Xylitol production by Cyberlindnera (Williopsis) saturnus, a tropical mangrove yeast from xylose and corn cob hydrolysate




Potential for xylitol production from xylose and corn cob hydrolysate by a tropical mangrove yeast.

Methods and Results

In the present study, 21 fungi were isolated from detritus-based mangrove wetlands along the Indian west coast. Of these, one yeast isolate had the ability to grow and assimilate xylose producing significant amounts of xylitol (38·63 g l−1). A maximum yield of 0·54 g g−1 was obtained after 144 h of growth on xylose (150 g l−1) and corn cob hydrolysate (CCH, containing 65 g l−1 xylose). Using biochemical and molecular methods, the yeast was identified as Cyberlindnera (Williopsis) saturnus. Preliminary characterization of enzymes in the cell-free extract revealed that while xylose reductase (XR) preferred NADPH to NADH as cofactor, xylitol dehydrogenase (XDH) was NAD specific.


Significant amounts of xylitol could be produced on CCH using C. saturnus isolated from tropical mangrove wetlands. The yeast has the potential to assimilate rather than ferment xylose as its XR has a preference for NADPH.

Significance and Impact of the Study

Microbes offer an economically viable and green approach for production of xylitol, an industrially important compound. A mangrove ecosystem with its battery of lignocellulolytic enzymes is an ideal location for isolating fungi capable of producing xylitol from agroindustrial waste such as CCH.


Mangrove forests are saline, detritus-based ecosystems found in tropical and subtropical regions of the world. The Indian Ocean region covers about 84 985 km2 (47% of total area of world mangroves) with India having 4871 km2 (approx. 6%; Kathiresan and Rajendran 2005). The rich and unique biodiversity in this ecosystem is attributed to constant fluxes in the tidal gradient, temperature, salinity, pH, low water potential, high rainfall, humidity and muddy sediments (Raghukumar 2008). Nutrient fluxes in these environments depend almost exclusively upon plant assimilation and microbial mineralization with nitrogen and phosphorus playing major roles in the growth of the ecosystem (Alongi 1996).

Mangrove fungi constitute the second largest ecological group of marine fungi with a number of fungi being documented in the Indian Ocean region (Sridhar 2004). Detritus from mangroves contributes an enormous amount of organic matter to the coastal waters and involvement of fungi in its degradation has been well documented as they play a fundamental role in the energy flow of this ecosystem (Newell et al. 1987; Wafar et al. 1997). These fungi possess enzymes that display unique characteristics with respect to temperature, pH and salinity of the medium (Raghukumar 2008). Aspergillus (Raghukumar et al. 2004) and Flavodon (D'Souza et al. 2006) genera from mangroves have been known to produce cellulase, xylanase and lignin degrading enzymes having high titres and specific activities. The high salt concentration and low water potential of the mangrove ecosystem favour the growth of microbes that can maintain a lower water potential than the surrounding saline waters. Mangrove fungi maintain this gradient by intercellular accumulation of polyols such as glycerol, mannitol, sorbitol and xylitol (Davis et al. 2000). These polyols especially xylitol find wide applications in a variety of industries such as food and pharmaceuticals.

Due to its anticariogenicity, tooth rehardening and remineralization properties, xylitol has been widely applied in odontological formulations. Being a low-calorie sugar substitute with high sweetening power and a metabolism independent of insulin, xylitol has been used in dietary foods especially for consumption by diabetics as it has a low glycemic index (GI; Prakasham et al. 2009). It is currently produced by chemical reduction of xylose obtained from wood hydrolysates under alkaline conditions. The drawbacks of this commercial chemical preparation are a low initial availability of sugar, harsh noneco-friendly purification and separation steps with about 50–60% conversion of xylose to xylitol (Dhiman et al. 2008). Microbial fermentation offers a greener and economically feasible alternative, and this high value-added product can be produced from acid hydrolysates of agroresidues containing hemicellulose composed of xylans (Chandel et al. 2011). Yeasts are known to be the best producers of xylitol and literature reports from Candida tropicalis HXP2, Candida guilliermondii FTI-20037, Debaryomyces hansenii, Pachysolen tannophilus and Pichia caribbica exist. Of all the yeast species, the Candida genera has been extensively studied, and the most commonly used xylitol-producing species such as C. tropicalis and Candida parapsilosis are known opportunistic pathogens, thereby limiting their use in the food and pharmaceutical industry (Granstrom et al. 2007) while nonpathogenic species such as C. guilliermondii and C.mogii have been used to produce xylitol from grass hydrolysates (West 2009). Hence, current research has been directed not only towards isolation of nonpathogenic strains capable of producing high yields of xylitol using various cheap agroindustrial residues but also towards optimization of their process parameters (Prakasham et al. 2009).

In the present work, fungi isolated from the mangrove forests of the Indian west coast have been screened for their ability to produce xylitol. We report here a nonpathogenic yeast, Cyberlindnera (Williopsis) saturnus, identified by molecular methods as a good producer of xylitol from this ecosystem. Further studies of the selected yeast for xylitol production on different concentrations of xylose and also on corn cob hydrolysate (CCH) were also carried out. The ability of the yeast to assimilate or ferment xylose was also determined by studying the cofactor specificity of its xylose reductase (XR) and xylitol dehydrogenase (XDH).

Materials and methods

Chemicals used

The chemicals for media were purchased from Merck India Ltd. while those used for enzyme assays were obtained from Himedia Chemicals, Mumbai, India. All other chemicals used were of AR grade and at least 98% pure according to the manufacturer.

Strains and growth conditions

Twenty-one fungal cultures were isolated from mangrove wetlands and maintained on Czapek's Dox agar (CDA) as described by Khot et al. (2012). An initial preinoculum of cells was grown in MXYP (g l−1), yeast extract, 3·0; malt extract, 3·0; peptone, 5·0; xylose, 20·0; and NaCl, 15·0 for 48 h before being inoculated into the fermentation medium. The fermentation media (g l−1) contained yeast extract, 4·6; (NH4)2SO4, 5·0; KH2PO4, 1·3; MgSO4.7H2O, 0·6; and NaCl, 15·0 (Ling et al. 2011) with varying xylose concentration (5–150 g l−1). CCH containing xylose 65 g l−1, glucose 13 g l−1 and arabinose 6·33 g l−1 was prepared and detoxified as described by Li et al. (2011). Media components were then added to the hydrolysate and filter-sterilized into 500-ml conical flask prior to inoculation. Fermentation runs were inoculated with cells or spores at a final concentration of 1 × 107 cells ml−1 (de Figueroa and Lucca 2001). Incubation conditions for both inoculum development and fermentation were 28°C at 180 rev min−1 on rotary shaker.

Wet mounts of the washed cells were prepared, and light microscopic observation using a Zeiss microscope (Axioskope 40) equipped with a photographic attachment was carried out and images acquired. All photographs with similar magnifications were chosen and 25 yeast cells from each test sample were measured for their cell diameter using AxioVision Rel 4.8 (Carl Zeiss Imaging Solutions, GmbH, Munich, Germany).

Characterization of selected isolate F12

The selected isolate F12 was characterized by conventional methods according to Kurtzmann and Fell (1999). Cellular and colony morphology, fermentation and assimilation tests, growth temperature, growth in vitamin-free medium, urease activity were assayed to identify the yeast to the genus level. Genomic DNA was isolated using Bangalore Genei Uniflex DNA isolation kit, and amplification of the fragment comprising the internal transcribed spacer 1 (ITS1)-5·8S-ITS2 rDNA was done using universal primers ITS1 (19-mer 5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (20-mer 5′-TCCTCCGCTTATTGATATGC-3′), respectively. The small subunit rDNA (SSU) and the D1/D2 domains of the large subunit (LSU) region were also amplified to confirm the identity of the yeast using the primers NS1 (19-mer 5′-GTAGTCATATGCTTGTCTC-3′) and NS4 (20-mer 5′-CTTCCGTCAATTCCTTTAAG-3′) for the SSU region and NL1 (24-mer 5′-GCATATCAATAAGCGGAGGAAAAG-3′) and NL4 (19-mer 5′- GGTCCGTGTTTCAAGACGG -3′) for the D1/D2 domains. PCR was carried out in a final reaction volume of 25 μl consisting of distilled water (15·3 μl), Bangalore Genei PCR buffer (2·5 μl, 1 × final concentration), dNTPs (2·5 μl, 0·2 mmol l−1), forward primer (1·25 μl, 1 pmol l−1), reverse primer (1·25 μl, 1 pmol l−1), Taq polymerase (0·2 μl, 1 U per reaction) and genomic DNA (2 μl). The PCR product was purified by the polyethylene glycol (PEG)-NaCl method and subjected to sequencing PCR in a 5 μl final volume consisting of Bigdye (2 μl), primers (forward and reverse, 1 μl) and PCR product (2 μl). The PCR product was then purified using sodium acetate method and sequenced using ABI prism 3730 DNA analyzer (Applied Biosystems, Forster City, CA, USA). The obtained sequence was then used to perform BLASTN searches of the NCBI database following which the sequences were downloaded and aligned using Clustal-W algorithm, and phylogenetic trees were generated by neighbour-joining method using MEGA software (ver. 5.0) (http://www.megasoftware.net/).

Fermentation media for xylitol production

For the screening of the 21 isolates, the developed inoculum was added to 50 ml of fermentation medium in 250-ml conical flasks at final cell or spore concentration mentioned earlier and a xylose concentration of 20 g l−1. The reaction was terminated at 72 h and estimation of polyol content in the cell-free supernatant was carried out using the method described by Sanchez (1998). For the selected isolate, batch experiments were performed in duplicates in 500-ml Erlenmeyer flasks containing 100 ml of medium with varying xylose concentrations or CCH. Aliquots (5 ml) were withdrawn every 24 h, cells were centrifuged at 10 000 g for 10 min at 4°C and the cell-free broth was filter-sterilized and stored at −20°C. This broth was used for determination of monosaccharide composition and pH while the pellet was used to determine cell dry weight (CDW, g l−1) by lyophilizing the cell mass until constant weight was reached.

Analytical methods

The broth from each time point was filter-sterilized through a 0·22-μm filter prior to injection and analysed using a SUPLECOGEL Pb column (300 × 7·8 mm) on a Shimadzu LC-Prominence system. The mobile phase was 0·22-μm filtered and degassed MilliQ water. The column was maintained at 80°C at a flow rate of 0·8 ml min−1. A RID-10A detector was used to detect the eluted compounds. Standard sugar solutions were applied to the column at concentration levels ranging from 0·1 to 1 mg ml−1 to construct a standard curve. The retention times (Rt) for glucose, xylose, arabinose and xylitol were 10·8, 11·4, 12·2 and 28·4 min, respectively. Product confirmation was also carried out by injecting 2 μl of the broth of selected isolate into a Q Exactive Mass Spectrometer (Thermo Scientific, Rockford, IL, USA) with heated electron spray ionization in a negative mode (100–800 a.m.u.).

Enzyme assay

The total intracellular protein was extracted from the lyophilized yeast biomass with the yeast protein extraction reagent (Y-PER) Thermo Scientific as per the instructions in the kit. Phenylmethylsulfonyl fluoride (PMSF) was added to the reagent at a final concentration of 1 mmol l−1 to prevent protease activity. Unless mentioned otherwise, all the steps were carried out at 4°C. Enzyme activity of aldose reductase was determined spectrophotometrically at room temperature by following the oxidation of either NADPH or NADH at 340 nm. The reaction mixture (1·0 ml) contained 50 mmol l−1 potassium phosphate buffer, pH 7·2, 0·17 mmol l−1 NAD(P)H and 100 μl of the enzyme. The reaction was started with the addition of either 170 mmol l−1 glucose or xylose (for XR) as substrates. For XDH, also determined spectrophotometrically at 340 nm, the assay contained 50 mmol l−1 potassium phosphate buffer, pH 7·2, 1·5 mmol l−1 of NAD(P),100 μl of enzyme and was started by addition of 160 mmol l−1 xylitol (Kumar and Gummadi 2011). Enzyme units were expressed in terms of μmoles of nicotinamide nucleotide oxidized or reduced min−1. Protein concentration was estimated by the Folin phenol reagent (Lowry et al. 1951), and specific activities were expressed as Units mg−1 protein.

Statistical analysis

All the experiments, enzymatic activities and other analytical measurements were performed in triplicates. Statistical analysis was performed using SPSS 16.0 software (SPSS Inc., Chicago, IL, USA). Means were compared and analysed using one-way analysis of variance (anova) with Tukey HSD post hoc multiple comparison test. Differences were considered statistically significant for P < 0·05.


Polyol production in screened cultures

Twenty-one fungal isolates with varied visible colony and cell morphology as verified by microscopy were obtained from the tropical mangrove wetlands. Of these, four were yeasts, while 17 were filamentous fungi. The cultures were maintained on CDA and stored at 4°C.

All strains showed significant growth on both solid and liquid medium with xylose as the sole carbon source and hence were also tested for their ability to produce polyol or sugar alcohol. As shown in Fig. 1, culture F12 exhibited maximal extracellular polyol production of 26 mmol l−1 when grown on 2% (132 mmol l−1) xylose in 72 h. The supernatant was also subjected to LC-MS and exhibited two peaks with m/z of 153·0757 and 175·0577 that were ascertained to be xylitol bound with a proton and sodium, respectively (Fig. 1 inset). Hence, because F12 exhibited maximal xylitol production, this culture was selected for further study.

Figure 1.

Polyol production of the 21 isolates obtained from mangrove wetlands of the Indian west coast. Inset: mass spectrum of the supernatant of the selected isolate Cyberlindnera (Williopsis) saturnus that indicates that the polyol is xylitol (m/z: 153·075 and 175·057).

Identification of the selected isolate

The sequences obtained by amplification and sequencing of the ITS, SSU and D1/D2 regions of the yeast were compared with sequences from the NCBI database using the BLASTN search tool. As can be seen from the dendograms in Fig. 2, the three phylogenetic marker gene sequences form a contiguous clade with previously sequenced genes from Cyberlindnera (Williopsis) saturnus. The sequences for the isolate have been deposited in NCBI database with the GenBank accession numbers JX.242996, JX.242997, JX.242998 for the ITS, SSU and D1/D2 domains of the LSU region, respectively. Thus, isolate F12 was identified as Williopsis saturnus that exhibited maximal growth at 34°C.

Figure 2.

Dendograms of the small subunit (a), the D1/D2 domains of the large subunit (b) and ITS (c) sequences of the isolate identified as Cyberlindnera (Williopsis) saturnus based on CLUSTAL-W alignment and the phylogenetic trees generated by neighbour-joining method using MEGA (5.0). Type strains are indicated (T) and bootstrap values are displayed on the nodes of each tree.

Xylitol production on different xylose concentrations

Growth on different initial xylose concentrations (5–150 g l−1) was measured as CDW at specified time intervals up to 144 h. The organism grew exponentially resulting in the formation of biomass or CDW. The specific growth rate (μ) in the exponential phase for each initial substrate concentration (S) was obtained by using CDW at 72 h, and the kinetic parameters were estimated by nonlinear regression analyses using Andrew–Haldane kinetics. Figure 3a shows the fit for growth rate as a function of initial substrate concentration with a R2 value of 0·95 suggesting a good fit. The yeast showed a slower growth rate at higher xylose concentrations with a μmax of 0·073 h−1, while the concentration at which half-maximal growth (Ks) was 1·98 g l−1 and Ki was 214·3 g l−1. Cell size for xylose grown cells as determined by the cell diameter was found to be 8·21 ± 1·8 μm (Fig. 3a inset).

Figure 3.

Growth of isolate Cyberlindnera (Williopsis) saturnus on xylose. (a) Specific growth rate of the selected isolate Csaturnus plotted as a function of the initial substrate concentration. Inset: cells of the isolate Csaturnus visualized by light microscopy (× 1000 magnification). The scale bar corresponds to 10 microns. (b) Xylitol production and (c) Xylose utilization of Csaturnus at different initial xylose concentrations of 5 (■), 10 (●), 25 (▲), 50 (▼), 75 (◄), 100 (►) and 150 g l−1 (♦).

The effects of initial xylose concentration on the ability of Csaturnus to produce xylitol were also investigated. Batch experiments were performed for xylose concentrations varying from 5–150 g l−1 as shown in Fig. 3b,c. Xylitol produced by the isolate accumulated in the medium only when the substrate concentration exceeded 50 g l−1 as at substrate concentrations below this, the amount of residual xylose and xylitol was negligible. At xylose concentrations of 75 g l−1 and above, an increase in the production of xylitol as a function of time (Fig. 3b) was noted. The xylitol was produced in concentrations ranging from 17·54 to 38·63 g l−1 for xylose concentrations 50–150 g l−1 with a rate that doubled from 0·18 g l−1 h−1 at 25 g l−1 to 0·36 g l−1 h−1 at 150 g l−1. Substrate utilization rates also doubled with increasing substrate concentration from 0·21 g l−1 h−1 at 5 g l−1 to 0·42 g l−1 h−1 at 50 g l−1, after which it remained relatively constant till 150 g l−1. Above 50 g l−1, the residual xylose concentrations were found to be 15·98, 38·14 and 81·7 g l−1 for initial xylose concentrations of 75, 100 and 150 g l−1, respectively (Fig. 3c). Table 1 summarizes the xylose consumed (Sc, g l−1), xylitol (P, g l−1) and biomass (X, g l−1) produced, xylitol yield (YP/S, g g−1), biomass yield (YP/X, g g−1), volumetric productivity (QP, g l−1 h−1) and specific productivity (qP, g g−1 h−1). The xylose-to-xylitol conversion efficiency increased with an increase in initial xylose concentration (5–150 g l−1) from 0·45 to 0·54 g xylitol per g xylose consumed. The volumetric productivity doubled from 0·13 g l−1 h−1 at 25 g l−1 substrate to 0·27 g l−1 h−1 at 150 g l−1 substrate. The xylitol produced was maximum (38·63 g l−1) at 144 h on 150 g l−1 with YP/S of 0·54 g g−1, YP/X of 7·45 g g−1, QP of 0·27 g l−1 h−1 and qP of 0·052 g g−1 h−1 . The above results indicate that the yeast Csaturnus exhibits the potential for xylitol production from medium- to high substrate concentrations and therefore might be a good candidate for xylitol production.

Table 1. Comparative study of yields, volumetric productivity, specific productivity, xylose consumed and xylitol produced at different initial xylose concentrations and CCH
Initial xyloseTimeBiomassXylose consumedXylitol producedXylitol yieldBiomass yieldVolumetric productivitySpecific productivity
g l−1HX, g l−1Sc, g l−1P, g l−1YP/S, g g−1YP/X, g g−1QP, g l−1 h−1qP, g g−1 h−1
  1. CCH, corn cob hydrolysate.

  2. The values are means of 3 independent determinations.

  3. Separate analysis was done for each column.

  4. *,†,‡,§,¶Mean values within a column with different superscript letters differ significantly and were determined using one-way anova followed by Tukey as mentioned in Methods (P < 0·05, n = 3).

25485·23 ± 0·30*,14·11 ± 1·03*6·47 ± 0·01*0·45 ± 0·03*1·23 ± 0·07*0·13 ± 0·001*0·025 ± 0·002
50726·09 ± 0·16*,29·96 ± 1·4014·89 ± 0·940·49 ± 0·01*,2·45 ± 0·220·21 ± 0·0130·034 ± 0·003
751448·57 ± 0·1860·61 ± 0·8631·56 ± 0·820·52 ± 0·01*,3·68 ± 0·020·22 ± 0·005,0·026 ± 0·001
1001206·32 ± 0·2056·70 ± 0·49,30·07 ± 1·020·53 ± 0·02*,4·76 ± 0·01§0·25 ± 0·008,§0·040 ± 0·001
1501445·19 ± 0·13*71·52 ± 0·28§38·63 ± 0·84§0·54 ± 0·017·45 ± 0·340·27 ± 0·005§0·052 ± 0·002§
CCH14414·16 ± 0·48§54·23 ± 2·0829·09 ± 2·260·54 ± 0·02*,2·05 ± 0·090·20 ± 0·0150·014 ± 0·001*

Xylitol production on corn cob hydrolysate (CCH)

Xylitol production on CCH (Fig. 4) proceeded at a slower rate as compared with the pure xylose media (Fig. 3). Although the hydrolysate contained (g l−1) xylose, 65; glucose 13; and arabinose 6·33, xylose was not completely consumed at the end of 144 h with a residual xylose concentration of 10·47 g l−1. Glucose was preferentially consumed within the first 24 h of the time course followed by xylose utilization that began at 24 h. Xylitol production began at 48 h and increased till 144 h. Arabinose concentration did not vary through the entire time course and remained relatively constant at around 6·5 g l−1 thereby indicating that this sugar is not assimilated. Of the 55·1 g l−1 xylose consumed, 29·1 g l−1 xylitol was produced with a yield of 0·54 g g−1 of xylose consumed while the rest was used for biomass generation that was similar to the xylitol yield when Csaturnus was grown on 150 g l−1 xylose as a carbon source. CDW (14·16 g l−1) at the end of the time course was higher than that obtained on any of the pure xylose fermentations indicating that the glucose and other hydrolysis products may have been consumed during the initial part of the fermentation contributing to the increase in biomass. The volumetric productivity was the same as that for pure xylose medium (50 g l−1) at 0·20 g l−1 h−1 while the qP (0·014 g g−1 h−1) was lower on CCH as compared with the fermentations on pure xylose (Table 1). Thus, it can be seen that the yeast was also able to utilize the agroindustrial waste, CCH, for xylitol production.

Figure 4.

Xylitol production on corn cob hydrolysate media showing utilization of glucose (■) and xylose (●), nonutilization of arabinose (▲), production of xylitol (○) and CDW (◊).

Enzyme activities

The specific activities of the key enzymes involved in xylitol production viz., xylose reductase (XR) and XDH were determined in the crude cell-free extracts of CCH and plotted in Fig. 5. It was observed that the specific activity of XR started at 0·05 U mg−1 protein at 24 h and reduced to 0·025 U mg−1 protein at 48 h before peaking at 0·16 U mg−1 protein at 120 h. The XDH activity was low in the initial time points and was found to increase at 48 h and exceeded the XR level at 72 h. However, at the time points of 96 h and beyond the maximum, specific activities observed did not exceed 0·087 U mg−1 protein.

Figure 5.

Specific activities of xylose reductase (image_n/jam12327-gra-0001.png), xylitol dehydrogenase (image_n/jam12327-gra-0002.png) and protein (image_n/jam12327-gra-0003.png) concentration of the intracellular cell extracts of Cyberlindnera (Williopsis) saturnus grown on corn cob hydrolysate media.

The cofactor specificity of the enzymes was also determined, where it was found that the XR of Csaturnus preferred NADPH while XDH preferred NAD. Negligible activity of XR and XDH could be detected with NADH and NADP as cofactors. This indicated that the yeast was a potential producer of xylitol as the NADPH-linked XR activity is a marker for yeasts with an ability to assimilate xylose rather than ferment it.


Yeasts, especially the Candida genera, have been reported throughout literature as the best producers of xylitol; but the more commonly used xylitol-producing species such as Candida tropicalis are often pathogenic, limiting their use in food and pharmaceutical production (Granstrom et al. 2007). It is preferable to use a Generally Regarded as Safe (GRAS) yeast for xylitol production. The yeast Cyberlindnera (Williopsis) saturnus isolated from the mangrove wetlands of the west coast of India was found to produce high yields of xylitol when grown on xylose. Neither ethanol nor any other sugar alcohol such as arabitol was detected as seen by LC-MS when C. saturnus was grown on xylose. The genus Cyberlindnera (Williopsis) has already been used in the food industry with proven applications in the production of papaya wine and prevention of growth of spoilage yeasts in cheese (Liu and Tsao 2009; Lee et al. 2010).Thus, the use of this yeast in the production of a compound in the food and pharma industry could be regarded as safe for human consumption.

Molecular identification of the culture was carried out by amplifying the regions of rDNA used for barcoding yeast species (Begerow et al. 2010). The internal transcribed spacer (ITS) sequence, the 18S sequence and the 26S sequence have all been used to identify fungi and yeast on molecular basis. It was found that all 3 sequences of C. saturnus when queried in the NCBI database using BLASTN showed maximum homology (up to 99%) with Cyberlindnera saturnus. Neighbour-joining (NJ) phylogenetic trees constructed using similar sequences also showed that the yeast was closely related to C. saturnus. Confirmation of the molecular result was also carried out by morphological and biochemical methods. Morphological characterization revealed that the culture belonged to Williopsis sp. as it is one of the few genera of yeast that produces a pellicle. Biochemical characterization also showed that the culture exhibited a characteristic fermentation and assimilation pattern of Williopsis saturnus and was identified as W. saturnus var saturnus because of its ability to utilize rhamnose and grow on vitamin-free medium.

In this study, the kinetics of growth and xylitol production by C. saturnus were affected by the initial concentration of substrate, that is, xylose. Substrate utilization rates remained relatively constant above 50 g l−1 at 0·4 g l−1 h−1 indicating that the yeast was saturated with xylose. This was also confirmed by the model fit of the data of specific growth rate vs. xylose as explained by the Andrew–Haldane model, which shows growth saturation above 50 g l−1. The xylose utilization rate for C. saturnus at 25 g l−1 was 0·33 g l−1 h−1 is higher than 0·26 g l−1 h−1 that was observed for Pichia stipitis YS-30 at similar substrate concentration of 22 g l−1 (Rodrigues et al. 2011). The xylose consumption rate on the CCH medium was almost twice that observed on pure xylose medium at 0·63 g l−1 h−1, which is in agreement with the observation made by Walfridsson et al. (1995).

Upon increasing the substrate concentration from 25 to 150 g l−1, the xylitol yield (YP/S) increased from 0·45 to 0·54 g g−1 xylose, YP/X from 1·23 to 7·45 g g−1 and productivity (QP) from 0·13 to 0·27 g l−1 h−1, with maximal xylitol production occurring at 150 g l−1. Debaryomyces hansenii (Prakash et al. 2011) and Candida athensensis (Zhang et al. 2012) show a similar increase in xylitol yield from 0·53 to 0·76 g g−1 (20–100 g l−1, xylose) and 0·69 to 0·82 g g−1 (100–250 g l−1, xylose). In studies with Candida guilliermondii, Candida parapsilosis (Nolleau et al. 1993) and Candida athensensis (Zhang et al. 2012), it was found that high substrate concentrations of 200 g l−1 and 250 g l−1, respectively, severely affected cell growth. As the Andrews–Haldane model for growth kinetics predicted a Ki of 214 g l−1 for Csaturnus, batch fermentations were carried out at a maximum xylose concentration of 150 g l−1. The optimum growth of the culture (CDW 10·37 g l−1) was seen on xylose-based medium at 50 g l−1. The growth rate of the Csaturnus reduced upon increasing concentrations of xylose beyond 50 g l−1 as the biomass generated showed a continuous decrease from 10·37 to 5·19 g l−1 of CDW for xylose concentrations from 50 to 150 g l−1, respectively. A similar finding was observed with Cguilliermondii wherein increasing substrate concentration led to an increase in xylitol yield with simultaneous decrease in biomass generation (Nolleau et al. 1993). This trend indicates that the optimum initial substrate levels for xylitol production and biomass generation are different, which need to be taken into consideration for subsequent optimization strategies to be designed using Csaturnus.

Acid-hydrolysate-based medium from agroindustrial wastes has often been used for the production of platform chemicals and fuels due to its availability, as well as its economy especially in a tropical country like India (Prakasham et al. 2009). C. guilliermondii when grown on wheat straw and rice straw resulted in differential xylitol yields of 0·59 g g−1 (Canhila et al. 2008) and 0·72 g g−1 (Mussatto and Roberto 2004). C. tropicalis (Ling et al. 2011) and D. hansenii (Prakash et al. 2011) have also produced similar yields of 0·73 g g−1 and 0·69 g g−1 on corn cob and sugar cane bagasse hydrolysates, respectively. Ctropicalis ATCC 750, C. mogii ATCC 18364 and Cguilliermondii ATCC 20216 have also been shown to produce xylitol from big bluestem prairie grass hydrolysate in yields ranging from 0·38 to 0·46 g g−1 (West 2009). In comparison, C. saturnus showed a marked increase in xylitol yield (0·54 g g−1) when grown in CCH medium, same as the medium containing 150 g l−1 xylose while the QP of 0·20 g l−1 h−1 was the same as media containing 50 g l−1 xylose. Rodrigues et al. (2011) reported a similar yield of 0·61 g g−1 and QP 0·18 g l−1 h−1 on diethyl oxalate (DEO) hydrolysate of corn stover using Pichia stipitis YS-30. On semi-defined fermentation media simulating the sugar composition of hemicellulosic hydrolysates, C. guilliermondii exhibited volumetric and specific productivities of 0·85 g g−1 and 0·08 g g−1 h−1, respectively (Mussatto et al. 2006). Similar volumetric and specific productivities to that of C. saturnus have been obtained from hardwood hemicellulose hydrolysates by Pachysolen tannophilus, D. hansenii and C. guilliermondii (Converti et al. 1999). The CDW (14·16 g l−1) of C. saturnus at the end of the CCH medium fermentation was higher than that on pure xylose (10·37 g l−1 CDW on 50 g l−1 xylose) indicating that glucose also present in the hydrolysate contributed to the increased biomass without affecting the productivity. Using optimized media, higher yields and productivities have been obtained by C. tropicalis HDY-02 with a maximum conversion of 0·73 g g−1 (Ling et al. 2011). Thus, media components along with other parameters like initial cell concentration, pH, temperature and salinities need to be optimized to obtain higher xylitol yields using Csaturnus. According to the classification system devised by Rao et al. (2007), the current yield of C. saturnus on CCH-based medium falls into the class of high xylitol-producing yeasts.

When C. saturnus was grown on xylose or CCH media and studied for its cofactor requirement, it was found that the culture preferred NADPH for XR while XDH was NAD specific. No aldose reductase activity, either NADH or NADPH, could be detected in either media in shake flasks under the given assay conditions. In a similar study, the XR from D. hansenii UFV-170 was shown to be NADPH dependent with a specific activity of 0·24 U mg−1 protein on pure D-xylose while the activity was 5–7 times higher when grown on cotton husk hydrolysates (Sampaio et al. 2009).

Thus, C. saturnus was found to produce xylitol on pure xylose medium with a maximum yield of 0·54 g g−1 at 150 g l−1 while maximum biomass of 10·37 g l−1 being produced at 50 g l−1, demonstrating that optimum substrate concentration for xylitol production and biomass accumulation were different. The isolate was also able to use xylose available in CCH medium to produce xylitol with a yield of 0·54 g g−1 along with a high biomass generation 14·16 g l−1. The NADPH-dependent XR activity of the yeast indicates that it is more suitable for xylose assimilation and xylitol production rather than xylose fermentation. The potential of this isolate to grow on lignocellulosic material coupled with the ability to produce xylitol from hydrolysates makes it a good candidate for further optimization of biotechnological xylitol production on an industrial scale.


The authors would wish to thank the Department of Biotechnology and Institute of Bioinformatics and Biotechnology, University of Pune for the financial support extended to this project. SK would like to thank Department of Biotechnology, Government of India for funding, Mrs. Santhakumari and Dr. Mahesh Kulkarni (National Chemical Laboratory, Pune) for use of ESI-MS facility for product confirmation and Dr. Yogesh Souche (National Centre for Cell Sciences, Pune) for ITS sequencing.

Conflict of Interest

No conflict of interest declared.