Diversity of the gut microbiome in three grasshopper species using 16S rRNA and determination of cellulose digestibility

Sample treatment

The collected and classified living grasshoppers were placed in cages without access to food for 2 days to remove their intestinal contents. The grasshoppers to be tested were washed repeatedly with sterile water, placed in a 75% alcohol solution for 2 min, washed with sterile water, irradiated with ultraviolet light for 3–5 min, and dissected grasshoppers under sterile conditions. The entire intestinal tract was removed, and the midgut and hindgut parts were separated; placed in labeled, sterilized 1.5 mL centrifuge tubes; and kept at −80 °C for later use.

Extraction of total DNA from the intestinal contents

Total DNA of the intestinal contents of grasshoppers was extracted using the PowerSoil DNA Isolation Kit according to the manufacturer’s protocol, and the quality and quantity of DNA were evaluated by the 260 nm/280 nm and 260 nm/230 nm ratios, respectively. DNA was then stored at −80 °C until further processing.

For each individual sample, the 16s rDNA V3 + V4 region was amplified using the 338 F (5′-ACTCTACGGAGAGCA-3′) and 806 R (5′-GGACTACHVGGGTWTCTAT-3′) primers (Mori et al., 2014). PCR was performed in a total reaction volume of 20 µL: H₂O, 13.25 µL; 10×PCR ExTaq Buffer, 2.0 µL; DNA template (100 ng/mL), 0.5 µL; primer1 (10 mmol/L), 1.0 µL; primer2 (10 mmol/L), 1.0 µL; dNTP, 2.0 µL; and ExTaq (5 U/mL), 0.25 µL. After an initial denaturation at 95 °C for 5 min, amplification was performed with 30 cycles of incubations for 30 s at 95 °C, 20 s at 58 °C and 6 s at 72 °C, followed by a final extension at 72 °C for 7 min. The amplified products were then purified and recovered using 1.0% agarose gel electrophoresis. Finally, all the PCR products were quantified by Quant-iT^™ dsDNA HS Reagent and pooled together. High-throughput sequencing analysis of bacterial rRNA genes was performed on the purified, pooled samples using the Illumina HiSeq 2500 platform (2 × 250 pairedends) at Biomarker Technologies Corporation, Beijing, China. Finally, library construction and sequencing were performed by Beijing Biomarker Technologies Co. Ltd.

Bioinformatics analysis

Bioinformatics analysis in this study was completed on the Biomarker Cloud Platform (Biomarker Biotechnology Co., Beijing, China). The original data obtained by sequencing were spliced by FLASH software. Then, raw tags were filtered and clustered. Sequences were removed from inclusion according to the following criteria: the average mass of bases was less than 20; the reads were low quality; the sequences contained primer mismatches; the sequences were less than 350 bp in length; and the sequences could not be spliced. UCHIME, a tool included in mothur (http://drive5.com/uchime), was used to remove chimeras and generate valid data. OTUs were taxonomically annotated based on the Silva (bacteria) and UNITE (fungi) taxonomic databases. The denoised sequences were clustered using USEARCH (version 10.0), and tags with similarity ≥97% were regarded as OTUs. Taxonomy was assigned to all OTUs by searching against the Silvadatabases (http://www.arb-silva.de.) using uclust within QIIME (Edgar, 2010).

Collection and treatment of samples

Grasshoppers collected in the field were separately packed in insect rearing cages, and each cage contained 10 individuals that were fed wheat seedlings (The wheat variety was Triticum aestivum Linnaeus, 1753). After consecutively feeding for 3 days (no dung was collected during the period, and the wheat seedlings provided sufficient nutrition), grasshoppers were fasted for 2 days. A layer of white plastic foam was spread on the bottom of the cages to facilitate the collection of excrement (Wang, Liu & Chen, 2008). During the experiment, the fresh weight of wheat seedlings fed each time was recorded, and the feces and residual wheat seedlings were dried to a constant weight at 70 °C and recorded (using an electrothermal constant temperature blast drying oven, DGG-9030A; Shanghai Flyover Experimental Instrument Co., Ltd.). The dry-fresh ratio of wheat seedlings was determined to calculate the dry weight of the wheat seedlings before the experiment (Wang, 1997). The collected feces were dried to a constant weight, pulverized, and filtered with a 40 mesh sieve.

The wheat seedlings were rapidly dehydrated by steam de-enzyming (Sun, 2014), dried at 70 °C until a constant weight, crushed, and filtered with a 40 mesh sieve for later use.

Determination of cellulose and hemicellulose content

Samples were prepared by weighing out 0.800 g of each sample, to which 8 mL 72% H₂SO₄ was added, followed by shaking. Samples were placed in a water bath at 30 °C for 1 h, followed by the addition of eight mL 4% H₂SO₄, and were then returned to the water bath for 45 min. Finally, 224 mL of distilled water was added, and the samples shaken well before being placed into conical flasks in an electric heating pressure steam sterilization pot (LS-30 type of Shanghai Bosun Industrial Co., Ltd.). Samples were then heated to a temperature of 121 °C for 1 h and filtered to obtain sample solutions.

One milliliter of this sample solution was diluted appropriately, and one mL of the diluted sample solution was added to one mL of anthrone reagent and three mL of 80% sulfuric acid, mixed well, and boiled at 100 °C for 5 min. After cooling to room temperature, absorbance at 620 nm was measured, with the sugar concentration calculated according to the glucose standard regression equation and then multiplied by 0.9 (Zhang et al., 2010).

One milliliter of the sample solution was diluted appropriately, and one mL of the diluted sample solution was add to two mL of A reagent and 0.134 mL of B reagent and boiled at 100 °C for 20 min after fully mixing. Absorbance at 660 nm was measured after cooling to room temperature, with the sugar concentration calculated according to the xylose standard regression equation and then multiplied by 0.88 (Zhang et al., 2010).

Calculation of the decomposition rates of cellulose and hemicellulose

The decomposition rates of cellulose and hemicellulose were calculated after the cellulose and hemicellulose contents of the adult grasshopper feces were determined by the above methods. Statistical analysis of digestibility data was done in SPSS 21.0 software using T-test.

Note: c is the sugar concentration (g/L) calculated according to the standard curve, m is the weighed sample mass (g). a is amount of cellulose fed on wheat seedlings (g), b is fecal cellulose content (g).

Evaluation of sequencing quality

A total of 702,445 paired-end reads were obtained by sequencing the nine pooled samples. and 512,109 clean tags were generated after splicing and filtering the paired-end reads. A minimum of 51,643 clean tags were generated for each sample, with an average of 56,901 clean tags. The proportion of effective sequences was 99.48%. The sequencing accuracy of the samples was high and met the standard requirements. Effective tags were the number of effective sequences after filtering chimeras from the clean tags. The number of sequences and the proportion for each sample are shown in Table 2 below.

OTU-venn analysis

To identify the number of common and unique OTUs among samples, a Venn diagram was used, which intuitively reflects the coincidence of OTUs among samples. As shown in Fig. 1A, there were 37 species of bacteria in the intestinal tract common to the three species of grasshoppers. There were six species specific to Shirakiacris shirakii, 11 species specific to Oedaleus decorus asiaticus and 13 species specific to Aiolopus tamulus. Further analysis of the identifies of the bacteria common to the three grasshopper species indicated that they were mainly composed of two families of Enterobacteriaceae and Enterococcaceae, as shown in Fig. 1B, with a relative abundance of 97.57%, indicating that these two families may form the core microflora in the grasshopper intestinal tract.

α-diversity analysis

As shown in Fig. 2A, the rarefaction curves of 9 samples tended to be flat over an increasing number of sequences. The Shannon, Simpson, Chao1 and ACE indices, as well as others, were used to express the α-diversity of the microorganisms in the samples. As shown in Table 3, the coverage of the nine samples was relatively high, reaching 99.97~99.99%. The above results show that the sequencing data were reasonable and that the vast majority of bacteria in the samples were detected. Different from the rarefaction curve, the species accumulation curve reflects whether the number of samples was sufficient and whether the information covered all the annotated species. As shown in Fig. 2B, as the sample number increased, the cumulative curve and the common quantity curve tended to be flat, which demonstrates that the new and common species detected in the sample were both approaching saturation, indicating that the sample size was sufficient and could be used for diversity and abundance analysis.

The α-diversity of nine samples varied according to the individual. In the three samples of Aiolopus tamulus, the Shannon index of At1 and At2 was much higher than that of At3, while the Simpson index of At3 was the opposite, indicating that the species diversity in samples At1 and At2 was higher than that in At3. Among the three samples of Oedaleus decorus asiaticus, the Shannon index of Od3 was much higher than that of Od1 and Od2, while the Simpson index of Od3 was much lower than that of the other two samples. For the three samples of Shirakiacris shirakii, the Shannon index and Simpson index were not significantly different, which may be related to the difference in the collection time (Table 1).

The average value of each index of three samples from the same species was calculated and then used to compare and analyze the α-diversity among the different species. The Chao1 index (Fig. 2C) of Shirakiacris shirakii was significantly higher than that of Oedaleus decorus asiaticus, and the ACE index (Fig. 2D) was the highest in Shirakiacris shirakii, followed by Aiolopus tamulus, which demonstrates that among the three species grasshoppers, the abundance of species in the intestinal tract of Shirakiacris shiraki was significantly higher than that of Oedaleus decorus asiaticus, with Aiolopus tamulus in the middle. The Simpson index (Fig. 2E) of Aiolopus tamulus was the smallest, while the index of Shirakiacris shiraki was the largest. The Shannon index (Fig. 2F) followed the opposite trend to the Simpson index, which indicated that the species diversity in the intestinal tract of Aiolopus tamulus was the highest, followed by the Oedaleus decorus asiaticus, with Shirakiacris shiraki as the lowest.

β-diversity analysis

Based on pyrosequencing data, PCoA and UPGMA clustering were carried out to determine β-diversity. As shown in Fig. S1, the smaller the distance between points in the figure, the smaller the difference in the intestinal flora structure, and vice versa. It can be seen from the figure that the difference in the intestinal microflora structure between the three samples of Shirakiacris shiraki and two of the samples of Aiolopus tamulus was relatively small, while difference in the intestinal microflora structure between one sample and the remaining two samples for both Oedaleus decorus asiaticus and Aiolopus tamulus was relatively large. The difference in the intestinal microflora structure among the three samples of Shirakiacris shiraki was not large. In addition, the hierarchical cluster tree (Fig. 3A) shows that the microbial communities of the three species grasshoppers are divided into three groups: (1) group I includes samples A1 and A2 and sample O3, (2) group II includes samples O1 and O2 and sample A3 and (3) group III includes all the samples of Shirakiacris shiraki. In addition, the distance between group II and group III was closer, that is, the composition of the intestinal microflora is more similar between those two groups. Taken together, these results show that the intestinal microflora of different species of grasshoppers vary from one another. The intestinal microflora of Aiolopus tamulus and Shirakiacris shiraki are more similar. At the same time, different sampling times will also lead to the recombination of microbial communities.

Nonmetric Multidimensional Scaling (NMDS) analysis can reflect the differences between groups or within groups according to the distribution of samples. As shown in Fig. 3B, the stress value is less than 0.01, which indicates that the analysis result is extremely reliable. In the figure, it can be seen that there is a large difference in the intestinal community between one sample the remaining two samples for both Oedaleus decorus asiaticus and Aiolopus tamulus, which is related to the different collection times of the samples, indicating that a difference in collection time leads to changes in the microbial community structure of the same species. The three samples of Shirakiacris shiraki along with two samples of Aiolopus tamulus are almost coincident, which indicates that the similarity of the intestinal microflora structure between the two groups was relatively high.

As shown in Fig. 3C, Ss1, Ss2 and Ss3 were grouped together; At1 and At2 were grouped together; and all (Ss1, Ss2, Ss3 and At2) were grouped with Od3. Od1 and Od2 were grouped with At3. The samples At1 and At2 of Aiolopus tamulus were relatively close to the three samples of Shirakiacris shiraki, which indicates that the intestinal community similarity between Aiolopus tamulus and Shirakiacris shiraki is high, that the difference of the microflora structure between them is relatively small, and that different collection times for the same species can lead to low similarity and large differences in the grasshopper intestinal microflora structure, which is consistent with the above results, indicating that a difference in collection time causes changes in the microbial community structure.

Intestinal microflora structure of the three species grasshoppers

High-quality sequences obtained from 16S rDNA identification were compared with the database, and a total of 54 genera of seven phyla, 12 classes and 20 orders were identified. The composition of each sample is shown in Table 2. Once the average relative abundance of different grasshoppers in the same treatment at each classification level is calculated, the average relative abundance can reflect the content of various intestinal microorganisms at the overall level.

Intestinal microflora structure at the phylum level

The nine samples At1, At2, At3, Od1, Od1, Od2, Od3, Ss1, Ss2 and Ss3 contained 85.65%, 83.51%, 93.45%, 89.51%, 91.43%, 87.32%, 86.92%, 87.33% and 87.35%, respectively, of the valid sequences that were able to be annotated at the phylum level. Seven phyla were detected in the nine samples. According to the annotation results of the samples at various classification levels (kingdom, phyla, class, order, family, genus and species), as shown in Fig. 4A, Proteobacteria accounted for the highest relative abundance in the three species of grasshoppers, Aiolopus tamulus, Oedaleus decorus asiaticus and Shirakiacris shiraki, at 94.10%, 90.72% and 93.94%, respectively. The second highest was Firmicutes, accounting for 5.72%, 8.94% and 5.31%, respectively.Actinobacteria accounted for a relatively high proportion of 0.52% in the intestinal tract of Shirakiacris shiraki, although less than 0.10% in the intestinal tracts of the other two species. Cyanobacteria was relatively abundant in the intestinal tract of Oedaleus decorus asiaticus, at 0.26%, while its abundance in the other two species was very small, accounting for 0.01%. Fusobacteria existed in the intestinal tracts of the three species grasshoppers in trace amounts, accounting for less than 0.10%. Bacteroidetes was found in trace amounts in Aiolopus tamulus and Oedaleus decorus asiaticus but was not detected in the intestinal tract of Shirakiacris shiraki. Tenericutes was only found in trace amount in the intestinal tract of Oedaleus decorus asiaticus, at 0.04%, but was not found in the intestinal tracts of the other two grasshoppers. Additionally, 0.20% unassigned microorganisms were present in the intestinal tract of Shirakiacris shiraki that have not previously been studied.

It is worth noting that the proportion of Firmicutes in At3 intestinal bacteria was 0.24%, which was much lower than that in At 1 (11.01%) and At2 (5.91%) treated with the same method. However, the proportion in sample Od3 (26.74%) was much higher than that in sample Od1 (0.03%) and sample Od2 (0.06%), while the proportion in the three samples Ss1 (3.72%), Ss2 (1.83%) and Ss3 (10.36%) of Shirakiacris shiraki was relatively constant, which could be related to their different collection times, indicating that the abundance of intestinal flora varied over different periods in the same species. Combined with α-diversity analysis, these results show that the diversity and abundance of intestinal microflora varied over different periods in the same species.

Intestinal microflora structure at genus level

At1, At2, At3, Od1, Od2, Od3, Ss1, Ss2 and Ss3 contained 54.57%, 68.35%, 93.40%, 89.35%, 90.93%, 87.31%, 86.08%, 85.45% and 87.30%, respectively, of the valid sequences that could be annotated at the genus level. A total of 54 bacterial genera were detected, of which 24 bacterial genera were common among the three species. As seen in Fig. 4B, Klebsiella accounted for the highest proportion of the microbial community in the three grasshopper species. The top 10 abundant bacterial genera (by average relative abundance) for each of the three species of grasshoppers after data standardization are shown in Tables 1–3. In the three samples of Aiolopus tamulus, the average relative abundance of Klebsiella, Enterococcus and Enterobacter was greater than 1%, which identifies them as the primary bacteria in the Aiolopus tamulus intestinal tract. The primary bacteria of Oedaleus decorus asiaticus were Klebsiella, Enterococcus, Pantoea, Wolbachia, Enterobacter and Lactococcus. Klebsiella, Lactococcus and Staphylococcus were the primary genera of Shirakiacris shiraki. Five bacterial genera were detected only in the intestinal tract of Aiolopus tamulus, namely, Anaerotruncus, Diaphorobacter, Morganella, Proteiniclasticum and Rikenellaceae_RC9_gut_group. The proportion of these five bacterial genera in the intestinal tract was not more than 0.1%. Among them, Morganella was not detected in the At3 samples, but was detected in the At1 and At2 samples, and the remaining four genera were detected only in the At3 samples but not in the At1 and At2 samples, indicating that there were significant differences in the intestinal microflora diversity of the same species from different time periods. The genera Sphaerotilus and Spiroplasma were only detected in the intestinal tract of Oedaleus decorus asiaticus, and Cronobacter was only detected in Shirakiacris shiraki. Therefore, the diversity of the intestinal microorganisms varied by grasshopper species.