Introduction

Biodiesel is a renewable, clean-burning diesel replacement. It is compatible with diesel engines when mixed with petroleum diesel and has a wide application in Europe and the USA. Biodiesel is produced by the transesterification of triglycerides, which converts triglycerides into long-chain alkyl esters as biodiesel. It can be made from a diverse mix of feedstocks including recycled cooking oil, plant oil, and animal fats. But, from the economic terms, these feedstocks account for more than 75% of the total production costs, making biodiesel more expensive than conventional fuels [1]. Meanwhile, the use of plant oil in biodiesel production also leads to food versus fuel controversy [2]. The high costs of lipid-producing feedstock make research groups search for alternative lipid source for biodiesel production, and single cell oils (SCOs) have emerged as potential candidates.

SCOs are the edible oils extracted from microorganisms. They are lipophilic compounds in microbial cells, generated for their cell metabolism and survival. While excessive carbon source is available due to the shortage of nitrogen or other nutrients, the cell growth becomes limited, and microbial lipids are accumulated as energy storage [3]. SCOs can be the source of oils and fats that serve as raw materials for oleochemical industries or as a source of polyunsaturated fatty acids (PUFAs) and other essential fatty acids (EFAs). Their compositions are similar to traditional vegetable oils [4], and the production of SCOs requires lower land resources compared with plant oil. However, SCO production largely depends on the microbial strain used for oil accumulation. Only a small group of microorganisms is identified as an oleaginous microorganism, which can accumulate high intracellular lipid content (> 20% of dry cell biomass weight). Moreover, it is also revealed that biodiesel quality depends upon the fatty acid composition of the oil feedstock [5]. The current studied oleaginous strains are not as efficient or robust as would be preferred for industrial applications [6]. Both the lipid content and the types of their intracellular lipids are important criteria for microorganisms to serve as a suitable feedstock for biodiesel production.

During the search of oleaginous fungi from oil-rich plants, one specific Fusarium equiseti UMN-1 strain was screened from soybean [7]. It is able to accumulate around 56% of lipid in the biomass when using glucose as carbon source in the optimized medium, and nearly all lipids extracted from cells can be converted to FAME. This strain is shown to utilize various types of carbon sources and possesses the capability to degrade plant tissue with considerable production of cellulase. These characteristics provide a possibility for this strain to accumulate a high amount of SCOs on inexpensive lignocellulosic materials.

Besides the selection of strains, accumulation of lipids by oleaginous fungi can also be affected by many factors, such as the nutrient level and culture conditions (e.g., temperature, pH, carbon and nitrogen source, and agitation) [8,9,10]. Fermentation substrate is an important factor in the development of SCOs. Screening of oleaginous microorganisms normally uses sugars as carbon source, but the high price of pure sugar limits its wide use in the industrial production of SCOs. To lower the cost of substrates, the adaption of inexpensive waste materials and agricultural residues for SCO production has been studied [11,12,13,14]. This requires oleaginous microorganisms to maintain a high lipid content during cell growth when utilizing these wastes or lignocellulosic materials as carbon source. Cultivation conditions can greatly affect microbial lipid content and lipid composition accumulated in fungi. Usually stressed conditions such as the limitation of nitrogen will lead to an enhancement of lipid production by fungi, and the excess of carbon is assimilated by the cells and converted into triacylglyceride (TAG) [3, 15]. C/N ratio is one of the most commonly studied stress factors to stimulate the oil accumulation for the oleaginous species [16]. Thus, the search for a suitable range of C/N molar ratio is also important to improve microbial lipid production. This research aimed to investigate the characteristics of this strain under pure culture and optimize cultivation conditions and to explore the promising properties of this strain for industrial application to biofuel production.

Materials and Methods

Strains and Culture Conditions

F. equiseti UMN-1 strain used in this study was isolated from healthy soybean samples collected from a farmland in Minnesota, USA [7]. Fungal hyphae were inoculated into a 100 mL modified potato dextrose (PD) medium (9.6 g/L PD broth and 12 g/L glucose) in 250-mL flasks to study the effects from temperature (20 °C, 27 °C, 35 °C, 42 °C) and agitation speeds (100 rpm, 150 rpm, 200 rpm). The rest of growth cultures were based on an improved growth medium (M3 medium) used for Fusarium fungi [17], including 0.75 g/L (NH4)2SO4, 0.1 g/L MgSO4·7H2O, 1.4 g/L KH2PO4, 0.68 g/L K2HPO4, 0.1 g/L KCl, 0.8 mg/L MnSO4·H2O, 0.4 mg/L CuSO4, 0.8 mg/L FeSO4·7H2O, 0.8 mg/L Na2MoO4·2H2O, 8 mg/L ZnSO4·7H2O, 0.04 mg/L Na2B4O7·10H2O, 0.75 g/L yeast extract, and 20 g/L glucose. The batch culture typically operated at 150 rpm for 6 days at 27 °C with triplicate flasks.

Central Composite Design for Optimized Cultivation Condition

The central composite design (CCD) was used to determine the optimum value of each variable that significantly affects the F. equiseti UMN-1 strain fungal growth and lipid production. It is one of the response surface methodologies to build a second-order (quadratic) model for the response variables. Each factor would be designed with five coded levels (− α, − 1, 0, + 1, + α), in which α is a function of the numbers of factors. The general equation for α value is α = (2k)1/4. In this research, the independent factors included temperature, carbon source concentration, and nitrogen concentration; thus, k = 3 and α value in this design was 1.682. Coded levels and the actual values for the independent variables are listed in Table 1. The lipid production was the dependent variable (the response for CCD). In Table 1, carbon source was glucose and nitrogen was the combination of (NH4)2SO4 and yeast extract. 0.2 g/L N in medium required 0.6 g/L (NH4)2SO4 and 0.6 g/L yeast extract, and other concentrations followed the same ratio. All media were prepared based on the optimized M3 medium as above. The design of the experiment matrix followed Design Expert 8.0 (Static Made Easy, Minneapolis, MN) to acquire a random order. After the designed experiment matrix was performed, biomass and lipid content were tested to obtain the value of lipid production. Linear regression was used to obtain the predicted optimum values of each variable. Then, the predicted model was validated with fungal cultivation. All fungal cultivation was carried out in the 100 mL optimized M3 medium in 250-mL flasks at 150 rpm for 6 days.

Table 1 Coded levels and true values of independent variables
Full size table

Analytical Methods

After cultivation, the culture broth was centrifuged at 9000 rpm for 5 min to obtain fungal biomass. The biomass and lipid content were determined following the methods in previous research [7]. Fatty acid methyl ester (FAME) was prepared by extraction–transesterification procedure [18] to analyze the fatty acid profile of lipids accumulated in fungal cells. Detailed fatty acid profile analysis followed the method in previous research [7]. The fungal culture broth after cultivation was centrifuged at 9000 rpm for 5 min to collect supernatants for cellulase activity analysis. Cellulase activity was tested by filter paper assay [19] and was represented in filter paper units per milliliter (FPU/mL). For residual sugar, the fungal culture broth was centrifuged at 9000 rpm for 5 min to collect supernatants. The residual sugar content in supernatants was analyzed by dinitrosalicylic acid (DNS) method [20].

Results and Discussion

Effect of Temperature on the Biomass and Lipid Production

Results in Fig. 1 showed that the optimum temperature for F. equiseti UMN-1 strain was around 27 °C. The highest lipid production was obtained at 27 °C, with 0.22 ± 0.02 g lipid/100 mL (Fig. 1). The highest lipid content of 49.74 ± 0.35% was obtained at 35 °C, but the biomass growth had a significant decline which leads to a reduction in the total lipid production. No fungal growth was detected at 42 °C, which suggests that 42 °C is beyond the temperature range of F. equiseti UMN-1 strain. This is similar to a report on the impact of temperature on Fusarium oxysporum, from which 30 °C was found as the optimum temperature and fungal growth drastically reduced below 15 °C and above 35 °C [21]. The FAME content in fungal lipids reached the highest ratio of 99.76% at 27 °C, indicating that nearly all lipids produced by F. equiseti UMN-1 strain at this temperature can be converted to FAME. It was significantly higher than the value achieved at 20 °C (65.27%) and 35 °C (43.47%) (Table 2). Meanwhile, the detailed FAME profile also changed with temperature. From 20 to 35 °C, the percentage of palmitic acid (C16:0) increased from 22.86 to 38.65%, and the percentage of stearic acid (C18:0) increased from 10.09 to 22.15%. Correspondingly, the percentage of linoleic acid (C18:2) received an obvious decrease. Generally, as growth temperature increases, fungi tend to produce more saturated fatty acids in lipids while reducing the proportion of unsaturated fatty acids. The degree of fatty acid unsaturation has been found to decrease with higher temperatures in many oleaginous microorganisms like fungi, yeasts, and bacteria [22,23,24].

Fig. 1
figure 1

Effect of temperature on biomass and lipid content

Full size image
Table 2 Effect of temperature on fatty acid profile
Full size table

Temperature is one of the most influencing factors for fungal growth. Impact of temperature on lipid accumulation has also been discovered; for example, lower temperature was discovered to trigger high levels of oil production on Metschnikowia pulcherrima [9], and the presence of stearidonic acid (C18:4) in Mortierella elongate at low temperature [25]. Fungi can live in a relatively large range of temperatures, but their growth rate and metabolism are different at different temperatures, which mostly relates to chemical reactions within fungal cells. The temperature at which fungi have highest biomass growth rate is normally accepted as the optimum temperature. It is considered that optimal growth temperature for oleaginous fungi is generally 20–28 °C [26, 27]. Such optimum temperature should allow the most efficient progression of chemical reactions necessary for growth. However, in this test, the biomass growth and lipid accumulation in cells has different optimum temperature. Similar phenomenon has been reported for Penicillium roqueforti that the best temperature for cell growth and substrate carbon conversion efficiency were different [28]. This is probably due to different metabolic reactions required for cell growth and generation of secondary metabolism products. In this test, the highest lipid production for F. equiseti UMN-1 strain was obtained at 27 °C and lipid content was 42.54 ± 0.46%. When temperature increased to 35 °C, lipid content can increase to 49.74 ± 0.35%. This higher lipid content could be interesting for oil production industry to consider higher culture temperature during continuous culture.

Effect of Agitation Speed on the Biomass and Lipid Production

Three agitation speeds were tested to explore the impact of agitation on lipid production. Results revealed that 150 rpm agitation speed was the best on both biomass production and oil accumulation, while either higher or lower speed had a negative impact compared to 150 rpm (Fig. 2). The FAME composition (98.08 ± 0.73%) was not significantly affected by the change of agitation speed, and relative percentages of each major component remained at a stable level. The oxygen in liquid substrate affects cell growth and metabolite biosynthesis. And usually, oxygen level in culture broth is related to the cultivation modes and agitation speed. During aerobic fermentation process in bioreactor, the generation of fermentation products is limited by oxygen availability. From an investigation on l-lysine fermentation in a continuous culture, it was found that l-lysine production was strongly influenced by the dissolved oxygen level, and 50% or above of dissolved oxygen was suggested to maximize the production of lysine [29]. Unlike bioreactor fermentation, external aeration is generally not applicable in flask cultivation. Due to the small surface area of growth medium in flasks, the oxygen transfer rate could be highly limited; this makes agitation a very important method to increase the dissolved oxygen in growth medium. Results in this test showed that from 100 to 150 rpm, both biomass growth and oil accumulation had a significant increase, which was probably due to higher dissolved oxygen content at 150 rpm speed. Besides increasing the oxygen content in liquid, agitation also facilitates to form uniform suspension of microbial cells in homogeneous medium and increase mass transfer rate. It was found that the mycelial morphology was significantly affected by agitation intensity [8]. However, higher agitation speed also leads to higher shear stress, which could be detrimental to mycelial growth [30]. During the cultivation of F. equiseti UMN-1 strain, both the biomass growth and oil accumulation decreased when agitation speed raised from 150 to 200 rpm.

Fig. 2
figure 2

Effect of agitation speed on biomass and lipid content

Full size image

Effect of C/N Ratio

The study of different C/N molar ratios used the improved M3 medium with the concentration change of both nitrogen sources of (NH4)2SO4 and yeast extract, and the results showed a significant impact on fungal lipid accumulation (Fig. 3). Fungal biomass production was quite high (1.14 ± 0.10 g/100 mL) when the C/N ratio was at 5. As the C/N ratio increased to 20, the biomass production decreased to 0.68 ± 0.01 g/100 mL. Further increase of C/N ratio did not reduce biomass production obviously, and in general, biomass production kept at a stable level (0.62–0.69 g/100 mL) from the C/N ratio of 20 to 100. Meanwhile, as the C/N ratio increased from 5 to 80, lipid content increased from 12.36 ± 1.91 to 36.45 ± 1.51%. Beyond the C/N ratio of 80, both biomass amount and lipid content began to decline at the C/N ratio of 100. Thus, the optimal C/N molar ratio for F. equiseti UMN-1 strain was 80, at which the highest lipid production was obtained. Since carbon (glucose) concentration was fixed at 20 g/L for all growth media, the C/N ratio actually reflected nitrogen concentration in medium. When the C/N ratio was at 5, nitrogen content in medium was 6 g/L (NH4)2SO4 + 6 g/L yeast extract. When increasing the C/N ratio to 20, the nitrogen content in medium decreased to 1.5 g/L(NH4)2SO4 + 1.5 g/L yeast extract, and such a big decrease of nitrogen content leads to the significant reduction of biomass production. In contrast, the nitrogen reduction when the C/N ratio increased from 20 to 100 was much smaller, and correspondently, a smaller reduction of biomass was received. Lipid accumulation by oleaginous fungi mostly happens when a nutrient in the medium becomes limited and excessive carbon source is present. Under this circumstance, fungal growth is inhibited and the synthesis of protein and nucleic acid tends to cease. The excessive carbon is preferentially channeled toward lipid synthesis, leading to the accumulation of TAG within intracellular lipid bodies [15]. Nitrogen limitation is considered as the most effective condition for most oleaginous fungal species in lipid accumulation [31], and an optimally high C/N molar ratio for lipid accumulation by fungi and yeasts was reported at the range of 65 to near 100 [6, 26].

Fig. 3
figure 3

Effect of C/N ratio on biomass and lipid content

Full size image

Time-Course Growth of F. equiseti UMN-1 Strain

With the improved M3 medium, the growth and lipid accumulation process of F. equiseti UMN-1 strain is shown in Fig. 4. Day 4 can be considered as the turning point for fungal growth as sugar in medium was consumed at a high rate before day 4 and biomass production had reached a high level on day 4. After that point, the utilization of sugar decreased to a low rate, and biomass in medium slowly increased to its peak value on day 8. Unlike biomass content, the intracellular lipid accumulation continued to increase steadily after day 4 and reached its maximum level of around 50% on day 10. It was normally recognized that lipid accumulation was triggered when cell growth was inhibited by nutrient limitation like nitrogen depletion [15], and F. equiseti UMN-1 strain followed a similar pattern in this growth curve. The lipid content reached 24.94 ± 0.96% on day 2, and at that time, biomass production remained at a high rate. There is high possibility that nitrogen in medium was nearly depleted before day 2, and the increase of fungal biomass after day 2 was mainly the result of lipid accumulation in cells, but not the result of cell proliferation. The high concentration of residual carbon source on day 2 triggered the lipid synthesis at a high rate. As the carbon concentration decreases, the lipid accumulation turns more slowly. More than 80% of sugar had been consumed on day 6, and from days 6 to 10, both biomass amount and lipid content stayed at a relatively constant level. Instead of longer culture time like 2 weeks, a short cultivation time (like 6 days to 8 days) would be considered to increase the lipid yield per unit of time in lipid production process. The overall growth curve of F. equiseti UMN-1 strain indicates this strain’s good capability to utilize simple sugar and rapidly convert that to intracellular lipids, which is a favorable characteristic for oleaginous fungal oil accumulation.

Fig. 4
figure 4

Time-course growth of F. equiseti UMN-1 strain in the optimized M3 medium

Full size image

Effect of Different Carbon Sources and Nitrogen Sources

The study on different source of carbon contained 20 g/L of the following: glucose, xylose, galactose, mannose, fructose, sucrose, lactose, starch, cellulose, cellobiose, xylan, and glycerol (total C was approximately 0.667 mol/L). The source of carbon can influence the fungal growth and lipid synthesis process and, hence, influence the efficiency of lipid accumulation. The results showed that F. equiseti UMN-1 strain can grow well on most of the carbon sources such as fructose, mannose, glucose, and xylose (Fig. 5). But, this Fusarium strain had much less biomass growth when using lactose as carbon source. Some other oleaginous fungi such as Mucor rouxii and Cunninghamella echinulata have also been described to have a poor growth on lactose [32]. Meanwhile, this strain was also able to synthesize a significant amount of lipids on most of the carbon sources explored in this research. Fructose and mannose lead to the highest lipid production, 0.39 ± 0.00 g lipid/100 mL and 0.37 ± 0.01 g lipid/100 mL, respectively. Glucose also gave high lipid production of 0.34 ± 0.03 g lipid/100 mL, and glucose is a more commonly used carbon source in fungal cultivation.

Fig. 5
figure 5

Effect of different carbon sources and nitrogen sources on biomass and lipid content

Full size image

Since this strain was found to have the capability to produce cellulase, the cellulase activity with different carbon substrates was also tested. Xylose, galactose, starch, and cellobiose can all significantly increase cellulase activity, and xylose obtained the best lipid production (0.26 ± 0.03 g lipid/100 mL) as well as the stimulation of cellulase production. They can be added as additional nutrients in fungal culture to enhance cellulase production when lignocellulosic biomass is adapted as carbon source. On the other hand, some carbon sources such as glucose, mannose, fructose, sucrose, and glycerol inhibited the cellulase production and nearly no cellulase activity was detected when F. equiseti UMN-1 strain grew on these substrates. The production of cellulase can be affected by the carbon source in medium, and some simple carbon sources (e.g., glucose) will repress cellulase in some fungi [33].

The oleaginous fungi also showed their preference on the utilization of nitrogen source for growth and lipid accumulation (Fig. 5). Different sources of nitrogen (total N was approximately 0.0178 mol/L) were tested, including (NH4)2SO4, NaNO3, NaNO2, NH4NO3, yeast extract, proteose peptone, urea, 0.956 g/L NaNO3 + 0.75 g/L yeast extract, 0.776 g/L NaNO2 + 0.75 g/L yeast extract, and 0.45 g/L NH4NO3 + 0.75 g/L yeast extract. (NH4)2SO4 was shown as the best inorganic nitrogen source for total lipid production. Fungal growth in medium with (NH4)2SO4 achieved higher biomass and lipid content than NaNO3, NaNO2, and NH4NO3 and obtained a lipid production of 0.27 ± 0.00 g lipid/100 mL. Urea was the best organic nitrogen source among yeast extract, peptone, and urea for total lipid production, which obtained 0.25 ± 0.03 g lipid/100 mL. Moreover, the combination of inorganic nitrogen and organic nitrogen achieved higher lipid production, and the highest lipid production was 0.34 ± 0.03 g lipid/100 mL by using (NH4)2SO4 and yeast extract. This combination was also set as the nitrogen source in the optimized M3 medium for F. equiseti UMN-1 strain. Different fungal species may have different preferences on using nitrogen compounds for growth and lipid accumulation. For example, Mucor rouxii had higher biomass and lipid production in a medium containing KNO3 [32], and C. echinulata had better growth with NH4NO3 and urea while KNO3 was best for its lipid production [34].

The source of carbon and nitrogen may have an influence on fungal growth, lipid accumulation, and also the type of fatty acids synthesized [32]. In this study, F. equiseti UMN-1 strain showed the capability to utilize various monosaccharides, disaccharides, polysaccharides, and glycerol. The wide range of carbon source utilization is a key feature for the oleaginous microbial species. And, it can use both hexose (6-C sugars) and pentose (5-C sugars). These characteristics indicated the strain’s good potential of using all available sugars in lignocellulosic biomass.

Effect of Initial Carbon and Nitrogen Concentrations

Different initial carbon and nitrogen concentrations were tested separately to explore their suitable ranges for lipid production. Study on the effects of initial carbon concentrations had the following initial glucose concentration: 10, 20, 40, 60, 80, 100, 120, and 140 g/L with the nitrogen concentration fixed with 0.75 g/L (NH4)2SO4 + 0.75 g/L yeast extract in the improved M3 medium. When carbon source (glucose) increased from 10 to 20 g/L, both fungal biomass and lipid content reached a higher level. When glucose increased beyond 20 g/L, biomass production raised steadily. On the contrary, lipid content did not get any improvement but gradually decreased to a low level (Fig. 6). This was probably correlated with the increasing C/N ratio as glucose increased. Lipids are generally accumulated under a condition of N starvation, and higher concentrations of sugar should be easier to create such circumstance. However, there is an optimum C/N ratio range for fungal lipid generation; extremely high values of C/N ratio will not promote lipid accumulation but decline lipid content. For the condition with 10 g/L glucose, the total lipid production was the lowest among all sugar concentrations tested while the rest of the concentrations produced a similar amount of lipids with a slight increase on the cultures with 140 g/L glucose. The possible reason for the low lipid yield is the limitation of oxygen during batch culture in flasks. Since a large amount of carbon source was provided, significant amount of sugars still existed in medium when nitrogen in medium was depleted. These sugars can be utilized for lipid accumulation, but oxygen is also required in the synthesis of saturated fatty acids and unsaturated fatty acids [35, 36]. The flask culture method had poor gas transfer efficiency, and the dissolved oxygen in medium was easy to be exhausted. Bioreactor is helpful to avoid this oxygen limitation by continuous supply of air, and the utilization of bioreactor was shown to achieve high lipid content and lipid yield [37]. When employing the flask cultivation method, a high sugar concentration will not be recommended. Results showed that a glucose concentration of around 20 g/L would be more suitable for the flask culture of F. equiseti UMN-1 strain.

Fig. 6
figure 6

Effect of initial carbon and nitrogen concentrations on biomass and lipid content

Full size image

The impact of different nitrogen concentration mixed with equal mass of (NH4)2SO4 and yeast extract fixed with 20 g/L glucose showed a similar trend. The increase of nitrogen source beyond 0.25 g/L N leads to a steady increase of biomass and a decrease of lipid content (Fig. 6). This means high levels of nitrogen will inhibit fungal lipid accumulation in cells. And, more nitrogen existed in medium, resulting in lower C/N ratio; extreme low C/N ratio will also reduce lipid content. 0.1 g/L N and 0.25 g/L N provided the highest lipid production of 0.26 ± 0.01 g lipid/100 mL and 0.25 ± 0.01 g lipid/100 mL, respectively. The suitable range of nitrogen concentration should be 0.1~0.25 g/L N for this strain. In the oleaginous fungal lipid production process, low levels of glucose and/or nitrogen can become a limiting factor since there is not enough carbon and nitrogen source in the medium, and high levels of glucose and/or nitrogen will have negative effects. To guarantee carbon and nitrogen supply for a high cell-density growth in batch culture and also avoid the inhibitory effects, the suitable ranges of carbon and nitrogen are important parameters for F. equiseti UMN-1 strain lipid production.

CCD for Optimized Cultivation Condition

Based on the studies of factors discussed in this experiment, a CCD was developed to determine the optimum values of variables that significantly affect lipid production in F. equiseti UMN-1 strain. The design of three independent variables (temperature, glucose concentration, and nitrogen concentration) in actual values and their response (lipid production) are presented in Table 3. CCD experiment results of lipid production at various conditions ranged from 0.57 to 3.80 g/L. A quadratic model was developed to calculate the optimum levels of temperature and glucose and nitrogen concentrations to determine the maximum lipid production corresponding to these factors. Table 4 shows the ANOVA results of the quadratic model. The F-test value for the overall model is 6.40, which implies the model is significant. And, there is only a 0.38% chance that a large F-test value could occur due to noise. In the quadratic model, A, B, A^2, B^2, and C^2 are significant model terms. The fitness of the model was checked by coefficient (R2 = 0.8520), suggesting that about 14.80% of the total variance could not be explained by this model. Multiple regression analysis on the experiment data provided the following second-order equation to explain the relationship between response variable and the tested variables:

Table 3 Central composition design (CCD) of three variables with lipid production as response value
Full size table
Table 4 ANOVA for the response surface quadratic model for optimization of lipid production
Full size table
$$ {\displaystyle \begin{array}{l}\mathrm{Lipid}\ \mathrm{production}=-22.35+1.45\times \mathrm{Temperature}+0.23\times \mathrm{Glucose}+39.77\times \mathrm{Nitrogen}-\\ {}0.0018\times \mathrm{Temperature}\times \mathrm{Glucose}-0.72\times \mathrm{Temperature}\times \mathrm{Nitrogen}+\\ {}0.015\times \mathrm{Glucose}\times \mathrm{Nitrogen}-0.026\times {\mathrm{Temperature}}^2-0.0026\times {\mathrm{Glucose}}^2-49.19\times {\mathrm{Nitrogen}}^2\end{array}} $$

Besides the regression equation, the two-dimensional (2D) contour plots and three-dimensional (3D) response surface curves provided a graphical solution to identify the types of interactions between every two test variables, located the optimum ranges of variables, and predicted the response value (Fig. 7). In the present study, each figure presented the effect of two test variables on lipid production, while the third variable was fixed at the 0 level. It is obvious from Fig. 7a, c, e that the 3D response surface curves are convex in nature. The optimum values for glucose, nitrogen, and temperature are all located within the ± 1 level range, implying that values for temperature, glucose concentration, and nitrogen concentration are well defined and fitted in this model. From the optimization result of CCD model, the optimal values of the tested variables were predicted as follows: 23.7 °C, 37.39 g/L glucose, and 0.236 g/L nitrogen. The maximum lipid production was predicted as 3.91 g/L. Validation of the experimental design was performed by cultivation of F. equiseti UMN-1 strain under the above optimal condition, and an average lipid production of 3.89 g/L with three replicates was obtained. This validation result was in close agreement with the model-predicted response of 3.91 g/L; therefore, the validation experiments confirmed the predicted values and the accuracy of the model equation. Several oleaginous yeast and fungal strains are capable of accumulating a high amount of lipids inside cells, and their lipid production capability is listed in Tables 5 and 6. Filamentous fungi typically can form cell aggregate pellets and therefore are much easier for the cell harvest, which can be a distinct benefit when applying for oil accumulation [46]. This can enable them to utilize a low concentration of carbons. Even though the strain found this study showed similar or lower biomass and oil production, rare species can directly utilize lignocellulosic biomass as the substrate since the lack of cellulase-producing capability [47]. Considering that this strain produced cellulase while accumulating lipids, this strain may have the capability to utilize lignocellulosic biomass for fungal growth and oil accumulation.

Fig. 7
figure 7

Three-dimensional response surface curves (a, c, e) and 2D contour plots (b, d, f) of oil production (g/L) versus the test variables (glucose and nitrogen at g/L in medium and temperature at °C)

Full size image
Table 5 Examples of fungi and yeast cultivation for oil accumulation
Full size table
Table 6 Fatty acid profile (%) in the lipids produced from F. equiseti UMN-1 strain and other oleaginous fungal strains
Full size table

Conclusions

This experiment explored the growth characteristics of F. equiseti UMN-1 strain, focusing on the improvement of growth condition for the production of intracellular lipids. With the high FAME percentage in fungal lipids, this strain has high potential to contribute in the renewable biodiesel production. Various factors such as temperature, agitation, carbon and nitrogen sources, and growth medium were observed to improve lipid production. This strain is able to achieve a high efficiency of 0.180 g/g sugar under the optimized M3 growth medium. Then, temperature, carbon concentration, and nitrogen concentration were identified as the most important factors in the lipid accumulation process, and a relatively high lipid production of 3.89 g/L was achieved under the optimized growth medium and growth conditions using the CCD model. Furthermore, this strain has the capability to utilize a variety of carbon sources and generate extracellular cellulase, showing a great potential to produce economically feasible lipid using lignocellulosic biomass. As a conclusion, this F. equiseti UMN-1 strain is a good candidate of microbial lipid production to be used in biodiesel industry.