Study of Periplaneta Americana Microbial Community Structure and Diversity by 16S rRNA High-Throughput Sequencing

Objective: The present study probes into the microbial community structure in Periplaneta americana under different breeding conditions, using 16S rRNA high-throughput sequencing technique, in the hope of finding the microbial community structure in Periplaneta americana and their diversity under different breeding conditions. Methods: In this study, we extract the microbial metagenomic DNA of 5 groups of Periplaneta americana which under different breeding conditions. Using lllumina Miseq sequencing platform, two-terminal sequencing of V3-V4 regions of 16S rRNA were sequenced; diversity of community structure was analyzed using the softwares such as fastqc, QIIME, PyNAST, fasttree and R language. Results: Shannon index of samples in SG group was lower than that of the other four groups, significantly lower than that of DB group (P<0.05), but not significantly different from other groups. This suggested that the intake of a mixed fodder with high sugar, high fat and high protein by Periplaneta americana can reduce the diversity of microbial communities. Our findings showed that breeding intervention with different fodders may cause differences in the contents of Bacteroidetes, Proteobacteria and Firmicutes in Periplaneta americana. Results showed that long-term intake of lots of sucrose and fat may increase the proportion of Bacteroidetes in Periplaneta americana; and long-term intake of lots of sucrose may reduce the proportion of Proteobacteria in Periplaneta americana; and long-term intake of lots of fat may reduce the proportion of Firmicutes in Periplaneta http://www.scholink.org/ojs/index.php/se Sustainability in Environment Vol. 2, No. 4, 2017 351 Published by SCHOLINK INC. americana. Two major dominant bacterial genera in all samples were Blattabacterium and Rickettsiella. But different feeding interventions can change the proportions of Blattabacterium and Rickettsiella. Conclusion: Periplaneta americana has a complex microbial community structure. Different feeding conditions may change the microbial community structure of Periplaneta americana. An important bacterial genus in Periplaneta americana, Blattabacterium is positively correlated with the intake of sucroseand fat-rich fodder. In the breeding process of Periplaneta americana, adding sucrose and fat to fodder may increase the content and proportion of Blattabacterium in microbial communities.


Introduction
Insects are an animal group in terrestrial ecosystems that has the most abundant biomasses, species and ecological habits, and also one of the groups with the highest biodiversity ( Erwin et al., 1982;Chapman et al., 2013). Due to diverse feeding characteristics and behaviors, almost all food resources on land (all substances that can be eaten by humans and provide nutrients and heat) can be consumed by insects (Shi et al., 2010). The success of insect diversity and evolution, to a certain extent, rests upon the contributions of various beneficial microbes ( . Symbiotic bacteria in insects have an important effect on the vital activities of insects (Babendreier et al., 2007;Mattila et al., 2012;Nikoh et al., 2011;Eichler et al., 2002;Philipp et al., 2012;Douglas, 1993;Hongoh et al., 2008). For example, the physiology and evolution of insects not only improves of hosts' nutrition and helps digest food ingredients, but also prevents invasion from predators, parasites and pathogens and helps interspecific and intraspecific communications, etc. ( Favia et al., 2012;Kaltenpoth et al, 2014). Insect microbes are highly dependent on each other, symbiotic with hosts and involved in regulating various vital activities of hosts, thereby affecting the health of insects indirectly Periplaneta americana were sequenced. The sequencing was directed. The target length of sequencing was about 460bp.

Data Analysis
After V3-V4 regions of 16S rRNA in 15 samples from the digestive tract of Periplaneta americana were sequenced, the quality of raw data were first tested with Fastqc software. Adaptors and low-quality base sequences were removed. The numbers of reads and data of raw and quality-controlled samples were counted. After that, double-end reads were spliced into effective sequences, according to one-end overlapping base.
Using QIIME (Version 1.50), high-quality sequences were clustered into operational taxonomic units (OTUs). OUT clustering were performed on high-quality sequences of all samples as per a similarity of 0.97. The longest sequence in each cluster was selected as a representative sequence of OTU. 1) In the study, we adopted QIIME (Version 1.50) software to conduct a correlation analysis of Alpha diversity of samples from the digestive tract of Periplaneta americana. According to values of derived Alpha diversity indices, i.e., Chao1, Shannon, PD whole tree and Simpson, a corresponding rarefaction curve was generated.
2) Flag sequences were selected from each OUT. Using Uclust double sequence alignment algorithm in PyNAST software (Caporaso et al., 2010), flag sequences were compared with a reference database (16S rRNA gene sequence). The positions of gaps in sequences were filtered. Then using Fasttree (Price et al., 2010) software, a phylogenetic tree was built, with OUT flag sequences as nodes. Then using UniFrac (Lozupone et al., 2012) algorithm, the distance matrix and PCoA of samples were calculated. According to the distance matrix of samples, a heatmap image was made using heatmap program in R language. The similarity among all samples was accounted for.
3) After the abundance of OUT in each species was calculated using QIIME software, the abundance matrix of OUT in each species was obtained. According to the derived data, generic-level taxonomic information was clustered as per sample and class and a heatmap image was generated, using heatmap program in R language. Among them, change in color represented two-dimensional matrix information.
Shade of color represented the value of data. Species of high and low abundances were clustered by blocks. The similarities and differences in generic-level community composition between multiple samples were reflected by color gradient and degree of similarity. 4) Principal Pomponent Analysis (PCoA) was carried out using QIIME. One point stood for a sample.
Points of the same color belonged to the same group. The closer two points were, the smaller difference between two samples in microbial communities. 5) Using UniFrac distance, a phylogenetic analysis including abundance was conducted. The abundance of bacterial genera, whose mean value was Top 100 among 5 groups of samples were selected, to make a heatmap and conduct a clustering analysis of 15 samples, so as to reflect intragroup and intergroup differences between samples.
Published by SCHOLINK INC.

Statistical Analysis
All experimental data involved in the present study were measured data. All average data were signified with Mean ± SD. As each breeding group had 3 sample data, Kruskal-Wallis H, a non-parametric test method with multiple independent sample rank conversion was chosen. A significance analysis was conducted of OUT number and diversity index of different breeding groups, as well as relative abundance of microbial communities in the digestive tract. After tested by Kruskal-Wallis H, the above data of each experimental group were different as whole. So a one-way ANOVA was adopted to make a pairwise comparison between all experimental groups. The statistical analysis in the present study should be implemented with R language statistical software. P<0.05 was regarded as statistically significant.

Feeding Results
Finish breeding, 40 Periplaneta americana were randomly selected from each group, to measure their length and weight. Results showed that the average weight, length and weight-length ratio (W/L) of periplaneta americana in PT group were 1.1999±0.2070 g, 3.29±0.16 cm and 0.36±0.05 g/cm respectively. The average weight, length and weight-length ratio of periplaneta americana in T group were 1.2126±0.2442 g, 3.32±0.18 cm and 0.37±0.06 g/cm respectively. The average weight, length and weight-length ratio of periplaneta americana in ZF group were 1.0407±0.1843 g, 3.21±0.16 cm and 0.32±0.05 g/cm respectively. The average weight, length and weight-length ratio of periplaneta americana in DB group were 1.0767±0.1963 g, 3.17±0.20 cm and 0.34±0.06 g/cm respectively. The average weight, length and weight-length ratio of periplaneta americana in SG group were 1.1690±0.2765 g, 3.24±0.21 cm and 0.36±0.09 g/cm respectively. Among them, the weight-length ratio in ZF group was significantly lower than that of PT group, T group and SG group (P<0.05). A correspondence analysis of the weight-length ratio of periplaneta americana between five groups showed that the weight-length ratio tended to be ≥0.4 in SG group, ≥0.35 and <0.4 in PT group, as shown in Table 1 and Figure 1.

Alpha Diversity Analysis
Using QIIME software, based on OTUs clustered as per a similarity of 97%, Alpha diversity value of samples was calculated. 4 diversity rarefaction curves, i.e., sparse index curves of PD whole tree rarefaction, Simpson, Chao1 and Shannon using R language software are shown in Figure 2. As shown in Tables 2 and 3, Shannon diversity indices of PT group, T group, ZF group, DB group and SG group were 4.73±0.71, 5.58±0.47, 5.01±0.50, 6.25±1.09 and 3.83±1.22 respectively. The diversity index of SG group was significantly lower than that of DB group (P<0.05), but not significantly different from other groups. And the coverages of all samples were greater than 95%.

Species Distribution
In this study, a total of 1,487,612 high-quality sequences were obtained from 5 groups of Periplaneta   DB group (31.70%, mean), ZF group (31.65%, mean), T group (12.15%, mean) and SG group (11.90%, mean), while changes in the proportion of Proteobacteria in 5 groups of samples were not significant (P>0.05); proportion of the third dominant bacterial phylum-Firmicutes in the total flora of DB group was the highest (23.44%, mean), followed by SG group (4.84%, mean), T group (4.59%, mean), PT group (4.29%, mean) and ZF group (2.44%, mean), while changes in the proportion of Firmicutes in 5 groups of samples were not significant (P>0.05).  Table 5

A Difference Analysis between Microbial Community Structures in the Digestive Tract of Periplaneta Americana
Through the above analysis of taxonomic levels of all floral, it was preliminarily judged that different analyze differences between floras in samples in terms of community complexity, using UniFrac software under QIIME platform, differences between samples were analyzed. PCoA and heatmap analyses were performed respectively.
As shown in Figure 6, from UniFrac PCoA analysis diagram under weighted UniFrac algorithm, the data of all samples in T group and SG group, as well as two samples in PT group and ZF group respectively are closely clustered, suggesting that intestinal floras between ten samples of four groups were similar and their flora structure were different from other samples. The data of DB group were dispersed, suggesting that individual differences in this group were large.
As shown in Figure 7, the clustering tree using weighted (OTU abundances considered) calculation method had two branches. Differences between 10 samples in T group, SG group, PT group and ZF group were small, but they were quite different from other samples. The results of clustering tree verified PCoA analysis.  Therefrom, in microbial communities in the digestive tract of Periplaneta americana in SG group, T group and ZF group, the proportion of Blattabacterium was higher, suggesting that an intake of sucrose-and fat-rich fodder may facilitate the growth and reproduction of Blattabacterium. Therefore, in the breeding process of Periplaneta americana, sucrose and fat were added to the fodder to increase the content and proportion of Blattabacterium in microbial communities and facilitate the production of all essential amino acids, multiple vitamins and other essential compounds in Periplaneta americana and thus improve the quality of medical material.

Conclusion
Periplaneta americana has a complex microbial community structure. Different feeding conditions may change the microbial community structure of Periplaneta americana. An important bacterial genus in Periplaneta americana, Blattabacterium is positively correlated with the intake of sucroseand fat-rich fodder. In the breeding process of Periplaneta americana, adding sucrose and fat to fodder may increase the content and proportion of Blattabacterium in microbial communities.