Lipid Production by Cryptococcus curvatus on

Appl Biochem Biotechnol DOI 10.1007/s12010-014-1007-y

Lipid Production by Cryptococcus curvatus on Hydrolysates Derived from Corn Fiber and Sweet Sorghum Bagasse Following Dilute Acid Pretreatment Yanna Liang & Kimberly Jarosz & Ashley T. Wardlow & Ji Zhang & Yi Cui

Received: 14 March 2014 / Accepted: 29 May 2014 # Springer Science+Business Media New York 2014

Abstract Corn fiber and sweet sorghum bagasse (SSB) are both pre-processed lignocellulosic materials that can be used to produce liquid biofuels. Pretreatment using dilute sulfuric acid at a severity factor of 1.06 and 1.02 released 83.2 and 86.5 % of theoretically available sugars out of corn fiber and SSB, respectively. The resulting hydrolysates derived from pretreatment of SSB at SF of 1.02 supported growth of Cryptococcus curvatus well. In 6 days, the dry cell density reached 10.8 g/l with a lipid content of 40 % (w/w). Hydrolysates from corn fiber, however, did not lead to any significant cell growth even with addition of nutrients. In addition to consuming glucose, xylose, and arabinose, C. curvatus also utilized formic acid, acetic acid, 4-hydroxymethylfurfural, and levulinic acid for growth. Thus, C. curvatus appeared to be an excellent yeast strain for producing lipids from hydrolysates developed from lignocellulosic feedstocks. Keyword Corn fiber . Sweet sorghum bagasse . Cryptococcus curvatus . Dilute acid pretreatment . Enzymatic hydrolysis . Fermentation

Introduction During recent years, microbial lipids or single cell oils (SCOs) have attracted extensive attention. Lipids produced by oleaginous microorganisms can be treated as vegetable oils and to be further processed to generate liquid transportation biofuels serving cars, trucks, and airplanes through hydrotreatment, decarboxylation, and other upgrading approaches [19]. To produce microbial lipids, oleaginous microbes, for example, yeasts, microalgae, fungi, and Y. Liang (*) : K. Jarosz : J. Zhang : Y. Cui Department of Civil & Environmental Engineering, Southern Illinois University Carbondale, 1230 Lincoln Dr., Carbondale, IL 62901, USA e-mail: [email protected] A. T. Wardlow Florida International University, 11200 SW 8th St., OE 304, Miami, FL 33199, USA

Appl Biochem Biotechnol

bacteria must be provided with certain kinds of carbons, either inorganic (CO2) or organic (e.g., sugars). Compared with autotrophic growth mode where CO2 serves as the only carbon source, heterotrophic growth of microbial species generally leads to higher yield of desirable products [10]. However, for this kind of cultivation, an organic carbon source must be supplied. As the most abundant biomass on earth, lignocellulosic materials can certainly be used as carbon sources to feed oleaginous microorganisms for the purpose of producing microbial lipids. This promises a renewable, domestic, and sustainable way to produce biofuels to meet the enormous fuel demand from basically every sector of our society. Currently, however, the biomass-to-fuel process is often cost inhibitory. From the perspective of biochemical conversion of lignocellulose to biofuels, this cost barrier is a result of the following: (1) expensive biomass pretreatment to disrupt the lignocellulosic structures; (2) costly enzymes for hydrolyzing cellulose and hemicellulose remaining in the pretreated materials; and (3) pricey fermentation operation of oleaginous microorganisms for lipid production [1]. To tackle these barriers, we seek to produce microbial lipids through investigating simple, low-cost, and efficient pretreatment techniques for corn fiber and sweet sorghum bagasse (SSB). The resulting hydrolysate will be used for cultivating a robust yeast strain that has high lipid productivity. Throughout literature, countless pretreatment approaches have been reported for different biomass feedstocks. Among those, dilute sulfuric acid has been studied extensively due to the fact that it can depolymerize carbohydrates effectively. Generally, the harsher the pretreatment conditions (higher acid concentration, higher temperature, and longer treatment time), the more degradation of cellulose and hemicellulose will occur. However, the harsher pretreatment tends to degrade sugars and lignin, too, resulting in high concentrations of potential inhibitors (acetic acid, furfural, and 5-hydroxymethylfurfural) in the hydrolysates. These chemicals have been demonstrated to negatively impact the performance of microorganisms for biofuel production, either ethanol or microbial lipids [8, 9, 15, 16]. Thus, for this project, we aim to compare three pretreatment conditions in terms of yield of total reducing sugars (TRS) and yield of cells and lipids from Cryptococcus curvatus grown on hydrolysates obtained from the pretreatment step. For this study, we chose corn fiber and SSB since they are renewable, abundantly available, and pre-processed already. Corn fiber is a by-product of the corn wet milling industry. It is a mixture of corn hulls and residual starch not extracted during the milling process. For every bushel of corn, about 4.5 lb of corn fiber is released [18]. Since oil and protein inside of the corn kernel have been extracted out, corn fiber is generally treated as a cheap animal feed ingredient. Typically, corn fiber contains approximately 70 % carbohydrates: 20 % starch, 14 % cellulose, and 35 % hemicellulose. SSB is the leftover material of sweet sorghum after juice is squeezed out of the stalk. As sweet sorghum being designated as a bioenergy crop, SSB containing typically 37 % cellulose and 18 % of hemicelluloses deserves to be studied on how microbial lipids can be produced from [12] . To produce lipids from released sugars from corn fiber or SSB, one oleaginous yeast strain, C. curvatus (ATCC 20509) will be used. As shown by our previous studies, this yeast strain grows fast on a broad range of substrates and accumulates intracellular lipids in the range of 40–70 % [4]. Besides high biomass and lipid productivities, C. curvatus can grow on xylose, a major five-carbon sugar of hemicellulose that is generally not consumed by ethanol-producing yeast strains [7]. In addition, this yeast strain has been reported to be able to tolerate degradation inhibitors, such as furfural and 5-hydroxymethyl furfural at certain concentrations [21]. Among several oleaginous yeast species, Rhodotorula glutinis, Rhodotorula toruloides, Lipomyces starkeyi, and Yarrowia lipolytica, C. curvatus shows the greatest promise in converting cellulosic sugars to microbial lipids [5].


To achieve this purpose, this study sought to (1) compare TRS yield from corn fiber and SSB pretreated under three conditions having different severities, (2) compare yield of yeast biomass on different hydrolysates obtained from the pretreatment step, and (3) understand how the yeast strain utilizes major chemicals in the hydrolysates, in particular, fermentable sugars, organic acids, and other compounds.

Materials and Methods Source of Sweet Sorghum Bagasse Fresh sweet sorghum stalks were collected from Sorghum Ridge Farms, Cobden, IL, USA. The remaining bagasse after juice expression was washed extensively with water, dried at 50 °C in a hot air dryer, ground by a cutting mill (Thomas Wiley Model 4, Arthur H. Tjomas Co., Phil, PA, USA), and passed through a 40 mesh screen. The resulting bagasse had particle sizes less than 420 nm, a moisture content of approximately 10 % and was stored in zip log bags at room temperature. Following a composition analysis protocol recommended by NREL, this bagasse was found to have 36.9 % of cellulose, 17.8 % of hemicellulose and 19.5 % of lignin. All of these percentages were based on dry weight [3]. Source of Corn Fiber Corn fiber samples were obtained from Cereal Process Technology which has a full-size drygrind fractionation system at a facility in Jefferson (Wisconsin, USA) and stored in a −20 °C freezer. Before it was used in this study, the corn fiber was ground in the same cutting mill as identified above to pass a 40 mesh screen. Similar to sorghum bagasse, moisture content of corn fiber was also around 10 %. Following the same procedure identified above, this corn fiber (dry weight basis) had a carbohydrate content of 66.5 %, which included cellulose (17.2 %), xylan (27.7 %), starch (8.8 %), and arabinan (12.8 %). Pretreatment Three different pretreatment conditions were employed for both corn fiber and SSB. For each condition, three replicates were set up. Pretreatment condition 1 used 0.5 % (v/v) sulfuric acid. To 100 ml of this solution in a 250-ml Erlenmeyer flask, 11 g of either corn fiber or SSB were added. These amounts of feedstock ensured that the dry weight was 10 g for each. Pretreatment lasted for 1 h at 121 °C. The second set of pretreatment was conducted by using 2 % (v/v) sulfuric acid solution at the same temperature with the same duration as those for pretreatment condition 1. The third pretreatment was performed using 2 % sulfuric acid (v/v) for 1 h at 134 °C. A total of 18 samples with either corn fiber or SSB were processed in this study. After pretreatment, the cooled samples were centrifuged at 4,000×g for 10 min. The liquid part was tested for pH and for concentration of TRS by using 3,5-dinitrosalicylic acid (DNS) reagent according to a protocol developed in our lab [13]. The same liquid samples were also used for yeast fermentation following pH adjustment to 5.0 and autoclave at 121 °C for 25 min. The solid fraction was washed with distilled and deionized water (DDW) three times with a total volume of 140 ml. Concentration of TRS in the wash water was also measured by using the same procedure. The washed solid was then air-dried in a fume hood. After 2 days, moisture content of each solid sample was measured by drying the subsamples in a conventional oven at 103 °C overnight.


Enzymatic Hydrolysis Washed solid samples of SSB were subject to enzymatic hydrolysis. In short, based upon moisture contents of different samples measured above, different amount of solids were transferred to citrate buffer (50 mM, pH 5.0) to satisfy a 2.5 % (w/v) solid content. After pH adjustment to 5.0, cellulase (C2730, Sigma-Aldrich, St. Louis, MO, USA) and cellobiase (C6105, Sigma-Aldrich) in the dose of 15 and 60 U/g solid, respectively, were added to the slurry samples. Enzymatic hydrolysis was carried out at 50 °C in a shaking incubator at 150 rpm. Yeast Fermentation An inoculum of C. curvatus (ATCC 20509) was set up by adding frozen stock of this yeast to a medium containing yeast extract (10 g/l), peptone (20 g/l), and glucose (10 g/l). After 3 days, this yeast culture was used to inoculate hydrolysate of corn fiber or SSB. These hydrolysates were those obtained after acid pretreatment, centrifugation, pH adjustment, and autoclave. Briefly, for SSB hydrolysate samples, an inoculum size of 10 % (v/v) of the final volume (40 ml) was directly added without any other nutrient supplementation. Regarding hydrolysate derived from corn fiber, the same inoculum size was adopted. In addition, a 10× minimal medium (4 ml) comprising (per liter): 27 g KH2PO4, 9.5 g Na2HPO4, 2 g MgSO4·7H2O, 1 g yeast extract, 1 g EDTA, and 25 g NH4Cl was supplemented to the hydrolysate together with 0.4 ml spore stock solution (100×, per liter): 0.4 g CaCl2·2H2O, 0.055 g FeSO4·7H2O, 0.052 g citric acid, 0.01 g ZnSO4·7H2O, 0.0076 g MnSO4·H2O, and 10 μl of 18 M H2SO4. All cultures were maintained at room temperature (26 °C) on a rotary shaker set at 120 rpm. On a daily basis, 1.5 ml samples were withdrawn from each culture. These samples were (1) observed under microscope to check for potential microbial contamination, (2) used to measure optical density at 600 nm for cell growth, and (3) centrifuged at 12,000×g for 5 min to separate liquid from cell pellet. The liquid fractions were subject to HPLC analysis as detailed below. The cell pellet was freeze-dried and weighed to attain cell biomass dry weight. The final day samples were freeze-dried and used for composition analysis as detailed below. Analysis Concentrations of glucose, xylose, arabinose, formic acid, acetic acid, levulinic acid, furfural, and 5-hydroxymethylfurfural (HMF) in samples were determined by HPLC (Shimadzu Scientific Instrument, Inc. Columbia, MD, USA) with a refractive index detector. An Aminex HP87 column (5 μm, 30 cm×4.6 mm, Bio-Rad, CA, USA) was used in an oven set at 50 °C. Sulfuric acid at 0.005 M was used as the mobile phase with a flow rate of 0.6 ml/min. The injection volume was 20 μl. Concentrations of the aforementioned chemicals were calculated based on calibration curves built for each compound using external standards. Before HPLC analysis, all samples were filtered through 0.2-μm filters to remove any potential particles. Analysis of cellular composition was conducted by following our published procedure [10]. Briefly, to obtain crude lipid content, dried cell pellet (0.05 g) was grounded to a fine powder and transferred to a 7 ml chamber of a bead-beater (BioSpec Products, Bartlesville, OK, USA). This chamber was filled with 0.5 mm zirconium beads to approximately 5 ml. Methanol was then added to fill the rest of the chamber. After cells were disrupted by bead-beating for 2 min, the entire content was transferred to a 50-ml glass centrifugation tube. The chamber was washed twice using methanol (total 10 ml) to collect the yeast residue. Chloroform was then added to the tube to make the chloroform/methanol ratio as 2:1 (v/v). The tube was vortexed


for 5 min and was allowed to stand for 24 h. After that, the tube was centrifuged at 4,000×g for 15 min to remove the beads and yeast solids. The supernatant was collected and the solvent was vaporized using a Rotovap. Oil left in the flask without solvent was weighed to calculate oil content. To analyze content of carbohydrates, dried cell pellet (0.1 g) was acidified by adding 20 ml HCl (2.5 N). The acidified solution was then hydrolyzed at 100 °C for 30 min and neutralized to pH 7. The volume was adjusted to 100 ml followed by filtration to remove cell residues. The filtered sample was subjected to assay using DNS [13]. The ash content was determined after igniting 0.1 g cells in a muffle furnace at 550 °C overnight. The protein content was then calculated by subtraction.

Results and Discussion Pretreatment Severity factor (SF) represented by Eq. 1 is commonly used to evaluate the severe degree of acid pretreatment [2]. As reflected from the equation, SF is affected by pretreatment temperature, duration, and final pH. As demonstrated in Table 1, the three pretreatment conditions had SF of 1.06, 1.84, and 2.24 for corn fiber and 1.02, 1.87, and 2.18 for SSB. Apparently, increasing pretreatment temperature and acid concentration increased SF. SF ¼ log

t expðT −100Þ − pH 14:75

ð1Þ

where t is the reaction time (minutes); T is the pretreatment temperature (°C); pH is the final pH of the slurry after pretreatment. Theoretically, severe pretreatment tends to disrupt the lignocellulosic structures more than those done under mild conditions. But severe pretreatment may also lead to degradation of released sugars from cellulose and hemicellulose. As shown in Fig. 1, for SSB, with the increase of SF from 1.02 to 1.87 and 2.18, mass of TRS per 10 g dry feedstock in the supernatant after pretreatment decreased significantly from 3.41 to 2.79 and 2.20 g, respectively. Similar trend was also observed for TRS in the wash water. Amount of TRS decreased from 1.62 to 1.38 and 1.20 g for samples with SF from low to high. Following washing, the air-dried solid samples with moisture contents ranging from 71 to 78 % were hydrolyzed by cellulase and cellobiase. As indicated by Fig. 2, most of TRS was released during the first 24 h. Samples with SF of 1.87 had the highest TRS release compared with those with SF of 1.02 and 2.18. Collectively, the total mass of TRS unlocked from 10 g SSB due to the combined effort Table 1 Severity factors for different pretreatment conditions Biomass feedstock

Pretreatment condition

Severity factor

Standard deviation

Corn fiber

121 °C, 1 h, 0.5 % sulfuric acid

1.06

0.03

Corn fiber


1.84

0.00

Corn fiber


2.24

0.02

SSB


1.02

0.02

SSB


1.87

0.03

SSB


2.18

0.06


7.0 Total TRS released

6.0

y = -1.2573x + 7.3129 R² = 0.9065

5.0 TRS in supernatant

TRS (g)

4.0 3.0

TRS in wash water

2.0 TRS released by enzymes

1.0

0.0 0

0.5

1 1.5 Severity factor

2

2.5

Fig. 1 Distribution of total reducing sugars (TRS) in the combined process of pretreatment and enzymatic hydrolysis for SSB

of pretreatment and hydrolysis was 5.96±0.20 g for SF of 1.02, 5.23±0.02 g for SF of 1.87, and 4.37±0.02 for SF of 2.18. Considering soluble sugars initially associated with bagasse, which is 3.73 g/100 g SSB and the theoretically available sugars which is 60.6 g/100 g SSB calculated based on the content of cellulose and hemicellulose in SSB, the net TRS recovery from pretreatment and hydrolysis was 92.2, 80.2, and 66.0 % for SF of 1.02, 1.87, and 2.18, respectively (Fig. 3). Thus, for SSB, the lowest severity treatment resulted in maximal release of sugars. This sugar yield is comparable to that 92 % reported for SSB treated by steam at 200 °C for 5 min plus impregnation with 2 % SO2 [20]. For samples with the lowest severity pretreatment, among 92.2 % of sugars freed from SSB, only 5.7 % was due to enzymatic hydrolysis (Fig. 3). The majority of TRS, which was 86.5 %, was present in the supernatant after pretreatment, the wash water, and remained as

Fig. 2 Time course of release of total reducing sugars (TRS) during enzymatic hydrolysis of pretreated and washed SSB


% of total sugars

100

92.2% 80.2%

80

66.0%

60 40 20 0

1.02

Supernatant

1.87 Severity factor Wash water

2.18

Released by enzymes

Fig. 3 Percentage recovery of total reducing sugars (TRS) for SSB at different conditions

soluble sugars associated with washed solids owing to the non-exhaustive washing process. Therefore, in light of the cost for enzymes, time for conducting enzymatic hydrolysis, and the low sugar yield from this process, this step could be eliminated if the main purpose is to recover most of sugars out of SSB. In this case, a fast and simple dilute acid pretreatment (0.5 % sulfuric acid) at 121 °C for 1 h is enough to release more than 86 % of potentially available sugars from SSB. The remaining material which includes unhydrolyzed cellulose, hemicellulose, and lignin can be processed further through thermochemical processes to produce bio-crude. Regarding corn fiber, a similar trend of TRS release with respect to SF was observed (Fig. 4). The lowest SF of 1.06 resulted in a recovery of 83.2 % of theoretically available sugars just by pretreatment. These sugars were present in the supernatant of the slurry after pretreatment/centrifugation and wash water from the incomplete washing process. Pretreatment of corn fiber with SF of 1.84 and 2.24 led to sugar yield of 80.7 and 67.5 %, respectively. Between the two lower SFs, percentages of sugar recoveries did not differ significantly. But pretreatment having the highest SF had the lowest sugar yield. Thus, for corn fiber, pretreatment at 121 °C for 1 h using 0.5 % of sulfuric acid is enough to unlock most of sugars out of the feedstock. Under the same condition of pretreatment but with an addition step of enzymatic hydrolysis, sugar yield between 85 and 100 % was reported by Saha and Bothast for corn fiber at solid loading of 15 % [17]. In view of the fact that insignificant amount of sugars will be released through the step of enzymatic hydrolysis, further hydrolysis of the washed corn fiber was not conducted for this study. Acid hydrolysates which were the supernatant of slurries after pretreatment and centrifugation of those derived from the two lower SFs for SSB and corn fiber were used to ferment C. curvatus. In this study, no detoxification of any hydrolysates was carried out. As shown by Fig. 5, SSB hydrolysates derived from SF of 1.02 supported growth of C. curvatus very well. Once yeast cells were added to the hydrolysates, they started to grow immediately with no lag phase. After 6 days, the cell dry weight was 10.8 g/l. In the case of hydrolysates developed from SF of 1.87, however, it took cells 3 days to acclimate to this material. A biomass density


100.0%

TRS recovrey (%)

83.2%

80.7%

80.0%

67.5%

60.0% 40.0% 20.0% 0.0% 1.06

1.84

2.24

Severity Factor Fig. 4 Percentage recovery of total reducing sugars (TRS) for corn fiber at different conditions. Only TRS freed by pretreatment was considered in calculation. This calculation did not include soluble sugars that were associated with the washed solids due to the non-exhaustive washing process

of 3.6 g/l was reached in 6 days. With regard to hydrolysates derived from corn fiber, no significant cell growth was detected though nutrients in the mineral medium and the spore’s solution were provided (data not shown). For this hydrolysate, detoxification may be needed to remove any potential inhibitors to yeast fermentation. Analysis of hydrolysates derived from SSB with SF of 1.02 by HPLC revealed the presence of three main sugars (Fig. 6a). During fermentation, glucose was the first one to be consumed.

12

Cell dry weight (g/l)

SF = 1.02 10 SF = 1.87

8 6 4 2 0 0

2

4 Time (day)

Fig. 5 Growth of C. curvatus on hydrolysates developed from SSB with two different SFs

6


a 30

Sugar conc. (g/l)

25

Glucose Xylose

20

Arabinose

15

10 5 0

0

1

2

3 Time (day)

4

5

6

b 1.2 HMF

Conc. ( g/l)

1 0.8

Acetic aicd

0.6

Levulinic acid

0.4

Formic acid

0.2 0

0

1

2

3 Time (day)

4

5

6

c

Sugar conc. (g/l)

30

Glucose

25

Xylose

20

Arabinose

15 10 5 0 0

1

2

3 Time (day)

4

5

6

d Levulinic acid

6

Formic acid

5 Conc. (g/l)

Fig. 6 Concentrations of chemicals changing with time during yeast fermentation. a Sugars, SF=1.02; b non-sugar compounds, SF=1.02; c sugars, SF=1.87; d non-sugar compounds, SF=1.87

Acetic acid

HMF

4 3 2

1 0

0

1

2

3 Time (day)

4

5

6


After its concentration dropped to 10.7 g/l at day 2, xylose and arabinose started to be utilized by C. curvatus. The hydrolysates also contained several other non-sugar chemicals, 4-HMF, acetic acid, levulinic acid, and formic acid in the order of concentration from high to low (Fig. 6b). Interestingly, C. curvatus consumed these four compounds simultaneously as their concentrations all decreased with time starting from day 0. Among the non-sugar compounds, HMF had the highest concentration of 1.5 g/l which is lower than the threshold of 3 g/l above which significant inhibition on C. curvatus growth and lipid production will take place. However, though HMF was consumed by C. curvatus in our study, this chemical remained unchanged at a concentration around 0.4 g/l throughout a 7-day fermentation using the same yeast strain but conducted by another research group [21]. Same kind of chemicals was observed in hydrolysates from SSB with SF of 1.87. During the first 2 days of fermentation, no sugars were utilized (Fig. 6c), which was consistent with the lag phase of cell growth demonstrated in Fig. 5. After day 2, glucose concentration decreased with time while not much change happened to xylose and arabinose within the 6day experiment. Similar trend was seen for other non-sugar chemicals (Fig. 6d). C. curvatus started to consume all of these compounds after 2 days. The utilization rates after day 3, however, were leveled off except that of HMF. Comparing sugars in hydrolysates of SSB obtained from SF of 1.02 and 1.87, concentrations of glucose and arabinose were basically the same. Concentration of xylose in hydrolysate from SF=1.87 was 41.2 % of that from SF=1.02. Thus, it indicated that severe pretreatment degraded xylose. As a result of this degradation, higher concentration of formic acid (1.6 g/l) was measured in hydrolysates from SF of 1.87 than that (0.1 g/l) from SF of 1.02. Increased formic acid could be formed from furfural and HMF from degradation of xylose and glucose, respectively [14]. In addition to formic acid, hydrolysate from SF of 1.87 contained levulinic acid at a concentration of 5.0 g/l which was much higher than that (0.3 g/l) from SF of 1.02. This acid has been reported to be a degradation product of HMF under strongly acidic conditions [6]. Thus, it seemed that more glucose was released under the harsher pretreatment condition, but owing to glucose degradation, same concentration of glucose was identified in both hydrolysates. From comparison of these two liquids, it is obvious that (1) mild pretreatment led to higher concentrations of sugars and smaller concentration of acids and HMF due to less degradation and that (2) the high concentration of levulinic acid and formic acid might be the reason for a slow cell growth during fermentation. In this study, furfural, a product of xylose degradation with a 47-min retention time on the column we used, was not detected by HPLC in any samples derived from SSB. To understand why C. curvatus did not grow on corn fiber hydrolysates, we profiled the chemicals in these hydrolysates, too through use of HPLC. As indicated by Fig. 7, corn fiber hydrolysates contained more xylose than glucose, especially for those obtained from SF of 1.06 and 1.84, and higher concentration of arabinose than those in SSB hydrolysates. This is consistent with the fact that corn fiber has a high content of hemicellulose (27.7 %) which includes xylan and arabinan while SSB contains 17.8 % of hemicellulose which is mainly xylan. The presence of high concentration of xylose and arabinose may limit cell growth as they are not the preferred sugars for C. curvatus. Among the three kinds of hydrolysates attained from different SFs, it is perceivable that (1) concentration of acetic acid which was released from hemicellulose was almost the same. This is agreeable with results from SSB; (2) xylose concentration decreased with an increase of SF from 28.1 to 21.9 and 15.2 g/l; (3) concentration of levulinic acid increased from 0 to 0.6 and 1.3 g/l with an increase of SF; (4) concentration of formic acid increased from 0 to 0.3 and 0.7 g/l with the same order of SFs; and (5) furfural with a concentration of 0.5 g/l only appeared in hydrolysates derived from SF of 2.24. Thus, it is apparent that harsher pretreatment increased the formation of non-sugar

Appl Biochem Biotechnol Fig. 7 Chemical profiles of hydrolysates derived from corn fiber. a SF=1.06; b SF=1.84; c SF= 2.24

a SF = 1.06 35.0 28.1

Conc. (g/l)

30.0 25.0 20.0

15.6

15.0

11.4

10.0 2.6

5.0 0.0 Glucose

Xylose

Arabionose

Acetic acid

b SF = 1.84 25.0

21.9

Conc. (g/l)

20.0 15.0

13.2 9.8

10.0

2.9

5.0

0.6

0.3

0.0

c SF = 2.24

Conc. (g/l)

20.0 15.0 10.0

5.0

14.8

15.2 7.6 2.8

1.3

0.7

0.5

0.0

compounds that could inhibit growth of C. curvatus. But the exact reason for why C. curvatus did not grow on corn fiber hydrolysates is not certain at this point since other chemicals that were not detected and quantified by HPLC could result in growth inhibition. In this case, to utilize corn fiber hydrolysates as substrates for cultivating C. curvatus or other yeast strains, detoxification through overliming and/or absorption is needed. Composition analysis of the day 6 C. curvatus cells obtained from fermentation of the SSB hydrolysate derived from SF of 1.02 revealed that the cells contained 40±0 % of lipid, 40.3± 0.2 % of carbohydrate, 11.7±2.6 % of protein, and 8±2.8 % of ash. As demonstrated in Fig. 8, this lipid content translated to a lipid yield of 3.03 g/100 g bagasse. This yield is lower than


Solid 100 g bagasse

Pretreatment

36.9 g glucan 17.8 g xylan 19.5 g lignin Soluble sugars: 3.73 g

Centrifugation

Wash with water

Wash water: TRS: 22.0 g Washed Solid

Supernatant: 63.24 ml TRS2: 34.1 g

Enzymatic hydrolysis

TRS: 3.44 g

Yeast fermentation

3.04 g total lipids

10.8 g/l, 40% lipid

Fig. 8 The overall process and mass balance demonstrating lipid yield from 100 g sorghum bagasse

8.47 g/100 g bagasse that was reported by our group through use of lime pretreatment followed by enzymatic hydrolysis [12]. However, it needs to be noted that only acid hydrolysates and only a fraction of sugars available in SSB were used for yeast fermentation. In addition, the calculated lipid yield from this study is lower than 4.7 g/100 g wheat straw published by Chen’s group [21]. However, in that study, Ca(OH)2 was used to raise pH of the hydrolysate to 5.5 and an extra step of filtration was included. Though that was not an overliming step, the addition of Ca(OH)2 followed by filtration could have removed some compounds out of the hydrolysates. As a result, higher cell density and lipid yield were obtained. In this study, however, we used NaOH to increase the pH to 5.0 which is suitable for C. curvatus. No filtration was used after that. Comparing total TRS concentration (34.1 g/100 g bagasse) in the acid hydrolysates with those (23.2 g/100 g bagasse) measured by HPLC, it was obvious that the acid hydrolysates contained reducing sugars that were not monomeric. Based on total monomeric sugar consumption during fermentation, the yield of cells and lipids was calculated as 0.35 and 0.14 g/g sugars, respectively. Again, these yields were lower than those from our previous studies [11, 12]. However, considering the fact that (1) acid pretreatment is fast and simple and (2) the hydrolysate can be directly used for yeast fermentation without conditioning or detoxification, the pathway presented in this study definitely merits further evaluation and optimization.

Conclusion Pretreatment of corn fiber and SSB through use of dilute sulfuric acid at mild conditions led to high sugar recovery and low concentration of non-sugar compounds. Hydrolysates of SSB derived from pretreatment with the lowest severity factor supported growth of C. curvatus which reached a cell density of 10.8 g/l with a lipid content of 40 % in 6 days. C. curvatus was able to consume all major monosaccharides and non-sugar chemicals in the hydrolysates. Corn fiber hydrolysates, however, did not result in significant cell growth even with the supplementation of nutrients. Explanation of differences in how cells perform in different samples was assisted by analysis of chemical compositions of the hydrolysates.

Appl Biochem Biotechnol Acknowledgments Kimberly Jarosz thankfully acknowledges the opportunity and financial support from the McNair Scholars Program at SIUC. Ashley T. Wardlow from Florida International University appreciates support from a NSF REU program (DMR 1157058) at SIUC. We also thank Dr. Sabrina Trupia at National Corn-toethanol Research Center for providing the corn fiber samples.

References 1. Alvira, P., TomÃ¡s-PejÃ, E., Ballesteros, M., & Negro, M. J. (2010). Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review. Bioresource Technology, 101, 4851–4861. 2. Banerji, A., Balakrishnan, M., & Kishore, V. V. N. (2013). Low severity dilute-acid hydrolysis of sweet sorghum bagasse. Applied Energy, 104, 197–206. 3. Choudhary, R., Umagiliyage, A. L., Liang, Y., Siddaramu, T., Haddock, J., & Markevicius, G. (2012). Microwave pretreatment for enzymatic saccharification of sweet sorghum bagasse. Biomass and Bioenergy, 39, 218–226. 4. Cui, Y., Blackburn, J. W., & Liang, Y. (2012). Fermentation optimization for the production of lipid by Cryptococcus curvatus: use of response surface methodology. Biomass and Bioenergy, 47, 410–417. 5. Cui, Y., & Liang, Y. (2014). Direct transesterification of wet Cryptococcus curvatus cells to biodiesel through use of microwave irradiation. Applied Energy, 119, 438–444. 6. Du, B., Sharma, L. N., Becker, C., Chen, S. F., Mowery, R. A., van Walsum, G. P., et al. (2010). Effect of varying feedstock–pretreatment chemistry combinations on the formation and accumulation of potentially inhibitory degradation products in biomass hydrolysates. Biotechnology and Bioengineering, 107, 430–440. 7. Hector, R. E., Qureshi, N., Hughes, S. R., & Cotta, M. A. (2008). Expression of a heterologous xylose transporter in a Saccharomyces cerevisiae strain engineered to utilize xylose improves aerobic xylose consumption. Applied Microbiology and Biotechnology, 80, 675–684. 8. Klinke, H. B., Thomsen, A. B., & Ahring, B. K. (2004). Inhibition of ethanol-producing yeast and bacteria by degradation products produced during pre-treatment of biomass. Applied Microbiology and Biotechnology, 66, 10–26. 9. Larsson, S., Palmqvist, E., Hahn-HÃ¤gerdal, B. R., Tengborg, C., Stenberg, K., Zacchi, G., et al. (1999). The generation of fermentation inhibitors during dilute acid hydrolysis of softwood. Enzyme and Microbial Technology, 24, 151–159. 10. Liang, Y., Sarkany, N., & Cui, Y. (2009). Biomass and lipid productivities of Chlorella vulgaris under autotrophic, heterotrophic and mixotrophic growth conditions. Biotechnology Letters, 31, 1043–1049. 11. Liang, Y., Tang, T., Siddaramu, T., Choudhary, R., & Umagiliyage, A. L. (2012). Lipid production from sweet sorghum bagasse through yeast fermentation. Renewable Energy, 40, 130–136. 12. Liang, Y., Tang, T., Umagiliyage, A. L., Siddaramu, T., McCarroll, M., & Choudhary, R. (2012). Utilization of sorghum bagasse hydrolysates for producing microbial lipids. Applied Energy, 91, 451–458. 13. Liang, Y., Yesuf, J., Schmitt, S., Bender, K., & Bozzola, J. (2009). Study of cellulases from a newly-isolated thermophilic and cellulolytic Brevibacillus sp. strain JXL. Journal of Industrial Microbiology & Biotechnology, 36, 961–970. 14. Mussatto, S. I., & Roberto, I. C. (2004). Alternatives for detoxification of diluted-acid lignocellulosic hydrolyzates for use in fermentative processes: a review. Bioresource Technology, 93, 1–10. 15. Palmqvist, E., & Hahn-HÃ¤gerdal, B. R. (2000). Fermentation of lignocellulosic hydrolysates. I: inhibition and detoxification. Bioresource Technology, 74, 17–24. 16. Palmqvist, E., & Hahn-Hägerdal, B. (2000). Fermentation of lignocellulosic hydrolysates. II: inhibitors and mechanisms of inhibition. Bioresource Technology, 74, 25–33. 17. Saha, B. C., & Bothast, R. J. (1999). Pretreatment and enzymatic saccharification of corn fiber. Applied Biochemistry and Biotechnology, 76, 65–77. 18. Saha, B. C., Dien, B. S., & Bothast, R. J. (1998). Fuel ethanol production from corn fiber current status and technical prospects. Applied Biochemistry and Biotechnology, 70, 115–125. 19. Serrano-Ruiz, J. C., Ramos-Fernández, E. V., & Sepúlveda-Escribano, A. (2012). From biodiesel and bioethanol to liquid hydrocarbon fuels: new hydrotreating and advanced microbial technologies. Energy & Environmental Science, 5, 5638–5652. 20. Sipos, B., Reczey, J., Somorai, Z., Kadar, Z., Dienes, D., & Reczey, K. (2009). Sweet sorghum as feedstock for ethanol production: enzymatic hydrolysis of steam-pretreated bagasse. Applied Biochemistry and Biotechnology, 153, 151–162. 21. Yu, X., Zheng, Y., Dorgan, K. M., & Chen, S. (2011). Oil production by oleaginous yeasts using the hydrolysate from pretreatment of wheat straw with dilute sulfuric acid. Bioresource Technology, 102, 6134–6140.