The plant cell wall is composed mostly of carbohydrate-based polymers, i.e., cellulose, hemicelluloses and pectins. All the plant cells that are in developmental expansion have a primary cell wall that is constantly remodeled, composed of cellulose fibers embedded in a hemicellulose-pectin matrix (Burton et al., 2010). At the end of the plant cell development, i.e., once the cell has ceased to expand, a secondary layer composed by cellulose, hemicellulose and lignin is deposited close to the primary cell wall; the architecture of such assembly varies among different cell types, being optimized to perform cell-specific functions (Keegstra, 2010; Höfte and Voxeur, 2017; Zhong et al., 2019). Lignocellulose from agricultural feedstocks is formed by clusters of secondary cell walls; in terms of energetic potential, lignocellulose appears as a deposit of reducing power stored in its complex structure and composition. Phytopathogens as well as saprophytes secrete a wide array of CWDEs to open a breach in the plant cell wall; the degradation of this complex matrix into simple sugars significantly contribute to the global carbon cycle (Gibson et al., 2011; Choi et al., 2013; Glass et al., 2013). CWDEs are divided in many categories and sub-categories depending on their substrate specificities toward the diverse cell wall polysaccharides. The most important CWDEs directed toward the plant cell wall polysaccharides are:

The bacterial cell wall is mainly composed of peptidoglycan, a heteropolymer consisting of amino sugars and oligopeptides. The backbone chain is formed by alternating residues of β-1,4(linked) N-acetylglucosamine (NAG) and N-acetylmuramic acid (NAM) (Vollmer et al., 2008; Dörr et al., 2019); NAM residues from adjacent chains, in turn, are interconnected by peptide bridges of variable length, ranging from three to five amino acids. Although the peptidoglycan composition is conserved among bacterial species, the overall cell wall architecture varies between Gram-positive, Gram-negative and acid-fast bacteria (Desvaux et al., 2018). The degradation of peptidoglycan is performed by amidases and glycosidases (Figure 1B; Alcorlo et al., 2017; Vermassen et al., 2019).

Fungal cell wall is a two-layered structure, with a hydrophobic and rigid inner layer, consisting of a branched β-1,3(linked) gluco-polysaccharide cross-linked to small portions of chitin (Kang et al., 2018); this layer is covalently linked to galactomannan and other branched glucans that, together, form the outer layer of the fungal cell wall (Gow et al., 2017; Kang et al., 2018). The outer layer is more hydrophilic and flexible than the inner layer due to the lack of chitin. Chitin is a homopolymer composed of β-1,4(linked) NAG residues. Notably, chitin is also a major component of the exoskeletons of arthropods, crustaceous shells, mollusc radula and cell walls of several microalgae species (Cauchie, 2002). Chitin naturally occurs in three different crystalline structures depending on the orientation of the constituting microfibrils. α-type and β-type chitin are characterized by antiparallel and parallel chains, respectively; γ-type chitin is a mixture of α-type and β-type chitin (Rudall and Kenchington, 1973). Chitin degradation is performed by chitinases, i.e., glycosyl-hydrolases that cleave the glycosidic bond between the NAG residues of chitin (Figure 1C).

Microalgae are found in a wide range of habitats and have adapted to a variety of environmental conditions by evolving a great genetic diversity, making them a precious source of interesting and useful metabolites (Benedetti et al., 2018a). However, the extraction of metabolites from microalgae represents one of the major bottlenecks limiting their potential in industrial applications. Enzymatic treatment is much less efficient than physical rupture, pointing to the necessity of developing CWDE blends effective in the degradation of the microalgal cell wall. It is worth noting that a univocal structural model of the microalgal cell wall cannot be established likely due to the broad adaptive diversification of microalgae. For example, the cell wall of the unicellular green alga Chlamydomonas reinhardtii is formed by seven layers of hydroxyproline-rich glycoproteins (Imam et al., 1985) while the cell wall of the marine microalga Nannochloropsis gaditana is composed by polysaccharides, i.e., cellulose and algaenan (Scholz et al., 2014). In this paragraph, we will focus on the cell wall of Chlorella vulgaris, one of the most employed oleaginous microalgae with applications ranging from the biofuel sector (i.e., production of biodiesel through lipid transesterification) to nutraceutical (Benedetti et al., 2018a; Dall’Osto et al., 2019). Although C. vulgaris is a well-studied organism compared to other microalgae species, the degradation of its cell wall is still a high hurdle. Preliminary studies demonstrated that the cell wall of C. vulgaris and other related microalgae species had rigid wall components embedded within a more plastic polymeric matrix. The acid-hydrolysis of this polymeric matrix revealed the presence of acid sugars, rhamnose, arabinose, fucose, xylose, mannose, galactose and glucose (Takeda, 1991). Subsequently, Gerken and collaborators showed the cell wall of C. vulgaris is constituted by a heterogeneous bilayer matrix; the inner layer is mainly composed of polysaccharides such as cellulose and pectin, while the external one is composed by a robust chitin-like glucan (Gerken et al., 2013).

Cell wall degrading enzymes from hyperthermophilic microbes (HCWDEs) represent a category of high industrial interest due to their peculiar enzymatic characteristics. These enzymes are also known as “Hot Extremozymes” since they are active at temperatures ranging from 70 to 100°C (Sarmiento et al., 2015). The high temperature required for optimal activity and stability of HCWDEs allows faster and more effective reactions (Yeoman et al., 2010). Moreover, elevated temperature prevents undesired growth of contaminating microbes during the catalysis, thus improving the conversion yield of cell wall polysaccharides into simple sugars. Proteinaceous CWDE-inhibitors, that are widely distributed in the plant cell wall as a defense mechanism (York et al., 2004; Juge, 2006; Mohammadzadeh et al., 2012; Kalunke et al., 2015), are inactivated by high temperature, thus avoiding interference with the enzymatic reaction. Another important feature of HCWDEs is protein stability that allows prolonged storage at room temperature and resistance to harsh conditions, e.g., the presence of aggressive chemicals, anionic/non-ionic detergents and extreme pH (Benedetti et al., 2019b), that can be exploited to deconstruct more efficiently cell wall recalcitrant material. Stability of HCWDEs also allows an efficient enzyme recycling over time, thus reducing the total enzyme loading in industrial practices. However, maintaining industrial processes at high temperature for a long time requires a great expense of energy, therefore a further step toward sustainability may imply the use of HCWDEs in industrial plants with excess heat that can be recycled in order to limit the additional heating cost. Up to now, not all the CWD-activities toward plant cell wall polysaccharides are available in their respective hyper-thermostable version. In particular, while hyper-thermostable orthologs have been isolated for mesophilic cellulases, hemicellulases and ligninases, the exo-polygalacturonases from Thermotoga maritima and Caldicellulosiruptor bescii (Kluskens et al., 2005; Chen et al., 2014) are the only pectinases isolated so far, and neither endo-polygalacturonases nor pectate lyases of hyperthermophilic nature have been identified yet. Similarly, LPMOs from the thermophilic bacterium Thermobifida fusca are the only available option for the degradation of crystalline cellulose at mid-high temperature (Moser et al., 2008). Other carbohydrate active enzymes with important industrial applications are amylases, employed in starch conversion, biofuel production, brewing, bakery, textile, detergent and paper industry. Well-known α-amylase producers are bacteria belonging to the genus Bacillus such as B. subtilis, B. licheniformis, and B. amyloliquefaciens (Jujjavarapu and Dhagat, 2019), while β-amylases are mainly obtained by plants such as barley (Hordeum vulgare) and sweet potato (Ipomoea batatas) (Saini et al., 2017). Unlike α-amylases from Bacillus species, β-amylases are less resistant to high temperature; in this regard, the use of a hyper-thermostable β-amylase may be a cost-effective choice for reducing the amount of enzyme used during the catalysis; to date, the only β-amylase of hyperthermophilic nature with proven activity was identified from Clostridium thermosulfurogenes (Nipkow et al., 1989).

Other HCWDEs of industrial interest are those degrading fungal and bacterial cell wall polysaccharides. A highly thermostable chitinase was isolated from Pyrococcus furiosus (Oku and Ishikawa, 2006); this enzyme showed marked degrading activity toward both the amorphous and β-type chitin, while it was less active toward α-type chitin. Noteworthy, highly thermostable lysozymes were also identified; they were isolated from hyperthermophilic bacteriophages such as the Pseudomonas phage φ KMV (Lavigne et al., 2004); the substrate specificity of thermostable lysozymes is not comparable to that of egg-white lysozyme commonly used in food processing, thereby precluding their exploitation in this field.

The industrial use of HCWDEs has been so far limited by the fact that they cannot be efficiently produced at large-scale by wild-type microbes. Production of recombinant HCWDEs may be carried out in E. coli by high cell-density fermentation. As an alternative expression system, the last advances in plant biotechnology allowed the large-scale production of recombinant HCWDEs also in tobacco plants (Castiglia et al., 2016; Schmidt et al., 2019) (see section “Yield and Cost Analysis of Plant CWDE-Producing Biofactory”). In Table 1, a selection of CWDEs with both consolidated and potential use in industrial applications is proposed. According to this selection, the use of HCWDEs is suggested for biofuel production rather than food processing since the high temperature of action may alter the organoleptic properties of food and beverages. The techno-economic aspects of different CWDE-expressing biofactories are discussed in section “YIELD AND COST ANALYSIS OF ENZYMES FROM DIFFERENT CWDE-PRODUCING BIOFACTORIES.”

Cellulolytic and pectinolytic microbes are mainly cultivated by submerged and solid-state fermentation. In submerged fermentation, bacteria and filamentous fungi are grown in liquid medium containing nutrients, macro and micro-elements under sterile conditions and continuous oxygen supplementation. The most appropriate carbon source is selected depending on the type of expressed CWDE, e.g., glucose and galacturonic acid for optimal production of cellulases and polygalacturonases, respectively. Submerged fermentation is usually operated in fed-batch or continuous culture. In fed-batch cultures, enzyme production occurs mainly during the first phase of biomass growth. Subsequently, the growth rate in the culture is maintained by adding nutrients at different time-points, thus reducing the risk of overflow metabolism. On the opposite end, continuous fermentation requires an open growth system where sterilized solutions of nutrients are added in continuum to the bioreactor at the same rate at which the fermented medium is recovered from the system. This procedure results in a steady-rate production of the fermentation broth. In order to maintain the fermentation as efficient as possible, parameters such as temperature, pH, oxygen and carbon dioxide levels must be continually monitored and adjusted.

Solid-state fermentation is used as an alternative to submerged fermentation. Solid-state fermentation occurs in the absence of free water, can usually reach greater volumetric productivity than submerged fermentation and is also characterized by an easier downstream process. In general, the solid-state fermentation involves a solid matrix (e.g., rice bran, wheat bran, steam exploded agricultural scraps) that is used as feed by the selected microbe. The substrate matrix is maintained for days at controlled temperature under constant or intermittent rotation allowing the growth of filamentous fungi under conditions that resemble their natural environment. Air flow rate must be monitored since it affects both the humidity and oxygen levels of the entire fermentation process. The use of sterilization procedures is not mandatory for solid-state fermentation except for the sterilization of the substrate at the beginning of the process. In general, solid-state fermentation has several advantages over submerged fermentation, such as the lower consumption of water and electricity, a lower waste efflux and a more concentrated enzymatic product (Zhuang et al., 2007). These three parameters impact the final cost of the enzyme-based product that will be cheaper than that obtained via submerged fermentation. At the end of the fermentation process, the CWDEs are separated from the residual materials by micro- and ultra-filtration procedures (Machado de Castro et al., 2010; Ferreira et al., 2018). Subsequently, the enzyme preparation will be concentrated at the desired level and stabilizing agents will be supplied to the preparation in order to increase the shelf-life of the enzyme-based product. Notably, if a high-value product is required as in the case of CWDEs for use in food processing and biomedical field, further purification procedures will be necessary in order to eliminate residual contaminants, thus increasing the final cost of the enzyme.

Pectinases, cellulases, xylanases, and α-amylases are produced by microbial fermentation. By adjusting the carbon source used for inducing CWDE production, different enzymes can be secreted in the same fermentation broth; e.g., Aspergillus awamori, a fungus with GRAS (Generally Recognized As Safe) designation, secreted simultaneously cellulases, xylanases and α-amylases when fed with the babassu cake, i.e., the residual material from babassu palm (Attalea speciosa) upon oil extraction (Machado de Castro et al., 2010).

One of the most relevant fungi for cellulase production is Trichoderma reesei. This filamentous fungus is employed as cellulase-producing bioreactor since 1980; a strain of this fungus named RUT-C30 was characterized by high level of cellulase production and, nowadays, is still exploited for the industrial production of cellulolytic enzymes such as cellobiohydrolases, endo- and exo-glucanases (Peterson and Nevalainen, 2012). The main weakness of T. reesei resides in the low level of secreted β-glucosidase, the enzyme responsible for the last step of cellulose degradation, i.e., the cleavage of cellobiose in two D-glucose units. Notably, cellobiose inhibits both endoglucanases and cellobiohydrolases by a product inhibition mechanism, thus pointing to the importance of β-glucosidase for an efficient enzymatic hydrolysis at industrial scale (Sørensen et al., 2013). In this regard, enzymatic blends from T. reesei often require supplementation with exogenous β-glucosidases. Important sources of pectinases, xylanases, α-amilases and β-glucosidases are fungi belonging to the genus Aspergillus. Amongst them, Aspergillus niger plays a prominent role in the industrial production of pectinases and cellobiase, i.e., a type of β-glucosidase that specifically cleaves cellobiose (Singh et al., 1990; Reginatto et al., 2017). Expression of cellulases, pectinases and α-amylases can be obtained by bacteria using submerged and solid-state fermentation as well. Amongst the many bacteria used as bioreactor, the most relevant in the production of cellulases and pectinases are B. licheniformis and B. subtilis. In addition to CWDEs from wild-type bacteria, recombinant CWDEs can be produced by E. coli through high cell density fermentation. Differently from other fermentation methods, high cell density fermentation is characterized by the highest operating costs (Ferreira et al., 2018). The advantage of using E. coli as bioreactor resides in five main factors: (i) rapid growth cycle, (ii) use of cheap growth media, (iii) capability of reaching high cell density through fermentation, (iv) possibility of using strains with low proteolytic activity, and (v) a thorough knowledge of its physiology and genetics. However, the translational machinery of bacteria cannot introduce post-translational modifications such as the formation of cytoplasmic disulphide bridges or glycosylation. Moreover, the poor ability of Gram-negative bacteria to secrete recombinant proteins in the culture medium implies expensive and time-consuming purification procedures, thus representing additional costs for their large-scale production (Table 2). To overcome this drawback, different approaches to enhance the secretion efficiency of recombinant CWDEs in E. coli have been developed. The yield of secreted recombinant cellulases and hemicellulases was increased by fusing the catalytic domain of the CWDE of interest to the CBM from another higher soluble cellulase (Murashima et al., 2003) or by reducing the temperature during expression in order to promote a correct protein folding (Song et al., 2012). Alternatively, expression hosts with higher secretion capabilities can be used such as the Gram-positive bacteria B. subtilis (Morello et al., 2008) or Lactococcus lactis (Pohl and Harwood, 2010), and yeasts; the latter have higher secretion efficiency and can perform post-translational modifications. Many CWDEs have been successfully expressed as secreted proteins in Saccharomyces cerevisiae (Margolles-Clark et al., 1996; Haan et al., 2007, 2013; Van Zyl et al., 2010; Zhou et al., 2015), Kluyveromyces lactis and Pichia pastoris (Bey et al., 2011; Chen et al., 2011; Akbarzadeh et al., 2014; Table 2). Compared to other yeasts, P. pastoris and K. lactis glycosylate the secreted proteins at lesser extent, reducing the side-effects of unwanted hyperglycosylation. Yeast has been genetically improved to increase CWDE expression levels by the use of synthetic (Blazeck et al., 2013) and constitutive promoters (Haan et al., 2007; Wang et al., 2012), codon-optimized genes (Akcapinar et al., 2011) and selection of multicopy transformants (Yamada et al., 2011; Mellitzer et al., 2012).

Plant expression of CWDEs can be a valid alternative to microbial-based bio-factories. Plants are characterized by a high [productivity/production cost] ratio and consume atmospheric CO₂ through photosynthesis, thus representing an eco-friendly expression system. Moreover, plants have GRAS status and are suitable for the production of high-value compounds at affordable costs. However, plant expression of CWDEs may have side-effects since CWDEs from lignocellulolytic fungi and bacteria are well-known pathogenicity factors whose accumulation can damage plant tissue and produce toxic effects (Benedetti et al., 2019a). Moreover, CWDEs can trigger plant defense responses independently of their enzyme activity, being recognized as dangerous non-host molecules through specific recognition mechanisms mediated by plant pattern recognition receptors (PRRs) at the apoplast/outer membrane interface (Poinssot et al., 2003; Ma et al., 2015; Zipfel, 2014; Choi and Klessig, 2016). In general, the expression of pectinases and cellulases in plants can be more insidious with respect to hemicellulases (Benedetti et al., 2015; Castiglia et al., 2016). It is worth mentioning that cell wall fragments derived from the partial breakdown of pectin and cellulose are powerful inducers of plant defense responses (Souza et al., 2017; De Lorenzo et al., 2019). Constitutively activated immunity may lead to detrimental effects due to the growth-defense trade off (Huot et al., 2014; Benedetti et al., 2015), resulting in impaired growth and eventually lethality. Recently, an enzymatic machinery capable of inactivating the elicitor activity of these cell wall derived fragments, namely oligogalacturonides and cellodextrins, has been identified (Benedetti et al., 2018b; Locci et al., 2019). Moreover, gene silencing events may also occur: the nuclear expression of the endoglucanase EGII from Acidothermus cellulolyticus in transgenic rice, for example, is suppressed during development from seedling to mature plant, resembling a self-defense mechanism (Li et al., 2019). Gene silencing is a well-known epigenetic mechanism in land plants that undermines the use of transgenic crops for field applications. Different plant expression strategies can be adopted to prevent these undesired event, such as compartmentalized expression and storage, inducible gene expression, expression of CWDEs with inducible activity (i.e., CWDEs with extreme pH- and temperature-dependent activities), use of hosts that are less sensitive to CWDEs (e.g., microalgae) or a combination of these strategies (e.g., chloroplast expression of hyperthermophilic CWDEs) (Benedetti et al., 2019a). The chloroplast expression of CWDEs in tobacco plants has several advantages: (i) high yields of recombinant protein since, unlike nuclear expression, chloroplast expression is not affected by gene silencing events, (ii) easier biocontainment of transgenes since the plastome is maternally inherited, thereby avoiding dispersion through the pollen by vertical gene transfer (Daniell, 2007; Adem et al., 2017), and (iii) the chloroplast expression of CWDEs in tobacco is well-documented, with a recent demonstration that CWDE production may be achieved in field condition without loss in enzyme yield for at least three consecutive growth cycles (Schmidt et al., 2019). Moreover, the chloroplast translational machinery catalyzes the formation of disulphide bridges, enabling the expression of eukaryotic CWDEs in their active form with no need of subsequent renaturation procedures (Taunt et al., 2018). However, the chloroplast system is not suitable for expressing cellulases that require glycosylation for proper activity and stability (Greene et al., 2015), although this trait could be favorable in other cases to avoid the undesired glycosylation of recombinant proteins. Tobacco can be grown in field with a production cost of 2€ kg DW^–1 and an annual productivity of about 8.1 t (ha^∗y)^–1 considering three growth cycle per year as the best achievable for tobacco plants under optimal latitude condition (45°N) (Maksymowicz and Palmer, 1997; Foreman, 2006; Faè et al., 2017; Schmidt et al., 2019; Table 2). Even if a high-density cultivation method has been reported for tobacco plants, transgenic plants expressing CWDEs have never been cultivated with this technology (Scott and Warren, 2012). Enzyme yields from transplastomic tobacco are very encouraging, on the other hand plants need arable land for their growth and longer times to produce significant amounts of recombinant proteins compared to microbes. In this regard, microalgae may represent a good alternative since they are characterized by faster growth rates and the capability of growing at higher CO₂ concentration than land-plants. Moreover, microalgae can be cultivated in closed growth systems, i.e., photobioreactors, that can occupy waste lands, avoiding the subtraction of arable lands to the agri-food sector (Slade and Bauen, 2013; Tredici et al., 2016; Clippinger and Davis, 2019). Photobioreactors are closed environments where microalgae are exposed to continuous mixing, air supplementation, controlled light and temperature; it comes that algal productivity increases in photobioreactors as well as the production cost. Here, the productivity of microalgae, expressed as biomass produced per year, can reach 60 t (y^∗ha)^–1, with a production cost of 3.2 to 3.8 € kg DW^–1 (Slade and Bauen, 2013; Tredici et al., 2016; Table 2). The production cost of photobioreactor-grown microalgae varies based on three main parameters: (i) the microalgal species of interest, (ii) the size, and (iii) the type of the photobioreactor used for the cultivation (Clippinger and Davis, 2019). The yield of CWDEs in the chloroplast of Chlamydomonas reinhardtii, i.e., the microalgae used as model organism, reached lower level compared to that of transplastomic tobacco plants (Faè et al., 2017; Richter et al., 2018); therefore, further optimization is still required for improving the expression stability and protein yield of CWDEs from microalgal-based biofactories (Benedetti et al., 2018a). However, it is worth noting that microalgae culturing may combine the expression of recombinant proteins to the production of valuable biomass (Benedetti et al., 2018a).

Trichoderma reesei is a filamentous fungus employed as cellulase-producing bioreactor secreting more than twenty different cellulolytic enzymes during fermentation. The cultivation of T. reesei can be carried out by submerged fermentation using both glucose and steam-exploded poplar as substrates for inducing the secretion of cellulases. By submerged fermentation, T. reesei produced up to 2.6 t (y^∗m³)^–1 of cellulases (Humbird et al., 2011), with a production cost comprised between 3.82 and 9.12 € kg^–1 (Humbird et al., 2011; Klein-Marcuschamer et al., 2011). It is worth noting that the commercial cellulase-based products are not only composed by cellulase(s). In general, other ingredients added on purpose and contaminants also contribute to the final weight of the cellulolytic blend. For example, some cellulolytic blends are composed by 15% (w/v) of cellulases and 20% (w/v) of glucose, the latter used as stabilizer to prolong the shelf-life of the enzymes (Rodrigues et al., 2015). By submerged fermentation, A. niger produced pectinases up to 9400 Units L^–1, with an enzymatic yield of about 7.7 ^∗10⁵ Units per kg DW mycelium (Reginatto et al., 2017). A techno-economic analysis on the production of xylanases and α-amylases by A. awamori is described in Machado de Castro et al. (2010). In this study, the large-scale production of α-amylases and xylanases and, at lower extent, of cellulases, was carried out by solid-state fermentation using babassu cake as substrate. The highest levels of production were reached upon 6–7 days of solid-state fermentation with values of 250.000- and 85.000-Units kg^–1 DW for xylanase and amylase activity, respectively. With respect to other filamentous fungi, Aspergillus awamori is characterized by the GRAS designation, positively impacting the production cost of the enzymes because the fermented substrate can be sold as feed, thus reducing the total costs of the entire process. According to this possibility, the initial production cost of the CWDE-based product (30.74 € kg DW^–1) could be reduced to 9.36 € kg DW^–1 upon selling the fermented babassu cake, reaching the same production cost of enzymatic blends from T. reesei. Production of a recombinant β-glucosidase in E. coli by high cell density fermentation reached a productivity of 0.88 t (y m³)^–1 and a production cost of 284 € Kg^–1 using a fermenter of 100 m³. At the end of the process, the β-glucosidase was available in a concentrated form (15 g enzyme L^–1) and ready to be supplemented in cellulolytic blends. Notably, the addition of β-glucosidase in cellulolytic blends in the ratio 0.9:10 [w_{β –glucosidase}: w_cellulases] would increase the final price of the enzymatic blend up to 137% (Ferreira et al., 2018). A techno-economic analysis concerning the production of thermostable cellulases by thermophilic bacteria was described in Zhuang et al. (2007). Here, the thermostable cellulases were produced by Clostridium thermocellum, a thermophilic bacterium with optimal growth at 60°C, using both the submerged and solid-state fermentation. The main characteristic of this thermophilic bacterium resides in the capability of producing cellulosomes, multi-enzyme complexes characterized by more than twenty different catalytic domains (Gold and Martin, 2007). Submerged and solid-state fermentation allowed the production of cellulases with a cost ranging from 14.1 to 36.3 € Kg^–1 for those obtained by solid-state and submerged fermentation, respectively. B. subtilis is also used as a pectinase-producer organism. Submerged fermentation of B. subtilis reached a maximal yield of pectinolytic activity around 66100 Units L^–1 (Oumer and Abate, 2018), about 7-times higher than that obtained from A. niger by using the same fermentation method (Reginatto et al., 2017); to date, techno-economic assessments of pectinase-expressing bio-factories have not been reported despite the high number of studies concerning the production of pectinases through different fermentation procedures (Garg et al., 2016).

Cell wall degrading enzymes production is generally confined to the chloroplast, thus avoiding the interaction with both PRRs and cell wall polysaccharides and preventing the formation of oligosaccharidic elicitors. Nonetheless, chloroplast expression of CWDEs preserved plant fitness at variable extents (Verma et al., 2010; Castiglia et al., 2016; Schmidt et al., 2019). Chloroplast expression of the hyper-thermostable cellulase from Sulfolobus solfataricus resulted in tobacco plants with stunted growth and pale-green phenotype despite the expression was compartmentalized and the enzyme activity was inducible by high temperature (Castiglia et al., 2016). Notably, transplastomic tobacco expressing the β-glucosidase from Pyrococcus furiosus accumulated 14.5 ^∗ 10⁶ Units kg DW^–1 plant biomass and a theoretical enzyme yield of 57 g kg DW^–1 in accordance with the specific activity of the enzyme (255 Units mg^–1). In this case, the enzyme productivity was 0.46 t (ha^∗y)^–1 while the production cost of the enzyme was 35 € kg^–1 in the tobacco leaf wet basis. Surprisingly, the specific activity of β-glucosidase from tobacco chloroplast was higher than that reported for the native enzyme from P. furiosus, i.e., 750 vs 446 Units mg^–1 as determined at 90°C using p-nitrophenyl-glucoside as substrate (Kengen et al., 1993; Castiglia et al., 2016). However, the transplastomic expression of β-glucosidase from P. furiosus affected the plant development at variable extent, resulting in both dwarf and wild-type like plants. Similar yield was obtained for the xylanase from Alicyclobacillus acidocaldarius; in this case, plant accumulated 1.6 ^∗ 10⁶ Units kg DW^–1 plant biomass and a theoretical enzyme yield of 15.2 g kg DW^–1 in accordance with the specific activity of the enzyme (105 Units mg^–1). Here, the enzyme productivity was 0.12 t (ha^∗y)^–1 while the production cost of the xylanase was 132 € kg^–1 in the tobacco leaf wet basis. In Faè et al. (2017), transplastomic tobacco plants accumulated 140 g kg DW^–1 of the cellulase CelK1 from Paenibacillus; notably, plants grew without morphological defect. The enzyme productivity was 1.13 t (ha^∗y)^–1, while the enzyme cost was 14.3 € kg^–1 in the tobacco leaf wet basis. However, the enzyme activity in the leaf tissue was very low (12^∗10³ Units kg DW^–1) indicating that CelK1 was not very active, with a theoretical specific activity of 0.08 Units mg^–1. A further study on the production of CWDEs from transplastomic tobacco plants in field-grown condition is described in Schmidt et al. (2019). According to this study, plants stably accumulated the thermostable endoglucanase Cel6A from Thermobifida fusca over different growth cycles reaching a maximum yield of 40 g DW^–1 plant biomass. The enzyme productivity was 0.32 t (ha^∗y)^–1, while the production cost of the enzyme was 50 € kg^–1 in the tobacco leaf wet basis. A comparison of activity would also be necessary, since data on Cel6A activity are not reported in this study. Healthy tobacco plants that accumulated high yield of active endoglucanase and two different pectin lyases are described in Verma et al. (2010). The enzyme productivities ranged from 0.81 to 1.14 t (ha^∗y)^–1 for the endoglucanase celD from C. thermocellum and pectin lyase PelD from Erwinia carotovora, respectively, with an enzyme production cost comprised between 14.2 and 20 € kg^–1 in the tobacco leaf wet basis.

All these results taken together suggest that plant-CWDE interaction is not easily predictable and therefore, it should be evaluated case by case. As an alternative, photosynthetic hosts less sensitive to plant CWDE could be used, such as the microalga C. reinhardtii. Transplastomic expression of endoglucanases has been obtained in C. reinhardtii although the enzyme yield was lower than that from transplastomic tobacco plants (Faè et al., 2017; Richter et al., 2018). The lack of a clear phytotoxic effect in C. reinhardtii toward plant CWDEs could be ascribed to the algal cell wall composition, mainly protein rather than polysaccharidic (Imam et al., 1985), along with a different host-microbe coevolutionary adaptation. In terms of yield, the chloroplast of C. reinhardtii may accumulate recombinant protein up to at most 5% (w/w) of the total soluble proteins (Manuell et al., 2007), corresponding to 1.2 % (w/w) of the algal biomass (Rasala et al., 2010). Cultivation of microalgae can be achieved at high productivity in closed growth systems such as photobioreactors; amongst the several techno-economic analyses performed on such topics, the best-case scenarios were reported by Slade and Bauen (2013) and Tredici et al. (2016). Considering these parameters, C. reinhardtii may theoretically produce a hypothetical CWDE up to 12 g kg^–1 DW of microalgal biomass, with an enzyme productivity of 0.72 t (ha^∗y)^–1 and enzyme production costs ranging from 267 to 317 € kg^–1 in the algal biomass wet basis.

Lysozyme isoforms from hen egg-white and human are the most relevant in biotechnological field; human lysozyme (HUL) is preferable over egg-white lysozyme (EWL) due to the higher specific activity (HUL: 201526 Units mg^–1 vs EWL: 70400 Units mg^–1) (Chiang et al., 1993; Huang et al., 2002b) and to the “safer” designation since HUL, unlike EWL, does not cause allergic reactions in humans (Ercan and Demirci, 2016). Notably, the use of lysozyme in medicine implies further purification procedures and, therefore, higher production costs compared to other CWDEs with application as low-value products (e.g., biofuels, paper industry). Due to its hydrolysing activity toward the cell wall of bacteria and filamentous fungi, expression in such organisms is unfeasible. Lysozyme is commonly obtained from hen egg-white upon extraction and purification. The amount of lysozyme is around 0.15 g per egg (Abeyrathne et al., 2013); given that the average weight of an egg is 65 g and that the production cost of a single egg is 0.05 € in a conventional hen-housing system, one kilogram of eggs contains about 2.3 g of lysozyme with a production cost of 0.75 € (Matthews and Sumner, 2015). Based on these estimations, the production cost of EWL is around 326 € kg^–1 in the egg wet basis. Wilken and Nikolov estimated that the purification procedure for obtaining EWL with a purity greater than 94% had an approximate cost of 1845 € kg^–1 (Wilken and Nikolov, 2011); such procedure was characterized by a purification yield of 68% indicating that about 600 kilograms of eggs are required to obtain 1 kg of pure lysozyme. According to this scenario, the production cost of a high-purity lysozyme preparation from egg-white is 2295 € kg^–1 (3 × 10⁷ Units €^–1). However, many lysozyme-based products such as those used in food processing are characterized by lower costs (expressed as € kg^–1 product) because the real lysozyme content does not correspond to the weight of the product. In these cases, customers should always refer to the Units per gram of product rather than to the sole weight. In this regard, it is worth noting that the enzymatic Units of lysozyme can be defined through diverse methods, making the comparison between different lysozyme-based products rather uncertain.

In addition to the hen egg-white, other important sources of lysozyme are recombinant yeast and rice. Lysozyme can be purified upon production by Kluiveromyces lactis through high-cell density fermentation with higher costs, i.e., 0.14 × 10⁷ Units €^–1 (Huang and Demirci, 2009). Importantly, the use of yeasts for expressing CWDEs with potential applications in the biomedical field requires great caution since the secreted CWDEs could be hyper-glycosylated. Glycosylated proteins from yeast-based biofactory cannot be used in food processing and medicine due to the possibility of triggering allergic reactions in humans. Fortunately, the amino acid sequence of lysozyme is not characterized by the presence of potential N-glycosylation sites. The most relevant bio-factory of HUL is transgenic rice (Oryza sativa). The expression level of lysozyme in the rice grain reaches 6 g kg^–1 DW and therefore, 1 kilogram of recombinant lysozyme is contained in about 167 kg of rice grains (Huang et al., 2002a). Bearing in mind that the annual rice productivity can reach about 8 t (ha^∗y)^–1 under the best climate and irrigation conditions² and the production cost of recombinant grain flour is 0.9 € kg^–1 DW, i.e., five-fold higher than the common grain flour (Wilken and Nikolov, 2011), then the annual productivity of HUL from transgenic rice is around 0.048 t (ha^∗y)^–1 with a production cost of 150 € kg^–1 in the rice flour wet basis. Wilken and Nikolov estimated that the purification procedure for obtaining HUL from recombinant rice with a purity greater than 91% had an approximate cost of 1971 € kg^–1. This procedure was characterized by a purification yield of 55% indicating that about 304 kg of rice grains are required to obtain 1 kg of pure lysozyme. According to this information, the production cost of pure HUL from recombinant rice is 2244 € kg^–1 (8.9 × 10⁷ Units €^–1).