Yeast cell factories for sustainable whey-to-ethanol valorisation towards a circular economy

,


Contents 1. Introduction
Cheese whey is the major by-product of cheese processing in the dairy industry.It corresponds to the watery portion remaining after the coagulation of milk, and it is free of fat and casein.Cheese whey has a high nutritional value, with riboflavin (vitamin B12) responsible for its yellowish colour (Ryan and Walsh, 2016).
During cheese manufacturing, 9 L of cheese whey is generated for every 1 kg of cheese produced.According to the Food and Agriculture Organization, the production of cheese whey has been highly enlarged over the past yearsan increased rate of 3.8% was observed between 2014 and 2018 (Smithers, 2015;FAO, 2020).In 2016, the worldwide production of whey from the cheese industry reached 190 million tonnes, and it is expected that its production will increase at the same rate of milk production, of 2% per year (Ryan and Walsh, 2016).With the rise in the consumption of cheese, higher volumes of cheese whey are also being formed, contributing to the intensification of industrial wastes.
The dairy industry sector is the main source of liquid waste in Europe, essentially due to the disposal of cheese whey (Demirel et al., 2005).The organic components of cheese whey are responsible for its highly polluting character.The Biochemical Oxygen Demand (BOD) and Chemical Oxygen Demand (COD) are key parameters used to characterize waste effluents.The COD value represents the amount of oxygen consumed in a particular biodegradation reaction, whereas BOD is related to the amount of dissolved oxygen necessary by aerobic biological organisms to degrade the organic matter of a mixture under specific conditions.The BOD and COD of cheese whey range between 30-60 g/L and 50-102 g/L, respectively (Guimarães et al., 2010;Yadav et al., 2015;Ryan and Walsh, 2016).Together, COD and BOD give the Index of Biodegradability (BOD/COD) that indicates if a mixture can be easily degraded in the environment.The Index of Biodegradability of cheese whey is between 0.4 and 0.8, which implies serious environmental problems.Moreover, the waste load of cheese whey is equivalent to 100-175 times the volume of domestic wastewater (Carvalho et al., 2013;Ryan and Walsh, 2016).Currently, most cheese whey surplus is used for animal feeding or spreading into the land, which constitutes an environmental problem.Cheese whey alters the soil composition and leads to an excess of oxygen consumption, impermeabilization, eutrophication, and toxicity in the environment where it is discarded.This compromises the growth of plants and decreases crop yields (Carvalho et al., 2013;Yadav et al., 2015).Furthermore, the available treatments to reduce the pollution load of cheese whey are not sufficient to efficiently reduce its environmental impact and imply high operational costs.
Due to the high volumes of cheese whey produced worldwide and the implied environmental hazards, it is urgent to develop new approaches to treat and/or reutilize cheese whey surplus in the dairy industry.The rising awareness of environmental problems and the increasing need for sustainability drive the research of recycling and reusing waste streams.This follows a circular economy system, where the products, materials, and resources are maintained in the economic system as long as possible to minimize the generation of waste (D'Amato et al., 2017; Álvarez-Cao et al., 2020).In Figure 1, a circular economy system for the valorisation of cheese whey surplus from the dairy industry is schematically represented.Novel integrated food biorefineries fed with cheese whey to generate functional foods with improved properties present solid perspectives and have been reviewed recently (Lappa et al., 2019).These food biorefineries follow the circular economy concept by allowing the reintroduction of produced bio-based food components in the food industry, aiming at zero waste discharge.In this sense, the high organic content of cheese whey represents a great potential to become an important resource that should be more broadly explored (Smithers, 2008).During the past years, several studies have addressed the valorisation of cheese whey through (1) the utilization of technology to recover valuable compounds, such as lactose, proteins, and derived compounds with different applications; and (2) the use of biological processes to generate value-added products such as ethanol, organic acids, single-cell proteins, enzymes, and bio-proteins (Prazeres et al., 2012;Mollea et al., 2013).Recently, cheese whey valorisation efforts gained a new driver with the need to develop biorefineries beyond the food industry.Biorefining is defined by the International Energy Agency as the sustainable processing of biomass into a spectrum of marketable products and energy.Biorefineries are classified in first, second and third-generation based on the different features of the feedstock.First-generation biomass  The basis of industrial biotechnology is the selection of optimal cell factories to establish a cost-competitive process.Significant progress has been made to develop yeast-based fermentation processes for converting cheese whey into fuels and chemicals.Some yeast strains, such as Kluyveromyces lactis, Kluyveromyces marxianus, Kluyveromyces fragilis (now considered a K. marxianus species, accordingly to the current nomenclature), and Candida pseudotropicalis, have been used as platform cell factories due to their natural ability to consume lactose (Guimarães et al., 2010).The yeast Saccharomyces cerevisiae, widely recognized as an ideal microbial cell factory to produce bioethanol (Cunha et al., 2020) and to be used in lignocellulosic biorefineries (Baptista et al., 2021), can also be tailored for the valorisation of cheese whey through lactose consumption (Guimarães et al., 2010).

Abbreviations
Table 1 shows the various features of the present review compared to those of the reviews on cheese whey valorisation published in the last five years (2017-2021).While the majority of these focused on the use of microbial fermentation, the use of metabolic engineering strategies using yeast for bioethanol production is practically uncovered.This review spans the valorisation of cheese whey effluent using yeast as cell factories, covering metabolic engineering strategies for cheese whey surplus utilisation, and its integration in lignocellulosic biorefineries, for the production of value-added products, with a special focus on bioethanol.

Types of cheese whey and composition
Cheese whey is mainly composed of water (95% of total whey volume), lactose (70-72% of total solids), whey proteins (8-10% of total solids) and minerals (12-15% of total solids) (Ryan and Walsh, 2016).The exact composition of cheese whey is variable depending on the type of cheese whey, source of milk, and processing technology (Panesar et al., 2007;Pescuma et al., 2015).According to the technique used to remove casein from milk, cheese whey can be classified into sweet or acid whey.The sweet whey (pH around 5.6) is obtained from a mixture of an industrial casein-clotting enzymatic complex called rennet (containing the protease chymosin), leading to the casein coagulation.This is the first step in cheese manufacturing, and it is used in most types of cheese production (Panesar et al., 2007).The acid whey (pH = 4.5) results from the precipitation of casein, using a mixture of organic (lactic acid) and mineral acids (hydrochloric or sulphuric acid), or with the action of specific bacteria (Jelen et al., 2003).These two types of whey present differences in minerals content, acidity, and protein bulk (Yadav et al., 2015;Ryan and Walsh, 2016).As a rule of thumb, acid whey has higher ash and lower protein content than sweet whey (Carvalho et al., 2013).
Lactose is the main constituent of cheese whey, followed by other proteins, including 50-55% of -lactoglobulin, 20-25% of -lactoalbumin, 10% of immunoglobulins, 5-10% of serum albumin, and small quantities of other proteins such as protease peptone, lactoferrin, lactoperoxidase and glycomacropeptide (Ryan and Walsh, 2016).In terms of minerals, more than 50% of the total minerals content is constituted by calcium, as well as sodium and potassium.Trace elements present in cheese whey include zinc and copper.Additional compounds can also be present in low quantities, like lactic and citric acids, non-protein nitrogen compounds (urea and uric acid), and B-group vitamins (Jelen et al., 2003;Ryan and Walsh, 2016).Thus, whey is a source of different and diverse compounds that can be directly used for several applications or be valorised through biotechnological processes.

Cheese whey processing and valorisation
Approximately 50% of the cheese whey produced worldwide is used as a supplement in animal feeding and as a land fertilizer or in food applications (ice creams, cakes, sauces, and derivatives) (Guimarães et al., 2010;Ryan and Walsh, 2016;Castillo et al., 2020).
The first step of cheese whey processing includes drying whey to obtain a whey powder form (González-Siso, 1996).The water content is reduced throughout a reverse osmosis process, followed by vacuum evaporation.After that, it is spray dried and crushed to obtain the powdered aspect.This allows to preserve the properties of whey for a long time, as well as to facilitate storage, manipulation, and transportation.
Proteins present in cheese whey (around 20% of total milk proteins) are important sources of essential amino acids, especially sulfur-containing amino acids like methionine and cysteine, which are reported to act as nutraceuticals (Lappa et al., 2019).They present valuable properties for food and cosmetic applications, such as high solubility, water absorption, gelatinization, and emulsifying capacities (Gunasekaran et al., 2007).Currently, whey proteins are extensively used as a food ingredient and nutritional supplement for muscle building.They also present health benefits associated with their antioxidant, antihypertensive, and somniferous properties (Audic et al., 2003;Korhonen, 2009;Modler, 2009).Another interesting application of whey proteins is for bioactive peptides production via enzymatic or fermentative processes (Yadav et al., 2015).Considering these advantageous properties, the application of technologies to recover the protein fraction, such as ultracentrifugation or membrane-based systems, is well established in the dairy industry (Pouliot, 2008).The resulting concentrated protein fraction can be treated to originate three different protein whey powders: whey protein concentrate (WPC), whey protein isolate (WPI), and whey protein hydrolysate (WPH) (Smithers, 2008;Álvarez-Cao et al., 2020).Due to its high nutritional properties, concentrate whey powder has a commercial value 3 to 40 times higher than whey powder (Mollea et al., 2013), and it can be subjected to several purification steps to obtain the purified proteins.
The permeate is a lactose-rich stream resultant from the protein fractionating process.This fraction contains 70% of the total solids of whey and still represents an environmental problem (Ryan and Walsh, 2016).Whey permeate can be directly used in human nutrition due to its oligosaccharides content (Barile et al., 2009;Mollea et al., 2013) or be subjected to a lactose recovery process by crystallization (Paterson, 2009).Lactose is used for baby milk formulation, as an excipient in the pharmaceutical industry, and also to produce added-lactose derivatives, such as lactulose, lactitol, lactobionic acid, lactosyl urea, galacto-oligosaccharides, and lactosucrose (González-Siso, 1996; Guimarães et al., 2010;Álvarez-Cao et al., 2020).Alternatively, lactose can be hydrolysed to originate a mixture of glucose and galactose for lactose-free dairy products or can be used as food sweeteners (Gänzle et al., 2008).Another application for lactose in cheese whey or whey permeate is the production of functional beverages.It can be mixed with fruit juice/pulp to obtain fruit-whey beverages -a mineral and vitamin-enriched product that combines the beneficial effects of whey and fruits (Zotta et al., 2020).Alternatively, whey or permeate can be converted into functional beverages through microbial fermentation using a consortium of lactic acid bacteria and yeast strains.The fermentation of lactose results in whey acidification (due to lactic acid production) and the formation of compounds such as acetaldehyde and diacetyl.These compounds are responsible for the aroma and flavour of beverages and improve beverages' shelf life and sensorial properties, respectively (Mudgil and Barak, 2019;Zotta et al., 2020).Furthermore, during fermentative processes, bioactive peptides are also produced, enhancing the functionality of the wheybased beverages (Brandelli et al., 2015).Cheese whey is also a potential raw material for the production of alcoholic beverages.Dragone et al. (2009) developed an alcoholic distilled beverage by continuous fermentation of cheese whey using K. marxianus.The volatile composition analysis of the produced drink identified compounds similar to those present in other alcoholic beverages (Dragone et al., 2009).Fermentation of whey has also been used to produce kefir-like whey beverages.The classical method to produce kefir involves milk fermentation using kefir grains, which contain a stable mixture of bacterial and yeast species (Magalhães et al., 2010).The utilization of these grains as starting culture for cheese whey and deproteinised cheese whey fermentation produced ethanol, acids (lactic and acetic), higher alcohols (2-methyl-1-butanol, 3-methyl-1-butanol, 1hexanol, 2-methyl-1-propanol, and 1-propanol), ethyl acetate, and acetaldehyde, in similar amounts of those obtained during milk fermentation (Magalhães et al., 2011a).These results and the beverages' acceptance in the sensory analysis (Magalhães et al., 2011b) demonstrate the potential of utilizing both cheese whey and deproteinised cheese to produce kefir-like beverages, which constitutes a new valorisation route.Besides its high lactose content, cheese whey and permeate present essential nutrients beneficial for microbial growth, and therefore can be used as a fermentation medium (Panesar et al., 2007).An overview of the processing of cheese whey, its applications, and valorisation paths is shown in  2).While this review focuses on yeast, microbial valorisation of cheese whey using diverse microorganisms such as bacteria and/or filamentous fungi have also been accomplished.Although these microorganisms are suitable to yield valuable chemicals, yeast-based fermentation processes have been widely explored for wheyto-ethanol valorisation.

Table 2.
Examples of value-added compounds obtained from cheese whey substrate by non-yeast fermentation.

Ethanol production by cheese whey fermentation using yeast as a metabolic cell factory
While ethanol is the most prevalent product from cheese whey, an alternative range of value-added compounds can also be produced from this substrate, and several studies have developed yeast-based processes in this sense (Table 3).The first reports for cheese whey valorisation for non-ethanol products using yeast (Kluyveromyces sp.) started more than three decades ago, aiming to produce citric acid (Abou-Zeid et al., 1983), glycerol (Jenq et al., 1989;Rapin et al., 1994), and the enzyme polygalacturonase (Donaghy and McKay, 1994;Gomez-Ruiz et al., 1988).Since then, the focus has mainly shifted to products with high industrial interest like lipids for biodiesel production (Castanha et al., 2014;Taskin et al., 2015;Vyas and Chhabra, 2019), organic acids like lactic acid (Plessas et al., 2008;Turner et al., 2017), or prebiotics, such as lactulose (de Freitas et al., 2020).
Most studies aim to utilize yeast naturally capable of fermenting lactose, therefore not having any genetic modifications.A wide range of microorganisms has already been used in this sense, like K. marxianus (Schultz et al., 2006;Bansal et al., 2008;Petrova and Kujumdzieva, 2010) or K. lactis (You et al., 2017), Yarrowia lipolytica (Taskin et al., 2015), Meyerozyma guilliermondii (Knob et al., 2020), among many others.Other approaches focus on the advantages of mixed cultures of K. marxianus and Candida krusei to produce single-cell proteins (Yadav et al., 2014) or the triple synergy between K. marxianus, Lactobacillus delbrueckii ssp.Bulgaricus and Lactobacillus helveticus to produce lactic acid (Plessas et al., 2008).
Alternatively, other studies have employed metabolic engineering strategies to attain the production of a plethora of compounds.These mainly focused on the introduction of heterologous pathways for the production of compounds like lactic (Turner et al., 2017)   Ethanol (C2H6O), naturally produced by the fermentation of sugars, is mainly used as raw material for pharmaceutical, cosmetical, and solvent industries.Under its potable form, it can also be used to produce alcoholic beverages (Ryan and Walsh, 2016).Bioethanol constitutes the alternative biofuel most used worldwide since it reduces the environmental impacts of fossil fuels, in particular greenhouse gas emissions (Balat, 2011;Cunha et al., 2018).The high lactose content of cheese whey suggests its utilization for bioconversion into ethanol, as shown in Equation 1.
Eq. 1 The theoretical yield is equal to 0.538 g ethanol/g lactose consumed.Thus, cheese whey represents another renewable feedstock for the production of biofuels.
Although cheese whey is not economically competitive with the substrates currently used for first-generation biofuels, this raw material presents numerous advantages.First of all, being a by-product that can be reused, it reduces competition for traditionally used food crops, such as corn-starch or sugarcane (Zhou et al., 2019).Secondly, its utilization decreases the pollution load associated with its disposal (Murari et al., 2019).Indeed, cheese whey is a low-cost and abundant substrate that does not require extensive pre-processing such as lignocellulosic materials (e.g., high temperature, acid treatments) (Ryan and Walsh, 2016).However, the direct fermentation of cheese whey for ethanol production is limited by the low lactose content of cheese whey, around 5%, which results in ethanol titers of 2 -3% (v/v), increasing distillation costs (Farkas et al., 2019).Thus, it is important to concentrate lactose in cheese whey, using ultrafiltration and/or reverse osmosis technologies (Guimarães et al., 2010).Another alternative proposed more recently is to increase the sugar concentration by mixing with other sugar-enriched materials, such as molasses (Álvarez-Cao et al., 2020).Conversely, cheese whey may also be used to increase the carbon content of lignocellulosic biomasses (Cunha et al., 2018;Cunha et al., 2021) and has been shown to positively contribute to lignocellulosic bioethanol production by nutritional supplementation (Kelbert et al., 2015).

Cheese whey-to-ethanol conversion processes by Kluyveromyces species
One of the most attractive aspects of Kluyveromyces sp. is their capacity to ferment lactose.This ability involves two proteins: lactose permease (encoded by LAC12 gene) and a -galactosidase (encoded by LAC4 gene).The lactose permease is a membrane protein with 587 amino acids responsible for lactose uptake into the cells (Boze et al., 1987).This  transporter is inducible by intracellular levels of lactose and galactose.It is also an active transporter system that requires energy and can transport lactose against the concentration gradient (Domingues et al., 2010).The galactosidase, also known as lactase, is a cytosolic protein with 1032 amino acids (Poch et al., 1992) that hydrolyses lactose into glucose and galactose.These two monosaccharides are easily metabolized through glycolysis and the Leloir pathway, respectively.In addition, Kluyveromyces sp. are generally regarded as safe (GRAS) thermotolerant yeast with less stringent catabolite repression (Guimarães et al., 2010).The catabolite repression is a mechanism of gene expression in response to carbohydrate availabilityin the presence of glucose, the genes for galactose metabolism (GAL genes) are not expressed; hence galactose is not used as a carbon source.Even though K. marxianus has more relieving catabolite repression, some studies addressed this effect to abolish it.The use of 2-deoxyglucose as a selective agent allowed the development of a K. marxianus strain (KD-15) which is less sensitive to glucose repression by adaptive evolution (Oda and Nakamura, 2009).Despite several research efforts devoted to the utilization of cheese whey as a source of lactose by Kluyveromyces sp., only a few works focus on ethanol production.The fermentation capacity of Kluyveromyces sp. is lower than S. cerevisiae, leading to lower ethanol yields and consequent higher distillation costs (Pescuma et al., 2015).Indeed, Kluyveromyces sp. is more sensitive to osmolytes, which constitutes a disadvantage.These strains are mostly used for heterologous protein production or other biotechnological-related applications, as shown in  (Farkas et al., 2019).This encompasses a great disadvantage since a reduced consumption of lactose inevitably leads to lower ethanol production.The osmotic sensitivity of Kluyveromyces sp. is not only noticeable at high substrate concentration but also at high product concentration.As ethanol amount increases, the growth rate and fermentative capacity decrease, leading to low ethanol productivity (Zafar and Owais, 2006).It was observed that the growth of K. marxianus CCT 7735 is severely reduced when ethanol concentration is higher than 4% (v/v) (Silveira et al., 2005;Costa et al., 2014).Also, an increase in the levels of valine and metabolites of the citric acid cycle was observed in K. marxianus cells under stress at high ethanol concentrations (Silveira et al., 2020).
Although most fermentations reported in the literature are performed in batch conditions, there are some interesting results with different fermentation systems, such as semi-continuous fermentation.The constant replacement of culture media by fresh media decreases the catabolic repression as well as the substrate inhibition effect, the central hindrances of ethanol production (Zohri et al., 2017).Another study reported a semi-continuous fermentation with K. marxianus CBS 712, resulting in higher ethanol titres (19.32 g/L) than batch fermentation (15.90 g/L).This improvement resulted from shortened lag phase and other non-ethanol productive phases in semi-continuous fermentation (Zoppellari and Bardi, 2013).
Nutrient availability can also play a key role in terms of fermentation performance.Koushki et al. (2012) observed an increase in ethanol production (from 5.1% to 5.8% v/v) and lactose utilization rate (almost 10%) in the presence of growth supplements (Table 4) (Koushki et al., 2012).Ergosterol and linoleic acid supplementation increased the fermentation capacity of Kluyveromyces sp., decreasing the fermentation time by almost 30 h (Janssens et al., 1983).This effect is in accordance with the role of sterols in ethanol tolerance (You et al., 2003;Aguilera et al., 2006).
Cell immobilization techniques have become an interesting strategy to increase fermentative capacity and surpass some batch-fermentation restrictions.The main advantage of using a matrix to immobilize cells is to overcome the product inhibition effect caused by higher ethanol concentrations.This approach allows a higher cell concentration per unit of reactor volume, reducing the fermentation time and increasing productivity.In addition, cells immobilized reactors showed long-term stability and an increased molecular selectivity ( marxianus DSMZ 7239 immobilized in sodium-alginate beads, reaching the highest reported productivity (4.5 g/L•h) (Table 4).Although most studies of immobilized cells are performed in alginate beads, other materials have also been exploited.Roohina et al. (2016) immobilized K. marxianus cells in carboxymethyl cellulose (CMC) polymer and synthesized a copolymer of CMC with N-vinyl-2-pyrrolidone (CMC-g-PVP).Ultimately, a continuous ethanol fermentation in a packed-bed immobilized cell reactor resulted in an ethanol production yield of 0.49 g/g, corresponding to 91.07% of theoretical yield (Table 4) (Roohina et al., 2016).
An advantage of K. marxianus in comparison with S. cerevisiae is its ability to ferment at high temperatures.Thermotolerance is an important factor for the biotechnology industry (Pinheiro et al., 2020).It reduces the costs associated with cooling and allows to obtain the ethanol directly from the broth by continuous evaporation at low pressure (Fonseca et al., 2008;Zoppellari and Bardi, 2013).While some K. marxianus can grow at temperatures as high as 52°C, efficient fermentation is reported to occur up to 45°C (Abdel-Banat et al., 2010).However, most of the K. marxianus studies refer to a range of 32 to 35 C as optimal temperature to maximize ethanol production (Dragone et  Although fermentation is an anaerobic process, some yeasts still ferment even in the presence of oxygen and high glucose concentrationa phenomenon known as the Crabtree effect, which occurs in S. cerevisiae.Other yeast strains, like Kluyveromyces sp., are Crabtree-negative, i.e., they favour respiration over fermentation when oxygen is available, even at low concentrations (Yadav et al., 2015).It is known that a small amount of oxygen is needed to maximize ethanol production, as oxygen provides the energy required to produce essential biomolecules, such as amino acids, fatty acids, and sterols, indispensable for fermentative metabolism.Thus, oxygen availability should be high enough to allow anabolism but low enough to avoid excessive cell growth and oxidative metabolism (Zara et al., 2009).Zoppellari and Bardi (2013) reported higher ethanol yield using K. marxianus CBS 712 in anaerobiosis than in aerobiosis conditions (Table 4).This effect was justified with the high nutritional value of cheese whey, which can supply the needs of the yeast (Zoppellari and Bardi, 2013).Several other studies addressed the influence of oxygen availability in the metabolism of either K. marxianus (Castrillo and 2009) tested the effect of different types of cheese whey (raw cheese whey, cheese whey permeate and ricotta cheese whey) as substrates in batch fermentation under anaerobic conditions.The highest ethanol yield (97% of the theoretical value) was reached by using ricotta cheese whey, in comparison with the 83% achieved from raw cheese whey (Table 4) (Sansonetti et al., 2009).The higher ethanol yields can be correlated with the lower amount of proteins in cheese whey from ricotta.A lower protein content drives the yeast metabolism more efficiently to fermentation instead of respiration.When the amount of protein is higher in the substrate, the yeast takes longer to start ethanol production (Sansonetti et al., 2009).Crude whey was also tested by Zafar and Owais (2006), but lower productivity of 0.1 g/L•h was reached (Table 4).The same effect was also observed by Zoppellari and Bardi (2013).
Using microbial consortia is becoming more popular due to the higher genetic variability that a group of different species represents.In fact, with the development of metagenomics and bioinformatics approaches, the utilization of a group of microorganisms isolated from a specific environment encompasses an attractive option (De Felice et al., 2012).A microbial consortium of yeast (mainly K. marxianus) and bacteria (mainly Streptococcus) that grew on cheese whey was isolated, and it was used in a batch fermentation with crude cheese whey, resulting in 7.60 % v/v ethanol, with a productivity of 1.25 g/L•h (Table 4) (De Felice et al., 2012).
Some efforts were also made to co-produce different value-added products beyond ethanol to increase process economics efficiency (Andrade et al., 2017;You et al., 2017;Sampaio et al., 2019).For example, the K. lactis strain B10, isolated from Canastra cheese, was reported to co-produce 19.81 g/L ethanol with roughly 39 volatiles aromatic compounds, emphasizing 3-methyl-1-butanol, octanoic acid and ethyl decanoate (Andrade et al., 2017).-galactosidase is one of the enzymes produced by Kluyveromyces sp. with high added value.Due to its capacity to hydrolyse lactose, this enzyme can be used to produce galacto-oligosaccharides and lactose-free products required for lactose-intolerance people (Gosling et al., 2010;Oliveira et al., 2011;Sampaio et al., 2019).However, the extraction and purification of galactosidase are expensive and considered a bottleneck for industry applications (Sun et al., 2016).In the past years, a new approach based on cells permeabilization by ethanol has been developed as an alternative to cell disruption and -galactosidase extraction (Faria et al., 2013) for possible process integration (Fig. 3).
Genetic engineering strategies have also been applied to Kluyveromyces sp., and multiplex tools like those involving the rescue of selective markers have been developed (Ribeiro et al., 2007).A study reported constructing a mutant strain lacking the KINDI1 gene that codifies for a dehydrogenase responsible for redox reactions in mitochondria (González-Siso et al., 2015).This deletion promotes a metabolic shift from a respiratory to fermentative metabolism, increasing the fermentation rate (González-Siso et al., 2009).The mutant strain KINDI1 produced 16.4 % more ethanol than the parental-wild type strain using cheese whey as a carbon source (González-Siso et al., 2015).
Evolutionary adaptation is an approach of yeast strain improvement that has been exploited during the past years.This strategy is highly used in industry to obtain more robust strains with the desired phenotype.Adaptive evolution follows the principle of variation and selection of the more advantageous organism to a specific environmental change.K. marxianus MTCC 1389 strain was subjected to high lactose concentrations (200 g/L) for 65 d (Saini et al., 2017a).This evolved strain showed an ethanol titre of 10% v/v, a 17.5% increase compared to the parental strain.Transcriptomic analyses showed that the adaptive evolution resulted in increased expression of GPD1, TPS1, and TPS2 genes.This suggests that the evolved strain accumulates more glycerol and trehalose than the parental strain to improve osmotolerance against lactose stress (Saini et al., 2017b).Despite the suitable characteristics of Kluyveromyces sp. for ethanol production mentioned above and its aptitude for lactose metabolization, the fermentative potential of this strain is limited when compared to S. cerevisiae, with the latter generally attaining better ethanol titres in the fermentation process (Díez-Antolínez et al., 2018).

Exploiting Saccharomyces cerevisiae for cheese whey-to-ethanol valorisation
S. cerevisiae is a GRAS microorganism widely used for bioethanol production (Cunha et al., 2020), being also applied in the production of valueadded chemicals (Baptista et al., 2021).The existing genetic toolbox allied to its high fermentative capacity facilitates the use of this yeast as a microbial cell factory.As mentioned earlier, S. cerevisiae cannot use lactose directly because it lacks the genes required for its transport (lactose permease) and hydrolysis (-galactosidase) (Álvarez-Cao et al., 2020).Still, it can transport galactose (by the permease encoded by the gene GAL2), which is directed to the Leloir pathway inside the cell (Nehlin et al., 1989).Nevertheless, the advantages for biotechnological applications of S. cerevisiae justify the attempts to circumvent this hindrance.
One of the most attractive characteristics of S. cerevisiae is its high osmotolerance, which enables the fermentation of high lactose concentrations (up to 200 g/L) (Guimarães et al., 2010).Besides this, its ethanol tolerance is also higher, and thus, higher ethanol titres can be reached (Silveira et al., 2020).In fact, some specific industrial strains have been employed in very high gravity ethanol fermentations (Gomes et al., 2021b), reaching ethanol titres higher than 19% v/v (Pereira et al., 2011a).Taking advantage of the innate capacity of S. cerevisiae to produce ethanol, several approaches have been exploited to obtain S. cerevisiae strains with the ability to consume lactose.To reach that goal, two main strategies have been addressed: (1) pre-hydrolysis of cheese whey and/or the utilization of S. cerevisiae in co-culture with lactose-positive microorganisms; and (2) the introduction of heterologous lactose genes for lactose utilization (Ryan and Walsh, 2016; Álvarez-Cao et al., 2020).
Regarding non-metabolic engineering strategies, the most used approach involves an initial step where lactose is pre-hydrolysed to glucose and galactose (Champagne and Goulet, 1988).The resulting sugar mixture is then used directly by S. cerevisiae for ethanol production.However, the presence of glucose hampers galactose metabolization leading to a diauxic growth, which decreases the ethanol yield.Another disadvantage is the high price of galactosidase enzyme, which increases the costs of the process (González-Siso, 1996; Domingues et al., 1999a).To overcome these drawbacks, several approaches have been designed.Bailey et al. (1982) used 2-deoxyglucose, a glucose molecule with an extra alcohol group, as a selection marker.This molecule cannot be hydrolysed, but it can still repress the expression of GAL genes (Sanz et al., 1994).Thus, the mutant cells that can surpass the catabolic repression and consume galactose will survive.The isolated mutant cells were then able to consume glucose and galactose simultaneously, almost with the same effectiveness.
An S. cerevisiae mutant able to produce 90 g/L ethanol in less than 37 h was successfully isolated (Table 5) (Bailey et al., 1982).This mutant was used in a continuous fermentation system, reaching ethanol productivity of 13.6 g/L•h (Table 5) (Terrell et al., 1984).The whey permeate was used as a substitute for water in wheat fermentation, and the ethanol yield was not affected, suggesting that cheese whey can be used as water replacement and as a zero-waste solution.In this study, a -galactosidase from Aspergillus oryzae was added to hydrolyse lactose (Parashar et al., 2016).Another study reported the utilization of barley-based mash to hydrolyse lactose as an alternative to commercial galactosidase to reduce the overall fermentation costs.The hydrolysis of lactose by mash led to the production of 3.23% v/v of ethanol (Lawton and Alcaine, 2019).The fermentation conditions for pre-hydrolysed cheese whey have also been optimized, using a hydrolysed cheese whey mixture in a bioreactor integrated with a direct contact membrane distillation (Tomaszewska and Białończyk, 2016).The continuous removal of the produced ethanol leads to a 1.9-fold increase in fermentation efficiency.Additionally, it was found that the salt present in the concentrated cheese whey might have decreased the process efficiency (Tomaszewska and Białończyk, 2016).
Cell immobilization techniques were also addressed for S. cerevisiae strains.Kokkiligadda et al. (2016) reported the hydrolysis of lactose with -galactosidase immobilized in chitosan alongside fermentation by S. cerevisiae immobilized in calcium alginate beads, with an ethanol production of 3.6 % v/v, corresponding to 53.4% of the maximum theoretical yield (Table 5).Glass raschig rings and alumina beds were reported as the best inorganic supports, that allowed a stable ethanol production for more than 1000 h operation, yielding 7.6% v/v of ethanol with a productivity of 1.09 g/L•h (Table 5) (Díez-Antolínez et al., 2018).On the other hand, gel supports can have stability problems, and organic supports usually require complex derivatization pre-treatments (Lee et al., 2012).Nevertheless, cell immobilization reactors have some disadvantages, such as the reduction of substrate accessibility and alterations in activity due to variations in biocatalyst conformations (Eş et al., 2015).
The utilization of mixed cultures to maximize the potential of different strains and take advantage of the possible synergistic effect has also been gaining interest.Guo et al. (2010) reported cheese whey powder fermentation by a mixed culture of K. marxianus TY-3 and S. cerevisiae AY-5, with the production of 4.6% v/v of ethanol, a higher value when compared to a monoculture of K. marxianus TY-3 (3.8% v/v) (Table 5).The utilization of immobilized cells in calcium-alginate beads for cheese whey fermentation was also tested.With a mixed culture and immobilized cells, better ethanol yields were achieved, with 5.3 % v/v of ethanol (79.95% of the maximum theoretical yield) (Guo et al., 2010).An important factor to point out regarding mixed cultures is the ratio between the microorganisms used.Farkas et al. ( 2019) tested several ratios of K. marxianus Y00963 and S. cerevisiae Levuline FB and an increase from 5.8% v/v to 7.6% v/v of ethanol was achieved with ratios of 1:1 and 3:1, respectively.Also, when coupled with a fed-batch culture fermentation, an ethanol titre of 10.34% v/v was attained (92% of the theoretical value), one of the highest concentrations described in the literature (Table 5) (Farkas et al., 2019).Interestingly, mixed cultures fermentation led to higher lactose consumption and, consequently, higher ethanol production when compared to monoculture fermentations, reducing substrate inhibition (Farkas et al., 2019).Different strategies could be coupled to increase ethanol productivity.Beniwal et al. (2018) constructed a controlled system for simultaneous lactose hydrolysis and ethanol fermentation.-galactosidase enzyme was immobilized in silicon dioxide nanoparticles, while a coculture of K. marxianus and S. cerevisiae (3:1 ratio) was immobilized in calcium alginate beads in the same fermentation vessel.While part of the lactose is being hydrolysed by -galactosidase, another part is already being consumed by K. marxianus.Besides this, galactose is metabolized faster by K. marxianus than by S. cerevisiae and, therefore, the expected diauxic growth is less prominent.As a result, the fermentation time decreased from 42 to 36 h and 8.1% v/v of ethanol was obtained with a productivity of 1.78 g/L•h (Table 5) (Beniwal et al., 2018).
As mentioned earlier, several factors could affect fermentation performance, and it can be hard to take them all into account (Diniz et al., 2014).In the past years, numerous mathematical software packages have been employed to correlate all factors and reach a general kinetic model equation to give better fermentation conditions (pH, temperature, substrate concentration) for a specific experiment.One example is the Response Surface Methodology (RSM) that combines mathematical and statistical functions to obtain empirical models for experiments optimization (Uncu and Cekmecelioglu, 2011;Diniz et al., 2014;Myers et al., 2016;Murari et al., 2019).According to the equation proposed by Murari et al. (2019) for K. marxianus URM 7404, the temperature is the factor that most influences productivity, followed by pH and lactose concentration.For an ethanol yield greater than 90%, the temperature must be between 32.5-35.0C, the pH between 4.8-5.3,and the lactose concentration in the 61-65 g/L range.
Despite all the strategies developed to use wild-type yeast strains for cheese whey fermentation, the construction of a lactose-consuming S. cerevisiae is a promising alternative for whey valorisation.In the past thirty years, two different paths were exploited: (1) the construction of recombinant S. cerevisiae strains able to produce and secret -galactosidase to the external medium; and (2) the construction of recombinant S. cerevisiae strains for lactose assimilation and further fermentation (Fig. 4).In the first approach, the production of an extracellular -galactosidase enables its recovery from the broth after fermentation, facilitating the downstream processing.Due to the high complexity and cost of this hydrolase production, coupling its production with ethanol fermentation constitutes an advantage (Becerra et al., 2001a).Several organisms can produce β-galactosidases, including bacteria, yeast, and fungi (Oliveira et al., 2011;Saqib et al., 2017).-galactosidase from the filamentous fungus Aspergillus niger is naturally secreted to the extracellular media and is particularly interesting for whey hydrolysis due to its high activity at acidic pH optimum and high temperatures (40-65C), minimizing the risk of contamination (Domingues et al., 2000a).Our group constructed a flocculant S. cerevisiae strain able to secret galactosidase of A. niger by the expression of lacA gene, using an episomal plasmid (Domingues et al., 2000b(Domingues et al., , 2002)).Due to the genetic instability of plasmid-based systems, multiple copies of lacA were integrated into the genome (Oliveira et al., 2007).Both systems were able to effectively produce the -galactosidase enzyme from lactose in batch and continuous fermentation while simultaneously co-producing ethanol (Table 5) (Domingues et al., 2002;Domingues et al., 2005;Oliveira et al., 2017).Also, a secretion pathway is always an attractive method for heterologous protein production (Becerra et al., 2001b) as it simplifies protein recovery and purification.
Another approach encompasses the expression of the lacZ gene coding for an Escherichia coli β-galactosidase, with a signal sequence to direct to the membrane.However, this is not straightforward, and several tags were tested without success (Venturini et al., 1997).The high molecular weight of -galactosidase and the oligomeric nature of this enzyme are hindrances to this goal (Becerra et al., 2001b).The best results were obtained with the membrane protein GgpI (glycosylphosphatidylinositol-containing protein) (Pignatelli et al., 1998).In another study, Porro et al. (1992) proposed the expression of -galactosidase from E. coli under the control of the upstream activating sequence of GAL genes.The GAL4 activator gene induced cell lysis and release of -galactosidase to the extracellular medium.The lysis phenomenon was correlated with high levels of Gal4p, which leads to a loss of plasma membrane integrity.Ethanol was then produced with low productivity, 0.1-0.2g/L•h, 73-84% of the maximum theoretical yield (Porro et al., 1992).The main disadvantage of using -galactosidase  secretion strategies is the catabolite repression effect that may lead to an undesirable diauxic growth (Guimarães et al., 2010).Interestingly, transformant cells did not show the expected diauxic growth.Gal4p is a regulatory protein related to glucose-repression genes of GAL genes, which could explain the phenomenon.High levels of Gal4p inhibit the repression effect and, therefore, cells consume galactose and glucose simultaneously (Porro et al., 1992).The control of -galactosidase release was also evaluated with the construction of thermosensitive autolytic mutants and cells permeabilization with chemicals agents, such as toluene or ethanol (Compagno et al., 1993;Becerra et al., 2004).The expression of the LAC4 gene from K. lactis, using a secretory plasmid, was also performed, and all the mutants were able to grow in lactose (Becerra, et al., 2001a).A hybrid protein between -galactosidase from K. lactis and A. niger was also constructed to increase the secretion capacity.In this study, the domains of -galactosidase that play an important role in the secretion process were determined.A higher yield in protein release and a better affinity for lactose was shown for the new hybrid protein (Rodríguez et al., 2006).
A different approach to combine the best characteristics of Kluyveromyces sp. and S. cerevisiae resorts to protoplast fusion, a widely used technique for genetic improvement in industrial yeasts (Guo et al., 2012).In this procedure, cell walls are removed with specific enzymes to obtain a hybrid organism that shares organelle, cytoplasm, and genetic material (Wolf et al., 1996).The possibility to create a hybrid yeast strain that includes a lactose-consuming Lac + yeast with high ethanol tolerance has been pointed out, and several insights are found in the literature.In a first attempt, a hybrid between S. cerevisiae and K. lactis was obtained.This strain was able to produce higher ethanol concentrations (5.5% v/v) compared to the parental strain (4.1% v/v) (Taya et al., 1984).Later, Farahnak et al. (1986) constructed a hybrid between S. cerevisiae and K. marxianus that was able to produce 30% more ethanol than the parental strain K. marxianus strain, with ethanol titre of 13% v/v and productivity of 1.3 g/L•h (Table 6) (Farahnak et al., 1986).Ryu et al. (1991) obtained fusants between these strains with a productivity of 2.42 g/L•h and explained the rationale behind ethanol tolerance (Table 6).Alcohols are soluble in lipids, and so they can penetrate the plasmatic membrane and alter membrane fluidity.As a consequence, the transport of nutrients across the membrane will be inhibited.S. cerevisiae, as well as the fusant obtained, has higher content in unsaturated fatty acids, particularly linoleic acid, which provide a higher ethanol tolerance, compared to others yeast species.Later, another hybrid strain derived from K. marxianus and S. cerevisiae was obtained, and a 3.8% v/v ethanol titre was reported by fermentation of cheese whey powder (Guo et al., 2012).
Another strategy consists of introducing heterologous genes for lactose assimilation to the cytoplasm and its hydrolysis inside the cells (Fig. 4b).The first attempt to express K. lactis LAC4 and LAC12 genes was reported by Sreekrishna and Dicksont (1985).The genome region of K. lactis, which includes both LAC4 and LAC12 genes, as well as an intergenic region codifying for the natural promoters of K. lactis, was expressed in S. cerevisiae.The resulting transformants grew very slow in lactose medium, with a doubling time of 6-7 h.Moreover, the authors also reported that the transformants integrated more than 15 tandem copies of the recombinant vector in a chromosome (Sreekrishna and Dickson, 1985).A second attempt was tried by Jeong et al. (1991), where a yeast integrative vector was used instead of a usual cloning vector.Nevertheless, the isolated transformants showed slow growth rates in the lactose medium.
Later, Rubio-Texeira et al. (1998) attempted to replace the native promoters of LAC4 and LAC12 genes with CYC-GAL promoters (a galactose-inducible hybrid promoter).Also, the recombinant DNA fragment was targeted for a ribosomal DNA region (RDN1 locus).After different selection methodologies, a fast-growing Lac + diploid S. cerevisiae strain was obtained.Still, in lactose medium, the obtained strain exhibited a respiro-fermentative metabolism, with high biomass yield and low ethanol titres, similar to K. marxianus (Table 6) (Rubio-Texeira et al., 1998).
Our research group constructed a flocculent S. cerevisiae Lac + strain (T1) based on the same plasmid (pKR1B-LAC4-1) used by Sreekrishna and Dickson (1985) (Domingues et al., 1999a).This strain was able to consume 10 g/L lactose, independently of the initial lactose concentration, and produced 2.0% v/v of ethanol (Table 6) (Domingues et al., 1999a).After a physiological adaptation process, the growth rate increased, and the lactose metabolization ability improvedthis strain consumed 50 g/L lactose in 40 h and produced 16 g/L of ethanol (Domingues et al., 1999b).In continuous operating airlift bioreactor experiments, it was possible to reach a productivity of 10-11 g/L•h, opening new perspectives in cheese whey valorisation (Table 6) (Domingues et al., 1999b and2001).However, the constructed strain lost its lactose metabolization capacity upon conservation.Thus, a long-term evolutionary engineering experiment was carried out, and a stable strain able to be frosted and defrosted without losing its lactose-metabolizing capacity, designated T1E, was isolated (Guimarães et al., 2008a).It was observed that the T1E strain had a deletion of 1593 bp in the intergenic region between LAC4 and LAC12 genes.Also, the evolved strain had a reduction of approximately 10-fold in the number of copies of the recombinant plasmid, about 10-fold, compared to the T1 strain (Guimarães et al., 2008b).A transcriptome analysis identified at least 173 genes with differential expression levels, mostly related to RNAmediated transposition, DNA repair, and recombination mechanisms (Guimarães et al., 2008b).These results suggest that the loss of the intergenic-region fragment leads to an alteration of the promoter structure that will, therefore, contribute to a better inducible lactose-promoter.This effect, accompanied by a lower plasmid copy number, leads to an improved overall phenotype in terms of lactose consumption (Guimarães et al., 2008b).
A multi-route non-structural kinetic model for the interpretation of ethanol fermentation of lactose using this recombinant strain was also developed (Juraščík et al., 2006).Batch fermentations with high-lactose concentrations were performed with T1E strain, attaining 8% v/v of ethanol from 150 g/L lactose in a synthetic medium with a productivity of 1.5-2.0g/L•h.When cheese whey was used as carbon source, 7% v/v ethanol titre was reached but the fermentation was slower (Table 6) (Guimarães et al., 2008b and c;Silva et al., 2010).Thus, the T1E strain was not just able to maintain the phenotype but also to efficiently ferment lactose from cheese whey with high ethanol titres (Guimarães et al., 2008c).Both T1 and T1E strains are flocculating yeast strains, a characteristic that takes several advantages for industrial applications (Domingues et al., 2000b).Zou et al. (2013) constructed two lactose-consuming S. cerevisiae strains based on the deletion of specific genes that could improve ethanol production.One of them, S. cerevisiae AY-51024A, has the LAC4 and LAC12 genes to enable lactose consumption introduced in the region of ATH1 and NTH1 genes, leading to the knockout of those genes.The accumulation of trehalose in the cells is associated with high-stress tolerance, and S. cerevisiae has two enzymes that hydrolyse this moleculea neutral cytosolic trehalase (codified by NTH1 gene) and an acidic vacuolar trehalase (encoded by ATH1 gene).Thus, the deletion of ATH1 and NTH1 will increase intracellular trehalose levels that could, in turn, increase ethanol titre.The other strain, S. cerevisiae AY-51024M, is based on the same lactose consumption genes mentioned above allied to the deletion of MIG1 and NTH1.MIG1 gene codifies for Mig1 protein that has an important role in the glucose repression effect.This protein is responsible for directing a complex protein to the consensus motif of the GAL4 promoter that inhibits GAL4 gene expression.GAL4 codifies for Gal4, a transcriptional activator protein of GAL genes.Consequently, GAL genes will not be expressed, and galactose will not be consumed.With Mig1 deletion, this repression effect is abolished, and glucose and galactose can be consumed simultaneously.An anaerobic shake flask fermentation was performed with both strains using cheese whey (with 150 g/L lactose) as a carbon source.AY-51024M strain produced 8.0% v/v of ethanol in 120 h, representing productivity of 0.53 g/L•h.On the other hand, AY-51024A produced 4.5% v/v of ethanol at the same fermentation time, but only 63.7% of initial lactose was consumed (Table 6) (Zou et al., 2013).Thus, it suggests that catabolic repression has a more powerful inhibitory effect than osmotolerance.In a subsequent study, the authors uncoupled glucose and galactose metabolism in S. cerevisiae by deletion of GAL80 and MGI1 again coupled to LAC4 and LAC12 expression.This resulted in the loss of diauxic growth and glucose repression, consequently leading to galactose consumption at higher rates, ultimately yielding higher ethanol titres.Likewise, lactose consumption and ethanol productivity rates also increased compared to AY-51024M when cultivated in cheese whey permeate powder solution (Zou et al., 2021).Liu et al. (2016) reported that an engineered S. cerevisiae strain to consume cellobiose is also able to ferment lactose.This means that the cellobiose transporter CDT-1 can transport lactose and -glucosidase can act as galactosidase and ferment lactose.The lactose fermentation ability was improved by artificial evolution, and an improved strain (EJ2e8), with a high copy number of CDT-1 and GH1-1 genes, was obtained.This strain could ferment lactose in a batch fermentation from cheese whey powder (150 g/L lactose).All lactose was consumed in 48 h, and 6.2% v/v of ethanol was produced (Table 6) (Liu et al., 2016).This study opened a new window of opportunities in terms of lactose-consuming S. cerevisiae strains.Genetic constructions beyond LAC4 and LAC12 genes may be explored for the attainment of Lac + S. cerevisiae strains.Also, the promoter influence should be evaluated, as it may play a central role in the activity of both the permease and β-galactosidase.While some studies successfully used the natural promoters of LAC4 and LAC12 (Sreekrishna and Dickson, 1985;Domingues et al., 1999b and2001;Guimarães et al., 2008c), the use of constitutive promoters of S. cerevisiae like ADH1 (Domingues et al., 2002 and2005) or PGK (Zou et al., 2013 and2021) has also been shown to be an effective strategy for lactose metabolization.Further investigation on the influence of different constitutive promoters and the optimum balance between the activity of the permease and β-galactosidase can also be valuable for improved lactose consumption.
Overall, the yeast S. cerevisiae shows itself as a suitable host for engineering lactose metabolizing traits.Some of the recombinant strains described so far are effective in converting lactose to ethanol, completely metabolizing the lactose available and reaching ethanol titres that can be economically competitive.

Incorporation of cheese-whey into lignocellulose biomass-to-ethanol processes
Lignocellulose-to-ethanol processes hold the promise of cleaner energy sources as lignocellulosic materials (LCMs) comprise complex sugars that can be converted by microorganisms to ethanol.LCMs are abundant, cheap, and do not compete with food crops.Still, the ethanol titres achieved in second-generation ethanol processes are typically below the 4% threshold commonly associated with economically viable processes (Gomes et al., 2021a).To increase ethanol titres and thus improve the process economic feasibility, increased overall solid loadings may be used with significant constraints on solid viscosity and mass transfer.An interesting alternative could be the use of low-cost sugar sources that could be economically viable on an industrial scale (Gomes et al., 2021a).Cheese whey and whey permeate are found among the most significant (and unavoidable) industrial waste streams and are constituted by a considerable amount of sugar (approx.50 g/L lactose).Besides, due to its recognised nutritional value, cheese whey has been applied as a supplement in yeast fermentation (Kelbert et al., 2015).Also, the improvement in yeast cell metabolism derived from nutritional supplementation in lignocellulose-to-ethanol processes has been demonstrated (Kelbert et al., 2015;van Dijk et al., 2020).Thus, the integration of cheese whey in multi-waste valorisation approaches, in particular, with LCMs as substrate, is a recently explored application where the integration of cheese whey increases the overall concentration of the carbon source, thus increasing ethanol titre with a consequent decrease in distillation costs, besides improving yeast cell metabolism and decreasing water usage.In fact, whey supplementation has been used just for water replacement in wheat straw fermentation (Parashar et al., 2016).Moreover, the mixture of cheese whey powder with Eucalyptus globulus wood was proposed to intensify simultaneous saccharification and fermentation at high solid loadings and high temperature, which boosted ethanol concentration by 1.5 folds compared to the results obtained without cheese whey powder (Cunha et al., 2018).When analysing the economic determinants on the implementation of a Eucalyptus wood biorefinery producing biofuels, energy, and high addedvalue compounds, the supplementation of low-cost cheese whey (Fig. 5) allowed a 52% increase of ethanol titres, leading to a clear improvement in the process net present value (NPV) from -14.4 to -3.4 million USD (Gomes et al., 2021a).
In addition, strategies used to metabolically engineer S. cerevisiae for lactose consumption can also be applied in this regard.In fact, a consolidated bioprocess was developed by cell surface display of cellulolytic enzymes, alongside β-galactosidase, in engineered thermotolerant industrial S. cerevisiae strains.This effort resulted in high ethanol titres (above 50 g/L), enabling simultaneous valorisation of corn cob and cheese whey and reducing costs (Cunha et al., 2021).For these multi-waste valorisation approaches comprising lignocellulosic residues, S. cerevisiae presents additional benefits against other yeast, such as the tolerance to the typical lignocellulosic inhibitors (Cunha et al., 2019), in particular, for some industrial isolated strains (Costa et al., 2017) besides the identification of specific tolerance molecular determinants (Pereira et al., 2011b).Still, Kluyveromyces sp. that naturally metabolizes lactose besides xylose, a relevant pentose in lignocellulose composition, are starting to be explored for lignocellulosic conversion processes and could be directly integrated into these multi-waste valorisation processes (Ferreira et al., 2015).

Conclusions and future perspectives
Cheese whey valorisation can substantially reduce the disposal problem, providing significant environmental benefits.According to the current environmental concerns, implementing circular bioeconomy systems to reduce by-products disposal is mandatory.Nevertheless, cheese whey has been recognized as a resource to be exploited rather than a waste requiring disposal.We can summarize cheese whey valorisation in two central branches.The first mainly involves physical processing to transform cheese whey into functional and nutritional fractions (WPC, WPI, WPH, and individual proteins).These fractions find several applications in the food industry (as nutritional supplementation, beverages, and as food ingredients) and also in the cosmetic and pharmaceutical industries.In the second valorisation route, lactose-rich cheese whey or cheese whey permeate can be used as a substrate in fermentation processes.The potential of natural lactose-consuming yeasts, especially Kluyveromyces sp., has been fully exploited, and many interesting value-added products have been produced, like citric acid, glycerol, enzymes like -galactosidase, organic acids, and ethanol, as the prevalent product.Nonetheless, despite the advances made in the last years, S. cerevisiae strains constitute the best choice for ethanol production.Its fermentative capacity allied to its high tolerance to osmotic stress and harsh industrial conditions has triggered the development of S. cerevisiae as a chassis microorganism for metabolic engineering aiming at the valorisation of cheese whey.Besides, due to its robustness and tolerance to fermentation inhibitors, a future central role is foreseen for S. cerevisiae in multi-waste valorisation approaches, where LCMs and whey residues may be combined to increase the overall concentration of the carbon source.This approach leads to higher ethanol titres, diminishing the distillation costs and thus, will contribute to fulfilling the worldwide needs for biofuel ethanol.Furthermore, the biorefinery concept, in which renewable carbon sources are converted into energy and multiple added-value products, is key to achieving a sustainable economy.An industrial whey to ethanol plant comprising the production of other value-added products, such as galactosidase that could be used in situ in the dairy industry, could represent an economic incentive to implement this biorefinery concept.Still, economic and sustainability analyses are needed for determining the best option.In any case, cheese whey valorisation has the potential to comprise multi-valorisation strategies that confer flexibility to the process, enabling it to be marketresponsive, a key aspect of the fast-changing times we live in.

Fig. 2 .
Fig. 2. Overview of cheese whey processing and its different applications.

Fig. 4 .
Fig. 4. Lactose consuming pathways by Saccharomyces cerevisiae based on genetic engineering techniques.a) Construction of a lactose-consuming strain (Lac + ), where the genes responsible for lactose metabolization, LAC4 and LAC12, were introduced in S. cerevisiae.Thus, lactose is transported into the cells, where it is metabolized into the monomers, glucose and galactose.b) A heterologous -galactosidase produced by S. cerevisiae is secreted to the external medium, where it hydrolyses lactose into glucose and galactose.The simple sugars are then transported into the cells and metabolized.Although glucose is directed to glycolysis, galactose is forwarded to the Leloir pathway.
(Cunha et al., 2018 and as suGomes et al., 2021a)se biomasses have a great sugar or starch content, being optimal to produce biofuels.However, the use of first-generation biomass is associated with socioeconomic and environmental issues such as price increases due to demand growth and soil deprivation due to over-fertilization (del Río et al., 2020).Lignocellulosic and algae biomass fed into second-and third-generation biorefineries, respectively, overcome the main disadvantage of first-generation feedstock, presenting a stable price with no need for extra land.Still, the lower carbon content compared to first-generation feedstocks becomes a bottleneck for biofuels production, particularly for bioethanol.Cheese whey, generated in high amounts and carbon content, has been recently applied in multi-waste valorisation approaches, boosting ethanol titres obtained in lignocellulosic biorefineries(Cunha et al., 2018 and 2021;Gomes et al., 2021a)

Table 1 .
Comparison of the coverage of the present review paper with those of the previously published reviews on the valorisation of cheese whey in the last five years (2017-2021).

Table 3 .
Main value-added compounds obtained from cheese whey substrate by yeast fermentation.
*Coupled with ethanol production.

Table 4 .
Fermentative parameters for ethanol production from cheese whey by Kluyveromyces species.

Table 5 .
The yeast Saccharomyces cerevisiae in the bioethanol valorisation process of cheese whey.Fermentative parameters for ethanol production from lactose, a mixture of glucose and galactose, and cheese whey substrates by native S. cerevisiae in pure or mixed cultures.

Table 6 .
The yeast Saccharomyces cerevisiae in the bioethanol valorisation process of cheese whey.Fermentative parameters for ethanol production from lactose and cheese whey substrates by engineered S. cerevisiae in pure or mixed cultures (n.r.not reported).