Non-Conventional Yeasts as Sources of Ene-Reductases for the Bioreduction of Chalcones

: Thirteen Non-Conventional Yeasts (NCYs) have been investigated for their ability to reduce activated C = C bonds of chalcones to obtain the corresponding dihydrochalcones. A possible correlation between bioreducing capacity of the NCYs and the substrate structure was estimated. Generally, whole-cells of the NCYs were able to hydrogenate the C = C double bond occurring in ( E )-1,3-diphenylprop-2-en-1-one, while worthy bioconversion yields were obtained when the substrate exhibited the presence of a deactivating electron-withdrawing Cl substituent on the B-ring. On the contrary, no conversion was generally found, with a few exceptions, in the presence of an activating electron-donating substituent OH. The bioreduction aptitude of the NCYs was apparently correlated to the logP value: Compounds characterized by a higher logP exhibited a superior aptitude to be reduced by the NCYs than compounds with a lower logP value.

Recently, whole-cells and enzymes catalyzing the C=C hydrogenation and C=O reduction of representative chalcones have been studied for obtaining compounds possessing noteworthy bioactivities. In particular, some dihydrochalcones (achieved via bioreduction of the C=C double bond) have been found to express antioxidant, UV-protective and pro-health activities, which could be interesting for pharmaceutical and cosmetic industries [25,30]. Moreover, their sweet taste make them attractive for producing sweeteners [31,32]. In addition, the dihydrochalcone obtained from bioreduction of (E)-1,3-diphenylprop-2-en-1-one has been isolated from the leaves of Leptoderris fasciculata [33], a woody liana used in traditional medicine for the treatment of dropsy, edema, pulmonary disorders and as a laxative [34].
Recent studies revealed that Non-Conventional Yeasts (NCYs) are able to express a number of promising biotechnological properties [35][36][37], including the ability to express important ERs activities. In this framework, due to the presence of cofactor-recycling systems for NAD(P)H at the level of cell metabolism, biotransformation processes catalyzed by whole-cells of NCYs could be considered as useful and cheaper alternatives in place of using purified enzymes for reducing α,βunsaturated alkenes including chalcones [38][39][40][41][42].
Aiming to identify new possible substrates for the NCYs expressing ERs activity, the present paper reports a study on the ability of lyophilized cells of NCYs to bioreduce the activated C=C double bonds of chalcones. A Structure-Activity Relationship (SAR) approach was used.

Yeast Strain
Thirteen NCY strains belonging to ascomycetous and basidiomycetous species (genera Candida, Cyberlindnera, Goffeauzyma, Hanseniaspora, Kazachstania, Naganishia, Pichia, Scheffersomyces, Solicoccozyma and Wickerhamomyces) were used. They were preliminarily selected from a few hundred of environmental strains isolated worldwide for their ability to catalyze the biotransformation of α,β-unsaturated alkenes [39][40][41]. All strains are conserved at the Industrial Yeasts Collection DBVPG of the University of Perugia, Italy. Salient information on strains are reported in Table 1 and are available on the DBVPG website (www.dbvpg.unipg.it). NCY strains were maintained in frozen form (−80 • C), while working cultures were routinely grown on YEPG agar slants at 20 or 25 • C, depending on their psychrophilic or mesophilic aptitudes.

Bio-Reduction Reactions
A total of 30 mg of lyophilized NCYs cells were resuspended in 25 mL sterile vials containing 4.5 mL of 50 mM phosphate buffer (pH 6.5). A total of 0.5 mL of 10% w/v glucose, acting as a cofactor-recycling system, was also added. As a final point, chalcone was added at a final concentration of 5 mM and the vials were incubated on an orbital shaker (120 rpm) at 20 or 25 • C (depending on their psychrophilic or mesophilic status) for 120 h. In order to determine whether chalcone was spontaneously reduced in the absence of the NCY cells, blank (cell-free) vials containing 50 mM phosphate buffer + 50 mM glucose and each chalcone were analyzed at 120 h. After incubation, vials were sealed and frozen (−30 • C) until GC-MS analysis.
All the results were expressed as biotransformation yield, i.e., a % of the substrate converted to a given derivative. The concentration of the substrate and product were measured by an internal standard method. All the results represented the average of three independent experiments, and the statistical significance of these average data was assessed via ANOVA.

LogP Calculation
The logP values of chalcones were calculated by the ACD/LogP v.14.06 program in the software package for ACD/Labs 2016 2.2 (Advanced Chemistry Development). Figure 1 reports the chemical structures of the substrates used for checking the NCYs' ERs activity. The first substrate, namely (E)-1,3-diphenylprop-2-en-1-one (1a, Figure 1A), was used as model compound to screen the ability of the lyophilized cells of the NCYs to reduce the α,β C=C double bond. The presence of the conjugate C=O double bond was also considered for assessing the chemoselectivity of the reduction.
All the results were expressed as biotransformation yield, i.e., a % of the substrate converted to a given derivative. The concentration of the substrate and product were measured by an internal standard method. All the results represented the average of three independent experiments, and the statistical significance of these average data was assessed via ANOVA.

LogP Calculation
The logP values of chalcones were calculated by the ACD/LogP v.14.06 program in the software package for ACD/Labs 2016 2.2 (Advanced Chemistry Development). Figure 1 reports the chemical structures of the substrates used for checking the NCYs' ERs activity. The first substrate, namely (E)-1,3-diphenylprop-2-en-1-one (1a, Figure 1A), was used as model compound to screen the ability of the lyophilized cells of the NCYs to reduce the α,β C=C double bond. The presence of the conjugate C=O double bond was also considered for assessing the chemoselectivity of the reduction.
Considering these encouraging results, the ability of NCYs to bioreduce chalcones substituted with both deactivating and activating groups on the B-ring ((E)-1-(4-chlorophenyl)-3-phenylprop-2-en-1-one 2a and (E)-1-(4-hydroxyphenyl)-3-phenylprop-2-en-1-one 3a, respectively) was also checked. The results are reported in Table 2. Worthy bioconversion yields were obtained when the substrate exhibited the presence of a deactivating electron-withdrawing Cl substituent on the B-ring (2a): all the NCYs exhibited the ability to reduce C=C double bond of 2a with bioconversion yields ranging from 8% to 98% (Table 2). On the contrary, in the presence of the activating electron-donating substituent OH (3a), no conversion was generally found, with the sole exception of Goffeauzyma gilvescens DBVPG 4712 (yield = 47.8%), and Cyberlindera amylophila DBVPG 6346, Kaz. spencerorum DBVPG 6746 and K. lactis DBVPG 6854, which totally reduced the chalcone 3a (yield = 100%) ( Table 1). Interestingly, the last three NCYs also exhibited worthy bioconversion yields of 1a and 2a (falling into the range from 94.0% to 100% and from 48.4% to 88.3%, respectively). On the contrary, Pichia kluyveri DBVPG 5826 showed a bioconversion yield of the chlorocalcone 2a ≥ 95%, but no or very low activity versus 1a and 3a. Taking into account the above few exceptions, the bioreduction aptitude of NCYs was apparently correlated to the logP value, which is an indirect measure of the lipophilic degree of a given compound: The substrates 1a and 2a, which were characterized by a higher logP (4.01 and 4.78, respectively), exhibited a superior aptitude to be reduced by NCYs than the chalcone 3a (logP = 3.65). This trend could be justified by considering how the different molecules can go across the yeast cell membrane. Due to the lyophilized nature of the whole cells of the NCYs herein used, the passage of molecules, including 1a, 2a and 3a, across the cell membrane to reach the intracellular ERs should be much simpler. In fact, some studies reported that dehydration-rehydration cycles can determine a significant decrease of cell sizes together with a strong folding of membranes, thus leading to an increased permeability in lyophilized cells [49,50]. In this framework, the hypothesis postulated by some authors [51,52] that higher lipophilic molecules (characterized by a higher logP, i.e., (E)-1,3-diphenylprop-2-en-1-one 1a and (E)-1-(4-chlorophenyl)-3-phenylprop-2-en-1-one 2a) could more easily go across yeast plasma membranes by using the free diffusion mechanism (thus easily reaching cytoplasm ERs) differently from the lesser lipophilic ones (characterized by a lower logP, i.e., (E)-1-(4-hydroxyphenyl)-3-phenylprop-2-en-1-one 3a) could justify the superior aptitude of 1a and 2a to be bioreduced by ERs occurring at the cytoplasm level of the lyophilized cells of the NCYs. In addition, the presence of an OH substituent on the B-ring of 3a (Figure 1) could affect its capability to form hydrogen bonds with hydrophilic components occurring in the cell surface, thus reducing the rate of passage across the membrane.
The degree of lipophilicity of a given molecule is just one of the parameters determining its passage across the membrane and its interaction with enzymes. Thus, the α,β-unsaturated ketones (3E)-4-phenylbut-3-en-2-one (4a) and (3E)-4-(4-chlorophenyl)but-3-en-2-one (5a) (Figure 1B), exhibiting lower logP (2.17 and 2.70, respectively), but also lesser steric hindrance and substituents unable to form hydrogen bonds, were studied for their ability to be bioreduced by NCYs (Figure 3). The results are reported in Table 2. Overall, with the sole remarkable exception of K. lactis DBVPG 6854 on substrate 4a (bioreduction yield = 99.1%), the aptitude of both substrates 4a and 5a to be bioreduced is significantly lower than those of the substrates 1a-3a. These results seem to confirm the importance of the lipophilicity and that the steric hindrance is a less important factor in the determining the aptitude to be bioreduced.  In close analogy with previous literature [53], the results herein reported underline that the use of lyophilized yeast cells could be considered preferable for enhancing the permeability of a membrane to the substrate. Indeed, as supposed by some authors [39,54], the ERs that reduced the α,β-unsatured carbonyl compounds could be exclusively associated with yeast biomass, with no release of extracellular enzymes.

Conclusions
Our study has shown that lyophilized NCYs whole cells are useful biocatalysts to obtain dihydrochalcones from corresponding chalcones. In particular, we have identified nine strains (namely, Cyberlindnera amylophila DBVPG 6346, Goffeauzyma gastrica DBVPG 4709, Hanseniaspora guillermondii DBVPG 6790, Kazachstania exigua DBVPG 6469, Kazachstania spencerorum DBVPG 6746, Kluyveromyces lactis DBVPG 6854, Naganishia diffluens DBVPG 6237, Scheffersomyces shehatae DBVPG 6850 and Wickerhamomyces canadensis DBVPG 6211) that exhibited high bioconversion yields, with an excellent repeatability and a low standard deviation The bioreduction aptitude of the NCYs was affected by both the structure of chalcones and the logP value of tested substrates: NCYs were able to better reduce compounds characterized by a higher logP than ones with a lower logP value. Due to their sweet taste, dihydrochalcones are interesting derivatives for the food industry for the production of new sweeteners. Furthermore, some dihydrochalcones have been found to express interesting biological activities, which could be important for the pharmaceutical and cosmetic industries.   In close analogy with previous literature [53], the results herein reported underline that the use of lyophilized yeast cells could be considered preferable for enhancing the permeability of a membrane to the substrate. Indeed, as supposed by some authors [39,54], the ERs that reduced the α,β-unsatured carbonyl compounds could be exclusively associated with yeast biomass, with no release of extracellular enzymes.

Conclusions
Our study has shown that lyophilized NCYs whole cells are useful biocatalysts to obtain dihydrochalcones from corresponding chalcones. In particular, we have identified nine strains (namely, Cyberlindnera amylophila DBVPG 6346, Goffeauzyma gastrica DBVPG 4709, Hanseniaspora guillermondii DBVPG 6790, Kazachstania exigua DBVPG 6469, Kazachstania spencerorum DBVPG 6746, Kluyveromyces lactis DBVPG 6854, Naganishia diffluens DBVPG 6237, Scheffersomyces shehatae DBVPG 6850 and Wickerhamomyces canadensis DBVPG 6211) that exhibited high bioconversion yields, with an excellent repeatability and a low standard deviation The bioreduction aptitude of the NCYs was affected by both the structure of chalcones and the logP value of tested substrates: NCYs were able to better reduce compounds characterized by a higher logP than ones with a lower logP value. Due to their sweet taste, dihydrochalcones are interesting derivatives for the food industry for the production of new sweeteners. Furthermore, some dihydrochalcones have been found to express interesting biological activities, which could be important for the pharmaceutical and cosmetic industries.