Lipomyces starkeyi: Its current status as a potential oil producer

Fuel Processing Technology 177 (2018) 39–55

Contents lists available at ScienceDirect

Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc

Review

Lipomyces starkeyi: Its current status as a potential oil producer a

b

c

Sylviana Sutanto , Siti Zullaikah , Phuong Lan Tran-Nguyen , Suryadi Ismadji

d,⁎

, Yi-Hsu Ju

a,⁎

T

a

Department of Chemical Engineering, National Taiwan University of Science and Technology, 43, Keelung Rd., Sec. 4, Taipei 106-07, Taiwan Department of Chemical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya 60111, Indonesia Department of Mechanical Engineering, Can Tho University, 3-2 Street, Can Tho City, Viet Nam d Department of Chemical Engineering, Widya Mandala Surabaya Catholic University, Kalijudan 37, Surabaya 60114, Indonesia b c

A R T I C LE I N FO

A B S T R A C T

Keywords: Fermentation L. starkeyi Microbial lipids Biodiesel production

Culturing of oleaginous yeasts has been studied extensively utilizing various substrates as a nutrient, such as industrial or agricultural residues. Despite many choices of oleaginous yeasts, attention should be given to specific species so that real application can be implemented, rather than on exploring new oleaginous yeasts with higher oil-producing ability. Lipomyces starkeyi is an oleaginous yeast that can be cultured using a wide range of feedstocks. It is worth noting that L. starkeyi can produce a high amount of lipids with a good proportion for biodiesel purpose and its ability to re-utilize small amount of its lipid, makes it superior compared to other oleaginous yeasts. This review offers a comprehensive summary of L. starkeyi, its characteristics and the type of nutrients it can assimilate, brief reviews of common fermentation modes used, and strategies for enhancing lipid accumulation will be discussed. Also, common transesterification methods, as well as possibility/future prospect of oleaginous yeast utilization to produce single cell oil will also be discussed. This review hopefully could help bridging the gap between theoretical and actual potentials of oleaginous yeasts in producing lipids as feedstock for biodiesel production.

1. Introduction

One way to help to reduce CO2 emission is by using renewable fuel (biofuels). The two most prevalent biofuels nowadays are bioethanol and biodiesel. Bioethanol (C2) mixed with gasoline (C4-C9) [5] can oxygenate the fuel, leading to better combustion hence reducing air pollution [6]. In the USA, 10% ethanol is used in the ethanol-gasoline blend for old vehicles [5] and 15% for new vehicles [6]. Bioethanol cannot be used at 100% to replace gasoline due to its low energy density and some other factors [5], while biodiesel consisting of C16-C22 alkyl chains [5], can replace diesel fuel without the need to modify diesel engine. Implementation of biofuels such as biodiesels is considered to be one of the best choices to reduce CO2 emission. Since biodiesel is derived from plants, the CO2 produced is not more than what the plant absorbed during growth, thus making it zero CO2 emission [7]. Most biodiesel currently in use is derived from the so-called the first generation feedstock (edible vegetable oils or oil from other food crops). This leads to the ‘food vs. fuel’ debate about the use of edible oil for biodiesel production which results in food and land competition;

In general, crude oil price shows an increasing trend within the past 50 years. It hit the lowest point at the end of 1998, which was less than $20 per barrel, then gradually went up to $157 in June 2008 [1]. Starting in mid-2014, crude oil price started to decline to ~$62 per barrel currently [1,2]. This fluctuation was likely due to most oil-producing countries, especially Saudi Arabia, refused to cut down production due to the fear of losing market share, which resulted in a surplus of crude oil and lower oil price. On the other hand, falling crude oil price is detrimental to the development of renewable energy. Increasing consumption of crude oil inevitably leads to CO2 accumulation and global warming. Also, world oil reserves are limited and will be depleted soon if alternative energy supplies cannot be found. In fact, as reported by EIA, global oil demand keeps increasing from 89.8 million barrels per day in 2011 to 98.5 million barrels per day by the end of 2017 [3]. World oil consumption is projected to increase to 112 million barrels per day by 2035 [4].

Abbreviations: 5-HMF, 5-hydroxymethylfurfural; ACL, ATP citrate lyase; AMP, Adenosine mono phosphate; ATP, Adenosine Tri Phosphate; ARA, arachidonic acid; DHA, docosahexanoic acid; EPA, eicosapentanoic acid; GLA, Gamma linoleic acid; C/N, Carbon to Nitrogen; CoA, Coenzyme A; DAG, diacylglyceride; FAs, Fatty acids; FAME, Fatty Acid Methyl Esters; GDH, glycerol-3-phosphate dehydrogenase; ICDH, isocitrate dehydrogenase; LCPUFA, long chain polyunsaturated fatty acids; LPA, lyosophosphatidic acid; NAD+, Nicotinamide adenine dinucleotide; NADPH, nicotinamide adenine dinucleotide phosphate; PDC, pyruvate dehydrogenase complex; PGDH, 6-phosphogluconatedehydrogenase; PHB, para-hydroxy benzaldehyde; PEG, polyethylene glycol; SCO, Single cell oils; TAG, Triacylglycerol; TCA, Tricarboxylic acid ⁎ Corresponding author. E-mail addresses: [email protected] (S. Ismadji), [email protected] (Y.-H. Ju). https://doi.org/10.1016/j.fuproc.2018.04.012 Received 30 January 2018; Received in revised form 9 April 2018; Accepted 11 April 2018 0378-3820/ © 2018 Elsevier B.V. All rights reserved.

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

ascopores then germinate and divide by budding [23]. Macromorphology of the colony shows that it is smooth and white-creamy in appearance with a mucoid texture [21]. This species reproduces asexually by multilateral budding resulting in round or oval-shaped cells [21]. The optimum temperature for L. starkeyi to accumulate lipid was found to be 25.5–29.5 °C [25]. The original strain of this yeast is CBS 1807; while mutation or modification results in CBS 1809, CBS 2512, CBS 6047, CBS 7536, CBS 7537, CBS 7544, CBS 7545, CBS 8064 [24] and other 10 strains [21]. L. starkeyi CBS 1807 has different collection number in each research center (The Netherlands-CBS 1807, USA-ATCC 58680, NRRL Y-11557, NRRL Y-1388, Germany-DSM 70295, Japan-IFO 1289 and JCM 5995, Taiwan-CCRC 21522 and BCRC 23408, Belgium-MUCL 39418, ItalyDBPVG 6193).

shortage in edible oils and the higher price of foods [8]. The search for feedstock to replace edible vegetable oils led to the development of the second generation feedstock such as waste cooking oil, animal fats, rice bran oil and Jatropha curcas oil [9,10]. The third generation feedstock currently under development [10] aims to produce biodiesel from oleaginous microorganisms such as yeasts, microalgae, and bacteria. The term ‘oleaginous’ means that the microorganism can accumulate oil > 20% of its dry weight, and can reach up to 60–70% in some cases [11]. Microbial oil is recognized as one of the potential biodiesel feedstock. Yeast is the preferred microorganism for producing microbial oil to microalgae or bacteria. In spite of advantages such as it requires no land to grow, (mixotrophic) microalgae has a limitation on carbon source utilization in the sense that it cannot convert starch to oil [12], and it has lower growth rate than yeast. Only some bacteria can be used for microbial oil production since most of the bacteria only produce lipoid complex like polyhydroxyalkanoate [13] in the outer membrane thus is difficult to be extracted [14]. Certain fungus can grow in starch [15], but it produces only a small quantity of biomass [12]. On the other hand, yeast such as Rhodoturula glutinis, Rhodosporidium toruloides, Cryptococcus curvatus, Trichosporon fermentans and Lipomyces starkeyi are able to grow in a broad range of substrates including hydrolysate of agricultural or industrial residues, and produces a high amount of lipids. The growth of yeast is relatively fast with no seasonal limitation, and the process is easily scaled up [11]. The growth of yeast is less susceptibility to viral infection, and bacterial contamination can be controlled easily by growing at low pH [10]. Single cell oil (SCO) has been known since the 1980s for use as a substitute for cocoa butter [16]. Also, researchers also focused on producing other beneficial FAs for human health such as γ-linoleic acid (GLA), docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA), and arachidonic acid (ARA) [17,18]. But utilization of SCO as a potential feedstock for biodiesel production just gained interests in recent years. However, high producing cost hinders its commercial application [11]. But hurdles are expected to be tackled with improving and maturing technologies [19], and in the future, it is possible that petroleum diesel price will be higher than what is now due to scarcity; allowing better chances of massive biodiesel utilization. This situation gives opportunities for researchers to explore SCO for cost-effective biodiesel production. This review focuses on oleaginous yeast L. starkeyi, about its classification, various carbon sources it could consume, and to a lesser extent about its other marginal products [20]. Various substrates-hydrolysis processes, lipid regulation, lipid extraction and conversion will be briefly explained. Many reviews have been published about lipid regulation mechanism and utilization of lignocellulosic or agro-industrial by-products for fermentation feedstock, however, specific yeast and overall practices needed for fermentation were seldom discussed. Thus, this review provides summaries and knowledge on producing SCO as biodiesel feedstock from L. starkeyi fermentation with the focus mainly on low-cost feedstock utilization, particularly agricultural biomass residues.

3. Physiological properties of L. starkeyi Yeast is composed of organic compounds, inorganic compounds and water [26]. After removing water which is the major part of yeasts, what left are organic and inorganic compounds. Organic compounds of yeasts are polysaccharides, lipids, and proteins, while inorganic compounds include cofactor, and trace metals [26]. The organelles in yeast cells are explained briefly [26]: - Cell wall, of which 80–90% is rigid polysaccharides (usually composed of glucan, mannan, and chitin) - Cell (selective) membrane, which is composed of the lipid bilayer (phospholipid and sterols) with membrane proteins (ATPase, transport proteins, etc.), cholesterol and glycolipid attached to it - Cytosol, liquid where other organelles are suspended inside the cell - Nucleus contains genetic material of the yeast - Ribosome, place where proteins are biosynthesized - Mitochondria converts food to form energy - Endoplasmic reticulum synthesizes (together with ribosome) and packages protein; transports molecules around the cells - Golgi apparatus, a place for protein modification and distribution - Vacuole, as a food storage Some important factors affecting lipid accumulation in yeasts include temperature, pH, dissolved oxygen and C/N ratio of the medium. An early report on L. starkeyi IAM 4753 grown aerobically in simple defined medium at 30 °C showed that lipid accumulation was low during cell growth (logarithmic phase) and increased significantly when the yeast was in early stationary phase, lipid accumulation was probably caused by a change in metabolism due to limited amount of dissolved oxygen left [27]. Suutari et al. [28] observed a similar phenomenon in which yeast started to accumulate lipid during the growth phase, and 28 °C was the optimum temperature for both biomass and lipid productivity. Investigation on L. starkeyi IAM 4753 showed that it could grow well in glucose mineral medium with pH 5 [29]. pH above 5 may inhibit the enzyme activity to produce biotin that promotes cell growth. Thus additional biotin in the medium was needed for pH 5.5–6.5 [29]. It was found that cations and anions were important substance for cells; Mg2+, Mn2+, and Zn2+ were important for fermentation, cell growth, and metabolism; Cu2+ and Fe2+ were needed as a cofactor; while phosphate and sulfate were essential for structural molecules and cell physiology, respectively [26]. Sufficient amount of Mn2+ was needed to increase biomass growth by 1.6 fold, while the low amount of Zn2+ was needed to obtain higher lipid content [25]. A subsequent study reported that the simultaneous addition of Mn2+, Zn2+and mono-potassium phosphate at stationary phase was able to induce biomass growth but not lipid production [30].

2. Taxonomy of L. starkeyi Genus Lipomyces is in Lipomycetaceae family, order Sacharomycetales, Saccharomycetes class (subclass Saccharomycetidae, subdivision Saccharomycotina [21]), phylum Ascomycota in kingdom Fungi [22]. To date, there are 16 species accepted as genus Lipomyces [23]. Among these species, L. starkeyi (together with L. lipofer to a lesser extent) is the most extensively studied yeast, chosen for its excellent ability to produce lipid. L. starkeyi (scientific name: L. starkeyi Lodder & Kregger van Rij) is a unicellular eukaryotic yeast-which has a nucleus and other organelles. This strain was originated from the soil in the USA and was isolated by R.L Starkey (Starkey's strain number 74) [24]. It reproduces sexually by developing 4–20 ellipsoidal [23] or round [21] ascospores (contains an oil droplet) per ascus. The light amber to brown 40

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

Fig. 1. Mechanism of lipid synthesis [37], blue arrows: TCA cycle, ACL: ATP Citrate lyase, ACC: acetyl-coA carboxylase, LPA: lyosophosphatidic acid, DAG: diacylglyceride, TAG: tricylglyceride. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

➢ Sugars (preferentially glucose) is converted into pyruvate in glycolysis metabolic pathway which occurs in the cytosol

4. How does oleaginous yeast accumulate lipids? A specific trait owned by oleaginous yeasts is the ability to continuously provide acetyl CoA and sufficient stock of NADPH which are important for lipid formation [31]. Sources that can be used for acetyl CoA production include fatty acids, amino acids, and sugars. Lipid biosynthesis in yeasts occurs in 2 ways, depending on a substrate: de novo and ex novo. In de novo biosynthesis, lipid accumulation is initiated with an excess of carbon source (sugars) and limited nutrient (usually nitrogen) in the substrate. Although in certain yeast such as Rhodosporidium toruloides, sometimes it requires phosphate or sulfate that was provided in the limited amount [32,33], but this review will discuss mainly nitrogen limited condition. The excess carbon source is no longer used for cell growth. Instead it is channeled to lipid accumulation by a series of cellular respiration. Cellular respiration is a biological step that occurs in all living organisms to convert food to energy. Cellular respiration consists of a few steps: glycolysis, Krebs cycle and electron transport chain. Cellular respiration takes place inside the cells as shown below [31,34], depicted in Fig. 1:

Glucose + NAD + 2 NADH + 2 ATP

ADP + 2

Pi → 2

pyruvate + 2

CO2 + 2

➢ Pyruvate is transferred to mitochondria and converted to acetyl coA with the help of enzyme PDC (pyruvate dehydrogenase complex) Pyruvate + NAD+ + Coenzyme A → Acetyl CoA + NADH + CO2 ➢ Meanwhile, a limited amount of nitrogen source increases the activity of AMP (adenosine mono phosphate) deaminase to break down AMP, generate ammonia to supply nitrogen for cells AMP → inosine 5′-monophosphate + NH3 ➢ AMP amount keeps decreasing and inhibits the formation of isocitrate dehydrogenase (ICDH) ➢ Conversion of isocitrate to alpha-ketoglutarate in TCA/Krebs cycle 41

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

does not occur since it is absent in ICDH ➢ Isocitrate is converted back to citrate by aconitase enzyme ➢ Citrate accumulated is transferred to the cytosol via citrate/malate shuttle, in an exchange with malate

○ ○

○ ○

Under limited nitrogen condition, the amount of citrate increased up to ~80% while the activity of ICDH enzyme decreased from 0.023 to 0.005 U/g [35].

Alpha-ketoglutarate, a precursor for amino acid and nucleotides Succinyl coA, a precursor for hemoglobin to supply oxygen in the blood Malate is used to produce pyruvate, and Oxaloacetate is used to produce glucose

In electron transport chain, NADPH acts as the electron donor and O2 is the final electron acceptor. Exchange of substrate and energy occurs in this stage in two different types [26]:

➢ Citrate in the cytosol is cleaved by ATP citrate lyase (ACL) to produce acetyl CoA

- Active transport

Citric Acid + CoA + ATP → acetyl-CoA + oxaloacetate + ADP + Pi

Transport of a substance from low concentration to high concentration thus requires ATP as energy to facilitate the transport. This mechanism occurs in proteins that are attached to the cell membrane.

➢ The newly formed of acetyl-CoA is carboxylated by enzyme ACC1 into malonyl-CoA ➢ Both acetyl-CoA and malonyl-CoA go to the next step for FA biosynthesis ➢ FAs formed are then linked to the glycerol produced by glycerol-3phosphate dehydrogenase to form TAG

- Passive transport ✓ Diffusion: transfer of a small molecule from high to low concentration, ATP is not needed, and the transport passes through the space of plasma membrane, e.g., O2, CO2 transfer. ✓ Facilitated diffusion: transport of a bigger molecule from high to low concentration, it passes through protein in the plasma membrane, occurs in 2 different porters: symport (amino acids and protons enter the cell together), or antiport (Na+/K+ pump, Na+ comes in, and K+ goes out while 1 ATP is formed).

Lipid accumulation requires acetyl-CoA: NADPH at 1:2 M ratio [36]. Generally, NADPH is provided by malic enzyme (as the main provider), and through the reaction of pentose phosphate pathway [36]. Meanwhile, elongation, desaturation of fatty acids into DAG and TAG occur in reticulum endoplasm [31,37]. Based on this step, most oleaginous yeasts theoretically would have 31.6% TAG yield from glucose [36]. But it was found that cytosolic malic enzyme in L. starkeyi prefers to take NAD+ as coenzyme rather than NADP+ [38], which theoretically does not supply NADPH in the cytosol. Thus, the source of NADPH in L. starkeyi is speculated only from pentose phosphate pathway, and the theoretical TAG yield from glucose declines to 27.6% [36]. Ex novo lipid biosynthesis usually occurs in hydrophobic substrates such as industrial fats, n-alkanes, vegetable oils and fish oils [8]. For yeast to assimilate these substrates, they need to be hydrolyzed with the help of lipase to produce free fatty acids [8,34,39,40]. Through active transport, fatty acids will be consumed by yeasts for growth or lipid production [39]. Since different yeasts may secrete different lipase, fatty acid profiles of the lipid accumulated may vary with the substrates used. Fatty acids will be converted into acyl-coA esters by acyl-coA synthetases (ACS) in the cytosol, then, degraded into smaller acyl-coA ester and acetyl-CoA by various acyl-CoA oxidases (Aox) in the β-oxidation process [39]. The products can be used for cell growth, maintenance, and other metabolites production. The whole process keeps repeating until all fatty acids in the substrates are completely degraded, depending on the length and concentration of substrates, the presence of acetyl-CoA and ratio of NAD+/NADH [39]. Producing high-value polyunsaturated fatty acids is the main interest in exploring hydrophobic substrates through ex novo lipid biosynthesis [39]. In ex novo biosynthesis, the presence of exogenous n-alkanes and fatty acids hinders fatty acid synthase (FAS) and ACL. Thus de novo and ex novo biosynthesis could not occur at the same time [39]. Storage lipids found in oleaginous yeasts are mostly in the form of the TAG (80–90%), and the rest is steryl esters (SE), which are not suitable for the development of phospholipid bilayer and are therefore accumulated in the lipid body [40]. In ex novo pathway, lipid production occurs simultaneously with cell growth; this is not related to nitrogen concentration in the medium [8,34] -which is unique to oleaginous microorganism [41]. In non-oleaginous microorganism, the excess carbon source is channeled into glycogen, glucans, and mannans [40,41]. TCA (tricarboxylic acid/Krebs) cycle, depicted in Fig. 1, is an intermediate reaction in cellular respiration. TCA cycle is amphibolic which means it can be used for both catabolic and anabolic reactions [26], because some intermediates in TCA cycle can be used as a precursor to produce other compounds, for example [26]:

Sugars uptake can occur in two types, free or passive diffusion, and facilitated diffusion or active transport [26]. Sugar assimilation by free diffusion may occur in small proportion because yeast membrane is not freely permeable for a polar compound like sugar. In ethanol producing yeasts (Crabtree-positive), facilitated diffusion of sugar uptake may be allowed to transport glucose to the cell, whereas in Crabtree-negative yeasts, like L. starkeyi, the high affinity of “proton symport glucosetransporters” is the one that transports sugars, therefore, sugar will not pass through facilitated diffusion [26]. During the transport of amino acids from medium to cells, protons are also carried away into the cell; protons build up in the cells tends to acidify the cell. Thus, some protons have to be secreted out of the cells to balance its pH. Secretion of protons to the medium by proton pump ATPase results in a decrease of the medium pH [26]. Aside from that, hydration and dissociation of CO2 (produced during fermentation) that yields H+ and HCO3−, and possible production of organic acids also contribute to acidification of the medium [26]. Acetyl CoA and malonyl CoA have a significant role in fatty acid synthesis to produce palmitic acid (C16:0), which is further elongated to become stearic acid (C18:0). Stearic acid is desaturated to produce oleic, linoleic and linolenic acid, which then encounter a series of elongation and desaturation to form saturated and unsaturated C20 and C22 fatty acids. These FAs will further react with glycerol to produce triglyceride [31]. 4.1. Enzymes affecting lipid biosynthesis in oleaginous yeasts Lipid accumulation is dictated by some enzymes such as 6-phosphogluconate dehydrogenase (PGDH), glycerol-3-phosphate dehydrogenase (GDH), isocitrate dehydrogenase (ICDH), ATP citrate lyase (ACL), acetyl-CoA carboxylase (ACC) and malic enzyme (ME). 6-PGDH is responsible for producing NADPH for later use in acylglyceride biosynthesis, while the latter 2 enzymes are the precursor for intermediates needed in the biosynthesis of acylglycerides [42]. Activities of those enzymes are high during lipid accumulation stage, and it can be increased by adding Mn+2 to the medium [42] to push the reaction forward. GDH, on the other hand, is very sensitive towards Zn+2, excess Zn+2 may inhibit GDH activity and lead to less lipid accumulation [42]. ICDH (an enzyme involved in Krebs cycle to change isocitrate to αketoglutarate) was also found to have high activity in oleaginous yeast 42

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

can be directly sterilized after characterization (if further pretreatment like removing impurities is not needed). Glycerol is the main by-product of biodiesel production, which accounts for 10 wt% of biodiesel product [55]. Since a lot of biodiesel is produced nowadays, glycerol could be used as a potential fermentation feedstock. Lignocellulosic biomass, such as agricultural residues, is the most abundant natural polymer that can serve as fermentation feedstock [56]. Plant cell wall of agricultural residues is made up of 3 main components: cellulose, hemicellulose, and lignin, with smaller amounts of pectin, proteins, extractives, and ash [56,57]. Extractives include lipophilic and hydrophilic components of plant cell wall, such as fats, waxes, steroids, terpenes, phenolic compounds which are a minor fraction of lignocellulosic biomass [56]. Cellulose is the major constituent of lignocellulosic biomass, comprises around 30–50% of biomass weight [56–58]. It has an unbranched and linear structure of repeated units of cellobiose (two glucose joined by β-1,4 glycosidic linkage), resulting in high molecular weight. Some linear chains of cellulose are linked together by hydrogen bonds, covalent bonds, or van der Waals bonds causing it to pack into cellulose microfibrils which are bundled together to form cellulose fibers [56,58]. Cellulose tends to arrange in parallel with a crystalline structure with less amorphous regions [56–58]. With this strong characteristic, cellulose is less soluble and less degradable [58]. The second major compound of lignocellulosic biomass is hemicellulose (~14–35% of biomass weight) [56–58]. It is linear polymers with branches, has lower molecular weight than cellulose, and typically is made up of hexoses, pentoses, and acetylated sugars. Lignin comprises 5–20% of biomass weight. It is an amorphous hetero-polymer and a complex molecule which is extremely resistant to chemical and enzymatic degradation, making it structural support of lignocellulose biomass [56–58]. Three main basic building blocks of lignin are guaiacyl, syringyl, and p-hydroxyphenyl moieties. Some other aromatic units exist in different types of woods [57]. Cellulose and hemicelluloses are the major sources of sugars from lignocellulosic biomass, and they both are referred to as holocellulose [56]. Hemicellulose is the most thermo-chemically sensitive part of lignocellulose biomass [56]. Examples of lignocellulosic biomass often utilized as fermentation feedstock are sugarcane bagasse, rice straw, rice bran, corn stover, corn stalk and wheat straw. Lignocellulosic biomass needs to be pretreated to release free sugars that can be consumed by oleaginous yeasts to produce lipid; this is because yeasts do not have cellulolytic activity [10]. Lignocellulose biomass pretreatment is needed to open up and separate the polymers for it to be accessible for hydrolysis process [10]. According to Balan [59], process pretreatment that has potential for commercialization should satisfy most of these criteria: process that can open up cell wall and separating lignin out of it and help improving long-term storage of the biomass; concentrate pretreated biomass, generate less inhibitors, possibility for scale-up, uses less energy and cheap chemicals, possibility to preserve lignin (to be transformed into useful product), uses less hazardous chemicals and operate at moderate conditions, and possibility to recover any catalyst used after pretreatment. Common pretreatment methods have been discussed in many reviews, which are summarized in Table 1. In the pretreatment of biomass to obtain simple sugars, biomass is treated either by chemicals or high temperature, or both, to break down the structure of the main components in it. This pretreatment, depending on the severity condition may result in the formation of inhibitors, which are the unwanted products that may give negative impact on the subsequent processes. Depending on the severity of hydrolysis condition, some inhibitors may form due to degradation of the cell wall component of lignocellulosic materials [10,63]:

grown in limited nitrogen medium [43]. A study on L. starkeyi AS 2.1560 showed that ICDH activity was high in the beginning of fermentation and decreased dramatically towards the end (48 h), while both intracellular and extracellular citric acid increased gradually during fermentation [42]. A specific factor inducing lipid accumulation is the presence of enzyme ATP citrate lyase (ACL). ACL is the prerequisite enzyme for lipid accumulation [44]. It cleaves citric acid to form acetyl CoA (as the precursor of lipid biosynthesis) and oxaloacetate [45]. Organisms that could accumulate lipid > 20% as TAG, was found to have ACL activity. However, a small number of yeasts was found to have ACL activity and produced only a tiny amount of lipids [41,46]. Therefore, the presence of ACL cannot be used as a standard for an organism to be oleaginous [45]. Malic enzyme (ME) is responsible for generating a pool of NADPH, which is then combined with acetyl-CoA to initiate lipid biosynthesis. If this enzyme activity is lost due to inhibition or mutation, it is no longer able to synthesise lipids. Thus, ME has a significant role in lipid accumulation [44,45]. On the other hand, the role of acetyl-CoA carboxylase (ACC) is to control metabolic efflux. Enzyme ACC helps in the formation of malonyl-CoA that initiates fatty acids biosynthesis [45]. 4.2. Metabolic engineering to enhance lipid production In general, some engineering modifications that were used successfully to enhance lipid accumulation in oleaginous yeasts are [47]: overexpression of enzymes in fatty acids biosynthesis pathway, overexpression of enzymes in TAG biosynthesis pathway, bypass TAG biosynthesis, blocking degradation pathways and the introduction of pathways from heterologous species. Cloning the enzyme that regulates lipid accumulation, or deleting enzymes that are responsible for degradation β-oxidation pathway [36] can be another method. To the best of authors' knowledge, there have been limited studies reported about the genetic or metabolic engineering of L. starkeyi to enhance fatty acids production or biofuels-related substance. An optimized transformation system of L. starkeyi has been established using modified lithium acetate (LiAc) transformation protocol. The results showed an increase in transformation efficiency by four orders of magnitude, compared to the initial transformation rate which was extremely low and required very high concentrations of DNA [48]. Meanwhile, Oguro et al. developed a system for integration of multiple copies of heterologous genes by spheroplast-polyethylene glycol (PEG) transformation. This feature is useful for overexpression of genes involved in lipid biosynthesis pathways [49]. Deletion of Δlig4 gene from L. starkeyi was also found to dramatically increase the recombinant efficiency (80%) compared to wild-type strain [50]. On the other hand, L. starkeyi transformation using Δ-15 desaturase gene from flax were able to improve LCPUFA produced. An increment in DHA and ALA in terms of ω-3 faty acids was noticed in the transformed yeast, screened by kanamycin resistance [51]. Wang et al. investigated an insertion of fatty acyl-coA reductase (far) gene, Maaqu_2220 in L. starkeyi enabled the production of fatty alcohols [52]. Another approach carried out was by engineered core fungal cellulases to L. starkeyi; the results suggest that L. starkeyi is capable of secreting fungal cellulases, therefore showing a potential candidate for consolidated bioprocess for biofuel production [53]. 5. Common fermentation feedstock and its pretreatment In biodiesel production, raw material accounts for > 70% of the total cost [54]. One way to reduce the cost is to utilize biomass residues which are available abundantly (low-value high-volume biomass) as the raw feedstock. If it is a plant-based material, ideally it should be available all year, instead of being a seasonal crop [11]. Thus, researchers came with ideas of utilizing agricultural residues and industrial by-products as fermentation feedstock to produce microbial oils. Industrial by-products such as glycerol, molasses or animals fats,

• Cellulose is degraded into 5-hydroxymethylfurfural (5-HMF) • Hemicellulose is degraded into furfurals, 5-HMF, which may further

be degraded into weak organic acids (formic acid, acetic acid, levulinic acid)

43

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

Table 1 Summary of pretreatment methods of lignocellulosic biomass [7,60–62]. Pretreatment method

Main action

Mechanical

Liquid hot water

Alkaline

Alters lignin structure, remove most of the hemicelluloses Alters some part of lignin structure, remove most of the hemicelluloses Hydrolyze cellulose and hemicelluloses, alters lignin structure Remove lignin and increase the surface area

Organosolv

Hydrolysis of lignin and some hemicelluloses

Wet oxidation

Reduce celluloses crystallization, remove lignin and some hemicelluloses Dissolution of (hemi)cellulose and lignin Degradate of hemicelluloses alters lignin structure

Dilute acid Strong acid

Ionic liquids Steam explosion Ammonium fiber explosion CO2 explosion Ozonolysis Biological

Remove lignin and hemicelluloses, decrystallize cellulose Remove hemicelluloses, decrystallize cellulose, does not alter lignin, increase the surface area Reduces lignin Remove/degrade hemicelluloses and lignin

Advantages

Disadvantages

No inhibitor or residue reduces cellulose crystallinity, applicable to all biomass feedstocks High sugar yield, low cost, applicable for all samples, applicable to pilot scale Suitable for low lignin content biomass, applicable to pilot scale Applicable for various biomass feedstock, applicable to pilot scale High sugar yield, low inhibitor formation, low investment cost High sugar yield, low inhibitor formation no residue, applicable to pilot scale No inhibitor or residue formed

High power consumption

Applicable to various samples High sugar yield, applicable to pilot scale, no residue formed, cost-effective High sugar yield, no inhibitor formed High sugar yield, no inhibitor or residue formation, cost-effective No residues Low energy requirements

• Lignin is degraded into phenolics, such as vanillin, syringaldehyde, 4-hydroxybenzoic acid, and ferulic acid. • Other derivatives from plant cell wall such as terpenes, tannins and

Inhibitor formed, long hydrolysis time Inhibitor and residue formed Inhibitor and residue formed, corrosive to the equipment, high cost, toxic Time-consuming, salt formation Solvents have to be drained and recycled, high cost, solvents used may inhibit cell growth High equipment cost, not applicable for pilot scale Inhibitor formed Incomplete destruction of the lignocellulosic matrix, inhibitor formed High cost, unsuitable for high lignin content biomass The high cost of equipment, not applicable for pilot scale Require large amount of ozone, high cost The slow rate of hydrolysis

instead [74]. Proteome analysis by Liu et al. [75] found that L. starkeyi was able to grow well (30 g/L biomass) with a lipid content of 46% after 96 h cultivation.

Hibbert ketones

6.1. Carbon and nitrogen uptake If the liquid product after pretreatment is being used as fermentation medium, detoxification is usually needed to remove inhibitors which may hinder the growth of microorganisms. Several detoxification methods commonly used are summarized in Table 2. Since this step will add up to the cost of biodiesel production, finding a microorganism that is resistant or has a high tolerance to inhibitors is important.

As can be observed in Table 3, most of the low cost-feedstocks used for L. starkeyi fermentation were hydrolysates of agricultural residues and wastewater. Angerbauer et al. [79] pioneered using sewage sludge with the addition of sugar as a feedstock for L. starkeyi fermentation and resulted in a lipid content of 75.2%, which was among the highest lipid contents ever reported for DSM 70295 utilizing waste. Following their result, other researchers also worked on different kinds of wastewater to grow L. starkeyi. Unfortunately, the lipid contents obtained were not > 25% [82,85,86,88,91,92,110], which is considered low, regardless of the yeast strain. Although resulted in low lipid content, it was found that the yeast could consume some organic materials present in the feedstock thus reducing BOD, protein or TOC content of the wastes. Other researchers utilized cheese whey [91] and potato starch wastewater [92] but lipid contents were < 5%; even with the addition of other nutrients, the highest lipid content was only 8.88% [92]. In contrast to previous results on utilizing wastewater as the medium, Tsakona et al. [96] utilized flour rich wastewater to grow strain DSM 70296 and obtained 40% lipid content after batch fermentation was

6. Nutrients assimilated by L. starkeyi Agricultural hydrolysates listed in Table 3 were mostly prepared by dilute acid hydrolysis [64–67], enzymatic hydrolysis [68], AFEX [20], steam explosion, or a combination of multiple steps [69,70] to produce fermentable sugars. The choice of treatment method (Table 1) used depends on the characteristic of raw material being processed. Different kinds of feedstock have been studied for culturing L. starkeyi. This yeast was reported to be able to assimilate a wide range of substrates, as can be seen in Table 3. L. starkeyi was found capable of consuming whey permeate [71], herbicides [72] and paraquat [73]. But it was not able to digest 1,2-propanediol and consumed its own lipid Table 2 Detoxification strategies [63]. Methods Physical methods Vacuum evaporation Membrane separations Chemical methods Calcium hydroxide Activated carbon adsorption Ion exchange resins Ethyl acetate extractions Biotechnological routes Microbial pretreatment Insitu microbial detoxification

• • • • • • Enzymatic clarification

Summary

Eliminates only volatile compounds (acetic acid, furfural), non-volatile compounds (extractives and lignin derivatives remained) Functional groups attached to the membrane surface to catch the inhibitors required organic phase to remove inhibitors Remove furfural and hydroxymethylfurfural, causing sugar loss (~10%) Effectiveness depends on process variables of activated carbon, cost-effective Remove lignin-derived inhibitors, acetic acid, and furfural, may cause sugar loss, not cost effective Remove acetic acid, furfural, vanillin, 4-hydroxybenzoic acid Lignin degradation by microorganisms after certain incubation time Detoxification employs yeast, fungi or bacteria or recombinant microorganisms expressing laccase or peroxidase; sugar loss depends on detoxification conditions Remove phenolic compounds

44

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

Table 3 Various substrates assimilated by L. starkeyi. Strain

Substrates

Mode

CDW (g/L)

Lipid content (% w/w)

Lipid productivity (g/L day)

Conversion (g oil/g sugar)

Optimum condition

Ref.

AS 2.1560 –

Glucose Molasses Glucose Ethanol Ethanol Ethanol Glucose Sweet whey permeate Glucose Sewage sludge

B f f FBB BF BF f f f f f f f f f f B f f FBB f f f f f f f f FBB f

46 55 35 54 43.1 46.6 10 36.9 68 (35.6) 72.3 (32.2) 75.2 61.5 21.8 37.2 22.4 30 40.3 31.2 62.4 64.9 20.8 21.5 16.47 18.6 80 24.73 52 35.02 43.85 2.3 0.79 2.81 3.27 8.8

3.45 – – – – – – 0.22 0.8 ND ND 2.52 ND ND 0.23 1.68 1.93 0.76 0.23 ND ND 0.07 ND ND 0.83 0.28 2.95 0.65 ND 0 0 0 0 0.06

0.20 – – – 0.07 0.15 ND ND ND ND ND ND 0.11 0.11 ND 0.1 0.16 ND 0.24 ND ND 0.04 0.04 0.02 ND ND 0.2 0.14 0.19 ND

C/N 133

[75] [76]

f

30 – – 153 – – – 6 9.4 9.27 6.91 20.5 9.35 6.16 10.4 11.7 10 14.7 13.3 104.6 5.34 8.6 8.5 8.6 12.5 4.61 25.5 12.32 76 5.26 15.03 4.66 4.26 2.59

f

2

25.3

0.51

f FBB FBB f f f f f f

12.76 27 60.4 22 19.7 12.28 21.69 17.2 8.2

35.65 51.7 60.1 58.3 30.8 47.3 29.5 47 42.7

1.52 2.79 3.42 2.64 1.20 1.16 0.80 1.01 0.32

0.18 0.16 0.17 0.18 0.12 0.13 0.08 ND 0.14

f f FBB B B FBB B B f f f f FBB f f f f f f f f f B

9.1 30.5 109.8 85.4 13.9 21.27 13.9 13.3 8.7 6.25 5.44 11.4 94.6 – – 5.74 17.1 23.5 18.28 24.63 13.6 16.5 21.2

46.2 40.4 57.8 48.9 26.7 32 26.62 17.29 20.3 21 17.6 60.4 37.4 ~40 ~20 50.49 37.3 33.3 54.85 38.07 22.79 30.3 52

0.6 1.44 9.6 ND ND ND ND ND ND 0.33 0.24 ND ND – – 0.58 1.28 1.57 1.44 1.18 ND ND ND

0.15 ND ND 0.10 0.22 0.06 0.21 0.13 ND ND ND ND 0.17 ND ND 0.13 0.12 0.13 0.17 0.14 ND ND ND

f f

9 34.4

32.7 35.9

ND ND

0.14 0.11

– IBPhM y-282 IBPhM y-694 IAM 4753 ATCC 12659 DSM 70295

AS 2.1560 2.1608 2.1390 – NRRLY-11557

Glucose & xylose Glucose

ATCC 12659 AS 2.1560

Olive mill wastewater Glucose Potato starch Wheat straw hydrolysate Glucose

HL

Fish meal waste water

NRRL Y-11557

Glucose syrup Protelan D-xylose MSG waste water Cellobiose, xylose, glucose Xylose Glucose-xylose Cheese whey

CBS 1807 GIM 2.142 AS 2.1560 DSM 70296 JAL 425 JAL 572 JAL 576 JAL 581 GIM 2.142 UCDFST 78–23 – ATCC 56304 AS 2.1560 CBS 1807 CH010 AS 2.1560

DSM 70296 DSM 70296 DSM 70296 DSM 70296 – – AS 2.1560 DSM 70296 NBRC 10381 NRRLY-11557 ATCC 56304

Potato starch wastewater + glucose Glucose Rice straw hydrolysate Xylose GSXF Glucose, mannose Spent cell hydrolysate Glucose, fructose, sucrose Sweet sorghum juice Corncob hydrolysate Willow wood sawdust hydrolysate Crude glycerol Flour rich waste Glucose, xylose Sugarcane bagasse hydrolysate Brazillian molasses Sugarcane bagasse hydrolysate Arundo donax hydrolysate Arundo donax hydrolysate Sorghum bicolor hydrolysate Xylose Xylose Starch solution

–

Raw glycerol Wheat bran hydrolysate Corn bran hydrolysate Glucose Corn stover hydrolysate Glucose Sunflower meal hydrolysate Glucose

DSM 70296

Crude glycerol

NRRL Y-11557 DSM 70296

ND

[77] [78] [78] [25] [71] [79]

C/N 30 C/N 150 C/N 100 C/N 60 G:X 2:1

[80] [81] [82] [83]

C/N 61.2 Nondetoxified 48 h ino 2-staged W/o nutrient Glucose + YE added As C source As N source Cinitial 100 g/L C40 X20 G10 C/N 50 C/N 50, G30X70

[65] [84] [85] [86] [87] [88] [89] [90] [91]

Glucose & ammonium sulfate added

[92] [93]

GSXF: glucose start xylose feed G23M47 C/N 190

[64] [94] [95] [68] [68] [66] [67] [67] [96]

Study 7 Study 10

[97] [97] [98] [99]

Continuous Batch C/N 58

[70] [69]

Two-staged 5% starch soln 10% starch soln

[100] [101] [12] [102] [103]

C/N 72 200 rpm

[20] [104]

Acid pretreated cells, Pederson method C/N 100 Ginitial 120 g/L

[105] [106]

(continued on next page) 45

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

Table 3 (continued) Strain

Substrates

Mode

CDW (g/L)

Lipid content (% w/w)

Lipid productivity (g/L day)

Conversion (g oil/g sugar)

Optimum condition

Ref.

DSM 70296

Glucose, xylose

48 44 12.3 51.2 49 36.3 64.3 62.3 58.9 64.4 17.8 25.4 34.5 26.1 27.3 57.4 60.1

ND ND ND ND ND ND 3.58 4 3.83 3.43 0.04 0.20 ND ND ND 3.36 3.36

ND ND ND ND ND ND ND ND ND ND 0.07 ND 0.11 0.14 0.18 0.17 0.17

[107]

Glucose

13 10 24.7 30.7 29.6 43.9 16.7 19.3 19.5 16 2.7 9.5 10.9 9.6 11.5 188.2 296.6

C/N 50, D 0.03/h C/N 50, D 0.06/h

Y-11557 Y-27495 Y-27494 Y-27493 AS 2.1560

B B f f f f f f f f f f f B C FBB FBB

C:N 80

[112]

f f f f FBB FBB

14.2 18.3 9.31 5.34 13 15.63

32.3 30.9 19.7 19.9 60.47 51.3

ND ND ND ND ND ND

0.17 0.17 ND ND 0.11 0.10

[113]

f f f f

9.32 9.43 13.5 23.1

50.4 56 52.59 56.11

ND ND 1.92 2.60

0.16 0.18 0.24 0.21

C/N 110 10% inoculum ADH 50% ADH 100% Xylose-acetate feeding Birchwood hydrolysate feeding G:X 2:1 G:X 1:2 1-staged Resuspension in DRBH (2staged)

NRRL Y-11557

DSM 70296 ATCC 56304

Myristic acid Palmitic acid Stearic acid Oleic acid Olive mill waste water (OMW) OMW + glucose Glucose Sugarcane bagasse hydrolysate

AS 2.1560

Glucose & xylose Biphasic (glucose then xylose) Crude glycerol

–

Arundo donax hydrolysate

CBS 1807

Glucose

ATCC 58680

Glucose & xylose

BCRC 23408

Rice bran hydrolysate

[108]

[109]

[110]

[111]

[114] [115]

[116] [117]

f: flask, FBB: fed-batch bioreactor, BF: batch fermentor, B: bioreactor, C: continuous. ND: no data reported.

another review [10].

carried out, and up to 57.8% when fed-batch method was applied. The difference in lipid content was probably due to the origin of wastewater with different nitrogen contents. Other wastes which had been explored for fermentation include molasses and crude glycerol. Vieira et al. [98] used molasses as the medium for yeast growth and was able to obtain a lipid content of 32%. Since molasses contains high protein, making it an optimal medium for biomass growth but not for lipid accumulation [98]. Addition of sugars could be one strategy to get high lipid accumulation, but this adds to an extra cost. Meanwhile, lipid contents of 30–50% were obtained by utilizing crude glycerol as the medium [67,102,106]. The discrepancy in lipid content was probably caused by the difference in composition of the feedstock used or could be due to the ability of different strains to consume certain medium. Viewing the results of biomass and lipid content of the yeast grown in oil mill wastewater (OMW) in Table 3 [82,110], it is clear that OMW was rich in nitrogen and contained less sugar, thus low lipid content was obtained. This is supported by the experiment done by Dourou et al. [110] which reported that after adding 30 g/L glucose, both dry cell weight and lipid content increased. It also showed that the presence of OMW did not give a positive effect on growth [110] since when L. starkeyi was grown in basal medium and glucose (without OMW) it yielded more biomass and higher lipid content. Also, L. starkeyi was not able to consume phenolic compounds in OMW since no color changes were observed [110]. Parameters such as lipid content, lipid productivity, and conversion are important criteria in selecting a yeast strain. Unfortunately, high lipid content and high biomass are usually not concomitant. Therefore, fermentation condition has to be adjusted to obtain high biomass as well as high lipid content, since lipid productivity is proportional to the biomass obtained (as described in Eqs. (1)–(4)), which in turn depends on the composition of nutrients available in the medium. Some differences in terminology may be used by researchers, as summarized in

Lipid content = weight of oil/dry cell weight biomass × 100%

(1)

Lipid production = %lipid content × biomass concentration = lipid titer = lipid yield (in g/L) Lipid productivity = lipid production/time × 24 h (in g/L. day)

(2) (3)

Lipid conversion = lipid coefficient (in g lipid produced /g total sugar consumed)

(4)

Lipid contents of yeasts grown in different hydrolysates were from 20.3 to 56%, with the highest value being obtained from rice bran hydrolysate [117]. Taking into consideration of the volume and its availability, wastes of edible plants such as rice bran, rice straw, corncob, and sugarcane bagasse actually have higher prospects in being used as the primary feedstock if microbial oil production is to be commercialized since those wastes were harvested along with the edible part of plants. These wastes residues could be utilized as fermentation feedstock. Thus the extra land is not needed to cultivate them. In addition to its low cost and resulted in good lipid production, better utilization as fermentation substrate is also seen as a help in solving waste disposal. In view of lipid conversion, theoretical lipid converted from glucose sugar in the medium is 0.34 g/glucose consumed and 0.33 g lipid/xylose consumed [118]. In Table 3, the highest lipid conversion (0.24 g lipid/g sugar consumed) was obtained when rice bran hydrolysate was used as the fermentation medium [117]; followed by sugarcane bagasse hydrolysate (0.22 g lipid/g sugar consumed) [97]. This value is seen to be the best lipid conversion that can be attained when ideal condition (chemostat) was employed [46]. Taking this into consideration, rice bran and sugarcane bagasse have the biggest potential to be used as the main feedstock for fermentation. In sugarcane-producing countries, 46

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

other hand, Zhao et al. [80] achieved a lipid content of 61.5% using a glucose-xylose ratio of 2:1 after optimizing the medium. The difference in lipid content probably was due to the difference in yeast strain used which might affect the metabolism of sugar uptake. Yeast was grown in mixed glucose-xylose medium resulted in higher lipid content than that grown in sole glucose, but lower than that grown in sole xylose, most likely due to poor oil production efficiency [112]. Observation on consumption of mixed sugars found that glucose was consumed first (due to glucose repression mechanism), followed by xylose or other sugars present in the medium [80,89,112,116]. Glucose repression occurs because glucose can repress the utilization of other sugars via catabolite repression mechanism [121]. Contrary to those findings, Hu et al. found that lipid produced by Trichosporon cutaneum was much higher when it was grown in glucose alone; while there were no significant changes in lipid produced when growing in mixed glucose-xylose at different ratios (2:1, 1:1, 1:2) [122]. Also, T. cutaneum was able to consume both sugars simultaneously and no diauxic growth was observed [122]. Diauxic growth sometimes can be observed in microorganism grown in multiple sugars in the presence of glucose, and lag phase occurred in between growth phases [123]. In the case of using other mixed-sugars as feed like cellobiose-xylose [89], L. starkeyi consumed both sugars simultaneously rather than sequentially. This phenomenon was also observed when L. starkeyi was fed by glucose-mannose mixture [95]. Since glucose is beneficial to cell growth and xylose seems to promote lipids production. Therefore Probst et al. [112] employed biphasic sugars as the medium. Glucose was provided at the start of fermentation and xylose was given after glucose was depleted in fed-batch fermentation which resulted in 60.1% lipid content [112]. This value is lower than the 80% obtained by sole xylose medium used by Oguri et al. [87], but using xylose as the only carbon source did not support cell growth well (only 2 g/L dry biomass obtained), leading to low lipid productivity [87]. Meanwhile, in biphasic fermentation, biomass could grow well by firstly consuming glucose, and high lipid content and productivity can be achieved by later addition of xylose. From a biological perspective, glucose metabolism results in 1.1 mol of acetyl-coA per 100 g of glucose utilized [118,124]. For xylose, after it is transformed into xylulose then xylulose-5-phosphate [59], it undergoes two common metabolic pathways in yeast and bacteria: pentose phosphate pathway and phosphoketolase pathway [59]. Xylose metabolism through phosphoketolase pathway (the most efficient pathway) produces 1.2 mol acetyl-coA per 100 g xylose consumed, whereas, metabolism through pentose phosphate pathway results in 1 mol [118,124]. Therefore, when all glucose or xylose in the medium was assimilated and channeled into lipid accumulation, the theoretical lipid conversion is 0.34 g/g glucose, and 0.33 g/g xylose consumed [118]. This shows that L. starkeyi could give high lipid content, lipid productivity, as well as lipid conversion if it is grown in a suitable medium, without the need of doing metabolic engineering. There is a possibility to find a low-cost medium using this idea for this yeast to perform as good as it was in a biphasic medium since biphasic medium consists of the pure glucosexylose mixture is expensive especially if it will be used in large-scale/ commercialized process. Another report on feedstock assimilation utilized cellobiose for culturing AS 2.1560 and produced 50% lipid content [89], but when CBS 1807 was used, lower lipid content of 33.3% was obtained [87]. This supports the assumption that each strain may have a preference for sugar it can assimilate which may affect its metabolism towards lipid accumulation. It should be noted that L. starkeyi strains DSM 70295/ CBS 1807, DSM 70296, AS 2.1560 and ATCC 56304 are among those yeast strains that yielded high lipid content (40–80%), although there were some cases that gave lower values probably due to different sugar types or nitrogen contents in the medium that may influence the metabolism of the yeast. An interesting experiment was done by Brandenburg et al. [115] to test the ability of L. starkeyi CBS1807 to assimilate xylose-acetic acid

sugarcane bagasse is available abundantly, and it could also be used as fermentation feedstock (along with bioethanol process through biorefinery). Other paddy residues (rice straw, rice husk) could also be explored as fermentation medium to grow L. starkeyi, since rice is the staple food for the majority of people in Asian countries. Therefore paddy residue must be available in abundance. Perhaps, it may be more profitable if we could utilize multiple feedstocks and integrate the processes [54] to obtain hydrolysate for culturing yeast. To do so, characteristic of each feedstock should be taken into account. For example, biomass with more lignin content should not be mixed with that having less lignin content, to ease hydrolysis treatment. From the viewpoint of the basal medium, lipid content of yeasts grown in basal medium ranges from 10 to ~83% (Table 3). This huge gap seems to be related to medium composition, type of sugars that can be assimilated by the yeast, fermentation mode, as well as the yeast strain. For example, CBS 1807 was able to grow in arabinose, mannose, and galactose resulting in lipid content of 63%, 54%, 51.72%, respectively [87]. But L. starkeyi strain DSM 70296 [99] and CH010 [66] were not able to assimilate arabinose. Meanwhile, Gong et al. [89] reported that cellobiose assimilation by L. starkeyi AS 2.1560 resulted in more biomass and higher lipid conversion compared to glucose or xylose assimilation at the same initial concentration. Another point worth discussing is sugars metabolism by yeast. Basal medium with glucose as the carbon source gave low lipid content (21–37%) and productivity [81,83,93], most likely due to low C/N molar ratio of the medium which was not suitable for lipid accumulation. Comparing those with data obtained by other groups [6,15,64,91], high lipid content (> 51%) can be achieved most probably due to the medium that induces lipid accumulation together with the fermentation mode used, despite using the same sugar-glucose as the carbon source. Similarly, when using xylose as the sole carbon source, Anschau et al. got 37.4% lipid content in fed-batch fermentation [101], while flask batch experiment by Oguri et al. could achieve ~80% lipid content [87]. This again may be due to the composition of the medium that was not suitable for lipid accumulation, i.e., high nitrogen content. It is interesting to note that L. starkeyi ATCC 58680 grown in sole glucose and sole xylose resulted in similar biomass amount, while lipid content was higher in sole glucose case [116]. On the other hand, using mixtures of sugars as the fermentation medium resulted in high lipid content (44–61.5%) [68,89,90,94,95,97,107,116]. Fermentation using mixed sugars as the carbon source was done as a comparison and to simulate actual fermentation employing newly-created-hydrolysate (such as spent cell hydrolysate [95,119], or sweet sorghum stalk [68], or hemicellulosic hydrolysate). Most mixed sugars fermentations were carried out by mixing glucose and xylose at different ratios since agricultural residue mostly comprises cellulose and hemicelluloses which break down into glucose and xylose after pretreatment (other sugars may be found in small amounts depending on the biomass used). By utilizing mixed glucose-xylose incorporated into the basal or minimal medium, high lipid content (> 44%) can be attained, although the experiments were carried out with different yeast strains and cultivation modes [80,90,94,97,107,112,116]. To simulate the composition of a hemicellulosic hydrolysate of sugarcane bagasse, Anschau et al. [97,107] (Study 6) and Tapia et al. [90], carried out fermentation of basal medium with mixed glucose-xylose 30:70% as C source. The lipid content obtained was slightly different (by ~5%) despite using the same strain of yeast and the same fed-batch mode. This perhaps was due to the difference in the amount of feeding pulse given as well as the composition of the feeding pulse. In contrast to previous data, according to Ye et al., sugar solution in hemicellulosic hydrolysate produced after hydrolysis, was a mixture of hexose and pentose- mostly glucose and xylose with a typical ratio of 2:1 [120]. Rahman et al. [116] carried out fermentation in the basal medium of mixed glucose-xylose. At a glucose-xylose ratio of 1:2, lipid content was higher than when the ratio was 2:1 (56 vs. 50.4%). On the 47

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

Pirozzi et al. [114] initially grew L. starkeyi in 50% Arundo donax hydrolysate (ADH), biomass was collected and some were used as inoculums for the next cycle carried out in 75% ADH. The biomass was again collected, and some were used as inoculums for fermentation in 100% ADH. This subsequent adaptation was successful for growing L. starkeyi in ADH without prior detoxification and resulted in 5.34 g/L biomass and 19.9% lipid content. This lipid content was the same as that obtained by 100% ADH after activated carbon detoxification (4.04 g/L biomass). In employing 50% ADH (~40 g/L initial reducing sugar) as a medium in the first cycle of fermentation, 9.31 g/L biomass with 19.7% lipid content was obtained, which was the highest biomass achieved compared to other conditions tested. Taking into consideration that when 100% ADH was used (with ~80 g/L initial reducing sugar), and less biomass was obtained, it was probably due to the high osmotic pressure that suppressed the yeast growth, as reported by Lin et al. [84]. If this was the reason, the utilization of 50% ADH was better since it could produce twice more lipid when all the hydrolysate was utilized. Xavier et al. [111] successfully grew L. starkeyi DSM 70296 in hemicellulosic hydrolysate of sugarcane bagasse without detoxification (contains 2.6 g/L acetic acid, 0.04 g/L HMF, 0.06 g/L furfural), and obtained a lipid content of ~27%. It was reported that L. starkeyi could tolerate and consume inhibitors during fermentation, at least in low concentration where inhibitors were not in toxic level for cell growth. This fact supports the previous finding that acetic acid can be metabolized to acetyl-coA, which holds an important role in FA biosynthesis [111]. They claimed that it increases the feasibility of sugarcane bagasse hydrolysate as a substrate for fermentation [111], which supports the idea that sugarcane bagasse could be one of the best candidates for producing lipid by looking at its lipid conversion-discussed in Section 6.1. Since sugarcane is not available abundantly everywhere, other lignocellulosic residues are still needed to support SCO production. A study on the effect of inhibitors on the growth of L. starkeyi suggested that concentrations of PHB (para-hydroxy benzaldehyde) and vanillin should be limited to ≤0.5 g/L; for HMF and syringaldehyde, concentration should be ≤1 g/L and ≤ 0.4 g/L for furfural [116]. Other inhibitors such as vanillin, formic acid, and levulinic acid were reported to lower both biomass and lipid content of L. starkeyi 2.1608 and 2.1390 [81]. In general, inhibitors formed after pretreatment of lignocellulosic biomass are usually, acetic acid, 5-HMF, and furfural (PHB and syringaldehyde are found in smaller amounts depending on the lignocellulosic biomass used and the pretreatment condition). Taking into consideration the effect of hydrolysis by-products on fatty acids profile of microbial oils, the presence of furfural seems help converting unsaturated fatty acids into saturated ones. 5-HMF and PHB seem to favor oleic acid production, whereas syringaldehyde and vanillin do not change fatty acid profile significantly [116]. The ability of L. starkeyi to consume a certain level of acetic acid, and tolerate the presence of 5HMF and furfural to some extent, shows that L. starkeyi has a good tolerance towards inhibitors and could be used for fermentation in hydrolysate without detoxification. However, this also depends on the severity of pre-treatment method. Other issues regarding microbial oil implementation are the disposals of large amount of water after collection of biomass and defatted cell mass. These concerns may be tackled by reutilizing the spent cell mass [95,119] and the spent water for the next fermentation process, as proposed by Yang et al. [119]. Although it was only limited up to 3 cycles [119], it can save water and make the fermentation process more economical.

mixture. Initially, the yeast was grown in birch wood hydrolysate with different dilutions (diluted with fermentation medium without additional carbon source), and growth was suppressed in the presence of ≤40% hydrolysate. In this case, perhaps it should be better to use birch hydrolysate alone as fermentation medium to obtain information about the feedstock. The authors [115] also carried out pH-regulated fedbatch cultivation (with glucose in batch phase) with two kinds of medium for the feeding pulse: xylose-acetic acid mixture (to simulate birch hydrolysate) and the hydrolysate itself. The highest lipid content was attained from xylose/acetate mixture (60.1%) while the highest biomass was obtained when using hydrolysate as the feeding pulse (15.63 g/L) due to more nutrients available in the hydrolysate. This result suggests that it is unnecessary to pre-cultivate L. starkeyi in xylose solution to obtain pH-regulated lipid production from lignocellulosic hydrolysate [115], and shows the superiority of L. starkeyi to produce lipid using various feedstock. Apart from basal medium or agro-industrial residue hydrolysates, L. starkeyi was also reported to be able to assimilate fatty acids and glycerol and resulted in non-negligible lipid content, depending on other nutrients added in the fermentation medium. Lipid biosynthesis in fermentation using substrates like fatty acids and glycerol follows ex novo lipid biosynthesis, in which hydrolysis of fatty acids should occur for the yeast to assimilate it. Using myristic acid (C14:0) and oleic acid (C18:1) as the carbon source, AS 2.1560 gave high lipid content (58–64%) with C14:0 and C18:1 as the predominant fatty acids in the lipid produced. However, if a stearic acid (C18:0) was used as the feedstock, the fatty acids in the lipid obtained was mostly palmitic acid (C16:0) and oleic acid [109]. In the case of using glycerol as the carbon source, lipid content ranged from 28 to 50%, regardless of the yeast strain. This could be due to the composition of the medium used, especially nitrogen. Liu et al. [113] found that organic nitrogen was more beneficial than inorganic nitrogen (NH4Cl and (NH4)2SO4) in promoting cells growth and lipid production. This was also observed by Matsakas et al. [68], and Azad et al. [64] in that addition of organic nitrogen (yeast extract or peptone) could help promoting cell lipid content. On the other hand, Zhao et al. [80] found that the addition of NH4Cl gave the highest lipid content, followed by addition of yeast extract in spite of employing the same strain of AS 2.1560 as that of Liu et al. [113]. This perhaps was related to other nutrients added in each fermentation medium such that synergistic effect may occur, thus affecting the lipid metabolism of yeast. The choice of nitrogen source for fermentation may affect lipid biosynthesis in the sense that nitrogen metabolism differs from one yeast species to another. Calvey et al. found an interesting phenomenon., in that urea was not a good N source for L. starkeyi because urease enzyme in L. starkeyi would hydrolyze it into ammonia at a faster rate than L. starkeyi could assimilate, resulting in a build-up of base in the medium and leading to retarded cell growth and premature death [20]. More researchers should be done to determine the most suitable nitrogen source for L. starkeyi in lipid production. 6.2. Necessity of detoxification for agricultural residue hydrolysate An important aspect that needs to be considered in producing hydrolysate is inhibitors. Methods commonly employed to detoxify hydrolysate are Ca(OH)2 overliming and activated carbon adsorption, as explained in Section 5. Chen et al. [81] reported that L. starkeyi strain 2.1608 was able to assimilate nutrients and accumulate lipid in the presence of up to 1 g/L HMF without significant effect on biomass produced. But the result could be different if other inhibitors are also present in the medium. Yu et al. [65] reported that L. starkeyi ATCC 12659 was resistant to HMF, furfural and acetic acid up to 0.05, 0.44 and 4 g/L, respectively, as biomass and lipid content obtained from non-detoxified wheat straw hydrolysate was not significantly different from those of the detoxified hydrolysate. The lipid content obtained was 31.2% and the productivity was 0.77 g/L day [65].

7. Factors affecting fermentation Fermentation of oleaginous yeasts under limited nutrient condition usually consists of three phases: the exponential phase where cells proliferate rapidly; lipid accumulation phase-where cells show minimum growth; and the stationary phase where lipid breakdown may 48

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

the medium. Average pH used in literature was 5–6.5, but it seems that L. starkeyi can grow well in a wide range of acidic pH, but not higher than 6.5 [79]. Angerbauer et al. found that the highest lipid accumulation of L. starkeyi grew in sewage sludge was 56% at pH 5, increasing pH to 6.5 led to a lower lipid content of ~50% with a corresponding higher biomass [79]. This suggests that lower pH which benefits lipid accumulation may be detrimental to cell growth. This phenomenon remains uncertain, and further studies should be conducted. pH 5 was also found to be optimal for L. starkeyi grown in MSG wastewater [88]. When using crude glycerol [113] and rice straw hydrolysate [64] as the medium, optimum pH for lipid accumulation was 6. In another study, L. starkeyi HL was reported to grow and accumulate lipid well at pH 4 in fishmeal wastewater medium [85]. In a defined minimal medium, pH of the fermentation medium was decreased from 5 to ~2.8 at the end of cultivation, and 54.85% lipid content was achieved at C/N = 72 [20]. Being able to live in slightly acidic pH is an advantage since it can suppress the growth of microbial contaminant which is important for industrial microbiology [11]. Based on those data, it seems that L. starkeyi can grow and accumulate lipid in a wide range of pH, which shows the potential of L. starkeyi as an oil producer; since the use of yeasts that is able to produce high lipid content despite of pH fluctuation would be preferable [10]. In addition, if it can be grown in slightly acidic condition, less base would be required for neutralizing hydrolysate, which means reducing the operation cost. Aeration/dissolved oxygen is important during fermentation, especially for yeast growth. Since higher oxygen transfer rate facilitates cell proliferation, increasing shaking speed leads to increase in dry cell weight/biomass. On the other hand, lipid accumulation works well in oxygen-limited medium since this condition blocks glucose oxidation pathway thus induces lipid accumulation [20]. This is in agreement with the report that lipid content found in Rhodotorula glutinis was higher at 25% dissolved oxygen (DO) while biomass was higher at 60% DO [126]. While oleaginous yeast has the advantage to be able to live in a broad range of temperature (10 to ~25 °C) with high total lipid produced, however, if the fermentation is carried out at large scaleextra cost will be needed for chilling [10]. Most fermentations were carried out at 30 °C [20,79,80,83,84,88,89,97,100,104], although there are some reported L. starkeyi fermentation at 28 °C [65,66,99] and 25–26 °C [103,104]. Lastly, concentration and age of inoculums are also considered as important factors in fermentation. Inoculum concentration employed for fermentation is usually between 5 and 10%. Higher inoculum concentration will dilute rate-limiting substrate concentration in the medium, leading to lower specific growth rate, while fewer inoculums would result in lag phase [127]. According to studies carried out in different fermentation mediums: potato starch wastewater [92], MSG wastewater [88] and crude glycerol [113], the optimum inoculums concentration to promote lipid accumulation was 10%. If higher inoculum was used, more nitrogen was introduced to the medium thus increasing N content which was not favorable for lipid accumulation [113]. In terms of inoculums age, the inoculum was usually taken at its exponential phase, and it may be different in each strain used. Investigation on inoculum age by Liu et al. found that older inoculum seems more mature and more potent in assimilating substrate and lipid accumulation than younger inoculum [84]. During fermentation, the sample is taken at certain times to check biomass growth, sugar and nitrogen left. Many data reported in literature did not monitor nitrogen content left in the medium. Ideally, nitrogen presents in the medium should be in the right amount to allow yeast to grow well. Since intracellular lipid accumulated is the actual food storage of oleaginous yeast, once nutrients are depleted in the medium, yeast will start to consume its own lipid (so-called ‘lipid turnover’). Thus, fermentation should be terminated rapidly after sugar is exhausted, approximately in the early stationary phase. Although L. starkeyi was reported to reutilize less lipid from its own than other oleaginous yeasts [128], yet it would be more profitable when it is

occur [125]. Some studies reported that there was a lag phase during fermentation of L. starkeyi [68,82,83]. Unlike other oleaginous yeasts that usually grow easily, L. starkeyi tends to have difficulty in growing, particularly if the medium is different from that of pre-culture. This lag phase will extend the cultivation time leading to low lipid productivity. A few ways may be used to reduce the lag phase such as slightly increasing inoculum concentration and preparing inoculum in the same medium as that will be used for lipid production. Important factors affecting fermentation includes the cultivation condition, along with the choice of cultivation mode that will be described below. 7.1. Cultivation condition Sitepu et al. [10] summarized factors affecting lipid accumulation which include yeasts species and strains, stage of growth, carbon, and nitrogen source, C/N ratio, other micronutrients, temperature, aeration/dissolved oxygen, pH, the presence of alcohol, and acclimatization to a carbon source. Some of the crucial aspects affecting lipid content of particular yeast will be discussed here include C/N molar ratio, aeration, pH, and temperature. As described previously, when balanced nutrient (low C/N) is available in the medium, biomass keeps proliferating, and when nitrogen content started to deplete, the remaining carbon source is channeled for lipid accumulation, as described in Fig. 1. Often, high C/ N ratio will result in low biomass with high lipid content (> 40%), though it is not always the case. For example, in batch mode fermentation, using glucose as the only carbon source in basal medium, 40% lipid content was obtained at C/N = 60, and lipid content of 68% was obtained at C/N = 150 [79]. When the composition of the basal medium was modified, Wild et al. [83] got only 30% lipid content at C/ N = 61.2. But when mixture of glucose, sucrose, and fructose was used in basal medium at C/N = 100, only 43.5% lipid was accumulated, and the highest lipid content of (47.3%) was obtained at C/N = 190; further increase of C/N ratio up to 250 resulted in declined lipid content as well as lipid titer [68]. The latter phenomenon was presumably because of too much sugar in the medium leading to high osmotic pressure [84] that stressed the yeast cell, even caused the burst of the cell. Another investigation by Calvey et al. [20] revealed that when yeast was grown in a defined minimal medium with C/N = 72, lipid content could reach 54.85%. When the medium favors lipid production, which means high sugar content (imbalance condition), it is unfavorable for biomass growth, since biomass growth needs a suitable sugar proportion (balance nutrients). Thus usually high lipid content is accompanied by low biomass produced, but this is not the case in a study which reported that 18.28 g/L dry cell weight was obtained with 54.85% lipid content at C/N = 72 [20] which is considerably higher than the 11.7 g/ L dry cell weight (30% lipid content) obtained at lower C/N of 61.2 [83]. This was probably because of difference in the initial concentration of sugar used. Higher concentration of sugars will lead to more biomass proliferate before it enters the stationary phase to accumulate lipid. It should be noted that the discussion above applies to the same strain (just different collection numbers stated in each article). Data reported on utilizing Arundo donax hydrolysate [70] showed low lipid accumulated (20.3%) at C/N = 58. Even after the addition of 3 cycles of ammonium sulfate to facilitate yeast growth, the highest biomass obtained was < 10 g/L. Unfortunately, information regarding initial sugar concentration and which strain of L. starkeyi used was not provided. Thus, it is important not to generalize the optimum C/N ratio as it would change with the medium and the strain used. In a biphasic fed-batch study, 60.1% lipid content was obtained using a minimal medium with C/N = 80 [112]. While in a study employing crude glycerol as the fermentation medium, the highest lipid yield and lipid coefficient was achieved at C/N 60 [113]. Based on those data, optimum C/N ratio for L. starkeyi seems to be in the range of 58–190, depending on the composition of the medium used. The next important aspect of fermentation is the pH and aeration of 49

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

polar solvents should be used for lipid extraction. Most researchers prefer to use chloroform and methanol for extraction. Since chloroform is toxic, it would be better to replace it by other solvents that are as effective. Other common solvent used is hexane, ethanol and diethyl ether [135]. The majority of oil globules found in the visualization of L. starkeyi IAM 4753 was TAG (85%), other minor parts were polar lipids (95% phosphatidyl ethanolamine) and protein [136]. Naganuma et al. quantified neutral lipid using a tower-shaped mixer and TAG assay kit. After cell disruption, they found that TAG was indeed the major component of lipid while phospholipids, FFA, ergosterol, and DAG were in the lesser amount [137]. On the other hand, Suzuki et al. (1973) investigated various lipid extraction methods on freeze-dried cells of L. starkeyi IFO 0678 and found that HCl treatment and Pederson's method gave the highest lipid content since those methods can break phospholipids and release fatty acids. Pederson's method (stirring the cells for 3 h in 40% chloroform: methanol (1:1), repeated 3 times) was found to be the best method in mild condition for extracting lipid [138]. Unsaturated fatty acids in the TAG, FFA and phospholipids were found to decrease at the latter stage of fermentation [141]. That result is slightly different from that found by Deinema and Landher [76]. Extending the fermentation up to 3 weeks resulted in an increase of oleic acid from 55% to 62%, but the little reduction of other fatty acids was noticed. They reported that FAs composition was similar throughout different fermentation phases (only oleic acid had significant increment) and no significant changes in the total amount of saturated FA (33.5%) and unsaturated FA (63%) was observed, regardless of cultivation stages.

harvested at the highest lipid content. 7.2. Cultivation mode Cultivation mode holds an important role in optimizing lipid productivity during fermentation; this will have a direct economic impact on the process [44]. Cultivation modes commonly used are the batch mode, continuous mode or fed-batch mode. In a batch operation, cells are cultured in a fixed volume of fermentation medium under specific conditions (nutrient type, temperature, agitation etc.). Many of the studies in literature performed batch operation since it allows researchers to determine the optimal condition for lipids accumulation [44]. In batch cultivation, C/N ratio holds an important role for bioprocess performance. Thus, initial C/N ratio should be optimized to maximize lipid production [129]. Basically, balanced-nutrient (low C/N ratio) favors cells growth while imbalanced nutrient (high C/N ratio) induces lipid accumulation. However, extremely high C/N ratio may result in the production of other metabolites such as organic acids and may cause a decrease in cells viability [129] thus become inhibitory to the culture. Therefore, batch mode is not feasible to achieve high-cell density culture for lipid accumulation [44]. In continuous mode, the fresh nutrient is continuously supplied to a well-stirred medium where products and cells are simultaneously withdrawn [44]. Fermentation was initially run as batch mode (constant C/N ratio at a given dilution rate) for certain time to condition the system before it starts as continuous mode. Under this condition, lipids accumulation highly depends on dilution rate (D) and C/N molar ratio of the culture medium [130]. A low D induces substrate utilization and lipid production. To maximize lipid production, both C/N ratio of medium and D have to be properly adjusted [129]. D < 0.06/h is usually required for optimal conversion [44]. On the other hand, in fed-batch mode, fermentation is prolonged by intermittent or continuous feeding of nutrients [44], which allows some control over specific growth rate and the flows of nitrogen and carbon utilization [131]. Thus it is possible to maintain microorganisms in the optimal metabolism state for growth, and lipogenic phase at a later stage for lipid production under limited nitrogen condition with constant C/N ratio to prevent acids production [131]. This mode is usually used to overcome substrate inhibition or catabolite repression by intermittent substrate feeding, thus improves fermentation productivity [44]. There are many studies [77,84,132] reported using fed-batch mode for microbial lipid production which resulted in high biomass concentration, lipid content and lipid productivity. After fermentation is terminated, biomass should be harvested (separating spent nutrient from biomass) rapidly to prevent lipid turnover that will reduce lipid production. Common methods to separate wet cells from spent nutrient include centrifugation, ultrafiltration, sedimentation or flocculation [133]. This biomass harvesting process can contribute 2 to 30% of the total biomass production cost [134]. After separation, wet cells are usually dried before extraction (in laboratoryscale procedure). Since drying is energy intensive, thus it would be of great interest to develop a less expensive method to dry yeast, or directly use of wet cells in in-situ transesterification to produce FAME.

9. Fatty acids (FAs) composition of lipids from L. starkeyi FAs of lipid of a yeast strain is highly dependent on the medium used for growth, and fermentation condition [125]. For example, the degree of saturation of yeast lipids is influenced by growth temperature [133]. Wu et al. [32] showed that higher carbon to sulfur (C/S) ratio favors the production of saturated FAs. Fatty acids profile can be obtained by completely converting fatty acids into its respective methyl esters. Since most yeast oil comprises TAG, a two-step method is usually needed; removing the unsaponifiable matter and methylation of fatty acid soaps. Methylation usually employs BF3-but considering its toxicity, the HCl-methanol method may be a better option. Direct use of an acid catalyst is not recommended since more acid will be needed to convert TAG to FAME. As can be seen in Table 5, the most abundant fatty acids of L. starkeyi-derived oil are C18:1 (oleic) and C16:0 (palmitic), regardless of strain. An exception is the strain AS2.1560 which has a peculiar fatty acid profile in that the most abundant fatty acids are palmitic and myristic with a small amount of oleic acid [85]. This could be due to the medium used which was not suitable for forming the desired fatty acids. Different fatty acids composition (even if the oil is derived from the same strain) of microbial oil might occur because of difference in culture medium and its growth condition [26,142]. It is interesting to note that both cultivation mode and composition of fermentation medium significantly affect FAs composition, as was reported by Wild et al. [83]. Comparing of FAs profiles in Tables 5 and 6 shows that L. starkeyi lipids have a similar composition to that of common vegetable oils, regardless of strain, medium and cultivation modes. This supports the idea that SCO obtained from L. starkeyi can be a potential candidate as a biodiesel feedstock in the future. A few reports on converting SCO into biodiesel are summarized as follows. An early attempt on direct methanolysis of dry biomass gave 98% FAME yield with a cetane number of ~56–59. The reaction was carried out at 70 °C at ambient pressure with the aid of dilute H2SO4 or HCl [145]. Unfortunately, long reaction time (20 h) and energy-intensive caused by drying was required. Another study used direct methanolysis (20 mL methanol/g dry biomass) with NaOH as the catalyst.

8. Extraction and characterization of lipids from L. starkeyi Some methods employed for extracting lipid from oleaginous yeasts are listed in Table 4. Since lipid is an intracellular product, it requires pretreatment to break cell wall before extraction. The pretreatment commonly used is HCl hydrolysis or sonication. The pre-treated yeast is then extracted using organic solvent. Other methods available for extraction of oleaginous yeasts include bead beating, cavitation, ultrasonication and microwave pretreatment [125]. It should be noted that the majority of lipids accumulated by yeast are non-polar (mainly TAGs), but cell wall contains polar phospholipids which can contribute to FAME production. Thus both polar and non50

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

Table 4 Commonly used methods for extraction and transesterification of lipid. Cells pretreatment

Extraction method

Transesterification

Ref.

1 M HCl, boiling water, 1 h – 4 M HCl 78 °C, 1 h

Petroleum ether, soxhlet, 5 h – Chlorofom: methanol (1:1 v/v)

[79] [65] [89]

–

Chlorofom: methanol (2:1 v/v)

– 2 N HCl (1:12.5 w/v) 80 °C, 1 h; enzymatic 30 °C, 3 h & inactivation 80 °C, 30 min 10 mL HCl 2.5 N, 100C, 30 min Ultrasonication 12 min, 50% power, 90% pulser 4 M HCl (1:12 w/v), 80 °C, 1 h Wet cells + conc HCl, 95 °C, 1 h Conc cells + chloroform, methanol & zirconia beads

Chlorofom: methanol (2:1 v/v) Chloroform: methanol (:water); hexane Chlorofom: methanol (2:1 v/v) Chlorofom: methanol (1:2 v/v) Chlorofom: methanol (2:1 v/v) Chlorofom: methanol (1:2 v/v) Chlorofom: methanol: water (1:2:0.8 v/v) Chlorofom: methanol: water [139] Chlorofom: methanol: water [139] Chlorofom: methanol (1:1 v/v) Chloroform: methanol: water [139] 10% methanol in chloroform

– 1 N KOH in MeOH, 1.5 h, 50 °C, methylated with H2SO4 1 mL wet cells +0.5 mL 5% KOH methanol, 65 °C, 50 min, 0.2 mL BF3 diethyletarate +0.5 mL methanol Li Lipid +2 M NaOH in MeOH, 60 °C, 15 min, 2.5 HCl in MeOH, 60 °C 30 min NaOH in MeOH, BF3 in MeOH, consecutively 5 min –

[68] [105]

Methanolic sodium and hydrochloric methanol – MeONa; HCl catalyzed esterification 1 mL anhydrous HCl in MeOH, 50 °C, overnight 3 M methanolic HCl 78 °C, 30 min

[106] [70] [96,104] [20] [103]

0.5 N methanolic sodium hydroxide; BF3-methanol

[102]

MeOH: HCl: CHCl3 (10:1:1 v/v/v)

[97,99]

2% H2SO4 in methanol, 85 °C, 1 h 2 M NaOH in methanol 60 °C, 15 min, 2.5 M HCl in methanol 60 °C, 30 min 1 mL wet cells +0.5 mL 5% KOH methanol, 65 °C, 50 min, 0.2 mL BF3 diethyletarate +0.5 mL methanol Dry cells dissolved in toluene and methanol, transmethylated with 8% methanolic HCl, 45 °C, overnight

[67] [85]

2 M HCl 2 M HCl, 80 °C 1 h 4 M HCl 78 °C, 1 h – Chlorofom: methanol (1:2 v/v) with ultrasonication 12 min, 50% power, 90% pulser –

HCl-catalyzed direct methylation

[88]

[82] [140]

(DU), oxidative stability (OS) and high heating value (HHV)/calorific value [125,133]. Empirical formulas for each property and its description are summarized in a review by Patel et al. [125].

The reaction was held at 50 °C for 10 h and the yield obtained was 97.7% [146]. Although still requires drying, but it was able to convert most TAG to FAME in a shorter time than the previous method. Another method to convert microbial oils into FAME was by hydrolyzing the oils using the hydrothermal method. Hydrolysis was done in a batch stainless steel reactor at 280 °C and an initial pressure of 500 psi for 1 h, the resulted fatty acids can be used for FAME production [147]. This method can eliminate drying and oil extraction steps, and perhaps could be promising for scale up, but the techno-economic study should be carried out before implementation. Anyway, both methods could support the feasibility of producing microbial oil as a biodiesel feedstock in the future.

11. Other metabolites produced by L. starkeyi Calvey et al. first reported the special trait owned by lipogenic yeast L. starkeyi strain NRRL Y-11557 to produce polyols such as mannitol, arabitol, 2,3-butanediol and ethanol [20]. L. starkeyi was also reported to be able to produce glucanhydrolase [148] and extracellular dextranase [149]. 12. Discussion and future prospect

10. Evaluation of biodiesel derived from SCOs With fatty acid composition similar to that of vegetable oils, SCO has the potential to be used as feedstock for biodiesel production, replacing vegetable oils and animal fats. Unfortunately, application of SCO-derived biodiesel has been hindered due to its high production cost, of which 75% came from the raw materials [150], making it not competitive with fossil fuels. The first step to reducing the production cost of biofuel is to utilize biomass wastes, such as agricultural residue

To be used in unmodified diesel engines, biodiesel must meet the regulatory standard given by ASTM D6751 (United States) or EN 14214 (European Union) for automotive fuel [125,133]. Some important properties need to be evaluated for SCO-derived biodiesel include density, kinematic viscosity, flash point, pour point, cloud point, cetane number (CN), cold filter plugging point (CFPP), degree of unsaturation Table 5 Fatty acids profile of various L. starkeyi strains. Strain

C14:0

C16:0

C16:1

C18:0

C18:1

C18:2

C18:3

C20:0

C20:1

C22:0

Ref

DSM 70295 AS 2.1560 – Y-11557 ATCC 12659 AS 2.1560 AS 2.1560 DSM70296 AS 2.1560 AS 2.1560 DSM70296 Y-11557 Y-11557 DSM70296

0.9 0.5 <1

55.93 36.6 19.1 39 36.2 36.5 45.99 38 33.8 35.8 27.8 39.3 0.9

1.85 4.3 0.5 3

13.8 6.2 8.5 3 4.5 5.4 9.37 4.2 11.5 7.5 3.7

25.89 48.9 49.1 55 46.3 52.8 8.99 51.7 49 43.8 48.2 7.8 10.3 5.41

< 0.1 1.1 18.8

0.12

0.48

0.18

0.48

3.5

0.2

11.05

8.38

0.8

0.6

[79] [80] [82] [83] [65] [84] [85] [89] [95] [67] [99] [102] [20] [104]

0.5 11.15 0.7

3.6 0.87 3.7 4.1 4 6 4.3 35.2 29.44

3 5.73

3.4 1.2 4.2 1.3 2.3 4.1 10.1 42.5 46.5 56

51

4.2 2.7 3.4

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

Table 6 Fatty acids profile of common vegetable oils [143,144]. Oils

C12:0

C14:0

C16:0

Palm

0.5 0.1

1.5 0.7 0.9

0.2

0.7

45.5 36.7 40.1 4.9 15.3 11.3 6.9 6.2

Rapeseed Soybean Sunflower

C16:1

0.1

0.1 0.1

C18:0

C18:1

C18:2

C18:3

4 6.6 4.1 1.6 3.9 3.6 6 3.7

38 46.1 43 33 17.8 24.9 26.5 25.2

10 8.6 11 20.4 34.4 53 66.5 63.1

0.5 0.3 0.2 7.9 4.7 6.1 0.2

C20:0

C20:1

C22:0

0.4

0.2

0.1

C22:1

C24:0

0.1

9.3

23

0.3

0.3

0.3

0.1

0.3

0.2

0.1

0.2

0.7

offers integrated biorefineries the possibility to select the most profitable combination of raw materials [152]. Aside from considering the amount and availability, the choice of a fermentation feedstock should also be based on whether it can be fully utilized. For example, rice bran can be hydrolyzed, and the hydrolysate was used as fermentation nutrient while lipids remained in rice bran residue was transesterified to produce biodiesel [117,153]. Similarly, hydrolysate obtained from sugarcane bagasse was used as fermentation medium to produce xylitol and the bagasse residue was used to produce xylanase and SCO which was transformed into FAME [154]. Hydrolysates from the dilute acid hydrolysis of corn stover [155] and wheat straw [156] were also utilized as fermentation medium to produce SCO, while solid residues were used to co-produce bioethanol via simultaneous saccharification and fermentation (SSF) process [155,156]. Biomass residues may be used as cattle feed or feedstock to produce methane by anaerobic digestion [44]. Ghosh et al. carried out experiment adopting biorefinery approach by producing bioethanol, furfural, and electricity [157]. Integrated biorefinery approach as mentioned above allows the effective utilization of lignocellulosic biomass to produce a sustainable product. In general, co-production of oleochemical products such as polyunsaturated fatty acids, lignin and microbe meal (a leftover of lipid extraction) could make SCO-derived biodiesel more economically feasible. This integrated biorefinery may have a bright future if technoeconomic as well as life cycle analysis [10] justify its feasibility.

or industrial waste. As previously discussed in Sections 6 and 7, L. starkeyi could assimilate a broad range of feedstock and it has an excellent ability to produce high lipid content in an environment that induces lipid accumulation, which is mainly affected by C/N ratio. High C/N ratio is favorable for lipid accumulation but too high C/N ratio leads to high osmotic pressure and stress the cells. The choice of carbon and nitrogen source is also an important part in the fermentation process. For example, utilization of basal medium or analytical grade chemicals (glucose, xylose, etc.) is convenient in laboratory scale since C/N ratio could be determined easily. But it cannot be used for large-scale production of SCOs since it is too costly. Utilization of lignocellulosic hydrolysate is by far the best economically viable way to produce SCOs. However, it all comes back to the choice of substrates and pretreatment process. Taking into consideration that severe pretreatment condition results in the formation of inhibitors in hydrolysate that will require detoxification process which entails an extra cost for microbial oils production. In addition, the choice of oleaginous yeasts that can tolerate inhibitors is also important to ensure feasibility of the process. In view of resistance to inhibitors, L. starkeyi shows good tolerance towards 5-HMF, and to some extent furfural. This yeast possesses the special feature of degrading the least amount of its lipid in the stationary phase. A techno-economic analysis in 2008 by Ratledge and Cohen [11] estimated microbial oils production cost at US$ 3,000/ton. In 2013, Huang et al. conducted a techno-economic analysis on the market price of SCO from lignocellulosic biomass, and the result was RMB 7500/ton [18]. Recently (2014) Koutinas et al. estimated that the price of yeast SCO would reach US$ 3400/ton excluding feedstock costs or US$5500/ ton when glucose was accounted with a biodiesel price of US$ 5,900/ ton [151]. This is not viable considering the market price of vegetable oils and commercial biodiesel at that time were US$ 800–900/ton and US$ 4/gal (US$1220/ton), respectively [19,151]. This value did not consider selling the co-product microbe meal (price US$400–800/ton), which if considered, would make the SCO-biorefinery process slightly more favorable [19]. Thus, it is important to find the cheapest feedstocks, preferably waste such as agricultural residues. In addition, cells harvesting, drying, lipids extraction and transesterification processes also add up to the extra cost. One way to tackle this financial issue is by employing the bio-refinery concept. Improvement in the production of SCO could result in a more economically viable process including faster microbial growth rate and higher cell density, higher lipid accumulation, conversion of residual cell mass into valuable co-products, and reduce the cost of waste disposal [10]. Biorefinery process will transform biomass into many industrial products such as transportation fuels, commodity chemicals, materials, specialty chemicals such as cosmetics and nutraceuticals, and food and feed production may also be incorporated [152]. There are three phases of biorefinery: phase one biorefinery which has one feedstock and one product. In phase two biorefinery, one feedstock is transformed into various end products (chemicals, energy, materials). In phase three biorefinery, it can produce a variety of energy and chemical products using various types of feedstock and processing technologies [152]. Multiple feedstocks process helps to secure feedstock availability and

13. Conclusion As one potential substance to replace fossil fuel, biodiesel is one of the hot topics in the energy sector. Biodiesel can be used directly in a diesel engine without modification. Food vs. fuel debates, a limited supply of non-vegetable oils and increasing global demand for biodiesel trigger the search to find new, sustainable feedstock for biodiesel production. SCO is such as feedstock with advantages such as it can be produced from a microorganism with fast growth rate, no area is needed, and no competition with the food supply. Many oleaginous yeasts with excellent ability in producing lipid have been found such as L. starkeyi, R. toruloides, R. glutinis, T. cutaneum, and C. curvatus. Among them, L. starkeyi has the lowest ability to degrade its own lipid. With its ability to assimilate a broad range of feedstock and results in oils that have a similar composition to that of vegetable oils, L. starkeyi is a potential lipid producer. Both hydrophilic and hydrophobic substrates can be assimilated by L. starkeyi, following mechanism of de novo and ex novo lipid biosynthesis, respectively. For the yeast to produce a high amount of lipid, crucial aspects have to be fulfilled such as high C/N ratio, optimum fermentation conditions (pH of the medium, time, temperature, dissolved oxygen). Explanations regarding acceptable limit for L. starkeyi growth during fermentation have been described in this review. Lipid produced by the yeasts is accumulated intracellularly as lipid body which mainly consists of TAG and sterols, with oleic and palmitic acids as the major fatty acids - similar to that of vegetable oils which is currently the main commercial biodiesel feedstock. This review describes overall processes involved 52

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

during fermentation of L. starkeyi as well as discussion related to the possibility of developing biorefinery process. Process improvement in L. starkeyi fermentation as previously described, combined with biorefinery approach to co-produce bioethanol, oleochemicals and microbe meal may be favorable and viable for future implementation.

102 (2011) 1803–1807. [33] S. Wu, C. Hu, G. Jin, X. Zhao, Z.K. Zhao, Phosphate-limitation mediated lipid production by Rhodosporidium toruloides, Bioresour. Technol. 101 (2010) 6124–6129. [34] S. Papanikolaou, G. Aggelis, Lipids of oleaginous yeasts. Part I: biochemistry of single cell oil production, Eur. J. Lipid Sci. Technol. 113 (2011) 1031–1051. [35] W. Tang, S. Zhang, Q. Wang, H. Tan, Z.K. Zhao, The isocitrate dehydrogenase gene of oleaginous yeast Lipomyces starkeyi is linked to lipid accumulation, Can. J. Microbiol. 55 (2009) 1062–1069. [36] C. Ratledge, The role of malic enzyme as the provider of NADPH in oleaginous microorganisms: a reappraisal and unsolved problems, Biotechnol. Lett. 36 (2014) 1557–1568. [37] M. Tai, G. Stephanopoulos, Engineering the push and pull of lipid biosynthesis in oleaginous yeast Yarrowia lipolytica for biofuelproduction, Metab. Eng. 15 (2013) 1–9. [38] W. Tang, S. Zhang, H. Tan, Z.K. Zhao, Molecular cloning and characterization of a malic enzyme gene from the oleaginous yeast Lipomyces starkeyi, Mol. Biotechnol. 45 (2010) 121–128. [39] E.J. Martinez, V. Ragvahan, F.G. Andres, X. Gomez, New biofuel alternatives: integrating waste management and single cell oil production, Int. J. Mol. Sci. 16 (2015) 9385–9405. [40] A. Beopoulos, J.-M. Nicaud, Yeast: a new oil producer? Dossier Lipochimie 19 (2012) 22–28. [41] C. Ratledge, Regulation of lipid accumulation in oleaginous micro-organism, Biochem. Soc. Trans. 30 (2002) 1047–1050. [42] T. Naganuma, Y. Uzuka, K. Tanaka, H. Iizuka, Differences in enzyme activities of Lipomyces starkeyi between cells accumulating lipid and proliferating cells, J. Basic Microbiol. 27 (1987) 35–42. [43] J.E. Holdsworth, M. Veenhuis, C. Ratledge, Enzyme activities in oleaginous yeasts accumulating and utilizing exogenous or endogenous lipids, J. Gen. Microbiol. 134 (1988) 2907–2915. [44] G. Christophe, V. Kumar, R. Nouaille, G. Gaudet, P. Fontanille, A. Pandey, C.R. Soccol, C. Larroche, Recent developments in microbial oils production: a possible alternative to vegetable oils for biodiesel without competition with human food? Braz. Arch. Biol. Technol. 55 (2012) 29–46. [45] I.K. Muniraj, S.K. Uthandi, Z. Hu, L. Xiao, X. Zhan, Microbial lipid production from renewable and waste materials for second-generation biodiesel feedstock, Environ. Technol. Rev. (2015) 1–16. [46] C. Ratledge, J.P. Wynn, The biochemistry and molecular biology of lipid accumulation in oleaginous microorganisms, Adv. Appl. Microbiol. 51 (2002) 1–51. [47] M.-H. Liang, J.-G. Jiang, Advancing oleaginous microorganisms to produce lipid via metabolic engineering technology, Prog. Lipid Res. 52 (2013) 395–408. [48] C.H. Calvey, L.B. Willis, T.W. Jeffries, An optimized transformation protocol for Lipomyces starkeyi, Curr. Genet. 60 (2014) 223–230. [49] Y. Oguro, H. Yamazaki, Y. Shida, W. Ogasawara, M. Takagi, a.H. Takaku, Multicopy integration and expression of heterologous genes in the oleaginous yeast, Lipomyces starkeyi, Biosci. Biotechnol. Biochem. 79 (2015) 512–515. [50] Y. Oguro, H. Yamazaki, S. Ara, Y. Shida, W. Ogasawara, M. Takagi, H. Takaku, Efficient gene targeting in non-homologous end-joining-deficient Lipomyces starkeyi strains, Curr. Genet. 63 (2017) 751–763. [51] D. Salunke, R. Manglekar, R. Gadre, S. Nene, A.M. Harsulkar, Production of polyunsaturated fatty acids in recombinant Lipomyces starkeyi through submerged fermentation, Bioprocess Biosyst. Eng. 38 (2015) 1407–1414. [52] W. Wang, E.K. Hui Wei, S.V. Wychen, Q. Xu, M.E. Himmel, M. Zhang, Fatty alcohol production in Lipomyces starkeyi and Yarrowia lipolytica, Biotechnol. Biofuels 9 (2016). [53] Q. Xu, E.P. Knoshaug, W. Wang, M. Alahuhta, J.O. Baker, S. Yang, T.V. Wall, S.R. Decker, M.E. Himmel, M. Zhang, H. Wei, Expression and secretion of fungal endoglucanase II and chimeric cellobiohydrolase I in the oleaginous yeast Lipomyces starkeyi, Microb. Cell Factories 16 (2017). [54] A.W. Go, S. Sutanto, L.K. Ong, P.L. Tran-Nguyen, S. Ismadji, Y.-H. Ju, Developments in in-situ (trans) esterification for biodiesel production: a critical review, Renew. Sust. Energ. Rev. 60 (2016) 284–305. [55] F. Yang, M.A. Hanna, R. Sun, Value-added uses for crude glycerol—a byproduct of biodiesel production, Biotechnol. Biofuels 5 (2012) 13. [56] D. Kumar, B. Singh, J. Kostrad, Utilization of lignocellulosic biomass by oleaginous yeas and bacteria for production of biodiesel and renewable diesel, Renew. Sust. Energ. Rev. 73 (2017) 654–671. [57] A.K. Chandel, F.A.F. Antunes, P.V. Arruda, T.S.S. Milessi, S.S. Silva, M.G.A. Felipe, Dilute acid hydrolysis of agro-residues for the depolymerization of hemicellulose: state-of-the-art, in: S.S. Silva, A.K. Chandel (Eds.), D-Xylitol, Springer-Verlag, Berlin Heidelberg, 2012. [58] S.I. Mussatto, J.A. Teixeira, Lignocellulose as raw material in fermentation process, A.M. Vilad Current Research, Technology and Education Topics in Applied Microbiology and Microbial Biotechnology, 2010. [59] V. Balan, Current challenges in commercially producing biofuels from lignocellulosic biomass, ISRN Biotechnol. 2014 (2014). [60] V. Menon, M. Rao, Trends in bioconversion of lignocellulose: biofuels, platform chemicals & biorefinery concept, Prog. Energy Combust. Sci. 38 (2012) 522–550. [61] P.F.H. Harmsen, W.J.J. Huijgen, L.M.B. López, R.R.C. Bakker, Literature Review of Physical and Chemical Pretreatment Processes for Lignocellulosic Biomass, University of Wageningen, 2010. [62] P. Kumar, D.M. Barrett, M.J. Delwiche, P. Stroeve, Methods for pretreatment of lignocellulosic biomass for efficient hydrolysis and biofuel production, Ind. Eng. Chem. Res. 48 (2009) 3713–3729. [63] A.K. Chandel, S.S.d. Silva, O.V. Singh, Detoxification of lignocellulosic

Declaration The authors declare no conflict or financial interests. References [1] Crude Oil Price History Chart, Macrotrends LLC, (2018). [2] Nasdaq, Crude Oil, (2018). [3] EIA, Global liquid fuels, Short Term Energy Outlook, US Energy Information Administration, USA, 2018. [4] BP Energy Outlook in, 2016. [5] L. d'Espaux, D. Mendez-Perez, R. Li, J.D. Keasling, Synthetic biology for microbial production of lipid-based biofuels, Curr. Opin. Chem. Biol. 29 (2015) 58–65. [6] Alternative Fuels Data Center, Ethanol Basics, Energy Efficiency & Renewable Energy, (2017). [7] J.-L. Wertz, O. Bedue, Lignocellulosic Biorefineries, CRC Press, 2013. [8] S. Papanikolaou, G. Aggelis, Lipids of oleaginous yeasts. Part II: technology and potential applications, Eur. J. Lipid Sci. Technol. 113 (2011) 1052–1073. [9] A. Yousuf, Biodiesel from lignocellulosic biomass – prospects and challenges, Waste Manag. 32 (2012) 2061–2067. [10] I.R. Sitepu, L.A. Garay, R. Sestric, D. Levin, D.E. Block, J.B. German, K.L. BoundyMills, Oleaginous yeasts for biodiesel: current and future trends in biology and production, Biotechnol. Adv. 32 (2014) 1336–1360. [11] C. Ratledge, Z. Cohen, Microbial and algal oils: do they have a future for biodiesel or as commodity oils? Lipid Technol. 20 (2008) 155–160. [12] A. Tanimura, M. Takashima, T. Sugita, R. Endoh, M. Kikukawa, S. Yamaguchi, E. Sakuradani, J. Ogawa, M. Ohkuma, J. Shima, Cryptococcus terricola is a promising oleaginous yeast for biodiesel production from starch through consolidated bioprocessing, Sci. Rep. 4 (2014) 4776. [13] A. Steinbüchel, B. Füchtenbusch, Bacterial and other biological systems for polyester production, Trends Biotechnol. 16 (1998) 419–427. [14] X. Meng, J. Yang, X. Xu, L. Zhang, Q. Nie, M. Xian, Biodiesel production from oleaginous microorganisms, Renew. Energy 34 (2009) 1–5. [15] H.-D. Jang, Y.-Y. Lin, S.-S. Yang, Effect of culture media and conditions on polyunsaturated fatty acids production by Mortierella alpina, Bioresour. Technol. 96 (2005) 1633–1644. [16] O.P. Ward, A. Singh, Omega-3/6 fatty acids: alternative sources of production, Process Biochem. 40 (2005) 3627–3652. [17] J.P. SanGiovanni, E.Y. Chew, The role of omega-3 long-chain polyunsaturated fatty acids in health and disease of the retina, Prog. Retin. Eye Res. 24 (2005) 87–138. [18] C. Huang, X.-f. Chen, L. Xiong, X.-d. Chen, L.-l. Ma, Y. Chen, Single cell oil production from low-cost substrates: the possibility and potential of its industrialization, Biotechnol. Adv. (2013) 129–139. [19] M. Jin, P.J. Slininger, B.S. Dien, S. Waghmode, B.R. Moser, A. Orjuela, L.d.C. Sousa, V. Balan, Microbial lipid-based lignocellulosic biorefinery: feasibility and challenges, Trends Biotechnol. 33 (2015) 43–52. [20] C.H. Calvey, Y.-K. Su, L.B. Willis, M. McGee, T.W. Jeffries, Nitrogen limitation, oxygen limitation, and lipid accumulation in Lipomyces starkeyi, Bioresour. Technol. 200 (2016) 780–788. [21] CBS, CBS Strain Database, Utrecht, The Netherlands, (1807). [22] Lipomyces starkeyi Lodder & Kreger-van Rij, Global Biodiversity Information Facility, Copenhagen, Denmark, (1952). [23] M.T. Smith, C.P. Kurtzman, Chapter 43: Lipomyces Lodder & Kreger van Rij, in: C.P. Kurtzman, J.W. Fell, T. Boekhout (Eds.), The Yeasts: A Taxonomic Study, Elsevier, United States of America, 1952–2011, pp. 545–560. [24] C. Kurtzman, J.W. Fell, The Yeasts - A Taxonomic Study, Elsevier Science, 1998. [25] T. Naganuma, Y. Uzuka, K. Tanaka, Physiological factors affecting total cell number and lipid content of the yeast, Lipomyces starkeyi, J. Gen. Appl. Microbiol. 31 (1985) 29–37. [26] G.M. Walker, Yeast Physiology and Biotechnology, John Wiley & Sons, England, 1998. [27] Y. Uzuka, T. Naganuma, K. Tanaka, K. Suzuki, Relation between neutral lipid accumulation and the growth phase in the yeast, Lipomyces starkeyi, a fat producing yeast, Agric. Biol. Chem. 49 (1985) 851–852. [28] M. Suutari, P. Priha, S. Laakso, Temperature shifts in regulation of lipids accumulated by Lipomyces starkeyi, J. Am. Oil Chem. Soc. 70 (1993) 891–894. [29] Y. Uzuka, T. Naganuma, K. Tanaka, Y. Odagiri, Effect of culture on the growth and biotin requirement in a strain of Lipomyces starkeyi, J. Gen. Appl. Microbiol. 20 (1974) 197–206. [30] T. Naganuma, Y. Uzuka, K. Tanaka, Using inorganic elements to control cell growth and lipid accumulation, J. Gen. Appl. Microbiol. 32 (1986) 417–424. [31] C. Ratledge, Fatty acid biosynthesis in microorganisms being used for single cell oil production, Biochimie 86 (2004) 807–815. [32] S. Wu, X. Zhao, H. Shen, Q. Wang, Z.K. Zhao, Microbial lipid production by Rhodosporidium toruloides under sulfate-limited conditions, Bioresour. Technol.

53

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

[64]

[65]

[66]

[67] [68]

[69]

[70]

[71] [72]

[73] [74]

[75]

[76] [77] [78] [79]

[80]

[81]

[82]

[83] [84]

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92]

[93]

yeast species, Bioresour. Technol. 144 (2013) 360–369. [94] K.V. Probst, Single Cell Oil Production Using Lipomyces starkeyi: Fermentation, Lipid Analysis and Use of Renewable Hemicellulose-Rich Feedstocks, Department of Grain Science and Industry College of Agriculture, Kansas State University, Manhattan, Kansas, 2014. [95] X. Yang, G. Jin, Z. Gong, H. Shen, Y. Song, F. Bai, Z.K. Zhao, Simultaneous utilization of glucose and mannose from spent yeast cell mass for lipid production by Lipomyces starkeyi, Bioresour. Technol. 158 (2014) 383–387. [96] S. Tsakona, N. Kopsahelis, A. Chatzifragkou, S. Papanikolaou, I.K. Kookos, A.A. Koutinas, Formulation of fermentation media from flour-rich waste streams formicrobial lipid production by Lipomyces starkeyi, J. Biotechnol. 189 (2014) 36–45. [97] A. Anschau, M.C.A. Xavier, S. Hernalsteens, T.T. Franco, Effect of feeding strategies on lipid production by Lipomyces starkeyi, Bioresour. Technol. 157 (2014) 214–222. [98] J.P.F. Vieira, J.L. Ienczak, C.E.V. Rossell, J.G.C. Pradella, T.T. Franco, Microbial lipid production: screening with yeasts grown on Brazilian molasses, Biotechnol. Lett. 36 (2014) 2433–2442. [99] M.C.A. Xavier, T.T. Franco, Batch and continuous culture of hemicellulosic hydrolysate from sugarcane bagasse for lipids production, Chem. Eng. Trans. 38 (2014) 385–390. [100] J. Lin, S. Li, M. Sun, C. Zhang, W. Yang, Z. Zhang, X. Li, S. Li, Microbial lipid production by oleaginous yeast in D-xylose solution using a two-stage culture mode, RSC Adv. 4 (2014) 34944–34949. [101] A. Anschau, T.T. Franco, Cell mass energetic yields of fed-batch culture by Lipomyces starkeyi, Bioporcess Biosyst. Bioeng. 38 (2015) 1517–1525. [102] F. Spier, J.G. Buffon, C.A.V. Burkert, Bioconversion of raw glycerol generated from the synthesis of biodiesel by different oleaginous yeast: lipid content and fatty acid profile of biomass, Indian J. Microbiol. 55 (2015) 415–422. [103] K.V. Probst, P.V. Vadlani, Production of single cell oil from Lipomyces starkeyi ATCC 56304 using biorefinery by-prodcuts, Bioresour. Technol. 198 (2015) 268–275. [104] D.E. Leiva-Candia, S. Tsakona, N. Kopsahelis, I.L. Garcia, S. Papanikolau, M.P. Dorado, A.A. Koutinas, Biorefining of by-product streams from sunflowerbased biodiesel production plants for integrated synthesis of microbial oil and value added co-products, Bioresour. Technol. 190 (2015) 57–65. [105] N. Bonturi, L. Matsakas, R. Nilson, P. Christakopoulos, E.A. Miranda, K.A. Berglund, U. Rova, Single cell oil producing yeasts Lipomyces starkeyi and Rhodosporidium toruloides: selection of extraction strategies and biodiesel property prediction, Energies 8 (2015) 5040–5052. [106] S.S. Tchakouteu, O. Kalantzi, C. Gardeli, A.A. Koutinas, G. Aggelis, S. Papanikolaou, Lipid production by yeasts growing on biodiesel-derived crude glycerol: strain selection and impact of substrate concentration on the fermentation efficiency, J. Appl. Microbiol. 118 (2015) 911–927. [107] A. Anschau, T.T. Franco, Continuous cultivations of the oleaginous yeast Lipomyces starkeyi, V Simposio de Bioquimica E Biotecnologia, Londrina, Brazil, 2015. [108] B.S. Dien, P.J. Slininger, C.P. Kurtzman, B.R. Moser, P.J. O'Bryan, Identification of superior lipid producing Lipomyces and Myxozyma yeasts, AIMS Env. Sci. 3 (2016) 1–20. [109] X. Yang, G. Jin, Y. Wang, H. Shen, Z.K. Zhao, Lipid production on free fatty acids by oleaginous yeasts under non-growth conditions, Bioresour. Technol. 193 (2015) 557–562. [110] M. Dourou, A. Kancelista, P. Juszczyk, D. Sarris, S. Bellou, I.-E. Triantaphyllidou, A. Rywinska, S. Papanikolaou, G. Aggelis, Bioconversion of olive mill wastewater into high-added value products, J. Clean. Prod. (2016) 957–969. [111] M.C.A. Xavier, A.L.V. Coradini, A.C. Deckmann, T.T. Franco, Lipid production from hemicellulose hydrolysate and acetic acid by Lipomyces starkeyi and the ability of yeast to metabolize inhibitors, Biochem. Eng. J. 118 (2017) 11–19. [112] K.V. Probst, P.V. Vadlani, Single cell oil production by Lipomyces starkeyi: biphasic fed-batch fermentation strategy providing glucose for growth and xylose for oil production, Biochem. Eng. J. 121 (2017) 49–58. [113] L.-p. Liu, M.-h. Zong, Y. Hu, N. Li, W.-y. Lou, H. Wu, Efficient microbial oil production on crude glycerol by Lipomyces starkeyi AS 2.1560 and its kinetics, Process Biochem. 58 (2017) 230–238. [114] D. Pirozzi, N. Fiorentino, A. Impagliazzo, F. Sannino, A. Yousuf, G. Zuccaro, M. Fagnano, Lipid production from Arundo donax grown under different agronomical conditions, Renew. Energy 77 (2015) 456–462. [115] J. Brandenburg, J. Blomqvist, J. Pickova, N. Bonturi, M. Sandgren, V. Passoth, Lipid production from hemicellulose with Lipomyces starkeyi in a pH regulated fedbatch cultivation, Yeast 33 (2016) 451–462. [116] S. Rahman, P. Arbter, M. Popovic, R. Bajpai, R. Subramaniam, Microbial lipid production from lignocellulosic hydrolyzates: effect of carbohydrate mixtures and acid-hydrolysis byproducts on cell growth and lipid production by Lipomyces starkeyi, J. Chem. Technol. Biotechnol. 92 (2017) 1980–1989. [117] S. Sutanto, A.W. Go, K.-H. Chen, P.L.T. Nguyen, S. Ismadji, Y.-H. Ju, Release of sugar by acid hydrolysis from rice bran for single cell oil production and subsequent in-situ transesterification for biodiesel preparation, Fuel Process. Technol. 167 (2017) 281–291. [118] C. Ratledge, Biochemistry, stoichiometry, substrates and economics, in: R.S. Moreton (Ed.), Single Cell Oil, Longman Scientific & Technical, Harlow (UK), 1988, pp. 33–70. [119] X. Yang, G. Jin, Z. Gong, H. Shen, F. Bai, Z.K. Zhao, Recycling microbial lipid production wastes to cultivate oleaginous yeasts, Bioresour. Technol. 175 (2015) 91–96. [120] X.P. Ye, L. Liu, D. Hayes, A. Womac, K. Hong, S. Sokhansanj, Fast classification and compositional analysis of cornstover fractions using Fourier transform near-

hydrolysates for improved bioethanol production, Biofuel Production-Recent Developments and Prospects, 2011. A.K. Azad, A. Yousuf, A. Ferdoush, M.M. Hasan, M.R. Karim, A. Jahan, Production of microbial lipids from rice straw hydrolysates by Lipomyces starkeyi for biodiesel synthesis, Microb. Biochem. Technol. (2014) 1–6. X. Yu, Y. Zheng, K.M. Dorgan, S. Chen, Oil production by oleaginous yeasts using the hydrolysate from pretreatment of wheat straw with dilute sulfuric acid, Bioresour. Technol. 102 (2011) 6134–6140. C. Huang, X.-F. Chen, X.-Y. Yang, L. Xiong, X.-Q. Lin, J. Yang, B. Wang, X.-D. Chen, Bioconversion of corncob acid hydrolysate into microbial oil by the oleaginous yeast Lipomyces starkeyi, Appl. Biochem. Biotechnol. 172 (2014) 2197–2204. R. Wang, J. Wang, R. Xu, Z. Fang, A. Liu, Oil production by the oleaginous yeast Lipomyces starkeyi using diverse carbon sources, Bioresources 9 (2014) 7027–7040. L. Matsakas, A.-A. Sterioti, U. Rova, P. Christakopoulos, Use of dried sweet sorghum for the efficient production of lipids by the yeast Lipomyces starkeyi CBS 1807, Ind. Crop. Prod. 62 (2014) 367–372. D. Pirozzi, A. Ausiello, A. Yousuf, G. Zuccaro, G. Toscano, Exploitation of oleaginous yeasts for the production of microbial oils from agricultural biomass, Chem. Eng. Trans. 37 (2014) 469–474. D. Pirozzi, G. Travglini, D. Sagneli, F. Sannino, G. Toscano, Study of a discontinuous fed-batch fermentor for the exploitation of agricultural biomasses to produce II-generation biodiesel, Chem. Eng. Trans. 38 (2014) 169–174. P. Akhtar, J.I. Gray, A. Asghar, Synthesis of lipids by certain yeast strains grown on whey permeate, J. Food Lipids 5 (1998) 283–297. K. Nishimura, M. Yamamoto, T. Nakagomi, Y. Takiguchi, T. Naganuma, Y. Uzuka, Biodegradation of triazine herbicides on polyvinylalcohol gel plates by the soil yeast Lipomyces starkeyi, Appl. Microbiol. Biotechnol. 58 (2002) 848–852. J.R. Anderson, E.A. Drew, Growth characteristic of a species of lipomyces and its degradation of paraquat, J. Gen. Microbiol. 70 (1972) 43–58. T. Suzuki, K. Hazegawa, Lipid composition and lipid molecular species of Lipomyces starkeyi cultivated in medium containing 1,2-propanediol, Agric. Biol. Chem. 40 (1975) 221–223. H. Liu, X. Zhao, F. Wang, X. Jiang, S. Zhang, M. Ye, Z.K. Zhao, H. Zou, The proteome analysis of oleaginous yeast Lipomyces starkeyi, FEMS Yeast Res. 11 (2011) 42–51. M.H. Deinema, C.A. Landheer, Composition of fats, produced by Lipomyces starkeyi, under various conditions, Arch. Mikrobiol. 25 (1956) 193–200. H. Yamauchi, M. H, K. T, S. S, Mass production of lipids by Lipomyces starkeyi in microcomputer-aided-fed-batch culture, J. Ferment. Technol. 61 (1983) 275–280. V.K. Eroshin, N.I. Krylova, Efficiency of lipid synthesis by yeasts, Biotechnol. Bioeng. 25 (1983) 1693–1700. C. Angerbauer, M. Siebenhofer, M. Mittelbach, G.M. Guebitz, Conversion of sewage sludge into lipids by Lipomyces starkeyi for biodiesel production, Bioresour. Technol. 99 (2008) 3051–3056. X. Zhao, X. Kong, Y. Hua, B. Feng, Z.K. Zhao, Medium optimization for lipid production through co-fermentation of glucose and xylose by the oleaginous yeast Lipomyces starkeyi, Eur. J. Lipid Sci. Technol. 110 (2008) 405–412. X. Chen, Z. Li, X. Zhang, F. Hu, D.D.Y. Ryu, J. Bao, Screening of oleaginous yeast strains tolerant to lignocellulose degradation compounds, Appl. Biochem. Biotechnol. 159 (2009) 591–604. A. Yousuf, F. Sannino, V. Addorisio, D. Pirozzi, Microbial conversion of olive oil mill wastewaters into lipids suitable for biodiesel production, J. Agric. Food Chem. 58 (2010) 8630–8635. R. Wild, S. Patil, M. Popovi, M. Zappi, S. Dufreche, R. Bajpai, Lipids from Lipomyces starkeyi, Food Technol. Biotechnol. 48 (2010) 329–335. J. Lin, H. Shen, H. Tan, X. Zhao, S. Wu, C. Hu, Z.K. Zhao, Lipid production by Lipomyces starkeyi cells in glucose solution without auxiliary nutrients, J. Biotechnol. 152 (2011) 184–188. L. Huang, B. Zhang, G. B, S. G, Application of fishmeal wastewater as a potential low-cost medium for lipid production by Lipomyces starkeyi HL, Environ. Technol. 33 (2011) 1975–1981. N.E.-A.A. El-Naggar, M.S. El-Hersh, H.A. El-Fadaly, W.I.A. Saber, Bioconversion of some agro-industrial by products into single cell oil using Candida albicans NRRL Y-12983 and Lipomyces starkeyi NRRL Y-11557, Res. J. Microbiol. 6 (2011) 784–795. E. Oguri, K. Masaki, T. Naganuma, H. Iefuji, Phylogenetic and biochemical characterization of the oil-producing yeast Lipomyces starkeyi, Antonie Van Leeuwenhoek 101 (2012) 359–368. J.-X. Liu, Q.-Y. Yue, B.-Y. Gao, Z.-H. Ma, P.-D. Zhang, Microbial treatment of the monosodium glutamate wastewater by Lipomyces starkeyi to produce microbial lipid, Bioresour. Technol. 106 (2012) 69–73. Z. Gong, Q. Wang, H. Shen, C. Hu, G. Jin, Z.K. Zhao, Co-fermentation of cellobiose and xylose by Lipomyces starkeyi for lipid production, Bioresour. Technol. 117 (2012) 20–24. E.V. Tapia, A. Anschau, A.L. Coradini, T.T. Franco, A.C. Deckmann, Optimization of lipid production by the oleaginous yeast Lipomyces starkeyi by random mutagenesis coupled to cerulenin screening, AMB Express 2 (2012). R.F. Castanha, L.A.S.d. Morais, A.P. Mariano, R.T.R. Monteiro, Comparison of two lipid extraction methods produced by yeast in cheese whey, Braz. Arch. Biol. Technol. 56 (2013) 629–636. J.-X. Liu, Q.-Y. Yue, B.-Y. Gao, Y. Wang, Q. Li, P.-D. Zhang, Research on microbial lipid production from potato starch wastewater as culture medium by Lipomyces starkeyi, Water Sci. Technol. 67 (2013) 1802–1808. I.R. Sitepu, R. Sestric, L. Ignatia, D. Levin, J.B. German, L.A. Gillies, L.A.G. Almada, K.L. Boundy-Mills, Manipulation of culture conditions alters lipid content and fatty acid profiles of a wide variety of known and new oleaginous

54

Fuel Processing Technology 177 (2018) 39–55

S. Sutanto et al.

[140] A. Tanimura, M. Takashima, T. Sugita, R. Endoh, M. Kikukawa, S. Yamaguchi, E. Sakuradani, J. Ogawa, M. Ohkuma, J. Shima, Cryptococcus terricola is a promising oleaginous yeast for biodiesel production from starch through consolidated bioprocessing, Sci. Rep. 4 (2014) 4776. [141] T. Suzuki, K. Hasegawa, Lipid molecular species of Lipomyces starkeyi, Agric. Biol. Chem. 38 (1974) 1371–1376. [142] C. Shene, A. Leyton, Y. Esparza, L. Flores, B. Quilodrán, I. Hinzpetera, M. Rubilar, Microbial oils and fatty acids: effect of carbon source on docosahexanoic acid (C22:6 n-3, DHA) production by Thraustochytrid strains, J. Soil Sci. Plant Nutr. 10 (2010) 207–216. [143] M.J. Ramos, C.M. Fernández, A. Casas, L. Rodríguez, Á. Pérez, Influence of fatty acid composition of raw materials on biodiesel properties, Bioresour. Technol. 100 (2009) 261–268. [144] S. Pinzi, D. Leiva-Candia, I. López-García, M.D. Redel-Macías, M.P. Dorado, Latest trends in feedstocks for biodiesel production, Biofuels Bioprod. Biorefin. 8 (2014) 126–143. [145] B. Liu, Z.K. Zhao, Biodiesel production by direct methanolysis of oleaginous microbial biomass, J. Chem. Technol. Biotechnol. 82 (2007) 775–780. [146] P. Thliveros, E.U. Kiran, C. Webb, Microbial biodiesel production by direct methanolysis of oleaginous biomass, Bioresour. Technol. 157 (2014) 181–187. [147] I. Espinosa-Gonzalez, A. Parashar, D.C. Bressler, Hydrothermal treatment of oleaginous yeast for the recovery of free fatty acids for use in advanced biofuel production, J. Biotechnol. 187 (2014) 10–15. [148] J.S. Park, B.H. Kim, J.H. Lee, E.S. Seo, K.S. Cho, H.J. Park, H.K. Kang, S.K. Yoo, M.S. Ha, H.J. Chung, D.L. Cho, D.F. Day, D. Kim, Optimization for novel glucanhydrolase production of Lipomyces starkeyi KSM 22 by statistical design, J. Microbiol. Biotechnol. 19 (2003) 993–997. [149] Y.-H. Chang, J.-H. Yeom, K.-H. Jung, B.C. Chang, J.H. Shin, S.-K. Yoo, Optimization of an extracellular dextranase production from Lipomyces starkeyi KCTC 17343 and analysis of its dextran hydrolysates, T. I. T. J. Life Sci. 19 (2009) 457–461. [150] S. Behera, R. Arora, N. Nandhagopal, S. Kumar, Importance of chemical pretreatment for bioconversion of lignocellulosic biomass, Renew. Sust. Energ. Rev. 36 (2014) 91–106. [151] A.A. Koutinas, A. Chatzifragkou, N. Kopsahelis, S. Papanikolaou, I.K. Kookos, Design and techno-economic evaluation of microbial oil production as a renewable resource for biodiesel and oleochemical production, Fuel 116 (2014) 566–577. [152] J.H. Clark, F.E.I. Deswarte (Eds.), The Biorefinery Concept–An Integrated Approach, 2008. [153] S. Sutanto, A.W. Go, S. Ismadji, Y.-H. Ju, Hydrolyzed rice bran as source of lipids and solid acid catalyst during in situ (trans)esterification, Biofuels (2017) 1–7. [154] S. Kamat, M. Khot, S. Zinjarde, A. RaviKumar, W.N. Gade, Coupled production of single cell oil as biodiesel feedstock, xylitol and xylanase from sugarcane bagasse in a biorefinery concept using fungi from the tropical mangrove wetlands, Bioresour. Technol. (2013) 246–253. [155] I. Kim, Y.H. Seo, G.-Y. Kim, J.-I. Han, Co-production of bioethanol and biodiesel from corn stover pretreated with nitric acid, Fuel 143 (2015) 285–289. [156] Y. Morikawa, X. Zhao, D. Liu, Biological co-production of ethanol and biodiesel from wheat straw: a case of dilute acid pretreatment, RSC Adv. 4 (2014) 37878–37888. [157] D. Ghosh, D. Dasgupta, D. Agrawal, S. Kaul, D.K. Adhikari, A.K. Kurmi, P.K. Arya, D. Bangwal, M.S. Negi, Fuels and chemicals from lignocellulosic biomass: an integrated biorefinery approach, Energy Fuel 29 (2015) 3149–3157.

infrared techniques, Bioresour. Technol. 99 (2008) 7323–7332. [121] H. Kawaguchi, A.A. Vertès, S. Okino, M. Inui, H. Yukawa, Engineering of a xylose metabolic pathway in Corynebacterium glutamicum, Appl. Environ. Microbiol. 72 (2006) 3418–3428. [122] C. Hu, S. Wu, Q. Wang, G. Jin, H. Shen, Z.K. Zhao, Simultaneous utilization of glucose and xylose for lipid production by Trichosporon cutaneum, Biotechnol. Biofuels 4 (2011). [123] J. Aduse-Opoku, W.J. Mitchell, Diauxic growth of Clostridium thermosaccharolyticum on glucose and xylose, FEMS Microbiol. Lett. 50 (1988) 45–49. [124] S. Fakas, S. Papanikolaou, A. Batsos, M. Galiotou-Panayotou, A. Mallouchos, G. Aggelis, Evaluating renewable carbon sources as substrates for single cell oil production by Cunninghamella echinulata and Mortierella isabellina, Biomass Bioenergy 33 (2009) 573–580. [125] A. Patel, N. Arora, Km Sartaj, V. Pruthi, P.A. Pruthi, Sustainable biodiesel production from oleaginous yeasts utilizing hydrolysates of various non-edible lignocellulosic biomasses, Renew. Sust. Energ. Rev. 62 (2016) 836–855. [126] H.W. Yen, Z. Zhang, Effects of dissolved oxygen level on cell growth and total lipid accumulation in the cultivation of Rhodotorula glutinis, J. Biosci. Bioeng. 112 (2011) 71–74. [127] Y.K. Lee, Bioprocess technology, in: Y.K. Lee (Ed.), Microbial Biotechnology : Principles and Applications, World Scientific Publishing, Singapore, 2006. [128] C.A. Boulton, C. Ratledge, Use of transition studies in continuous cultures of Lipomyces starkeyi, an oleaginous yeast, to investigate the physiology of lipid accumulation, J. Gen. Microbiol. (1983) 2871–2876. [129] M. Rossi, A. Amaretti, S. Raimondi, A. Leonardi, Getting lipids for biodiesel production from oleaginous fungi, in: M. Stoytcheva, G. Montero (Eds.), Biodiesel – Feedstocks and Processing Technologies, InTech, Croatia, 2011. [130] C. Ratledge, Yeasts, moulds, algae and bacteria as sources of lipids, Technological Advances in Improved and Alternative Sources of Lipids, Blackie Academic and Professional, London, 1994. [131] A. Beopoulos, J. Cescut, R. Haddouche, J.-L. Uribelarrea, C. Molina-Jouve, J.M. Nicaud, Yarrowia lipolytica as a model for bio-oil production, Prog. Lipid Res. 48 (2009) 375–387. [132] Y. Li, Z.B. Zhao, B.F. Wu, High density cultivation of oleaginous yeast Rhodosporidium toruloides Y4 in fedbatch culture, Enzym. Microb. Technol. 41 (2007) 312–317. [133] C.J. Chuck, F. Santomauro, L.A. Sargeant, F. Whiffin, T. Chantasuban, N.R.A. Ghaffar, J.L. Wagner, R.J. Scott, Liquid transport fuels from microbial yeasts – current and future perspectives, Biofuels 5 (2014) 293–311. [134] T.M. Mata, A.A. Martins, N.S. Caetano, Microalgae for biodiesel production and other applications: a review, Renew. Sust. Energ. Rev. 14 (2010) 217–232. [135] A.C. Guedes, H.M. Amaro, F.X. Malcata, Microalgae as sources of high added-value compounds-a brief review of recent work, Biotechnol. Prog. 27 (2011) 597–613. [136] Y. Uzuka, T. Kanamori, T. Koga, K. Tanaka, T. Naganuma, Isolation and chemical composition of intracellular oil globules from the yeast Lipomyces starkeyi, J. Gen. Appl. Microbiol. 21 (1975) 157–168. [137] T. Naganuma, Y. Uzuka, K. Tanaka, Quantitative estimation of intracellular neutral lipids of the yeast, Lipomyces starkeyi, Agric. Biol. Chem. 46 (1982) 1213–1217. [138] T. Suzuki, A. Takigawa, K. Hasegawa, Lipid extraction methods for Lipomyces starkeyi, Agric. Biol. Chem. 37 (1973) 2653–2656. [139] E.G. Bligh, W.J. Dryer, A rapid method of total lipid extraction and purification, Can. J. Biochem. Physiol. 37 (1959).

55