Different Cultivation Environments Affect the Yield, Bacterial Community and Metabolites of Cordyceps cicadae

Zhaoying Zeng1,2,3† , Dan Mou1† , Li Luo1 , Wenlin Zhong1 , Lin Duan1 and Xiao Zou1,2*

1 Institute of Fungal Resources, College of Life Sciences, Guizhou University, Guiyang, China, 

2 Key Laboratory of Plant Resource Conservation and Germplasm Innovation in Mountainous Region (Ministry of Education), Guizhou Key Lab of Agro-Bioengineering, College of Life Sciences/Institute of Agro-Bioengineering, Guizhou University, Guiyang, China, 

3 School of Ecology and Environmental Science, Yunnan University, Kunming, China

Cordyceps cicadae is an entomogenous fungus with important uses in traditional Chinese medicine. However, its wild resources have not met consumers’ demand due to excessive harvesting practices. Artificial cultivation is therefore an important alternative, but research on cultivating C. cicadae in natural habitats has not been reported. In this study, we aimed to explore the viability of cultivating C. cicadae in a natural habitat, in the soil of Pinus massoniana forest. We assessed and compared the yield, metabolite contents and bacterial community composition of C. cicadae grown in the Antheraea pernyi pupae at different growth stages, and under different cultivation conditions, in the soil of a natural habitat and in sterile glass bottles. Our results showed that cultivating C. cicadae in a natural habitat is feasible, with up to 95% of pupae producing C. cicadae fruiting bodies. The content of nitrogen compounds (amino acids) in C. cicadae cultivated in a natural habitat was significantly higher than in glass bottles, while the yield and carbon compound (mannitol and polysaccharide) and nucleoside (cordycepin and adenosine) contents were lower. Different bacterial genera were enriched in C. cicadae at different growth stages and cultivation environments, and these bacterial genera were closely related to metabolites contents during growth. This study demonstrated the viability of a novel cultivation method of C. cicadae, which could be used as an alternative to wild stocks of this fungus. These findings provided new insights into the growth mechanism of C. cicadae and its interaction with soil microorganisms.

Keywords: Cordyceps cicadae, cultivation in natural habitat, artificial cultivation, metabolite, bacterial community composition, PICRUSt

Edited by:

George Tsiamis, University of Patras, Greece

Reviewed by:

Hyung-Kyoon Choi, Chung-Ang University, South Korea

Yan-Qun Liu, Shenyang Agricultural University, China

*Correspondence:

Xiao Zou xzou@gzu.edu.cn † These authors have contributed equally to this work and share first authorship

Specialty section: 

This article was submitted to Microbial Symbioses, a section of the journal Frontiers  in Microbiology

Received: 19 February 2021

Accepted: 08 April 2021

Published: 11 May 2021

Citation: 

Zeng Z, Mou D, Luo L, Zhong W, Duan L and Zou X (2021) Different Cultivation Environments Affect the Yield, Bacterial Community and Metabolites of Cordyceps cicadae. Front. Microbiol. 12:669785. doi: 10.3389/fmicb.2021.669785

INTRODUCTION

Cordyceps cicadae is an entomogenous fungus belonging to the family Claviciptaceae that has been used as a traditional Chinese medicine for over 1600 years (Liu, 2012). It has several host cicadas, such as Cicada flammata Distant, Platypleura kaempferi Fabricius, Cryptotympana pustulata Fabricius, Platylomia pieli Kato, and Oncotympana maculatieollis Motsch (Zeng et al., 2014, 2017), and can form fruiting bodies on the surface of such cicada nymphs after they have been parasitized and killed by Isaria cicadae, the anamorphic form of this fungus (Cheng et al., 2006). C. cicadae mainly grows in the soil of bamboo, broad-leaved  and evergreen broad-leaved forests, but also grows in mixed coniferous and broad-leaved forests. It is geographically widely distributed in  provinces  south  of  the  Qinling-Huaihe  River  in China, including Guizhou, Yunnan, Sichuan, and Zhejiang,  as well as in Japan, South Korea, Australia, Brazil, and other countries (Liu, 2012; Chen Z.A. et al., 2014; Zeng et al., 2015).

In China, C. cicadae is used to treat chronic kidney diseases, palpitations and  children  with  seizures,  and  is  also  used  as  a food and tonic (Hsu et al., 2015; Li et al., 2019). Recent studies have found that C. cicadae contains metabolites with important pharmacological functions. Examples of these include the carbon compound mannitol, which is a diuretic, exhibits anti- free-radical activity, and improves cerebral microcirculation and blood flow (Lin et al., 2016; Zhang et al., 2019); polysaccharides, which are also essential carbon compounds, have antioxidant properties and play a role in treating diabetes and improving immune regulation (Lu et al., 2016; Zhang et al., 2018; Wang   et al., 2019); adenosine, a nucleoside with anti-inflammatory, neuroprotective and anti-convulsive properties,  which  has  been shown to improve cell viability and prevent and treat neurodegenerative diseases (Latini and Pedata, 2001; Nakav et al., 2008; Olatunji et al., 2016b); and cordycepin, another nucleoside with anti-cancer, immune regulation, and antioxidant functions (Olatunji et al., 2016a; Zhang et al., 2019). Owing to its use      as a medicine and tonic, wild C. cicadae has been harvested at unsustainable levels, resulting in a significant population reduction in its natural habitats (Liu et al., 2008). The demand for C. cicadae cannot be met using these practices, and artificial cultivation is necessary to be an alternative to wild C. cicadae and ensure an adequate supply (Hsu et al., 2015; He W.W. et al., 2019; Sun et al., 2019; Nxumalo et al., 2020).

Artificial cultivation techniques for C. cicadae include liquid fermentation (Cheng et al., 2006, 2012) and cultivation on cereal medium (Feng, 2002; Zhang et al., 2012; Ke and Lee, 2016) or pupae (Hu et al., 2009; He Y.Q. et al., 2019). Hyphae can only be produced by liquid fermentation, and fruiting bodies only by cereal medium. Cultivation on pupae has become more popular in recent years, because C. cicadae with insects and fruiting bodies can be harvested by this way (He W.W. et al., 2019). However, the cultivation of C. cicadae on pupae has mainly been conducted in sterile environments, such as sterilized glass bottles, sand, or soil (Hu et al., 2009; Liu et al., 2018); little research has been conducted on cultivating C.  cicadae in natural habitats, despite  a preference among consumers for more “natural” production methods (Huang, 2008).

Wild C.  cicadae  grows  in  complex  natural  habitat  that  are rich in vegetation including Camellia japonica, Camellia oleifera, Pyrus  pashia,  and  Myrica  nana  (Liu,  2012;  Zeng  et al., 2015). Researches have shown that the habitat soil, sclerotia, and external mycelial cortices of C. cicadae enrich abundant microbial communities and share some of the same bacterial genus, and some ectomycorrhizal fungi have also been detected  in  sclerotia  (Zeng  et  al.,  2019;  Mou  et  al.,  2021).These results suggested that wild C. cicadae populations are ecologically linked to the surrounding plants and soil through microorganisms, which may impact C. cicadae growth and production of metabolites. Previous research found that the composition of microbial communities in host varies between different growth environments (Mueller et al., 2020), and changes in microbial community composition in Cordyceps spp. can affect growth and metabolite production of host (Zhang et al., 2010; Qu et al., 2019a,b). Cultivating C. cicadae in natural habitats, rather than under sterile conditions, may therefore be a helpful experimental approach to exploring soil microbial interactions and their potential impact on growth and metabolite production in C. cicadae. In addition, we previously demonstrated that the relative abundance of fungi other than Isaria in the sclerotia of C. cicadae was less than 1%, while the bacterial community  was more abundant and stable (Zeng et al., 2019; Mou et al., 2021). Given their relative abundance compared with fungi, exploring the composition of bacterial communities in the sclerotia of C. cicadae is a specific priority in this study.

This study therefore aimed to explore the feasibility of cultivating C. cicadae in natural habitats, the bacterial community composition of C. cicadae in different cultivation environments, and the relationship between bacterial communities in sclerotia of C. cicadae and metabolite production in this medicinally important fungal species. To achieve this, we compared the yield, bacterial community composition and metabolites content of C. cicadae cultivated in aseptic glass bottles versus in the soil of a natural Pinus massoniana forest habitat. 

MATERIALS AND METHODS

Cultivation of Cordyceps cicadae in Two Cultivation Environments

The Isaria cicadae strain GZUIFR_DJS1 (preserved in Institute of Fungal Resources of Guizhou University) was activated and purified on potato dextrose agar (PDA, 200 g potato extract,  20 g glucose, 20 g agar, 1,000 mL distilled water) solid growth medium. Purified colonies were inoculated in the center of the PDA medium and cultured upside down at 25℃ for 5–7 days, until they produced spores. Spore suspensions [concentration (C) = 5×107 spores/mL] were prepared with sterile ultrapure water containing 0.05% Tween 80 viscous liquid.

Pupae of Antheraea pernyi Guérin-Méneville, sourced from the Yuxi City Artificial Sericulture Base (Liaoning, China), were selected as host insects (cultivation medium) for C. cicadae. Pupae were soaked and scrubbed with 75% alcohol for 10 s for body surface sterilization; pupae were rinsed three times with sterile water to remove any residual alcohol, and  then  dried with sterile tissue. An inoculation of 0.02 mL spore suspension was made into each pupa at the intersection of the wing and    the third somite (Zhang et al., 2016). Inoculated pupae were cultivated in sterile glass bottles at 25℃, each bottle contained   a pupa and added a sterile moist cotton ball for maintaining humidity. After a week, the  pupae  were  completely  infected by I. cicadae and had become mummified. Each pupa  was  given one of two possible treatments for 2 weeks (Figure 1):either their cultivation in glass bottles was continued (hereafter referred to as CB), or they were cultivated in the soil of natural habitat under a Pinus massoniana forest (CS). The experimental cultivation site for natural habitat used in this study was located in the West Campus of Guizhou University (Guizhou, China) and measured approximately 100 m2; a detailed description of the habitat is in Supplementary Table 1. Mummified pupae were cultivated in natural habitat under a layer of covering soil approximately 1.5 cm thick, and an interval between any two pupae of 20 cm or more.

FIGURE 1 | A comparison of Cordyceps cicadae cultivation in sterile glass bottles and in the soil of natural habitats.

Experiment 1: Effects of Two Cultivation Environments on the Yield of C. cicadae

In April 2019, 150 pupae were inoculated and cultivated in sterile glass bottles for 1 week. After removing non-mummified pupae, 120 mummified pupae were randomly divided into six groups of 20; three groups in CB and three in CS. Under both treatment conditions for 2 weeks, C. cicadae fruiting bodies did not grow anymore and began to produce spores, at this point, C. cicadae were considered mature; then we harvested all C. cicadae and washed any soil from CS pupae using tap water. Then, the proportion of pupae producing C. cicadae fruiting bodies (i.e., the number of pupae producing fruiting bodies divided by the total number) was calculated. The average proportion of pupae with fruiting bodies across all three groups was then calculated for CB and CS separately. After removing any non-mummified pupae and any C. cicadae damaged in the process of washing soil from CS pupae, the fresh weight, fruiting body number, length and diameter of 53 out of 55 C. cicadae in CB and 39 out of 57 in CS were measured. Finally, we compared the relative yield between in CB and CS by 100 cicadas i.e., 100× the proportion of pupae producing C. cicadae fruiting bodies× mean fresh weight per C. cicadae).

Experiment 2: Effects of Two Cultivation Environments on the Bacterial Community and Metabolites of C. cicadae

In September 2019, 250 pupae were inoculated and cultivated in sterile glass bottles for 1 week. After removing non-mummified pupae, 200 mummified pupae were randomly divided into two groups and cultivated for 2 weeks under CB and CS treatment conditions, respectively. Since the pupae were inoculated with spore suspension, the sclerotia of nine randomly selected pupae were collected for bacterial community analysis each week, with three pupae combined in each of three replicates. Only pupae in the CB treatment condition were taken in week 1; in subsequent weeks, pupae in both CB and CS treatment conditions were taken. Samples were coded according to the week they were taken, e.g., samples from the second week were SCB_2 and SCS_2. Any soil on the CS pupae was washed off with tap water before the external mycelial layer and fruiting body were stripped off, leaving only the mummified pupae. These pupae were soaked in 75% alcohol for 2 min, then soaked in 2.5% sodium hypochlorite solution for 20 s, washed with sterile water five times, and dried with sterile tissue (Xia et al., 2015). Each pupa was crosscut in half on a clean laboratory bench, and the C. cicadae sclerotia were picked out with an inoculation needle and placed in sterile centrifuge tubes. Each sample was flash-frozen with liquid nitrogen, ground, and  stored  at  80℃  until  the  DNA  could  be  extracted.  In the third week, the remaining C. cicadae in the two treatments were harvested, washed to remove soil, and then dried by dry oven at 80℃. Fruiting bodies were collected, ground and sifted through a 100 mesh screen, and used to detect the contents of five metabolites in C. cicadae.

Detection of Metabolites

In this study, we measured the content of adenosine, cordycepin, polysaccharides, and mannitol in C. cicadae and  the  content and type of  amino  acids  in  C.  cicadae  fruiting  bodies. Amino acids were determined using the Hitachi L-8800 high speed amino acid analyzer, according to the national standard protocol GB 5009.124-2016. Mannitol content was determined by colorimetry according to a previously established method (San et al., 2010): 50% ethanol was used to extract mannitol using ultrasonic method, then sodium periodate solution, L-rhamnose solution and Nash reagent were used to perform the coloration reaction; absorption values were then detected at 412 nm using an ultraviolet spectrophotometer. Polysaccharide content was determined using a sulfuric acid– phenol method according to the national agricultural standard NY/T 1676-2008. Polysaccharides were precipitated in ethanol, and the resulting furfural derivatives of the polysaccharides  were dehydrated in concentrated sulfuric acid and condensed with phenol, forming orange-red compounds with a color intensity proportional to the concentration of polysaccharides   in the  solution.  Colorimetric  quantification  was  carried  out  at 490 nm by an ultraviolet  spectrophotometer.  Cordycepin  and adenosine were determined using high performance liquid chromatography according to national agricultural standard NY/T 2116-2012. The chromatographic column used in this study was the Shim-pack VP-ODS 5 micron 250 mm×4.6 mm. The mobile phase was water and acetonitrile with 95:5 volume (v/v) ratio at a mobile speed of 1.0 mL/min. The temperature of chromatographic column was maintained at 35℃. The contents of cordycepin and adenosine were measured using a UV detector at 260 nm. We conducted three replicates for the measurement of each metabolite.

Bacterial DNA Extraction, PCR Amplification, and Sequencing

Bacterial  DNA  was  extracted using the E.Z.N.A.QR®  Soil DNA  Kit  (Omega  Bio-tek),  and  the  concentration  and  purity of DNA were assessed by electrophoresis on 1.0% weight/volume agarose  gels.  The  primers  used  to  amplify  the V4 variable region of the bacterial 16s RNA gene were:  515F (5' -GTGCCAGCMGCCGCGGTAA-3' ) and 806R (5' - GGACTACHVGGGTWTCTAAT-3' ) (Zhou et al., 2020).

DNA amplification was carried out using the ABI GeneAmp 9700 polymerase chain reaction (PCR) instrument. The PCR parameters were as follows: initial denaturation for 3 min at 95℃; denaturation, annealing and extension at 30 s at 95℃, 30 s at 55℃, and 45 s at 72℃ for 30 cycles; then final extension for 10 min at 72℃, then 10℃ until the reaction stopped. The amplification products were sent to Shanghai Majorbio Bio- pharm Technology Co., Ltd. (Shanghai, China) for sequencing on the Illumina MiSeq sequencing platform.

Bioinformatics and Data Analysis

Fast Length Adjustment of SHort reads (FLASH) and Quantitative Insights Into Microbial Ecology (QIIME) software programs were used to merge and  quality-filter  raw  reads  from the original DNA fragments (Caporaso et al.,  2010;  Magoc and Salzberg,  2011).  Effective  reads  were  obtained  by using a UCHIME algorithm to identify  and  remove chimeric sequences. Quality sequences were then clustered into operational taxonomic units (OTUs) with a >97% similarity cutoff value using the UPARSE algorithm (Edgar et al., 2011). The Basic Logical Alignment Search Tool (BLAST) was then used to compare representative sequences for each OTU with sequences in the SILVA database to obtain the taxon of each OTU (Quast et al., 2013).

α-diversity of bacterial communities in each sample was represented by Shannon, Chao1 and Coverage indices, calculated after reads were normalized to the minimum reads (30, 665 reads) in each sample. Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt)1 software was used to predict the functional profile of bacterial communities using 16S ribosomal ribonucleic acid (rRNA) marker gene sequences (Langille et al., 2013). We then calculated the abundance of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and functions at the OTU level in each sample. 

The Kolmogorov–Smirnov tests were used to verify the normality of data. Student’s t-tests and non-parametric Mann– Whitney U-tests were used to identify differences in yield and abundance of KEGG pathways and functions between CB and CS treatment conditions that the data were normal or not, respectively. One-way analysis of variance (ANOVA) and post hoc comparisons were performed using Duncan’s Multiple Range test to compare α-diversity among the different groups. The Spearman’s rank correlation coefficient between contents of metabolites mentioned above and the top 30 most abundant bacterial genera was calculated and then visualized by the “pheatmap” package in the R software program. Visualizations of the yield index, the α-diversity index and the bacterial community composition were produced using GraphPad Prism 7 (GraphPad Software Inc., San Diego, CA, United States).