NC State
Coman, G., Leuştean, I., Georgescu, L., and Bahrim, G. (2012). "Optimization of protein production by Geotrichum candidum MIUG 2.15 by cultivation on paper residues, using response surface methodology," BioRes. 7(4), 5290-5303.


Response surface methodology (RSM) based on the 23 factorial central composite design (CCD) was used to optimize the biotechnological conditions for growth and protein production by a selected fungal strain Geotrichum candidum MIUG 2.15, by solid-state cultivation on a semisolid medium based on a mixture of paper residues, i.e. office paper, newspaper, and cardboard, mixed in a ratio of 1:1:1(w/w), supplemented with cheese whey waste and complex manure. Three independent variables, the solid:liquid ratio, the concentration of complex manure, and cultivation time, were evaluated to determine their correlative effect on biomass production and protein biosynthesis. The optimal conditions for obtaining a maximum protein yield of 9.53% w/w dry mass were the following: the complex manure concentration of 0.5%, the solid:liquid ratio of 1:5, and the growth time of 10 days.

Download PDF

Full Article


Gigi Coman, Iuliana Leuştean, Luminita Georgescu, and Gabriela Bahrim*

Response surface methodology (RSM) based on the 23 factorial central composite design (CCD) was used to optimize the biotechnological conditions for growth and protein production by a selected fungal strain Geotrichum candidum MIUG 2.15, by solid-state cultivation on a semisolid medium based on a mixture of paper residues, i.e. office paper, newspaper, and cardboard, mixed in a ratio of 1:1:1(w/w), supplemented with cheese whey waste and complex manure. Three independent variables, the solid:liquid ratio, the concentration of complex manure, and cultivation time, were evaluated to determine their correlative effect on biomass production and protein biosynthesis. The optimal conditions for obtaining a maximum protein yield of 9.53% w/w dry mass were the following: the complex manure concentration of 0.5%, the solid:liquid ratio of 1:5, and the growth time of 10 days.

Keywords: Geotrichum candidum; Protein production; Solid-state fermentation; Paper residues; Response surface methodology (RSM)

Contact information: “Dunărea de Jos” University of Galati, Faculty of Food Science and Engineering, Bioengineering Department, 111 Domnească Street, 800201, Galaţi, Romania

* Corresponding author:


Bioconversion processes have been developed for the utilization of renewable resources to produce useful chemicals and feed stocks. The use of renewable resources and, in particular, of hemicellulosic biomass, as substrate in biotechnological processes has broad economic and environmental implications.

Lignocellulosic wastes from different sources have varying composition of hemicellulose, cellulose, and lignin. Some sources of lignocellulosic material are wood from angiosperms and gymnosperms, grasses, leaves, wastes from paper manufacture, sugarcane bagasse, wheat straw, wheat bran, rice bran, groundnut shell, and other agricultural wastes (Bahrim 2004).

Fungal single cell protein (SCP) production can be a very attractive alternative to valorize a variety of agro-industrial by-products and municipal wastes, including molasses, hydrocarbons, lignocellulosic materials, waste from wood industry, paper residues, cheese whey waste, wastewaters, etc.

Few applications for SCP production have been realized at the laboratory and pilot-scale levels by using paper residues as fermentative substrate. The composition of paper is from 40% to 80% cellulose, 20% to 30% lignin, and 10% to 30% hemicellulose and xylosans (

Prior to use, the paper residues need preliminary treatment; the polymers are hydrolyzed in order to obtain simple components which can be then easily metabolized by cultivated microorganisms (Leuștean et al. 2012). After hydrolysis, carbon is the main chemical component. For this reason, mixing the paper residues with organic and inorganic nitrogen sources (cheese whey waste, complex manure, etc.) contributes to balancing the composition of the fermentative medium from a nutritional point of view. Whey contains carbon and nitrogen compounds assumable by the microorganisms and is widely available. Therefore it could be used for the production of microbial biomass. Valorization of whey is also interesting because its chemical oxygen demand is elevated by high concentrations of soluble material that are extremely polluting for the environment, and there is an associated need to protect the environment from the extremely large quantities of whey that are generated by the industrial production of cheese (Spălățelu 2012).

The protein obtained from the microorganisms is not only cheap but also may provide a balanced nutrition for humans and animals (Rajoka et al. 2006). SCP is normally considered to be a valuable source of protein, but it also contains nucleic acids, carbohydrates, cell wall material, lipids, minerals, and vitamins (Ugalde and Castrillo 2005).

The valorization of paper residues and cheese whey waste as abundant and low-cost raw materials in the production of fungal biomass which could be used as a protein supplement (SCP) for animal feed production would not only be economically viable but would also solve problems caused by the accumulation of organic wastes and protect the environment (Pandey 2003; Silva et al. 2011).

Most of the research on SCP production has been focused on the use of the yeast Candida utilis (formerly Torulopsis utilisTorula utilis) strains. The mold Geotrichum candidum has also been the subject of biochemical and physiological studies due to common occurrence and its biotechnological value. The strains are able to grow on different substrates and produce many enzymes such as: cellulases, xylanases, lipase, proteases, or peroxidases (lignin peroxidase and manganese peroxidase) (Boutrou and Guéguen 2005; Witkowska and Piegza 2006; Asses et al. 2009). The production of enzymes varies greatly from one strain to another. Based on this multiple enzymatic potential, Geotrichum candidum strains are able to grow well on solid and liquid organic wastes in order to obtain organic decontamination and also to produce fungal biomass rich in proteins (Zara 1999; Asses et al. 2009).

Response surface methodology is a three-factorial design method, which provides the relationship between one or more measured dependent responses and a number of input (independent) factors. The response surface method is advantageous because it requires a small number of experiments, it is suitable for multiple factor experiments, it seeks relativity between multiple factor experiments, and it finds the most suitable correlation and forecast. Therefore, it finds the optimum values of the factors under investigation, and it predicts the response to the optimum conditions (Popa et al. 2007).

Limitations of the single-factor optimization can be eliminated by employing response surface methodology (RSM), which is used to explain the combined effects of all the factors in a fermentation process (Popa et al. 2007; Zheng et al. 2008; Montgomery 2005).

In the present work, a central composite design (CCD) of response surface methodology (RSM) has been used to optimize the biotechnological parameters of the solid-state growth and protein biosynthesis for Geotrichum candidum MIUG 2.15 strain by cultivation in stationary conditions on a semisolid medium based on paper residues supplemented with cheese whey waste and complex manure.


Materials and Methods


All chemicals were purchased from Sigma-Aldrich and used without further purification. The complex manure containing N-P-K, in ratio of 7:4:5 w/v, used to enhance the nutritive value of the substrate, is a commercial biofertilizer purchased from a specialized market from Galati, Romania. Cheese whey waste was obtained from Galacta SA dairy factory from Galati, Romania.

Mold strain

Geotrichum candidum MIUG 2.15 strain is member of the Collection of microorganisms of Bioaliment Research Platform of “Dunărea de Jos” University of Galati, Romania. It was maintained on agar slants supplemented with (% w/v): glucose 2, peptone 1, yeast extract 0.5, and agar 2, pH = 5.0, and kept at 4 °C.

Paper residues treatment

The cellulosic waste materials used as the solid substrate of the culture medium consisted of a mixture of three components based on paper residues: office paper, newspaper, and cardboard, mixed in a ratio of 1:1:1 (w/w). The materials used were milled using a vibratory ball mill (Janke and Kukel GmbH and Co. Ika Labortechnik), and dried to 80% dry matter, at 60 °C, for 6 hours, using a laboratory drying oven (MMM Group, GmbH Germany). The chemical characteristics of waste cellulosic material used as substrate were determined according to standard procedures in accordance with the AOAC method (Table 1).

Table 1. The Chemical Composition of Cellulosic Waste Substrate Based on Paper Residues

Culture conditions and protein yield determination

A semisolid medium was prepared, containing 4 g of milled untreated mixed cellulosic material based on three paper residues and cheese whey waste in a solid:liquid ratio varying between 1:3.3 and 1:6.7. It was then supplemented with complex manure, containing 33.5% total nitrogen (16.7% ammonium nitrogen and 16.8% nitrogen in nitrates). The physico-chemical characteristics of whey were evaluated using Milk-Lab UK Ltd (Oldham, UK) and are shown in Table 2.

Table 2. Physico-chemical Characteristics of Whey

The medium was prepared in Petri dishes and was then sterilized at 120 °C for 30 minutes. The sterile medium was inoculated with 107 CFU/g substrate suspension of Geotrichum candidum MIUG 2.15 conidiospores, and then the cultivation was performed in stationary conditions at 20 °C. After cultivation, the total nitrogen content (N) expressed as % w/w dry mass of solid-state fermented crude product was measured by using the Kjeldahl system (VELP, Italia). The crude protein value was expressed as N x 6.25. The assays were performed in duplicate.

Response surface methodology for the optimization of protein production

A factorial central composite design with three factors, at the central point and star points, was used for the investigation (Agaie et al. 2009). The used independent variables were: complex manure concentration as inorganic nitrogen and phosphorous source, solid:liquid ratio, and time of cultivation, each at five coded levels (-α, -1, 0, +1, +α) as shown in Table 3. The response value (Y) in each trial is the average of duplicates.

Table 3. The Independent Variables and Their Levels for the Central Composite Experimental Design

The chosen independent variables used in this experiment were coded according to the following equation,

xi = (X– Xo) / X, for i = 1, 2, … , k (1)

where, xi is the dimensionless value of a variable, Xi the actual value of a variable, X0 the value of xi at the center point, and ∆X is the step change.

The second order polynomial coefficients were calculated and analyzed using the ‘Design Expert’ software statistical package ( The general form of the second degree polynomial Eq. (2) is:


where, Y is the predicted response, Xi and Xj are the input variables, β0 is the intercept term, βi is the linear effect, βii is the squared effect, and βij is the interaction term.

Statistical analysis of the model was performed to evaluate the analysis of variance (ANOVA). This analysis included the Fisher’s F-test (overall model significance), its associated probability p(F), the correlation coefficient R, and the determination coefficient R2 which measures how well the regression model fits the data. For each variable, the quadratic models were represented as contour plots (3D), and response surface curves were generated using Design Expert 7.0 software version 6.0, Stat-Ease Inc., Minneapolis, USA (Kim et al. 2008).


Previous studies based on the experimental Plackett–Burman design made it possible to verify the variables that had an effect on paper residues biotransformation. Among the variables tested, the solid:liquid ratio in the fermentative medium composition has a positive influence, because the optimum water content is essential for the growth and metabolic activity of the microorganism (Leuștean et al. 2010).

The physiological and biochemical characteristics of Geotrichum candidum MIUG 2.15 strain on cultivation on media based on whey were also investigated previously (Zara 1999; Palela et al. 2008, 2010).

In the present study, the correlative effect of three variables, solid:liquid ratio of the semisolid medium, concentration of complex manure, and time of cultivation, upon Geotrichum candidum MIUG 2.15 growth and protein production by cultivation in solid-state system, were investigated by statistical optimization using response surface methodology (RSM), in order to find the optimum conditions to increase the crude protein content of the final fermented product.

The Central Composite Design consisted of 15 experimental trials and is shown in Table 4.

The final model equation is shown as (Eq. 3):

Y = 4.02– 0,055A + 0.80B – 0.98C – 1.49AB + 1.42AC – 0.23BC +0.11A2 + 0.31B2 + 0.55C2 + 1.35ABC – 0.17A2B + 1.23A2C (3)

The statistical significance of the model equation was checked using F-test analysis of variance (ANOVA). The fitness of the models was also expressed by the coefficient of determination, R2, for the quadratic model, which was found to be 0.9797 for the production of protein. This value indicates there was 97.97% of response variability in protein production.

Table 4. The Experimental Design of the Biotechnological Conditions (Independent Variables) of Protein Production (Response) by Cultivation of Geotrichum candidum MIUG 2.15 Strain in Solid State Fermentation System on Cellulosic Waste Based on Paper Residues

Table 5. Analysis of Variance (ANOVA)

The Model F-value of 17.24 implies the model is significant. There is only a 0.27% chance that a “Model F-Value” this large could occur due to noise. “Model P-value” less than 0.0500 indicates that the model terms are significant. In this case B, C, AB, AC, C2, ABC, A2C, and A2B2 are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms, the model reduction may improve the model. The significant lack of fit is considered to be bad if one wants the model to fit (Table 5).

In Table 6, the negative “Predicted R-Squared” implies that the overall average is a better predictor of the response than the current model. The “Adequate Precision” measures the signal to noise ratio. A ratio greater than 4 is desirable. The ratio of 14.709 indicates an adequate signal.

Table 6. Statistic Parameters for Central Composite Design Model

Fig.1. Response surface plot (a) and contour plot (b) showing the effect of complex manure concentration and solid:liquid ratio on protein production by Geotrichum candidum

As can be seen in Fig. 1 (a, b), the protein yield increased when the complex manure concentration increased to 0.5%, and when the solid:liquid ratio increased up to the level of 1:4.

A small increase can be observed in Fig. 2 (a, b) for the percentage of 0.5% complex manure concentration over a time of cultivation of 10 days.

Fig. 2. Response surface plot (a) and contour plot (b) showing the effect of complex manure concentration and time of cultivation on protein production by Geotrichum candidum MIUG 2.15

Figure 3 (a, b) shows that the protein content increased at a solid:liquid ratio of 1:4 and a cultivation time of 6 days. The contour plots shows that two by two, the factors influence each other, the complex manure concentration, the solid:liquid ratio and time of cultivation.

Fig. 3. Response surface plot (a) and contour plot (b) showing the effect of the time of cultivation and solid:liquid ratio on protein production by Geotrichum candidum MIUG 2.15

Figure 4 also shows that the protein production was influenced by the time of cultivation (C) and by the solid:liquid ratio. The corresponding slopes are more pronounced and much more extended compared to those related to complex manure concentration, which are almost linear.

Fig. 4. Perturbation evolution for three factors involved in protein production by Geotrichum candidum MIUG 2.15. A – complex manure concentration, B – solid:liquid ratio, C – time of cultivation

Analyzing the results in the cube representation (Fig. 5), one can see a maximal protein yield (12.079 % w/w dry biomass) corresponding to the following conditions of cultivation conditions: 0.1% complex manure concentration, solid:liquid ratio of 1:4, and time of cultivation of 6 days.

Fig. 5. Cube plot showing the influence of the relevant factors involved in the protein production by Geotrichum candidum MIUG 2.15 in solid-state fermentation system on paper residues A – complex manure concentration, B – solid:liquid ratio, C – time of cultivation

The small differences between the experimental data and the prediction of the model chosen can be observed (Fig. 6).

Fig. 6. Parity plot showing the distribution of experimental vs. predicted values

The numerical method ( was used to solve the regression equation. The model validation was conducted using the same bioconversion conditions, as well as the comparative model to the predicted values (Table 7).

Table 7. Model Validation (Biotechnological Conditions, Responses) For Protein Production by Geotrichum candidum MIUG 2.15 by Cultivation in Solid State Fermentation System on Paper Residues

To confirm the predicted optimum biotechnological conditions for protein production by Geotrichum candidum MIUG 2.15 by solid-state cultivation on paper residues, the theoretical optimum conditions were used. The model predicted that the maximum yield of protein would be 9.066 % w/w dry mass. The experimental results of 9.53 % w/w dry mass were in close agreement with the model prediction. The predicted yield of protein value was calculated through the model equation (Eq. 3).

Rao et al. (2010) reported obtaining of 46% microbial protein production by cultivation of Penicillium janthinellum selected strain on medium containing bagasse hydrolysate, ammonium sulphate, and potassium dihydrogen phosphate. Miller and Srinivasan (1983) obtained 23 to 38% protein content of biomass by cultivation of a thermotolerant strain of Aspergillus terreus on cellulose with a rate of biotransformation of the substrate of 78 to 84%. Khan and Dahot (2010) investigated the effect of agricultural wastes (sugar cane bagasse, orange peel, wheat straw, and rice husk), some plant seeds (cotton seeds, cajanus cajan seeds, and castor beans), and pure sugars (mannose, glucose, fructose, galactose, maltose, lactose, lactose, sucrose, starch, cellulose) on the production of SCP by Penicillium expansum strainThe maximum yields of biomass (1.64 g/l) and protein (18.25%) were obtained when rice husk was used as carbon source. Higher yields of biomass (1.92 g/l) and protein (21.36%) production were obtained when acid hydrolysates of rice husk and cotton seeds were mixed.

Although the results reported in the literature relating fungal crude protein production by biovalorisation of agro-industrial wastes are higher than those highlighted in this study, our study is original compared with the few data existing in scientific literature, regarding the wastes and fungus strains used in biotechnological processes. Therefore, the biotechnological process proposed by this study is very important because the paper residues are used as solid substrate without other treatments (pretreatments, chemical or enzymatic saccharification), the Geotrichum candidum MIUG 2.15 strain has a good enzymatic potential and whey and complex manure play an important role in fungal metabolic activity stimulation.

The utilization of two abundant, recalcitrant, and low-cost raw materials (paper residues and whey) for the production of crude protein has great economic impact for high grade animal feed production and it also solves problems caused by organic wastes accumulation and environmental pollution.


The growth ability and protein biosynthesis of Geotrichum candidum MIUG 2.15 strain by solid-state cultivation on a medium based on a mix of three paper residues and whey were studied. The following conclusions were drawn from the results:

  1. The statistical optimization of growth of the mold strain on paper residues in the solid-state fermentation system has been successfully carried out by using RSM, based on the 23factorial CCD, in order to increase the yield of proteins in controlled biotechnological conditions, such as time of cultivation, complex manure concentration, and solid:liquid ratio.
  2. The optimal conditions for obtaining a maximum protein yield of 9.53% w/w dry mass were the following: a complex manure concentration of 0.5%, a solid:liquid ratio of 1:5, and a cultivation time of 10 days.
  3. After cultivating the Geotrichum candidum MIUG 2.15 strain on the paper residue substrate supplemented with whey and complex manure, the cellulose content was 53% lower, while the protein content was increased by 6.7%.
  4. The results suggest that the paper residues supplemented with cheese whey waste could be valuable substrates for SCP production with the Geotrichum candidum selected strains, resulting in a fermented product with useful application in animal feeding and that could also be a good opportunity to minimize the environmental pollution caused by the agro-industrial and municipal by-products.


The authors acknowledge material and financial assistance from the Integrated Center for Research and Education, Biotechnology Applied in Food Industry – Bioaliment(


Agaie, E., Pazouki, M., Hosseini, M. R., Ranjbar, M., and Ghavipanjeh, F. (2009). “Response surface methodology (RSM) analysis of organic acid production for kaolin beneficiation by Aspergillus niger,”Chemical Engineering Journal 147, 245-251.

Bahrim, G. (2004).“Agricultural biowaste as resources for fodder yeast additives development,”Roumanian Biotechnological Letters 9, 1751-1756.

Boutrou, R., and Guéguen, T. M. (2005). “Interests in Geotrichum candidum for cheese technology,” International Journal of Food Microbiology 102, 1-20.

Asses, N., Ayed, L., Bouallagui ,H., Rejeb, B. I., Gargouri, M., and Hamdi, M. (2009). “Use of Geotrichum candidum for olive mill wastewater treatment in submerged and static culture,” Bioresource Technology 100, 2182-2188.

Khan, I. M. Y., and Dahot, M. U. (2010), “Effect of various agriculture wastes and pure sugars on the production of single cell protein by Penicillium expansum,” World Applied Sciences Journal (Special Issue of Biotechnology & Genetic Engineering) 8, 80-84.

Kim, K. J., Oh, B., Shin, H., Eom, C., and Kim, S. (2008). “Statistical optimization of enzymatic saccharification and ethanol fermentation using food waste,” Process Biochemistry 43, 1308-1312.

Leuștean, I., Coman, G., and Bahrim, G. (2010). “The Plaket-Burman model – An improved alternative to indentify the significant factors implied in the bioconversion of the complex cellulosic waste to ethanol,” Innovative Romanian Food Biotechnology 7, 55-60.

Leuştean, I., Coman, G., and Bahrim, G. (2012). “Statistical optimisation of ethanol production from a cellulosic mixture based on paper residues,” Environmental Engineering and Management Journal 11, 1037-1044.

Miller, T. F., and Srinivasan, V. R. (1983). “Production of single-cell protein from cellulose by Aspergillus terreus,” Biotechnology and Bioengineering 25, 1509-1519.

Montgomery, D. C. (2005). “Design and Analysis of Experiments,” Wiley, New York.

Palela, M., Ifrim, G., and Bahrim, G. (2008). “Microbiological and biochemical characterization of dairy and brewery wastewater microbiota,” The Annals of the University Dunărea de Jos of GalaţiFascicle VI–Food Technology 31, 23-30.

Palela, M., Ifrim, G., Barbu, M., Bahrim, G., and Caraman, S. (2010). “Strategies for the aerobic biological treatment of the dairy wastewaters in controlled conditions,” Environmental Engineering and Management Journal 9, 399-405.

Pandey, A. (2003). “Solid state fermentation,” Biochemical Engineering Journal 13, 81-84.

Popa, O., Babeanu, N.,Vamanu, A., and Vamanu, E. (2007). “The utilization of the response surface methodology for the optimization of cultivation medium and growth parameters in the cultivation of the yeast strain S. cerevisiae 3.20 on ethanol,” African Journal of Biotechnology 6, 2700-2707.

Rajoka, M. I., Khan, S. H., Jabbar, M. A., Awan, M. S., and Hashmi, A. S. (2006). “Kinetics of batch single cell protein production from rice polishing with Candida utilis in continuously aerated tank reactors,” Bioresource Technology 97, 1934-1941.

Rao, M., Varma, A. J., and Deshmukh, S. S. (2010). “Production of single cell protein, essential amino acids, and xylanase by Penicillium janthinellum, BioResources 5, 2470 -2477.

Silva, C. F., Arcuri, S. L., Campos, C. R., Vilela, D. M., Alves José, G. L. F., and Schwan, R. F. (2011). “Using the residue of spirit production and bio-ethanol for protein production by yeasts,” Waste Management 31, 108-114.

Spălățelu, C. (2012). “Biotechnological valorisation of whey,” Innovative Romanian Food Biotechnology 10, 1-8.

Zara, M. (1999). “Biovalorization of cheese whey waste,” PhD Thesis, Dunărea de Jos University of Galati, România.

Zheng, Z. M., Hu, Q. I., Han, J., Xu, F., Guo, N. N., Sun, Y., and Liu, D. H. (2008). “Statistical optimization of culture conditions for 1,3-propanediol by Klebsiella pneumoniae AC15 via Central composite design,”Bioresource Technology 99, 102-105.

Witkowska, D., and Piegza, M. (2006). “Capability of Geotrichum candidum yeasts for cellulases and xylanases biosynthesis,” Electronic Journal of Polish Agricultural Universities 9, 41 (available online:

Ugalde, U. O., and Castrillo, J. I. (2005). “Single cell proteins from fungi and yeasts,” In: Khachatourians, G. G., Arora, D. K., and Berka, R. M. (Eds.), Applied Mycology and Biotechnology. Agriculture and Food Production, Vol. 2, Elsevier, London, pp. 123-150. – access 20.04.2010 – access 14.08.2012

Article submitted: March 30, 2012; Peer review completed: August 9, 2012; Revised version received: August 30, 2012; Accepted: August 31, 2012; Published: September 12, 2012.