In recent years, mushrooms have become a valuable source with various bioactive ingredients [1,2]. Cordyceps militaris (C. militaris ), belonging to the class Ascomycetes and Dong Chong Xia Cao group [3], has been used as a folk tonic food and an important medicinal mushrooms in Asia extensively [4]. Cordycepin (3'-deoxyadenosine Figure 1), one of naturally isolated nucleoside antibiotics, was the major active metabolite of C. militaris [5]. Recent studies have demonstrated that cordycepin exhibited multiple pharmacological actions, to be specific, immunological regulation [6-8], antivirus [9], antihyperlipidemia [10], antifungus [11], antileukemia [12-14], anticancer effects [15]. Recently, with the advance of the standard of living, there is an increasing requirement for large amounts of cordycepin. Due to strict requirement for host and living conditions, coupling with over harvesting, natural C. militaris are facing with extinction along with the change of environment [16,17]. Therefore, the limited natural resources cannot meet the demand for health food application or herbal medicine. It is urgent to find the effective methods to produce a great deal of cordycepin.

Figure 1: Chemical structure of cordycepin

Although cordycepin could be chemically synthesized absolutely. The synthesis process was complicated, and a large volume of organic solvents was discharged, which was adverse to human health and environment, so it was not widely applied to industrial production [18,19]. Recent studies indicated that solid and liquid fermentation were used to produce cordycepin [20]. However the solid cultivation need to take several months to get a fruiting body with a lower productivity of bioactive ingredients. Liquid fermentation yields potential dominant of shorter time and higher mycelial production. What's more, There were similar chemical ingredients and pharmacologic effects between fermentation broth and wild C. militaris [21]. The process of liquid fermentation also could be optimized to achieve a higher productivity. Therefore, liquid cultivation was viewed as a promising field to generate artificial cordycepin production. Many scholars studied a good deal of strategies to meet the increasing demand of cordycepin since 1960s and made remarkable achievements. They focused their attention on the three aspects: strain improving, optimizing ingredients of medium (selection of carbon and nitrogen sources, precursors, mineral ion) and optimizing culture conditions. The aims of this review were to describe the cordycepin biosynthesis and summarize the strategies for increasing cordycepin production of C. militaris by liquid fermentation.

Strain improving

Different strains had different productivity. Strain improving maybe enhance the production of secondary metabolites dramatically. So many researchers studied a series of paths to obtain mutations which had higher productivity of cordycepin. Mutants were generated by spontaneous mutation, ultraviolet light, ionizing radiation, (X-rays, gamma-rays, ion beam), chemical mutagens (ethyl methane sulfonate) and biological mutagens (transposon tagging, T-DNA insertion).

Das et al. [22] applied a high-energy (~MeV) proton beam which provided a higher mutation frequency and wider mutant spectrum to irradiate C. militaris NBRC9787 and obtained a high-yielding mutant G81-3. After optimizing ingredients of medium (Bacto yeast extract (YE) 45 g/L, glucose 50 g/L and a few major inorganic salts), the cordycepin production by G81-3 was up to 3.1 g/L, which was 72% higher than that of basal medium (1.8 g/L).

Optimizing ingredients of medium

There are a variety of ingredients in the medium: carbon source, nitrogen source, phosphorus source, sulfur source, mineral ion, growth factor, precursor, inducer, accelerant and inhibitor. Each of them may be significant to the growth of cells and the formation of any cultivation products.

Selection of carbon and nitrogen sources: Carbon and nitrogen sources, two major nutrient ingredients of medium, were essential for cell proliferation and metabolite biosynthesis. There were many studies on optimizing carbon sources, nitrogen sources and their optimal concentrations for mycelial growth and metabolites formation.

Mao and co-workers [23] investigated various carbon sources (lactose, sucrose, glucose, fructose, galactose, maltose and xylose) and found that glucose was most favourable to cordycepin production. They also studied the effect of carbon/nitrogen ratios to improve the accumulation of cordycepin. The maximum production of cordycepin they obtained was 345.4 ± 8.5 mg/L with 42.0 g/L glucose and 15.8 g/L peptone. Mao [24] investigated masses of nitrogen sources, including YE, casein enzymatic hydroilsate, peptone, casein acid hydrolysate, combination of YE and peptone at 1:1 (w/w). Their investigation showed that all of nitrogen sources could support the growth of cells, but the effect on cordycepin production were different. YE was beneficial to growth of mycelial, while peptone was best for cordycepin biosynthesis. Another report by Masuda [25] showed that the mixture of peptone and YE (peptone:YE = 1:3) were the preferable nitrogen sources for generation of cordycepin, and the peak concentration of cordycepin they obtained was 0.64 g/L when glucose/mixed nitrogen source was 2/1. Both studies of Mao and Masuda showed that peptone was best for the synthesis of cordycepin. It might because that peptone, which composed by 20 kinds of amino acids and NH4⁺ [26,27], was decomposed and consumed by mushroom to form the secondary metabolites [27,28].

In addition to organic nitrogen sources, inorganic nitrogen sources also had positive impact on growth of organisms. Ammonium was the most common inorganic nitrogen sources. Mao and Zhong [24] examined the effects of ammonium feeding on the cordycepin production by submerged cultivation of C.militaris. About 70% increase in maximum cordycepin production was achieved in feeding of NH4⁺ (40mM) on day 7, reaching to 420.5 ± 15.1 mg/L. Their experiment showed that ammonium feeding was a simple and effective strategy for increasing the production of cordycepin in mycelial cultures. Leung [29] also studied the effects of ammonium feeding on the cordycepin production in mycelial culture of Cordycepes Sinensis HK1. The yield of cordycepin with 10 mmol/L NH₄CL corresponds to four times than that of control group.

Adding additives

Precursors: So far, there were several reports related to cordycepin biosynthesis of C. militaris and it has not been completely clarified. Adding some precursors may increase cordycepin significantly, so it was very important to know the biosynthetic pathway of cordycepin. Previous studies on incorporation ³H-labeled ribose and ¹⁴C-labeled glucose, adenine, adenosine into cordycepin demonstrated that most of them acted as underlying precursors [30,31]. Cordycepin was intracellularly converted into its 5'-mono-, di- and triphosphates that inhibited the activity of 5-phosphoribosyl-1-pyrophosphate amidotransferase and ribose-phosphate pyrophosphokinase in the de novo biosynthesis of purines [32-35]. Lennon’s investigation showed that the ³H: ¹⁴C ratio of the AMP isolated from the RNA of C. militaris was identical to that of cordycepin. They also showed that adenosine converted to cordycepin by a reductive mechanism without hydrolysis of the N-riboside bond. Taking the above studies into consideration [30-35], the cordycepin biosynthesis may be shown in the figure 2. Glucose firstly turned into glucose-6-phosphate (G-6-P) and then transformed into ribose-5-phosphate (R-5-P) by pentose phosphate pathway (pp pathway), R-5-P was the starting material of the de novo purine nucleotide pathway. The de novo purine nucleotide pathway involved serial conversions of phosphoribosyl pyrophosphate (PRPP) to IMP and then to AMP and GMP. Glutamine and Glycine also took part in the serial conversions. But, the mechanisms that AMP, adenine and adenosine converted into cordycepin were unknown, which maybe became research focus in the days ahead.

Figure 2: The de novo purine nucleotide pathway and possible biosynthetic pathway of cordycepin. Question mark indicates those reactions are unknown.

Mina Masuda [36] added a great deal of purine-related compounds to increase the production of crodycepin by a surface culture of C. militaris NBRC 9787. The peak value of cordycepin they obtained was 2.5 g/L under the condition of 1 g/L adenine and 16 g/L glycine. Their study also showed that adding of L-glutamine, glycine, adenosine, adenine and L-aspartic were effective methods to enhance the cordycepin production. Das [37] also explored the effects of different precursors (glycine, adenosine) on the production of crodycepin using the mutant G81-3 in a surface liquid culture. The highest crodycepin production reached 8.57 g/L when adding 6 g/L adenosine in medium (glucose 86.2 g/L, YE 93.8 g/L), about 28.10% increase (from 6.69 g/L to 8.57 g/L) in crodycepin production was observed after adding adenosine. Similarly, glycine also had influence on the production of crodycepin, the best production (6.80 g/L) was obtained when adding glycine with YE in weight percent ratio of 90/10, which was 12.40% higher versus the control group (6.05 g/L). They demonstrated that adenosine had a much better effect on cordycepin production than that of glycine.

Mineral ion: Mineral ion, one of the constituents of culture medium, also played a pivotal role in the growth of cells. There were some studies on the effects of mineral ion on the generation of secondary metabolites of C. militaris.

A recent report by Fan [38] showed that some metal ions (Cu²⁺, Ca²⁺, Mn²⁺, Fe³⁺ and Fe²⁺) could markedly enhance the production of cordycepin, among of them Fe²⁺ was the most effective. While Zn²⁺ dramatically reduced cordycepin production. As a result, adding 1 g/L ferrous sulfate on day 0, the production of cordycepin could reach 596.59 ± 85.5 mg/L, which was 70% higher than the group that without ferrous sulfate. Another research [39] showed that 0.1mM Mn²⁺ can strikingly promote synthesis of some nucleosides like adenosine and guanosine, which may be benefit to synthesis of cordycepin.

Optimizing culture conditions

The modes of propagation: As everyone knows, oxygen supply was crucial for aerobic organisms to cells growth and formation of secondary metabolites. Due to poor solubility of oxygen in water, different modes of propagation (submerged culture, surface liquid culture and the repeated batch culture) were investigated to control oxygen availability in medium.

Some authors investigated the effects of dissolved oxygen (DO) on cordycepin formation by C. militaris. Mina Masuda et al. demonstrated that the repeated batch culture was a simple method to increase cordycepin production [36]. Shih et al. [40] combined shakeflask with static culture to facilitate the production of cordycepin. The maximum production they obtained was 2214.5 mg/L by C. militaris CCRC 32219, the optimized conditions were at PH 6, YE 45 g/L, 8.0 days of the shake cultivation followed by 16.0 days of the static culture. Their investigation also showed that two-stage dissolved oxygen control was good for cordycepin formation. Perhaps, in the early time, cells needed more oxygen to growth, but cordycepin was synthesized in hypoxia state.

Other factors: Living environment (light condition, PH, temperature), inoculum size, incubation time and seed age also had an impact on the growth of organisms and the formation of metabolites.

Tang et al. [41] optimized the fermentation conditions and the ingredients of medium for enhancing the production of cordycepin with Placket-Burman, single-factor experiment in static culture. The maximum cordycepin yield (7.35 g/L) was obtained under followed conditions: incubation temperature of 27.1, inoculum size of 10%, seed age of 3 days, YE 9.00 g/L, peptone 17.10 g/L. In the progress of cultivation of organisms, culture time also had a great influence on the yield of cordycepin. Mastering the law of cordycepin formation and controlling the point of culture time also could obtain much higher output. The investigation by Masuda et al. [42] showed that the day when cordycepin production up to the peak by mutant strains was latter than that of wild strains about 10 days. The maximum cordycepin production they obtained was 8.6 g/L of mutant G81-3 by the repeated batch culture. Up to now, this was the highest report of crodycepin product.

Table 1 summarized the production disparity of cordycepin under the different conditions of liquid fermentation. We could conclude that the productivity of mutant stronger than that of wild C. militaris . The concentrations of carbon and nitrogen sources also had influence on production of cordycepin. Adding additives could enhance cordycepin production. The formation of cordycepin preferred surface liquid culture or the repeated operation to submerged culture.

Table 1: The effects on Cordycepin production by some impact factors.

The aims of strain improving were to break the normal metabolism mechanism, terminate microorganisms own regulating mechanism and increase accumulation of target products. It can be achieved by mutation breeding. The ion beams could induce nuclear DNA alterations such as transversion, inversion, translocation and large deletions rather than point mutations, so it could produce various types of mutants with broad-spectrum mutation [44,45]. A high-yielding mutant (G81-3) of C. militaris was obtain by irradiation of ion beams. The ion beams also could be used in other fungi research.

The types and concentrations of carbon and nitrogen sources had effects on yield of secondary metabolites. So it was very necessary to optimize the compositions of medium. Carbon and nitrogen sources, two major ingredients of medium, consisted of basic skeleton of cultivation products. Carbon sources also had influence on respiration of microorganisms, glycometabolism and growth of cells. In mushroom cultivation of Tremella mesenterica, Wasser et al. [46] demonstrated that various kinds of nitrogen sources influenced polysaccharide production and growth of cells markedly. Another report by Cho showed that peptone-YE was best for the mycelial growth of mushroom Paecilomyces sinclairi , while meat peptone was most favorable to its red pigment production [47]. Glucose was the best carbon source for cordycepin, which may be the starting material of the biosynthesis pathway (Figure 2). Peptone was optimum nitrogen source for the formation of cordycepin.

In many studies [48,49], NH4⁺ showed negative effects on secondary metabolic pathways for nitrogen catabolic repression on the microbial secondary metabolism. Cho [47] also found similar report in cultivation of Paecilomyces sinclairii. However, the investigation by Mao [24] showed that NH4⁺ could promote cordycepin production. They added inhibitor of plasma membrane H⁺-ATPase (diethylsilbestrol or sodium orthovanadate) to explore related mechanisms of NH4⁺ in cordycepin biosynthesis. They found that adding enzyme inhibitor led accumulation of intracellular ATP and decrease of cordycepin significantly (Figure 2). They supposed that NH4⁺ maybe stimulate the activity of H+-ATPase to promote cordycepin formation. There are similar phenomena in cultivation of Aspergillus niger [50] and Penicillium cyclopium [51].

There were some reports about the effects of trace metals on the cultivation of fungus [52,53]. Fan et al. [38] also studied the effect of ferrous sulfate addition on cordycepin production. Their experiment showed that feeding of Fe²⁺ could increase cordycepin production by raising the transcription level of adenylosuccinate synthetase (purA) and cutting the transcription level of IMP cyclohydrolase (purH) and IMP dehydrogenase (guaB) (Figure 2).

In conclusion, both the strain improving and additives had effect on cordycepin production predominately. Fe²⁺ and NH4⁺ were found as efficient inducers for cordycepin biosynthesis. Adding purine-related compounds (precursors), feeding of Fe²⁺ , NH4⁺ and the repeated batch culture were simple and effective strategies for increasing cordycepin production of C. militaris by liquid fermentation. This study may be useful for other fungi research.

This study was supported by Program for New Century Excellent Talents in University (NCET-10-0273), Beijing Nova Program (2010B035), National Natural Science Foundation of China (81373780).