Enzyme-treated chicory for cosmetics: application assessment and techno-economic analysis

Chicory material

Forced roots of Belgian endive (Cichorium intybus var. foliosum) (sample #1) were provided by Flanders Research Institute for Agriculture, Fisheries and Food (ILVO). Industrial chicory roots (Cichorium intybus var. sativum), before (sample #2) and after (sample #3) inulin extraction, were provided by CoSucra Groupe Warcoing (Warcoing, Belgium). Unprocessed chicory roots (sample #2) were washed and cut into cossettes before inulin extraction using a countercurrent hot water extractor. This produced raw juice containing inulin and exhausted chicory cossettes, which were pressed to yield bagasse containing ~ 25% dry matter (sample #3). All samples were dried in a hot air oven at 60 °C for 6–8 h until the moisture content fell below 10%. Dried samples were processed in a ZM 200 ultracentrifugal mill fitted with a 0.5-mm sieve mesh (Retsch, Haan, Germany). The composition of each sample is summarized in Additional file 1: Table S1.

Enzymatic treatment

Samples were suspended in 0.05 M acetate buffer (pH 5.0) to dry matter content of 5% (w/v) with a total volume of 20 mL, and were mixed at 50 °C for 24 h in an HT Minitron temperature cabinet (Infors, Bottmingen-Basel, Switzerland). We used six enzymes alone or in combinations: inulinase (Fructozyme L), pectinase (Pectinex Smash), cellulase (Cellic CTec2) and β-glucosidase (Novozym 188), all from Novozymes (Bagsværd, Denmark), as well as xylanase (Depol 40L) and ferulic acid esterase (Depol 740L) from Biocatalysts (Cardiff, UK). The enzyme dose was calculated based on the protein content. At the end of each treatment, the enzymes were inactivated by boiling the sample for 15 min, followed by centrifugation (2600 ×g, 15 min, 23 °C). The supernatants were stored at − 20 °C.

Quantification of reducing sugars

The sugars released by enzymatic hydrolysis were analyzed using a reducing sugar assay based on the dinitrosalicylic acid (DNS) colorimetric method in a 96-well format (Silveira et al. 2014).

Antimicrobial activity assay

Antimicrobial activity was measured using the liquid culture method in a total volume of 1 mL (Puupponen-Pimiä et al. 2016). We screened the following microbial strains: Staphylococcus aureus VTT E-70045 (ATCC 6538), Pseudomonas aeruginosa VTT E-84219 (ATCC 15692), Lactobacillus rhamnosus GG VTT E-96666 (ATCC 53103) and Streptococcus thermophilus VTT E-96665 (ATCC 19258). S. aureus and P. aeruginosa were cultivated aerobically in BD Difco nutrient broth (Thermo Fisher Scientific, Waltham, MA, USA) at 37 °C, shaking at 150 rpm. L. rhamnosus and S. thermophilus were cultivated in De Man Rogosa Sharpe (MRS) broth (Oxoid, Basingstoke, UK) at 37 °C without agitation. Microbial stock cultures were stored at − 80 °C. Before cultivation, they were grown on solid media for 1–2 days as described above for each strain. Single colonies were transferred to liquid media, incubated for 20–24 h, and used as the source of inoculum for antimicrobial activity tests. Microbial cultures incubated with enzyme(s) in 100 µL acetic acid buffer without antimicrobial samples were used as positive growth controls. Cultures were incubated with 50 µg/mL chloramphenicol in 100 µL acetic acid buffer as a negative control. The inhibitory effects of chicory samples suspended in 100 µL acetic acid buffer were evaluated by comparison with the control growth curves.

Sugar analysis

Sugars were extracted as previously described (Muir et al. 2009) with minor changes. Briefly, 100 mg of powder was mixed with 8 mL distilled water in a 10-mL volumetric flask by sonication for 15 min at 80 °C in a Branson 3510 device (Marshal Scientific, Hampton, NH, USA). After cooling to room temperature (25 °C), the volume was adjusted to 10 mL with distilled water and the sample was centrifuged (3000 ×g, 10 min, room temperature). The supernatant was passed through a 0.22-µm sterile Millex filter (Merck, Overijse, Belgium) and analyzed by size exclusion chromatography (SEC) using an Acquity H-Class ultra-high performance liquid chromatography (UPLC) system (Waters, Milford, MA, USA) with a refractive index detector and external column oven. We used two TSK gel G2500PWXL columns (Tosoh, Tokyo, Japan) at 80 °C and a guard column (Bio-Rad, Temse, Belgium) under isocratic conditions, with distilled water as the mobile phase at a flow rate of 0.5 mL/min. The injected volume was 10 µL, and each run lasted 40 min. Sugars were identified by retention time and quantified against calibration curves of sucrose, fructose, glucose, raffinose and stachyose standards (Merck).

Sesquiterpene lactones

Sesquiterpene lactones (STLs) were extracted as previously described (Kips, 2017) with minor changes. Briefly, we added 74 µL of the internal standard (10 µg/mL santonin) to 50 mg of powdered sample, followed by extraction with 1.405 mL deionized water containing 0.1% formic acid. The sample was incubated for 15 min at 30 °C, shaking at 1300 rpm on a thermomix comfort (Eppendorf, Rotselaar, Belgium). It was then centrifuged (20,817 × g, 15 min, room temperature), and the supernatant was passed through a 0.22-μm Millex filter. STLs were separated on the Waters Acquity UPLC system using a BEH C18 column (150 mm × 2.1 mm, 1.7 μm). The mobile phase was a mixture of deionized water plus 0.1% formic acid (solvent A) and acetonitrile plus 0.1% formic acid (solvent B) at a flow rate of 0.350 mL/min. The gradient started at 5% B for 5 min, followed by a linear increase from 5 to 53% B in 20 min, a hold for 1 min, with a final phase at 100% solvent B for 3 min. The column was then re-equilibrated to 5% B for 4 min before the next injection. The column temperature was 40 °C and the injection volume was 5 µL. STLs were detected by high-resolution mass spectrometry (HRMS) using a Waters Synapt G2-S quadrupole time-of-flight (QTOF) instrument in positive electrospray ionization (ESI +) MSE mode. Before sample analysis, the HRMS was calibrated (50–1200 Da) using sodium formate solution. The HRMS was operated in resolution mode (20,000 FWHM) and a leucine-enkephalin solution (200 pg/µL) was constantly infused during analysis as lockmass.

Four compounds were quantified against reference standards (Extrasynthese, Genay, France): lactucin, lactucopicrin, dihydrolactucin and dihydrolactucopicrin. For oxalates and glycosides, no standards are commercially available so we based our analysis on relative peak areas. Data were recorded using MassLynx v4.1 and integrated using TargetLynx v4.1 (Waters).

Phenolic compounds

Phenolic compounds were extracted as previously described (Kips 2017). Briefly, 75 mg of sample powder was mixed with the internal standard (1 μg/g daidzein) and extracted with 5 mL 100% methanol followed by 5 mL 20:80 (v/v) methanol:water. Each extraction step consisted of 1 min vortexing and 15 min ultrasound-assisted extraction using a Transsonic Digital S device (Elma Schmidbauer, Singen, Germany), followed by centrifugation (3000 ×g, 15 min, room temperature). The supernatants were combined and passed through a 0.22-μm Millex filter before separation on the Waters Acquity UPLC system fitted with a BEH C18 column (150 mm × 2.1 mm, 1.7 μm). The mobile phase was a mixture of water plus 0.1% formic acid (solvent A) and acetonitrile plus 0.1% formic acid (solvent B) at a flow rate of 196 µL/min. The gradient increased in a linear fashion from 1 to 24% B in 9.9 min, then to 65% B at 18.5 min and to 99% B at 18.7 min, followed by a hold at 99% B until 20.7 min. The column was then restored to 99% A from 20.8 to 23 min before the next injection. The column temperature was kept at 40 °C and the injection volume was 5 µL. Fractions were analyzed using a Xevo TQ-S tandem mass spectrometer (Waters) in electrospray ionization negative (ESI-) mode with multiple reaction monitoring (MRM). Phenolic compounds were quantified based on relative peak areas and against the reference standards chlorogenic acid, chicoric acid, quinic acid, caffeic acid and 4-OH-phenylacetic acid (Merck). Data were recorded and integrated as described above.

Phenolic compounds were also evaluated by non-targeted Fourier transform mass spectrometry (FTMS) as previously described (Cankar et al. 2021). Briefly, 300-µL samples were extracted with 700 µL methanol containing 0.1% formic acid by vortexing and sonication as above, followed by centrifugation (21,000 ×g, 15 min, room temperature). The supernatants were then fractionated on an LC-PDA-LTQ-Orbitrap-FTMS system (Thermo Fisher Scientific) comprising an Acquity H-Class UPLC fitted with an Acquity elambda photodiode array (PDA) detector (220–600 nm) connected to an LTQ/Orbitrap XL hybrid mass spectrometer with an ESI interface. Raw LC-Orbitrap-FTMS data were processed using MetAlign (Lommen 2009) to correct for background and noise and to align the chromatographic peaks from the different samples. Values lower than the detection threshold (s/n ≥ 3) were removed using MetAlign Output Transformer (Houshyani et al. 2012), and the remaining mass peaks were clustered into centrotypes using MSClust (Tikunov et al. 2012) based on original retention times and peak intensities across all samples. For sample #3A (chicory biomass treated with pectinase plus xylanase), the most differential mass peaks compared to the other treatments were selected using Pearson’s correlation with a cut-off of 0.98. To exclude the least abundant metabolites, mass peaks with a tenfold lower mass peak area than average area were omitted.

For gas chromatography mass spectrometry (GC–MS), 0.5-mL samples were extracted with 1 mL dichloromethane by vortexing and sonication as above, followed by centrifugation (240 ×g, 10 min, room temperature). The supernatant was dehydrated in sodium sulfate and analyzed on GC–MS 7890A device coupled to a 5975A mass-selective detector (Agilent, Santa Clara, CA, USA). Analytes from 1-μL samples were separated on a Phenomenex (Torrance, CA, USA) 30 m × 0.25 mm ZB-5 column (0.25 mm film thickness) using helium as the carrier gas at a flow rate of 1 mL/min. The injector was used in splitless mode with the inlet temperature set to 250 °C. The initial oven temperature of 45 °C was increased after 1 min to 300 °C at a rate of 10 °C/min and held for 5 min at 300 °C.

Cosmetic formula challenge test

Extracts of chicory by-product samples #3A (chicory biomass treated with pectinase plus xylanase) and #3D (untreated biomass) were added to a cosmetic cream formula (Additional file 1: Table S2) and tested for antimicrobial activity against S. aureus (VTT E-70045, ATCC 6538) and P. aeruginosa (VTT E-96728, ATCC 9027) at cell densities of 105–106 cfu/g formula according to ISO 11930 (Preservative Effectiveness Test), with some modifications. Chicory extracts were added at a concentration of 4% (v/v). We transferred 10 g of the sterile formula to a 50-mL Falcon tube and added the chicory extracts, which were filter sterilized and diluted with an equal volume of sterile glycerol before mixing. As a positive control, we used the commercial preservative Euxyl PE 9010 (containing phenoxyethanol and ethylhexylglycerol) at a concentration of 0.5% in the formula. As a negative control, we used the cream formula without preservative or chicory extract.

Microbial cultures were incubated at 37 °C for 18–24 h on tryptone soy agar (TSA) and were subcultured once before the experiment. A loop of bacterial biomass was collected from the plate and transferred to 5 mL of peptone-saline solution in a 20-mL tube with glass beads, and cell aggregates were disrupted by vortexing. The density of the suspension was adjusted with peptone-saline to Mac Farland 0.5 using densitometer (corresponding to 107–108 cfu/mL). Microbial inoculum (100 µL) was then added to the formula by vortexing and incubating at 25 °C. Samples (0.5-g) were collected from each culture (including controls) weekly for 4 weeks, resuspended in 4.5 mL peptone-saline and serially diluted before plating on TSA. The plates were incubated at 37 °C and colonies were counted after 24, 48 and 72 h.

Techno-economic analysis

A conceptual level techno-economic analysis was carried out using sample #3 (Additional file 1: Table S1). In the baseline case (scenario 1), we processed 1500 tons of dry pulp with a dry matter (DM) content of 90%. The process is shown in Fig. 1. We also considered three alternative scenarios (Additional file 1: Table S3) in which the extract concentration step was omitted (scenario 2), fresh pulp was used instead of dry pulp (scenario 3), and the treatment was integrated with an inulin plant (scenario 4).

Fig. 1figure 1

Baseline process (scenario 1) for the production of antimicrobial extracts from chicory biomass following inulin extraction

The process operating and performance parameters used in the mass and energy calculation are listed in Additional file 1: Table S4. The power demand of the enzyme solution pump was considered insignificant, whereas the pumps linked to the evaporator and rotary dryer (where included in the process) were incorporated in the power consumption estimate. A conceptual level estimate for fixed capital investment (FCI) costs was calculated using the chemical engineers’ factorial method. A Lang factor of 2.63 was applied based on purchased equipment costs (Towler and Sinnoit 2008). Total capital investment (TCI) was estimated by adding working capital to FCI. The TCI was fixed at 15% of the FCI.

Purchase cost estimates for major equipment were based on the Aspen Process Economic Analyzer, in-house knowledge and other sources (Towler and Sinnoit 2008; Coleman 2015; Koehorst 2020) and were scaled from the reference equipment cost using the exponential scaling equation (Eq. 1) if the reference equipment size was not equal to the designed process unit size (Peters et al. 2003). Here, cost scaling factors (Si) were derived from the mass balances. The reference costs (C0), reference scaling factors (S0) and cost regression indexes (exponent k) were then used to estimate the purchase cost (Ci).

A regression index of 0.6 was used when no better knowledge was available (Towler and Sinnoit 2008). Equipment costs were inflated to the value in 2019 euros using the Chemical Engineering magazine’s Plant Cost Index (Jenkins 2020). If necessary, we used the average currency exchange rate of 2019, where €1 = US$ 1.12 (Statista 2020).

Variable operating costs and revenues were calculated from the mass and energy balances and unit costs of raw material, chemicals, and utilities, which were based on publicly available cost data, in-house knowledge or data from project partners and collaborating companies. The price of by-product was set 60% lower than the price of the raw material because the quality of the solid residue after enzymatic treatment is yet unclear.

Fixed operational costs included labor and maintenance. Labor costs were estimated by assuming two operators per shift over five shifts, and included overheads. For scenario 1 (11 months operation), we assumed a total of 10 person-years. For scenarios 3 and 4, we assumed a labor requirement of 6 months per year, equating to 6 person-years. Annual maintenance and materials costs were assumed to be 2% of FCI, excluding chemicals used for cleaning-in-place (Peters et al. 2003).

To compare several concepts and scenarios, and to identify the most significant cost factors, the production cost was calculated (€/kg extract DM) using Eq. 2.

$$Production \, cost =\frac$$

(2)

Taxes, interest, depreciation and amortization were excluded. The economic and financial parameters and process operation strategy are summarized in Additional file 1: Table S5.

留言 (0)

沒有登入
gif