에 게재된 논문(2022년 2월)으로 표시되어 있습니다.

이것은 식물성 낙농 대체품의 경우 혁신 곡선 초기에 있는 과학 분야에서 지식 격차가 존재한다는 것을 의미합니다.

미생물 발효 유제품 대용품이 프로토타입으로 등장했으며(Madsen et al., 2021a, Madsen et al., 2021b), Tangyu et al.(2019)에 의해 추가 검토되었습니다.

그러나,

유제품 대용품에 대한 세포 외 단백질 분해 활성의 영향에 대한 지식은

없습니다.

1. Download: Download high-res image (124KB)
2. Download: Download full-size image

Fig. 2. The number of publications in animal-based and plant-based dairy

The graph shows the number of publication per year in the Pubmed databased for the two search terms “animal dairy” and “plant dairy”.

Therefore, we review the current knowledge on proteolytic microbes, which show extracellular proteolytic activity with plant proteins as substrates. The focus is on microbial species of LAB, yeast and filamentous fungi, which have a dominating history of use in food fermentation. Hereby, we aim to provide an overview of the application potentials of extracellular proteases targeting plant proteins related to the establish research field of microbial proteolysis of animal-based protein substrates, such as in the animal-based dairy fermentation.

그림 2. 동물성 및 식물성 유제품에 관한 출판물 수
이 그래프는 “동물성 유제품”과 “식물성 유제품”이라는 두 가지 검색어를 대상으로 Pubmed 데이터베이스에 수록된 연간 출판물 수를 보여줍니다.

따라서, 식물 단백질을 기질로 하는 세포 외 단백질 분해 활성을 보이는 단백질 분해 미생물에 대한 현재의 지식을 검토합니다. 식품 발효에 주로 사용되어 온 미생물 종인 락토바실러스, 효모, 사상균에 초점을 맞춥니다. 이를 통해 동물성 유제품 발효와 같은 동물성 단백질 기질의 미생물 단백질 분해 연구 분야와 관련된 식물 단백질을 표적으로 하는 세포 외 프로테아제의 응용 가능성에 대한 개요를 제공하고자 합니다.

2. Well known extracellular proteases from food microorganisms

Microorganisms perform proteolysis for the purpose of survival and growth, but the remaining peptides and amino acids can be useful to humans in food processing. This section explains how food microorganisms use extracellular proteases as part of their proteolytic system, as well as how well-known microbial extracellular proteases have been used in food development.

LAB play a prominent role in many food fermentations, and they dominate the research on the biochemistry of food fermentations. LAB are typically fastidious microorganisms with complex requirements that are auxotrophic for a range of amino acids. Several LAB possess proteolytic systems to satisfy their amino acid requirements. The proteolytic system in some strains includes a cell envelope proteinase (CEP), which allows extracellular proteins to be broken down into peptides short enough to be taken up by peptide transport systems. The CEP enzymes of LAB have been intensively researched, and the CEP of LAB species used in the dairy industry in particular have been thoroughly characterized. Roland Siezen has authored several papers and reviews on CEPs and subtilisins (Siezen, 1999; Siezen and Leunissen, 2008), and Ji, Ma, Xu, and Agyei have recently reviewed the biochemical features and biotechnological applications of CEPs from LAB (Ji et al., 2021). A modeled structure of a dairy associated CEP of Lactococcus lactis (Hansen and Marcatili, 2020) has led to a proposal for how this protease could use an entire casein micelle as substrate rather than individual protein molecules. The model has also led to novel hypotheses about the functions of the domains A, H, and W (Hansen and Marcatili, 2020). The current status is that the majority of research has been conducted on dairy-associated LAB strains, and casein has become the universal substrate for protease research.

However, this choice of substrate might lead to that proteases with alternative specificities are overlooked.

Fungi are another type of microorganism commonly used in food fermentation. Fungi have a proteolytic system based on proteases that can generate small peptides and amino acids for subsequent transport into the cell (Sabotič and Kos, 2012). Depending on the genus, different types of fungal proteases have been described. Aspartic proteases, also known as aspartic acid proteases, are one type of proteolytic enzymes synthesized by many filamentous fungi that can also be produced by some yeast species. Although these enzymes can be considered as virulence factors in fungal pathogenicity, they serve important functions in nutrition, health, and the food industry (Rao et al., 1998). Fungal aspartic proteases are synthesized as inactive precursors (zymogen) to protect against proteolysis. Changes in pH transform them into active enzymes with two aspartic residues at the catalytic site (Horimoto et al., 2009). The lengths of the mature proteins depend on each fungus, where the molecular weights range between 30 and 45 kDa (Mandujano-González et al., 2016). From the chemical point of view, the best activities of these proteases have been documented at low pH values (pH 3–4), and their specificity against aromatic or bulky amino acid residues on both sides of the peptide bond have been reported (Rao et al., 1998; Yegin et al., 2011). Aspartyl proteinases can be produced by Aspergillus spp., Penicillium spp., Rhizopus spp., Neurospora spp., Mucor spp. and Endothia spp., and they can be categorized as pepsin-like and rennin-like enzymes (Sumantha et al., 2006). It is also known that some fungal genera, like Aspergillus or Rhizopus, can synthesized neutral or alkaline proteases, as it is the case of serine and metalloproteases. The group of serine proteases is characterized by having a serine residue in the active site, a low molecular mass (18–35 kDa) and an optimum working pH between 7 and 11 (Gupta et al., 2002). In contrast, metalloproteases constitute a group of diverse proteases that require divalent metal ions for being active (Rao et al., 1998).

Fungal proteases derived from GRAS [generally recognized as safe] microorganisms have traditionally been associated with food bioprocessing. Aspartic proteases play an important role in cheese-making industry, especially as milk-coagulating agents for cheese processing because of their high activity and stability at low pH values (Mandujano-González et al., 2016). The coagulant marked has been reported to be dominated by recombinant produced chymosin (55–60 %), fungal proteases (25–30 %), and the traditional animal rennets constituting the remaining share (10–20 %) (Andrén, 2021). Commercially available proteases used in the dairy industry for milk-clotting are mainly mucorpepsins produced by Mucor pusillus, Mucor miehei or recombinant mucorpepsins synthesized in Aspergillus oryzae, although many fungal species have been investigated as putative sources (e.g. Endothia parasitica, Penicillium oxalicum, Saccharomycopsis fibuligera). Moreover, other fungal proteases, such as neutral proteases from Rhizopus oryzae or Aspergillus niger proline-specific endoprotease, are used in cheese-making for debittering or accelerating cheese ripening times (Feijoo-Siota et al., 2014). Other extended applications of these enzymes are as clarification tools in the wine and beer industries. When beers or wines are not stable, haze appears, which may be caused by the interaction of proteins and polyphenols extracted from plant tissues (Lopez and Edens, 2005; Van Sluyter et al., 2015). Mamo and Assefa reviewed various fungal proteases that could be suitably used to degrade the turbidity complex (Mamo and Assefa, 2018). Acid proteinases produced by S. fibuligera combined with Saccharomyces cerevisiae activity, have been effective in avoiding hazy beers, whereas aspergillopepsins I and II and Botrytis cinerea aspartic protease BcAP8 have successfully eliminated haze-forming proteins in white wines (Mamo and Assefa, 2018).

3. Crops and proteins of key importance in food production

Proteins derived from animals and plants have demonstrated extensive functionality in food technology, not only because of specific physicochemical properties such as emulsification, solubility, gelation and flavor binding, but also because they provide nutritional values and bioactivities (Foegeding and Davis, 2011; Kinsella, 1976; Ma et al., 2022). Although the structure-functionality relationships of animal proteins have been extensively researched and used in food industry, the growing interest in plant proteins as a sustainable alternative has demonstrated challenges regarding their physiochemical and functional properties (Day, 2013; Ma et al., 2022; Qamar et al., 2020). Additionally some plant proteins are associated with toxicity and low nutritional value, among other things, which make them difficult to use in high-quality foods. Nevertheless, enzymatic hydrolysis of plant proteins can provide a way to overcome these challenges (Qamar et al., 2020; Wouters et al., 2016).

This section is based on a review of the literature on the use of plant proteins as substrates for extracellular proteases from microorganism that have traditionally been used in food fermentation, as well as their roles in food product development and improvement. The focus is on LAB, yeast and filamentous fungal species and their extracellular proteases, which have been reported to target plant proteins.

Our main search is performed against the Scopus database using the keyword “lactic acid bacteria” in a parallel search with the keywords yeast, mold and “filamentous fungi”. The two parallel searches include the search terms proteolysis, protease, “proteolytic activity” or proteinase. The search is performed within different groups of plants related to foods and plant-based food products including legume, pea, soy, lentil, bean, rice, cereal, sourdough, bread, seed, quinoa, flour, fruit, grape, berry, apple, wine, juice, seaweed, leaves, grass, vegetable, cabbage, root, potato, carrot, pumpkin, squash, tomato, kimchi, kraut, kvass, chutney, sauce, tea, nuts, and chili. Keyword filters related to microbial species, protein components, and protein hydrolysis have been used to refine the search results before manually assessment of the literature.

Our literature search reveals that certain groups of plants dominate the field of plant protein hydrolysis, which can be divided into four categories: cereals, seeds, fruits and legumes (Fig. 3). Other groups of plants such as nuts have not appeared based on our search criteria though we have conducted extra searches with more specifying keywords on plant sources e.g. almond.

1. Download: Download high-res image (1MB)
2. Download: Download full-size image

Fig. 3. Plant substrates and derived properties for microbial proteolysis

Strains of the listed microbial species display extracellular proteolytic activity targeting plant proteins. Only plant proteins derived from the visualized plants have been studied as substrates for microbial extracellular proteases. The plant protein substrates excite as either a component of a food matrix or as pure protein. Properties of the proteolytic processes have been reported to increase (↑) or decrease (↓) functionalities of plant proteins. Plants are assigned a number if they are among the 15 higest ranked crops in Europe according to FAOSTAT, 2020. The current taxonomy of microbes is used, including the resent taxonomy reorganization of Lactobacillus genus (Zheng et al., 2020). Abbreviation: Angiotensin-converting enzyme (ACE). The majority of the plant figures are from BioRender.com.

3.1. Cereals

Cereals are cultivated grasses of the Poaceae family that provide edible grains with protein contents from 6 to 14 % (w/w) (Coda et al., 2012). The majority of the cereals produced, such as wheat, rye, rice, oat, maize and barley, are used for food or feed production. They are important protein sources in a variety of global diets (Wrigley, 2017). The proteins of cereals provide important properties to the food products and production processes because they aid in the milling, dough forming and baking processes (Schofield, 1994). Fermentation of cereal grains and flours has been proven to improve the sensory and nutritional value of cereal food products, as well as provide longer shelf-life and increase bio-activities (Nout and Motarjemi, 1997). However, there is still a controversy about how fermentation affects quality improvement as well as protein and amino acid availability (Blandino et al., 2003). The microbiology of fermented cereal-based products is complex since spontaneous and industrial fermentations involve mixed cultures.

In sourdough, LAB co-exist with yeast at a general ratio 100:1, where LAB strains can represent a variety of LAB species including Fructulactobacillus sanfranciscensis, Lactiplantibacillus plantarum, Levilactobacillus brevis, Limosilactobacillus pontis, Companilactobacillus paralimentarius, Furfurilactobacillus rossiae, Leuconostoc mesenteroides and Enterococcus faecalis (Gänzle et al., 2008; Gobbetti, 1998; Reale et al., 2021). Lactobacilli species are often present in sourdoughs and recognized for supporting sourdough development (Gänzle et al., 2008; Luti et al., 2020). Peptides of the native cereal proteins are released during the fermentation process as a result of a synergetic system of proteolytic activities. The acidifications of LAB activate endogenous cereal proteases as most of these proteases reach highest activity around pH 4–6 (Gänzle et al., 2008). The proteolysis of these endogenous cereals are known as the primary proteolytic step, which may be followed by microbial proteolysis in a secondary proteolytic step.

Peptides can be released from cereal proteins by chemically acidification (Coda et al., 2012), but LAB strains may contribute to additional protein and peptide degradation processes, increasing the functional properties of cereal proteins. Lactobacilli isolated from wheat sourdough can facilitate the release of antioxidant peptides in a wide range of cereals, including wheat, rye, rice, and oat doughs (Coda et al., 2012). Hereby, LAB strains demonstrate the ability to induce functionalities in some cereal proteins by allowing the release of bioactive peptides embedded in cereal proteins during fermentation.

LAB strains can degrade cereal proteins differently, which is connected to their extra- and intracellular proteolytic activities, among other LAB properties (Reale et al., 2021). These activities may be strain specific and not necessarily connected to species or isolated origin as observed in a clustering analysis of 131 Lactobacilli strains originating from sourdoughs (Galli et al., 2018). Strains of different LAB species have been shown to degrade cereal proteins such as gluten and gliadin (Reale et al., 2021; Stefańska et al., 2016). A plate based assay that uses wheat gluten or gliadin as the sole nitrogen source is effective in detecting extracellular proteolytic activity of a Levilactobacillus brevis strain (Kunduhoglu and Hacioglu, 2021). Clearing zones appear after bacterial growth when the plates are colored with the protein binding dye Coomassie Blue Brilliant, suggesting a quite efficient strain-specific protease activity. This elegant and simple plate assay excludes the contribution from endogenous proteases, but the enzymatic activity of the strain has not been elucidated further. The degradation of cereal proteins in e.g. sourdoughs appears often to be relatively modest. Peptide profiling of sourdough fermented wheat proteins reveals that the cereal proteins are cleaved at the surface and in their more flexible protein structures (Reale et al., 2021). Because even a modest degradation of the sourdough proteins may be enough to degrade epitopes, proteolytic LAB strains may facilitate the reduction of allergenicity of cereal proteins such as gluten (Stefańska et al., 2016). In other cases, epitopes can be targeted first by the protease activities of communities of LAB strains or communities of LAB strains combined with yeast or molds (Fu et al., 2021).

Protein hydrolysis can also occur during yeast fermentation in sourdoughs and alcoholic cereal-based beverages, with diverse effects on the final product depending on the cereal species and the yeast species. The degradation of allergens has been a major focus of wheat fermentation research. Although yeast species can have lower proteinase activity than LAB, some S. cerevisiae and Torulaspora delbrueckii strains present a synergistic effect with P. acidilactici, suggesting that co-cultures may accelerate protein degradation and reduce allergen content in the final wheat product (Fu et al., 2021, Fu et al., 2020). Furthermore, the combination of proteases from sourdough, LAB and fungal proteases has permitted for the elaboration of bread with reduced gluten content and with improved protein digestibility including higher bioavailability of essential amino acids (Rizzello et al., 2014). The protease utilization during long fermentation process has allowed for elimination of toxicity in wheat flour (Rizzello et al., 2007). Moreover, this interaction of LAB (Lactiplantibacillus plantarum) and fungi (R. oryzae) in other fermented cereals such as wholegrain oats has resulted in a higher content of soluble proteins and small peptides with ACE inhibitory activity, which could improve the nutritional and health profiles of wholegrain products (Wu et al., 2018).

Fungal proteases are also applied in the production of cereal based alcoholic beverages. As previously mentioned, proline-specific proteinases derived from A. niger are used to prevent haze formation in beers by degrading proline-rich peptides and proteins (Lopez and Edens, 2005; Van Schaick et al., 2021). Additionally, its combination with novel beer clarification tools appears to improve the colloidal stability and may present a solution for long-terms storage (Cimini and Moresi, 2018). Regarding other cereals, Asia has a great variety of rice wines, which are usually obtained by a simultaneous scarification and fermentation of steamed glutinous rice that has been inoculated with traditional starters, such as Qus, Koji, composed of a complex fungal and bacterial microbiota (Chen et al., 2021). These mixed starters are the most important driving force of sensorial development, but they pose a food safety risk, so specific fungal species have been selected for the elaboration of stable industrial starters (Zhang et al., 2019). Certain strains of A. oryzae, A. niger and R. oryzae species have been used as mixed cultures to enhance key flavor compounds as a result of their synergistic enzymatic activities. Interestingly, three types of proteases (acidic, alkaline and neutral proteases) are present in this multi-strain starter, but only neutral proteases increase protein decomposition efficiency by producing abundant short peptides, whereas acidic proteinases release the highest amounts of free amino acids (Yu et al., 2021). Fusel alcohols, which are characteristic flavor compounds in Asian fermented beverages, are derived from the amino acids catabolism, and their presence is associated to the strong proteolytic activity observed when S. fibuligera is used as the unique fungal starter. Nevertheless, a greater sensorial complexity is described when is combined with A. oryzae strains (Son et al., 2018).

3.2. Seeds

Seeds are embryonic plants, which are protected by outer covering. Nevertheless, as previously stated, seeds are not classified as fruits in terms of food technology. Raw seeds contain anti-nutritional factors and toxic compounds as chemical defenses against other living beings, which can be removed or reduced though technological treatments including fermentation (Gänzle, 2020). Fermentation of some seeds, such as cocoa and coffee beans, is essential for development of aroma precursors. Fermentation may also assist the removal of mucilaginous pulp, which draws out the drying process and leads to the development of spoilage microorganisms. Microbial enzymes from LAB, yeasts and other fungi play a major role during fermentation together with endogenous and native bean enzymes (Haile and Kang, 2019; Ho et al., 2014).

Filamentous fungi have been monitored during coffee and cocoa bean fermentation due to their negative effect during post-harvest processes. However, specific species have been proposed as new starter cultures based on their interesting proteolytic activity. A. oryzae is able to secrete a serine carboxypetidase, which significantly enhances the sweet fruity notes of coffee by increasing the concentration of alkyl pyrazines. Because the production of these compounds is strongly related to amino acids degradation and aminoketone synthesis, it confirm‎s that this protease has an important role in the improvement of the flavor profile of coffee (Murthy et al., 2019). Similar approach has been used in cocoa beans fermentation with an aspartic protease secreted by A. oryzae. In this case, the protease treatment of dry cocoa beans reduces the bitterness and increases the fruity and sweet flavors caused by elevated concentrations of pyrazines and acetates derived from amino acids (Murthy et al., 2020). Green coffee fermentation using Rhizopus oligosporus is also related to aroma modulation and a concentration increase of amino acids, especially alanine, glutamic acid and aspartic acid. Nonetheless, it seems that the change in aroma and amino acid levels is more of a combined effect of protein hydrolysis and R. oligosporus amino acid metabolism during fermentation than a specific impact of the fungal proteases (Lee et al., 2016).

3.3. Fruits

Fruit is defined in food technology as the edible part of a plant, tree, bush, or vine that contains the seeds and pulpy surrounded by an external tissue with a sweet, acid, or sour taste (World Health Organization and International Agency for Research on Cancer, 2003). Fruits are not associated with a high protein content, typically ranging between 0.1 and 1.5 % (Belitz et al., 2009). Most of the protein fraction is constituted of enzymes, and the fruit proteins may be limited to the non-edible parts of the fruit (Belitz et al., 2009; Cejudo-Bastante et al., 2022). Despite their low protein content, the proteins in some fruit-based foods are associated to food quality, shelf-life, and health benefits.

Proteins are of interest in grape juice and wine as precipitation of unstable proteins can result in haze formation over time, reducing shelf-life and quality of the juice and wine. Wine proteins are derived from both the grapes and the yeast in the wine. The hydrolysis of the peptide bonds by extracellular microbial proteases has been studied as a potential alternative to bentonite, which is commonly used to remove unstable proteins. Aspartic proteases (Aspergillopepsin I and II) produced by A. niger are active at wine pH and at temperatures where wine proteins remain unfolded.A combination of these proteases and a heat treatment (75 °C, 1 min) prior to fermentation has been proven to reduce nearly 90 % haze proteins in white wines (Marangon et al., 2012). Other proposed methods include using yeasts such as Wickerhamomyces anomalus and Metschnikowia pulcherrima during white must fermentation since some strains are able of secreting acidic proteases while the process is ongoing (Schlander et al., 2017). Prior to the fermentation, incubation of grape juices for 24–48 h with a protease produced by a S. cerevisiae strain reduces notably the presence of thermo-sensitive (thaumatin-like/27–28 kDa) and thermal-resistant (endochitinases/34 kDa) proteins, confirm‎ing its potential application for hydrolyzing some of the proteins implicated in haze wine formation (Younes et al., 2013).

LAB may also hydrolyze grape proteins during wine fermentation, though the contribution is less elucidated. The LAB can reach population levels similar to yeast in wine fermentation and include the genera Lactobacillus, Leuconostoc, Pediococcus and Oenococcus. These LAB are key contributors in the malolactic fermentation of wine, which is a decarboxylation process converting tart-tasting malic acid into softer-tasting lactic acid. Interestingly, a genetic screening of 120 Lactobacillus strains from wine indicate that genes coding for PrtP-like serine proteases are not rare as they are represented in half of the tested strains and across Lactobacillus species (Mtshali et al., 2010). This study shows a potential of LAB as proteolytic operators in wine fermentation, but their actual role needs further examinations. The proteolytic effects of Oenococcus oeni strains with wine origin are best understood among LAB, and also here the proteolytic abilities appear to be strain specific (Cappello et al., 2010). Four O. oeni strains X2L, L2, m and ST secrete two extracellular proteases with activity against grape juice proteins. These proteases are produced at two different time points, corresponding to the early and late growth phases (Rollán et al., 1993). Protease activity of the LAB strains differs according to strain origin and protease type in terms of stability, optimal conditions and substrate specificity among other things (Rollán et al., 1995, Rollán et al., 1993). Divalent cations affect the activities of the proteases differently, and the protease of the early growth phase appears to be more substrate selective than the protease of the late growth phase. The specific activity of the X2L strain also show higher protease activity against the nitrogenous macromolecular fraction of red wine compare to white wine, which may be related to the lower protein concentration in red wine (Manca De Nadra et al., 1999). This observation is related to how nutrition and energy starvation can induce extracellular proteolytic activity in O. oeni X2L (Rollán et al., 1998). The protease of this proteolytic strain has been partially purified and characterized as an aspartic, dimeric exoprotease whose activity for autoclaved grape juice protein substrate is abolished by Pepstatin A inhibition (Farías and Manca de Nadra, 2000). The dimeric protease has a size of 33 kDa, and it has an optimum activity at pH 4.5 and 25 °C. Other O. oeni strains show extracellular protease activities, but their specificities against plant proteins stand unexamined (Folio et al., 2008; Remize et al., 2005).

Fruit-based by-products, such as seeds, peels and pulps residues, are also protein sources, which are normally discarded during industrial processing. Fermentation of citrus residues with a non-Saccharomyces yeast, Candida utilis, has increased the content of soluble proteins as well as essential and non-essential amino acids, especially leucine and phenylalanine, although protease activity is higher when a co-fermentation is carried out with Bacillus subtilis. Reval‎orization of citrus residue is complex since it requires the participation of several active enzymes such as pectinases, xylanases and celullases. Enzymatic activities have been positively correlated with an increase in soluble protein content. Thus, after fermentation, the proteolytic activity is fostered, and the nutrient composition is improved (Huang et al., 2021). Filamentous fungi have also been tested for their ability to improve low-quality fruit by-products. The protein content increases in pineapple peels and olive cakes when they are fermented by Trichoderma viride and a mixed fungal cultures, respectively. In fact, the combination of different mold species allows for higher protein content (+95 %) than single specie fermentation (+15 %). Although it is assumed that this improvement is produced by the greater presence of proteases in mixed cultures, there is a lack of information regarding the determination and characterization of the specific enzymes implicated (Sabater et al., 2020).

3.4. Legumes

Legumes belong to the family Fabaceae and include among others beans, soybeans, chickpeas, lentils, lupines and peanuts. This group of plants has both important agroecological and nutritional properties why legumes provide a basic pillar of human nutrition since ancient times (Martín-Cabrejas, 2019). Legumes form symbiotic relationships with nitrogen-fixing bacteria by which high-quality biomass of legumes can be obtained without requiring a large amount of nitrogen fertilizer. The protein contents of legumes are high and range between 20 and 35 % (w/w) of the total dry seeds (Martín-Cabrejas, 2019), which make them interesting as food ingredients as well as an important group of croups when changing to a more plant-based diet.

Among the legume proteins, soy proteins have particular interest as they are connected to allergenicity (Meinlschmidt et al., 2016b). Fermentation using Lactobacillus helveticus can significantly reduce the immunoreactitivy of the soluble soy protein β-conglycinin (Meinlschmidt et al., 2016b). Meinlschmidt and coworkers suggest that the reduced immunoreactivity of β-conglycinin may be caused by a combination of acidic protein denaturation and strain specific proteolytic activity. The potential proteolytic activity of a L. helveticus strain has recently been elucidated, and this strain shows indeed specific degradation of several soluble soy proteins, including β-conglycinin, glycinin and albumin (Shirotani et al., 2021). The tested strains show different proteolytic activities for the total soy protein content, with proteolytic activity decreasing after 3 days of fermentation as measured by the release of free amino acids. However, the most proteolytic strains display only moderate activity and appear to preferentially target surface assessable cleavage sites or disordered protein regions. Other LAB and microbial strains also show relatively low proteolytic activity against legume proteins (Rizzello et al., 2019; Sharma et al., 2018), whereas dairy LAB strains of S. termophilus and Lacticaseibacillus rhamnosus show substantial proteolytic activity against soy milk proteins as a dairy alternative (Hati et al., 2018). Mold species as R. oryzae and Actinomucor elegans do not show any effect on major soy allergens as glycinin and β-conglycinin when comparing protein patterns of untreated and fermented soy hydrolysates by SDS-PAGE. Although, their exoproteases can reduce the bitterness of the soy protein hydrolysates to make them more sensory attractive as food ingredient. Pretreatment of the soy protein with commercial microbial proteases permits the complete hydrolysis of the allergen β-conglycinin and the acidic subunit of glycinin by which subsequent fermentation with some molds may allow the production of low-allergen and sensory attractive soy hydrolysates (Meinlschmidt et al., 2016a, Meinlschmidt et al., 2016b). Protease treatments can also increase soluble proteins in plant-based beverages (Manus et al., 2021; Verni et al., 2020). Prior to LAB fermentation, a lentil-based beverage uses a mixture of proteases derived primarily from plants to release peptides and free amino acids (Verni et al., 2020). Further proteolysis occurs during fermentation, but the contribution of LAB strain-derived extracellular proteases is unknown.

In some food matrixes, the microbial growth potential depends on their extracellular proteolytic activities as recently observed for a proteolytic dairy Streptococcus thermophilus strain in soy milk (Boulay et al., 2020). Proteins involved in nitrogen metabolism predominate in numbers among the identified proteins in the proteomic profiling of S. thermophilus grown in soy milk as well as in cow milk. These predominating proteins include proteases and transporters of amino acid and peptides. Their abundance reflects the growth requirement for an extracellular nitrogen source. The extracellular protease PrtS plays an important role of the soy milk fermentation as mutations of this gene disturb the growth of this strain (Boulay et al., 2020). Supplementations of amino acid or low molecular casein hydrolysate do not stimulate growth of S. thermophiles in soy milk, but the PrtS mutant does restore growth at these conditions. In comparison to soy milk, the proteolytic strain still grows best in cow milk, where the highest lactate level is reached during fermentation (Boulay et al., 2020). The lactate production is not limited by the pH of the soy milk, which reaches the same level as for cow milk fermentation. Hereby, the growth of this strain may be limited by other factors than pH and proteolytic activity in soy milk.

Soy proteins, unlike most other plant proteins, have been used as a substrate for proteases discovery and characterization. In this respect, various molds associated with soy environments have emerged as an intriguing source of these enzymes, which are essential in the sensorial and technological development of soy-based foods. The bean curd product tofu can be fermented in a variety of ways to produce sufu, which is a soft cheese-like and highly flavored soybean-based product (He et al., 2022; Yin et al., 2020). Natural fermented tofu and tofu fermented with specific mold strains have a higher level of volatile compounds, which may be related to fungal protease activities. The volatile compounds can be linked to the enrichment of free amino acids produced during A. elegans fermentation of tofu (Yin et al., 2020). Interestingly, A. elegans secretes proteases and contains anabolic pathways for most of the amino acids. Mucor flavus produces a considerable amount of proteases at 15 °C during sufu production, and the rate of the proteolysis and amino-type nitrogen formation are comparable or even greater than that of other molds incubated at normal temperatures (Cheng et al., 2009). Fungal proteases may be used in combination with other enzymes to guide and control the flavor development of sufu production, though the proteolytic effects need to be investigated further (Cheng et al., 2009; He et al., 2022). In contrast, natto is traditional Japanese fermented whole soybean food, and natto fermented with a combination of Mucor wutungkiao and B. subtilis shows higher protease units than other mixtures, resulting in a significant better flavor. In addition, M. wutungkiao is able to reduce the content of various biogenic amines based on its ability of consume polyamines as a nitrogen source (Lan et al., 2020). The synergistic effect of diverse fungal species is also related to the biotransformation process of some soy byproducts like the okara from soy milk production. The yeast Yarrowia lipolytica can secrete proteases that hydrolyze proteins from the soy cell wall in the okara. The proteolysis may improve the action of carbohydrate-cleaving enzymes produced by R. oligosporus during fermentation besides releasing peptides and free amino acids. This microbial and enzymatic combination meliorates the nutritional and flavor properties of fermented okara, which could expand its valorization (Vong et al., 2018).

Because of the activity of microbial proteases, many small peptides are released during soy fermentation. Some of these peptides have demonstrated interesting therapeutic properties such as anti-hypertensive, antioxidant, antimicrobial. Traditional fermented soy products present small peptides with ACE inhibitory ability for treating high blood pressure and hypertension. Different studies highlight that the concentration of peptides with ACE inhibitory activity increases with fermentation and ripening time of soy products like douchi, sufu, mao-tofu and fermented soy milk (Sanjukta and Rai, 2016). Although the amino acid sequences of most soy foods fermented by molds have not been characterized, it is known that peptides with molecular weight less than 10 kDa are responsible for higher ACE inhibitory activity. Soy fermented with Aspergillus egyptiacus for producing douchi (fermented black soybeans) show a separated ACE inhibitor peptide fraction with an amino acid composition of phenylalanine, isoleucine and glycine in the ratio 1:2:5. (Lijun et al., 2003; Ma et al., 2013). In contrast, a different amino acid composition are found for ACE inhibitory peptides derived from soy milk fermented with the LAB species Lacticaseibacillus casei sub. Pseudo Plantarum (Vallabha and Tiku, 2014). Here, glutamic acid and threonine are highlighted to play an important role in ACE inhibition. Peptides derived from tempeh (fermented soybean cake-like product) produced with Rhizopus spp. or R. oligosporus have minor therapeutic effects. Antioxidant activity is found in peptides with aromatic amino acid and His residues that contribute to radical scavenging activity by stabilizing free radicals. Anticancer properties are associated to the peptide lunasin, which is not hydrolyzed by R. oligosporus, unlike other fungi use as starters, such as A. oryzaee (Handoyo and Morita, 2006; Iwai et al., 2002).

4. Applications and challenges of plant protein hydrolysis

Microbial facilitated proteolysis is a well-known phenomenon in traditional fermented food products such as cheese, sausages, and cured ham, where the primary focus has been on the hydrolysis of animal-based proteins. This bioprocessing strategy of food appears to be transferable to plant proteins, but with other challenges due to the physicochemical differences of plant and animal-based proteins. Diverse LAB, yeast and mold species show extracellular protease activity that targets specific or various groups of plant proteins (Fig. 3). Their ability to ferment plants will facilitate the ongoing transition to a more plant-based diet (Fig. 1). Nonetheless, as previously discussed, the identification and characterization of the proteases involved in specific plant proteins hydrolysis, as well as their protein targets, require further and in-depth analysis (Table 1).

Table 1. Overview of extracellular microbial proteases that have been identified and characterized to target plant substrates.

Food substratesProtease TypeE.C. NumberSpecies (strain)Protein targetRoleReference

Beer	Acid proline-specific endoprotease	3.4.21.26	Aspergillus niger	Proline rich peptides and proteins	Haze reduction	Lopez and Edens, 2005; Van Schaick et al., 2021
Rice wine	Aspartic protease	3.4.23.-	A. niger YF2; Rhizopus oryzae YF1; Aspergillus oryzae SU-16	–	Increment of free amino acids	Yu et al., 2021
	Neutral protease	3.4.24.-	A. oryzae SU-16	–	Increment of short peptides
	Alkaline protease	3.4.21.-	A. oryzae SU-16	–	–
Coffee bean	Serine-type carboxypeptidase	3.4.16.-	A. oryzae KX522630	–	Increment on the concentration of alkyl pyrazines	Murthy et al., 2019
Cocoa bean	Aspartic protease	3.4.23.-	A. oryzae CPO 025	–	Increment on the concentration of pyrazines and acetates	Murthy et al., 2020
Wine	Dimeric aspartic protease	3.4.23.-	Oenococcus oeni X2L	Peptides and glycosylated proteins from grape	LAB growth	Rollán et al., 1993; Manca De Nadra et al., 1999
	Aspergillopepsin-1	3.4.23.18	A. niger var. macrosporus	Chitinases and thaumatin-like proteins	Haze reduction	Marangon et al., 2012
	Aspergillopepsin-2	3.4.23.19
	Aspartic protease	3.4.23.-	Wickerhamomyces anomalus 227	–		Schlander et al., 2017
	Aspartic protease	3.4.23.-	Metschnikowia pulcherrima 446	–
	Aspartic protease	3.4.23.-	Saccharomyces cerevisiae PlR1	Chitinases and thaumatin-like proteins		Younes et al., 2013
Citrus pulp	Aspartic protease	3.4.23.-	Candida utilis GIM2.9	–	By-product reval‎orization	Huang et al., 2021
	Neutral protease	3.4.24.-
Soybean	Serine protease (PrtH and PrtH2)	3.4.22.-	Lactobacillus helveticus LH88	β-conglycinin α subunit 1, β-conglycinin α'subunit, glycinin G1, and 2S albumin	Allergen reduction and volatile compound production	Shirotani et al., 2021
	Serine protease (PrtS)	3.4.21.110	Streptococcus thermophiles	–	Soy milk fermentation	Boulay et al., 2020

– Not specified.

4.1. Proteolytic conditions in the fermented plant-based food matrix

Food fermentation provides a dynamic environment for proteolytic events, to which both the raw food material and microbial community contribute (Fig. 4). Additionally, the proteolytic output can be affected by physical and/or chemical treatment before, during or after the fermentation process.

1. Download: Download high-res image (364KB)
2. Download: Download full-size image

Fig. 4. Parameters for microbial-based hydrolysis of plant proteins

Proteolytic parameters are outlined around the triangle and connected to three major groups: environment, substrate, and protease. These groups are highly interconnected, and together they define the reaction time and degree of proteolysis. Besides of external factors, plant and microbial material define the environmental matrix for proteolysis. The substrate and protease originate from plant material and microbial material, respectively, but they are also defined by their protein-protein interactions. Abbreviation: post-translational modifications (PTMs).

The plant cell wall protects a large portion of plant proteins, such as storage proteins (Souza Cândido et al., 2011). Proteins from raw plant material can therefore be spatially protected from protease treatments. In contrast, proteins in some raw animal-based materials are not protected by cell walls such casein and whey proteins in milk. Lysis of plant cells may be required prior to proteolysis or as a step to speed up the fermentation process.

Different endogenous enzymes and inhibitors are common in a broad range of raw plant materials, and their activities can have a strong impact on both the degree and specificity of the protein hydrolysis during fermentation. The endogenous enzymes, including proteases, are inactive in the native plant matrix, but a drop in pH activates some endogenous enzymes (Gänzle et al., 2008). An acidified environment may also activate aspartic proteases from both yeast and molds, as aspartic proteases have working pH optima in the range of 3–6 (Mandujano-González et al., 2016; Purushothaman et al., 2019). Because acidification is a part of LAB fermentation, it can be a strategy to achieve a peptide pool in a plant-based environment. The addition of microbial proteases with low pH optima may facilitate further hydrolysis of plant proteins by which availabilities of essential amino acids can be increased to support additional microbial fermentation with both proteolytic and non-proteolytic LAB and yeasts strains. Changes in the pH can also provide changes in plant protein solubility, which can make the proteins more or less assessable for proteolysis during acidifying fermentation (Gänzle et al., 2008; Shirotani et al., 2021). Soy protein has low solubility and forms gelation at acidic conditions, which may limiting their accessibility to proteolytic degradation during LAB fermentation (Shirotani et al., 2021). However, hydrolysis of soy proteins can increase their solubility by using aspartic proteases even at pH 4 (Purushothaman et al., 2019). Proteolytic properties have a potential to increase protein solubility, which interesting property could be used as a technological tool for stabilizing and produce complex plant-based products.

The contribution of endogenous enzymatic activities may also be removed by heating the raw plant material before fermentation. Thermal treatment may destabilize the endogenous enzymes as well as protease inhibitors along with other proteins in the food matrix. For example, when rice is steamed previous to a fungal fermentation to make rice wine, the content of free amino acids increases, which is related to flavor improvement (Yang et al., 2021). Therefore, heat treatment may change the structure of target plant proteins for proteolysis. A non-thermal protein pre-digestion can also be applied as it has been done in the sourdough matrix. The addition of fungal proteases (A. niger and A. oryzae) before lactic acid fermentation significantly reduces the gluten content of bread. The previous proteolysis may increase peptide availability for LAB and result in bread with higher digestibility, nutrition quality and similar structural and sensory features than non-treated dough (Rizzello et al., 2014).

Other intrinsic factors like the presence of ethanol after alcoholic fermentation led by Saccharomyces spp. may induce partial denaturation and precipitation of some of plant proteins (Nikolaidis and Moschakis, 2018). However, some oenological LAB secrete proteases that can increase the rate of amino acid liberation in presence of ethanol, implying that adapted microorganisms may generate adapted proteases for specific substrates conditions (Manca de Nadra et al., 2005). Other environmental changes can lead to structural changes in protein substrates, making them more or less accessible to proteases by which proteolytic activities may be increased or reduced.

Different groups of proteases require different salts, particularly divalent cation ions, in order to support their proteolytic activities. The structures of the CEPs are stabilized by the high Ca2+ level in milk by which they stay bound to the cell envelope of LAB (Exterkate and Alting, 1999). CEPs can be released in Ca2+ deficient media without losing their abilities to hydrolyze casein. Other cations have a direct impact on protein hydrolysis like Mg2+ and Mn2+, which are essential for casein degradation (Guo et al., 2016). The extracellular microbial proteolytic activity targeting plant proteins also show to be differently affected by cations (Rollán et al., 1993). Sodium chloride stabilizes the aspartic protease of A. niger and protects the protease from thermal inactivation (Purushothaman et al., 2019). In contrast, increasing sodium chloride concentration reduces the unspecific proteolytic processes of rice koji derived A. oryzae in dairy fermentation, just like the salt reduces proteolytic activity of Penicillium roqueforti in blue cheese (Dank et al., 2021). Also, Ca2+ and Mn2+ salts can increase the activity of a neutral protease secreted by a A. oryzae strain isolated from fermented broad beans, whereas Mg2+, Ba2+ and Zn2+ salts present an inhibitory effect (Ao et al., 2018). Hereby, salts can be used to stabilize and adjust proteolytic activity in food developments (Fig. 4).

Protease expression‎ can be quite energy consuming for microorganisms, as it is for the CEPs of LAB because CEPs are typically synthesized as pre-proteins of approximately 2000 amino acids (Ji et al., 2021). CEPs of LAB are only expressed when the environment is deficient in nitrogen sources and essential amino acids (Ji et al., 2021). In this way, the environment dictates the proteolytic activity by regulating the expression‎ levels of some proteases besides the environmental impact on the structure and activity of the proteases and their substrates (Fig. 4).

4.2. Protein degradation in cross-over fermented and co-fermented food matrixes

The microorganisms demonstrate strain-specific proteolytic abilities against plant-based proteins. Different strains of the same microbial species can have different proteolytic activities, and not all strains of the same microbial species do necessarily have extracellular protease activity (Mtshali et al., 2010). Dairy LAB strains indeed show strain-specific proteolytic patterns of their homologous extracellular proteases where CEPs from different Lactococcus lactis subsp. cremoris strains both show differences in cleavage patterns as well as substrate selectivity for e.g. caseins. The underlying mechanism for these differences in proteolytic specificity and selectivity remains unclarified (Hansen and Marcatili, 2020). This highlights the challenges in predicting microbial proteolytic activity against target substrates, such as plant-based proteins, because their extracellular proteolytic potential cannot be linked to microbial species or isolated origin. One the other hand, this suggests that these food microbes contain a broad catalog of microbial extracellular protease variants, which could be used in combinations as part of a fermentation process and/or across food protein substrates.

The concept of cross-over fermentation has recently been defined where a microorganism of a traditional fermentation process is introduced to a new substrate and/or new microbial partner in a mixed culture (Dank et al., 2021). Although, this is not a novel concept, the technical term provides a good framework for studies focusing on non-traditional substrates for microbial-based proteolysis. LAB strains of well characterized dairy starters can target plant proteins using their extracellular enzymatic properties, as observed for L. brevis and S. thermophilus strains, which hydrolyze wheat and soy proteins, respectively (Boulay et al., 2020; Kunduhoglu and Hacioglu, 2021). Probiotic supplement with three LAB strain has also been used to ferment plant protein-enriched beverages containing soy, hemp and/or rice protein concentrates (Manus et al., 2021). Although the protein degradation of the co-culture fermentation has not been directly linked to microbial extracellular proteases, the probiotic co-culture increases the soluble protein fraction and the levels of dipeptides and free amino acids in the plant-based beverages.

Several studies have also established that LAB strains originating from plants can utilize animal derived proteins as casein, which is a popular substrate in screenings of proteolytic food microbes. As an example, a plate-based screening assay has identified a L. lactis strain and an Enterococcus hirae strain as caseinolytic, and proteolytic potential of these microbes have been followed up by a second screening with different substrates (Pérez Borla et al., 2010). Among the caseinolytic LAB strains, only the L. lactis strain generates clearing halos on plates with other animal-based proteins, but also wheat gluten. Gluten is the only tested plant protein of this study why its ability to hydrolyze cabbage proteins is unexplored. This work shows that extracellular proteases of LAB can exhibit varied substrate selectivity across and between plant and animal food matrixes, implying that proteolytic microbes may bring unique features in future cross-over fermentations. A similar selection approach is provided in Ben-Harb et al., 2019 study, which creates a mixed plant-based product (e.g., 50 % milk – 50 % pea protein) and a pure pea protein emulsion. However, the proteolytic activity of caseinolytic L. lactis strain remains selectivity against the substrates reported (Pérez Borla et al., 2010). The strong substrate selectivity of some extracellular LAB proteases raised the question of whether proteolytic screenings overlook other, possibly novel, proteolytic activity by using casein as a substrate. One way to circumvent this concern is to test non-caseinolytic strains against other substrates, as Pérez Borla et al., 2010 did by using random checks to reduce the uncertainties for overlooking targeting proteolytic bacteria.

Another aspect of the plant protein degradation is the occurrence of posttranslational modifications (Fig. 4). Wheat germ agglutinin, a cereal lectin protein, have 16 disulfide bridges to stabilize its monomeric structure. This protein stability can be challenged by microbes with glutathione reductases, which can change the redox potential during fermentation (Rojas Tovar and Gänzle, 2021). Transglutaminase activity of microbes may also cross-link different cereal proteins by isopeptide bonds during sourdough fermentation. This create structural networks with dough structure qualities, but it can also enhance the formation of volatile compounds (Scarnato et al., 2016). However, it is unclear how these and other posttranslational modifications influence on proteolytic processes in fermentation where proteases of the microbial community and plant matrix may synergetic contribute to plant protein degradation.

A few studies have looked into the synergetic proteolytic effects between microbes of co-cultures, which can promote proteolytic activity in plant fermentation (Fu et al., 2021; Wu et al., 2018). Fermentations with co-cultures result in complex changes to the food matrix including the proteins as a result of microbial competition and synergetic activities (Fig. 5). In rich nutrient environments such as sourdoughs or grape juices, LAB and yeast growth are stimulated by the lack of competition for the nitrogen source and the excretion of amino acids, peptides and micronutrients by the yeasts. Yeast proteases can release some essential amino acids such as valine and isoleucine and small peptides that allow specific LAB species to grow (Gobbetti et al., 1994). Because of the accelerated metabolism of LAB, monosaccharides are easily available and permit yeasts to uptake glucose, which induce the secretion of amino acids from the cell. The efficient use of these peptides and amino acids produced by yeast provides a competitive advantage for the LAB, which contribute to the stability of the fermented food (Gobbetti, 1998). Hereby, growth performance of microbes can be supported by a complex synergetic interplay in a microbial consortia. A L. plantarum strain isolated from cereal has improved growth performance when co-cultured with R. oryzae in whole-grain oat fermentation (Wu et al., 2018). The improved growth may be related to protein metabolic processes as the co-inoculated fermentation provides higher protein solubility, degree of hydrolysate and content of smaller peptides. RNA-sequencing shows transcriptomic changes of P. acidilactici when co-cultures with the yeast species S. cerevisiae and T. delbrueckii (Fu et al., 2021). These changes indicate that yeast can enhance the protein metabolism of P. acidilactici, which correlates with accelerated protein degradation and reduce immunoreactivity of insoluble wheat proteins. Furthermore, the proteolytic activity of Y. lipolytica might have made carbohydrates, such as cellulose and hemicellulose, more accessible to R. oligosporus carbohydrases. These carbohydrate-cleaving enzymes would transform insoluble dietary fibers into disaccharides and monosaccharides, which could then be further catabolized to support microbial growth (Vong et al., 2018). This synergistic metabolism between Y. lipolytica and R. oligosporus may be responsible for the majority of the conversion from non-digestible polysaccharides to soluble fibers, which can improve the nutritional value of some plant-based foods.

1. Download: Download high-res image (713KB)
2. Download: Download full-size image

Fig. 5. Synergistic effects of proteolytic food microbes in plant fermentation

Principal synergistic effects (arrows) are observed in food fermented with co-cultures. The figure center depicts the main compounds derived from proteolytic activity of plant proteins, including peptides and amino acids (aa). The synergetic effects of the microbial proteases (scissors) may further facilitate the proteolytic process, but degradation of carbohydrates (pentagon and hexagon) may also facilitate microbial proteolysis of plant proteins. The data reviewed in this article support interactions between lactic acid bacteria (LAB) and fungi based on the only plant fermented foods that have been studied, such as sourdoughs, wines, and soy products, as well as products resulting from proteolytic activity. The figure does not include synergetic contributions of endogenous enzymes within the plant matrix.

4.3. Hydrolyzing plant proteins in food product developments

Endogenous proteases found in plant-based matrices may provide enough free peptides and amino acids for microbial growth, which is why these proteases enable development of fermented food products without the use of microbial extracellular proteases (Fig. 4). Therefore, traditionally plant-based food products and spontaneously fermented plant-matrixes are not necessarily sources of microbes with extracellular proteases, but examples of microbes with plant origin have been identified with extracellular protease activity (Cappello et al., 2010; Fu et al., 2021; Reale et al., 2021; Shirotani et al., 2021). Microbial assisted proteolysis can further enhanced proteolysis, which may increase protein digestibility (Rizzello et al., 2019). Hereby, the microbial proteases may expand the functional catalog of proteolysis by introducing potential increased and modified cleavage events. Monocultures of LAB used as dairy starter cultures perform not as well in soy milk compared to cow milk (Hati et al., 2018), but the proteolytic capacity is not the limiting growth factor for the PrtP containing S. thermophilus strain (Boulay et al., 2020). Nevertheless, the extracellular proteolytic capacity of S. thermophilus is directly linked to its growth potential in soy milk (Boulay et al., 2020). The soy proteins are not fully degraded but just enough to support growth. Among 276 LAB strains, the S. thermophilus strain acidifies soy milk most efficiently (Harlé et al., 2020); whether this difference is related to their extracellular proteolytic potential is unknown. However, microbial assisted proteolysis of plant-based proteins can be required for some strains and support plant protein functionalities, which is why food microbes with extracellular proteases have the potential to facilitated plant-based food development in a similarly way that they are used in animal-based foods (Fig. 1).

To link the protease with the microbial proteolytic capacity, a specific extracellular protease has been mutated in a S. thermophilus strain. Such research studies are uncommon, leaving a knowledge gap in our understanding of microbial-assisted proteolysis of plant proteins (Fig. 6). The proteolytic activities of microbes have been studied in diverse plant-based food matrixes (Fig. 3) using different methods, the most prominent of which are growth experiments (Fig. 6). Spectrometric assays such as enzyme linked immunosorbent assay (ELISA) and o-phthalaldehyde (OPA) assay are frequently applied to quantify proteolytic processes of plant-based substrates. Proteolysis is also frequently quantified by measuring levels of free amino acids with liquid chromatography-mass spectrometry (LC-MS) or by considering the substrate degradation patterns by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The qualities of plant-based protein substrate hydrolysis is less frequently assessed. Peptide profiling has provided valuable insights into the moderate cleavage of plant-based substrates, as the microbial extracellular proteases mostly cleaved the surface assessable surface areas and intrinsic disorder regions of their target substrates (Reale et al., 2021; Shirotani et al., 2021). Microbial proteases with plant protein targets have been studied with either bioinformatics or biochemical approaches (Fig. 6). Genome mining and also polymerase chain reaction (PCR) show the presence of plant origin homologs of proteases used in dairy while SDS-PAGE has characterized proteases in proteolytic fractions. Future studies are requested to bridge the knowledge gap between proteolysis and the proteases to match, for example, by supplementing proteolytic studies with bioinformatics searches and characterization of potential microbial extracellular proteases. Extracellular proteolytic activities of LAB strains are not well characterized with a few known extracellular proteases with plant protein targets (Table 1), as observed for O. oeni with different suggested proteases in wine fermentation (Section 3.3). Extracellular protease activities of yeast and filamentous fungi are better characterized since there appear to be few species and protease types that are useful in plant protein hydrolysis (Table 1). Nonetheless, more LAB species can be selected as candidates for proteolytic activity targeting plant proteins since it seems that studies with molds commonly focus on the well-known species (e.g. A. oryzae, R. oryzae, and A. niger), which is more notorious in the case of yeasts (e.g. S. cerevisiae and S. fibuligera). Future studies may focus on the proteolytic potential of other non-investigated fungal species as well as non-conventional yeasts. However, food fermentation procedures are associated with potential health risks, which is why microorganisms and their metabolites should be handled with caution in both spontaneous fermentation and innovative co-cultures, as well as cross-over fermentation with applied starter cultures (Capozzi et al., 2017, Capozzi et al., 2020; Verni et al., 2020).

1. Download: Download high-res image (183KB)
2. Download: Download full-size image

Fig. 6. A methodological gap in the study of microbial extracellular proteolysis of plant proteins

The schematic illustration shows some methods, which are applied to study proteolysis of plant proteins using microbial extracellular protease. The methods are listed in decreasing order of relative use from left to right. The frequency is related to the purpose of the methods, which is to characterize proteolysis as a process or the protease itself. The connection between the proteolytic process and the protease is rare, leaving a gap illustrated by the “broken” triangle. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) has been used to study proteolytic processes and characterize proteases by focusing on the substrate or the protease, respectively. Other abbreviations: liquid chromatography-mass spectrometry (LC-MS), Polymerase chain reaction (PCR).

Characterization of microbial proteases is required to predict and steer proteolytic events in food developments in order to obtain target protein functionalities. Plant proteins are generally moderately hydrolyzed, leaving the hydrolysate with larger peptide fragments and the majority of proteins in their native states (e.g. A. Reale et al., 2021; Shirotani et al., 2021). Defining the optimal conditions for a target microbial driven proteolysis are challenging where both substrate, protease and microorganism must tolerate the fermentation conditions, which can change over time (Fig. 4). Extracellular microbial proteolysis has only been analyzed in a relatively few plant-based protein sources, which only represent four groups of crops. These groups are only represented by a few crops of which six are among the 15 major crops in Europe (FAOSTAT, 2020) (Fig. 3). Other crops have not, to our knowledge, been studied in relation to extracellular microbial proteolysis, despite the fact that major crops are clearly missing. Together with wheat and sugar beets, maize and potatoes are two of Europe's four major crops. Maize is a cereal with 8–12 % (w/w) protein of its kernel. These proteins embed bioactive peptides with antioxidant, antihypertensive, anticancer and antimicrobial properties that are primarily released through Alcalase hydrolysis (Díaz-Gómez et al., 2017). Antioxidant activity has not been increased by using a specific combination of different Lactobacilli strains, which release antioxidant peptides from other cereal proteins (Coda et al., 2012). This strain specific proteolytic activity of microbes highlights the importance for further searching of proteolytic microbes, which can provide proteases with desired proteolytic activities against diverse plant protein substrates. Potato is a root containing 1–2 % (w/w) proteins, and it is essential as core component in diet and in industrial starch production. Potato protein accumulated in industrial side streams provides a valuable source for bioactive peptides with emulsifying and antioxidant activities, which activities are verified using corresponding synthetic peptides (García-Moreno et al., 2020b). Hereby, plant proteins of diverse plant groups may embed valuable bioactive peptides with important functionalities for food developments. The release of bioactive peptides require specific proteolytic cleavage where food microbes with modest extracellular protease activity may be ideal for their release. The specific cleavage is challenged by the protease selectivity and specificity which determined if and where it cleavages its substrate. Therefore, novel extracellular proteases and protease composition are required to fully explore the proteolytic potential of LAB, yeast and molds.

5. Conclusion

Versatile extracellular proteases are found in traditionally applied food microbes of LAB, yeast and mold species, the origins of which cover a wide range of animal- and plant-based food matrixes. The protease activity is specific for the microbial strain and not to the species or the isolated origin of the microbe. This indicates that these food microbes have a diverse set of extracellular proteases, which can be used on substrates across food matrixes. Extracellular proteases of LAB, yeast and mold have only been shown to target plant proteins from four crop groups, with a very limited representation of plants within each group. The low representation of plant substrates in proteolytic studies is a challenge. Although casein is the most commonly used substrate for proteolytic screenings, this approach runs the risk of overlooking novel proteolytic activities because extracellular proteases can display versatile substrate selectivity across and between plant and animal proteins. We are facing a knowledge gap where the proteolytic process is rarely connected to the protease itself, and vice versa. This knowledge gap makes it challenging to predict and steer the proteolytic capacity of microbes. However, the hydrolysis of the four plant groups reflects a promising application potential for developing plant-based foods as dairy alternatives. Microbial proteolysis can improve plant protein digestibility, just as their extracellular proteolysis can target protein-based off-flavors, anti-nutritional factors, toxins, and allergens. Extracellular proteolysis of plant proteins appears to be quite modest, which may explain why the desired properties of the plant protein hydrolysate may be facilitated by a complex synergetic interplay of microbes in co-cultures. Specific cleavage of proteolytic microbes may also release bioactive peptides, which provide valuable functions to foods. Hereby, the search for plant-targeting proteolytic microbes is not limited to supporting microbial growth in plant fermentation, where endogenous proteases can cover the amino acid requirements for some microbes. The vision is to explore and utilize the full proteolytic potential of food microbes in order to facilitate the development of high quality food products as part of the ongoing transition to a plant-based diet.

Declaration of competing interest

The authors have no conflicts of interest to declare.

Acknowledgement

LFC and EBH were funded by the Danish national grant: Innovation Fund Denmark project Provide [grant number 7045-00021B]. BGBB was supported by the University of Castilla – La Mancha and the European Social Fund who kindly funded the postdoctoral grant as well as the international research stay. CHBB has been supported by Food & Bio Cluster Denmark projects: ApplePom [grant number RFDK-19-0021] and “Fra Rapskage til proteinrig fødevareingredients” (Rapeseed) [grant number 1022-00005B]. The funding organizations had no involvement in the preparation, reviewing or the decision to publish the manuscript.

References

1. Aiking and de Boer, 2020

   H. Aiking, J. de Boer

   The next protein transition

   Trends Food Sci. Technol., 105 (2020), pp. 515-522, 10.1016/j.tifs.2018.07.008

   View PDFView articleView in ScopusGoogle Scholar

2. Andrén, 2021

   A. Andrén

   Milk-clotting enzymes

   A.L. Kelly, L.B. Larsen (Eds.), Agents of Change: Enzymes in Milk and Dairy Products, Springer International Publishing, Cham (2021), pp. 349-362, 10.1007/978-3-030-55482-8_14

   View in ScopusGoogle Scholar

3. Ao et al., 2018

   X.Lin Ao, X. Yu, D.Tao Wu, C. Li, T. Zhang, S.Liang Liu, S.Juan Chen, L. He, K. Zhou, L.Kou Zou

   Purification and characterization of neutral protease from Aspergillus oryzae Y1 isolated from naturally fermented broad beans

   AMB Express (2018), p. 8, 10.1186/s13568-018-0611-6

   Google Scholar

4. Belitz et al., 2009

   H.D. Belitz, W. Grosch, P. Schieberle

   Fruits and fruit products

   Food Chemistry, Springer, Berlin, Heidelberg (2009), pp. 807-859, 10.1007/978-3-540-69934-7

   Google Scholar

5. Ben-Harb et al., 2019

   S. Ben-Harb, A. Saint-Eve, M. Panouillé, I. Souchon, P. Bonnarme, E. Dugat-Bony, F. Irlinger

   Design of microbial consortia for the fermentation of pea-protein-enriched emulsions

   Int. J. Food Microbiol., 293 (2019), pp. 124-136, 10.1016/j.ijfoodmicro.2019.01.012

   View PDFView articleView in ScopusGoogle Scholar

6. Blandino et al., 2003

   A. Blandino, M.E. Al-Aseeri, S.S. Pandiella, D. Cantero, C. Webb

   Cereal-based fermented foods and beverages

   Food Res. Int., 36 (2003), pp. 527-543, 10.1016/S0963-9969(03)00009-7

   View PDFView articleView in ScopusGoogle Scholar

7. Boulay et al., 2020

   M. Boulay, M. Al Haddad, F. Rul

[치토스]