Effect of Different Host Plants on the Diversity of Gut Bacterial Communities of Spodoptera frugiperda (J. E. Smith, 1797)

Simple Summary The gut microbes are important to the insect physiology and behavior. Spodoptera frugiperda (J. E. Smith) is a worldwide migratory invasive and polyphagous pest, which causes great damage to various host crops in China. The structures of gut bacterial community of S. frugiperda feeding on different host plants are largely unexplored. In this study, high-throughput sequencing was used to compare the gut bacterial communities of S. frugiperda reared on different host plants. The results confirmed that the host plants can significantly affect the gut bacterial diversity and structure of S. frugiperda. This study lays a foundation for further studies on the function of gut bacteria of S. frugiperda and its adaptive mechanism to the host. Abstract Intestinal symbiotic bacteria have formed an interdependent symbiotic relationship with many insect species after long-term coevolution, which plays a critical role in host growth and adaptation. Spodoptera frugiperda (J. E. Smith) is a worldwide significant migratory invasive pest. As a polyphagous pest, S. frugiperda can harm more than 350 plants and poses a severe threat to food security and agricultural production. In this study, 16S rRNA high-throughput sequencing technology was used to analyze the diversity and structure of the gut bacteria of this pest feeding on six diets (maize, wheat, rice, honeysuckle flowers, honeysuckle leaves, and Chinese yam). The results showed that the S. frugiperda fed on rice had the highest bacterial richness and diversity, whereas the larvae fed on honeysuckle flowers had the lowest abundance and diversity of gut bacterial communities. Firmicutes, Actinobacteriota, and Proteobacteria were the most dominant bacterial phyla. PICRUSt2 analysis indicated that most of the functional prediction categories were concentrated in metabolic bacteria. Our results confirmed that the gut bacterial diversity and community composition of S. frugiperda were affected significantly by host diets. This study provided a theoretical basis for clarifying the host adaptation mechanism of S. frugiperda, which also provided a new direction to improve polyphagous pest management strategies.


Introduction
Most insect guts contain a multitude of commensal microorganisms, which form an interdependent symbiotic relationship with insects after long-term coevolution [1,2]. With advances in high-throughput sequencing technology, the focus of research has gradually transferred to the composition and functions of microorganisms related to insects. Bacteria are the dominant group in most insect microbial communities and contribute to various insect life processes, such as digestion and absorption of nutrients, immune defense response, degrade exogenous biological toxins, pesticide resistance, reproductive regulation, and host adaptability regulation [3][4][5][6][7][8][9][10][11].
Previous studies indicated that the composition of the gut microbiota communities of insects is affected by multiple endogenous and exogenous factors. Firstly, different species and developmental stages of insect hosts had distinctive intestinal microbial compositions [12][13][14].
Spodoptera frugiperda was obtained from the Henan Academy of Agricultural Sciences (Zhengzhou, China) and reared on maize seedlings for more than 3 years at 26 ± 1 • C, 60-70% relative humidity, and 16 L:8D photoperiod in the artificial climate incubator.
Newly hatched larvae were reared with maize (Zea mays L. Zhengdan 958), wheat (Triticum aestivum L. Jimai 22), rice (Oryza sativa L. Zhenghan 10), the flower of honeysuckle (Lonicera japonica Thunb. Juhua 1), the leaves of honeysuckle (Lonicera japonica Thunb. Juhua 1) and Chinese yam (Dioscorea oppositifolia Turczaninow Xiaobaizui) until fifth instar. Then, 70% ethanol was used to sterilize the larvae for 1 min followed by three sterile water washes. Under a stereoscope, the midgut was dissected in sterile phosphate buffered saline. Larvae fed each different diet were obtained from three different batches and each batch contained 15 to 20 individuals.

DNA Extraction, Library Preparation, and High-Throughput Sequencing
The genomic DNA was extracted from all samples using DNeasy ® PowerSoil ® Pro Kit (QIAGEN, Dusseldorf, Germany), according to the instruction. DNA quality and quantity were detected with a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and the integrity was evaluated using 1% agarose gel. The total DNA was stored at −80 • C. The universal primer pair 338F and 806R was used to amplify the hypervariable regions V3-V4 of the bacterial 16S rRNA gene. The PCR products were detected and purified on 2% agarose gels. The recovered products were quantified using Quantus™ Fluorometer (Promega, Madison, WI, USA) and homogenized to form a sequencing library. Sequencing was conducted on Illumina Miseq PE300 (Illumina, San Diego, CA, USA). Raw reads were submitted to the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) database (accession No. PRJNA869195).

Statistical and Bioinformatics Analysis
FLASH (version 1.2.7) was used to truncate the barcode and primer sequences and reassign the paired-end reads, the criteria of overlap was greater than 10 bp and a had mismatch rate of less than 0.02 [34,35]. Then, the high-quality sequences were classified into multiple operational taxonomic units (OTUs) based on 97% similarity truncation value with UPARSE (version 7.1) [36]. The representative sequences from each OTU were screened and annotated against the 16S rRNA database (Silva database version 138) by RDP classifier (version 2.2) with a confidence threshold at 70% [37]. OTU abundance information was normalized by the smallest sequence for subsequent analyses. Alpha diversity analysis was executed by Mothur (version 1.30.2) software to evaluate the abundance and diversity of microbial communities [38]. We used QIIME (version 1.9.1) to conduct principal coordinate analysis (PCoA) based on the Bray-Curtis similarities index, which was usually applied to reveal the differences in microbial community structures among different treatment groups [39]. To intuitively show the composition and differences of communities, the columnar cumulative diagram and heat map were generated using R vegan package [15]. Linear discriminant analysis (LDA) effect size (LEfSe) was used to identify the species with statistical differences in an abundance of microbial communities. PICRUSt2 was used to annotate pathways of OTUs against the KEGG databases to predict the abundance of microbiota functions [40,41]. All statistical analyses were performed by SPSS (version 22.0) with a one-way ANOVA analysis followed by Tukey's test (α = 0.05).

General Profile of 16S rRNA Sequencing Data
A total of 960, 641 high-quality reads with an average length of 428 bp were obtained (Table S1). The sample rarefaction curve indicated that the sequencing volume and depth of the samples were saturated ( Figure S1). Furthermore, the Good's coverage reflected sequencing integrity. In this study, the coverage of each sample was above 99%, showing that most species in the samples were detected (Table S2). The high-quality sequences were divided into 431 OTUs by cluster analysis, including 21 phyla, 43 classes, 101 orders, 167 families, 292 genera, and 380 species.

Comparison of the Bacterial Communities
The alpha diversity analysis reflected significant differences in the bacterial communities of S. frugiperda larvae feeding on six different host plants ( Figure 1; Table S2). The Ace and Chao1 indices indicated the abundance of microbial community, and the Shannon and Simpson indices reflected the diversity of microbial community ( Figure 1; Table S2). Spodoptera frugiperda larvae fed on rice had the highest Ace and Chao abundance values, the highest Shannon diversity values, and the lowest Simpson diversity values. In contrast, the larvae fed on flowers of honeysuckle had the lowest abundance and diversity of gut bacterial communities ( Figure 1). Ace and Chao1 indices indicated the abundance of microbial community, and the Shannon and Simpson indices reflected the diversity of microbial community ( Figure 1; Table  S2). Spodoptera frugiperda larvae fed on rice had the highest Ace and Chao abundance values, the highest Shannon diversity values, and the lowest Simpson diversity values. In contrast, the larvae fed on flowers of honeysuckle had the lowest abundance and diversity of gut bacterial communities ( Figure 1). We identified the gut microbial communities of each treatment at different taxonomic levels. The main phyla of S. frugiperda larvae were Firmicutes, Actinobacteriota, Proteobacteria, Cyanobacteria, and Bacteroidota, respectively. Among them, Firmicutes was the main component of gut microbes, the relative abundance accounting from 80.80 ± 12.71% to 98.97 ± 0.44%. Firmicutes were more abundant among S. frugiperda larvae fed on maize, honeysuckle flowers and leaves, and Chinese yam than those fed on wheat and rice. The relative abundances of Actinobacteriota in S. frugiperda larvae fed on wheat (13.33 ± 3.89%) and rice (11.90 ± 2.26%) were significantly higher than those fed on other host plants. In addition, the relative abundance of Proteobacteria was much higher in S. frugiperda larvae fed on honeysuckle leaves (15.17 ± 0.87%) than those fed on honeysuckle flowers (Table 1; Figure 2). At the family level, Enterococcaceae was prevalent in all samples, accounting from 68.50 ± 9.75% to 97.48 ± 2.07%. Relative abundances of Staphylococcaceae in S. frugiperda larvae fed on rice (34.31 ± 11.58%) was significantly higher than those fed on other host plants (Table 1; Figure 2). We identified the gut microbial communities of each treatment at different taxonomic levels. The main phyla of S. frugiperda larvae were Firmicutes, Actinobacteriota, Proteobacteria, Cyanobacteria, and Bacteroidota, respectively. Among them, Firmicutes was the main component of gut microbes, the relative abundance accounting from 80.80 ± 12.71% to 98.97 ± 0.44%. Firmicutes were more abundant among S. frugiperda larvae fed on maize, honeysuckle flowers and leaves, and Chinese yam than those fed on wheat and rice. The relative abundances of Actinobacteriota in S. frugiperda larvae fed on wheat (13.33 ± 3.89%) and rice (11.90 ± 2.26%) were significantly higher than those fed on other host plants. In addition, the relative abundance of Proteobacteria was much higher in S. frugiperda larvae fed on honeysuckle leaves (15.17 ± 0.87%) than those fed on honeysuckle flowers (Table 1; Figure 2). At the family level, Enterococcaceae was prevalent in all samples, accounting from 68.50 ± 9.75% to 97.48 ± 2.07%. Relative abundances of Staphylococcaceae in S. frugiperda larvae fed on rice (34.31 ± 11.58%) was significantly higher than those fed on other host plants (Table 1; Figure 2).
The dynamic patterns in the gut bacterial communities of S. frugiperda feeding on different host plants heat map performed with the top 30 abundant genera is shown in Figure 3. The columns represent the samples and the rows represent the bacterial at the genus level. The bacterial communities of S. frugiperda larvae feeding on honeysuckle leaves and Chinese yam were converged to the same branch, while that on rice was in the outermost branch. Enterococcus was the absolute dominant genus of the gut microbiota in all samples. Staphylococcus only had a high abundance of S. frugiperda larvae fed on rice.
The PCoA scatter plot revealed that the principal coordinates represent the two eigenvalues that contribute to the largest variation between the samples ( Figure 4). Based on Bray-Curtis algorithm, the principal coordinates explained 76.55% and 11.06% of the data variation, respectively. The ANOSIM analysis showed that the structure of gut bacterial com-Insects 2023, 14, 264 5 of 12 munity among different groups was significantly different (R = 0.4066, p = 0.001) ( Figure 4A,C). According to weighted Unifrac algorithm, the influence degrees of data changes were 60.99% and 18.33%. The ANOSIM analysis also revealed significant differences in the structure of gut microbiota in different samples (R = 0.2856, p = 0.002) ( Figure 4B,D). The results of both algorithms showed that the bacterial communities of S. frugiperda larvae fed on rice and wheat were dispersed with other larvae groups.   The dynamic patterns in the gut bacterial communities of S. frugiperda feeding on different host plants heat map performed with the top 30 abundant genera is shown in Figure 3. The columns represent the samples and the rows represent the bacterial at the genus level. The bacterial communities of S. frugiperda larvae feeding on honeysuckle leaves and Chinese yam were converged to the same branch, while that on rice was in the outermost branch. Enterococcus was the absolute dominant genus of the gut microbiota in all samples. Staphylococcus only had a high abundance of S. frugiperda larvae fed on rice.
The PCoA scatter plot revealed that the principal coordinates represent the two ei- data variation, respectively. The ANOSIM analysis showed that the structure of gut bacterial community among different groups was significantly different (R = 0.4066, p = 0.001) ( Figure 4A,C). According to weighted Unifrac algorithm, the influence degrees of data changes were 60.99% and 18.33%. The ANOSIM analysis also revealed significant differences in the structure of gut microbiota in different samples (R = 0.2856, p = 0.002) ( Figure  4B,D). The results of both algorithms showed that the bacterial communities of S. frugiperda larvae fed on rice and wheat were dispersed with other larvae groups.   The LEfSe was used to analyze the main bacterial community of different taxa at various levels (phylum to genus) in different groups (LDA threshold was 4) ( Figure 5A). Simultaneously, we drew an evolutionary branching to further understand the difference in bacterial community between different treatments ( Figure 5B). As shown in Figure 5, the The LEfSe was used to analyze the main bacterial community of different taxa at various levels (phylum to genus) in different groups (LDA threshold was 4) ( Figure 5A). Simultaneously, we drew an evolutionary branching to further understand the difference in bacterial community between different treatments ( Figure 5B). As shown in Figure 5, the main gut microbial community of S. frugiperda larvae feeding on Chinese yam were Firmicutes and Actinomycetes, including Lactobacillales, Enterococcus, Bacilli, and Streptosporangiales. Only one dominant microbial species was detected in the honeysuckle flower-feeding group, which belonged to the Brevibacillus genus of Firmicutes. The Bacillales was a dominant taxon of S. frugiperda feeding on the leaf of honeysuckle. For rice-feeding group, Staphylococcaceae and Brevibacteriaceae were the main bacterial communities. There were five microbial species of S. frugiperda larvae feeding on wheat, four of which were exclusive to Firmicutes and the other belonging to Actinomycetes ( Figure 5).

Functional Prediction of Microbiota
The PICRUSt2 results showed that the functional prediction categories of the gut bacteria of S. frugiperda feeding on different hosts were focused on metabolism, environmental information processing, genetic information processing, human diseases, cell transformation, and organismal systems, and the relative abundance of the metabolic pathway was the highest ( Figure 6A). In addition, there were significant differences in the relative abundance of the secondary classification level of metabolic pathways among each treatment ( Figure 6B). Specifically, the relative abundance of amino acid metabolism, energy metabolism, cofactors, and vitamins metabolism in the gut bacteria of S. frugiperda feeding on rice were significantly higher than those of other treatments ( Figure 6B).

Functional Prediction of Microbiota
The PICRUSt2 results showed that the functional prediction categories of the gut bacteria of S. frugiperda feeding on different hosts were focused on metabolism, environmental information processing, genetic information processing, human diseases, cell transformation, and organismal systems, and the relative abundance of the metabolic pathway was the highest ( Figure 6A). In addition, there were significant differences in the relative abundance of the secondary classification level of metabolic pathways among each treatment ( Figure 6B). Specifically, the relative abundance of amino acid metabolism, energy metabolism, cofactors, and vitamins metabolism in the gut bacteria of S. frugiperda feeding on rice were significantly higher than those of other treatments ( Figure 6B).

Discussion
Insect gut microbes are important to the expansion of host populations and they enhance the adaptability to external adverse factors [13,15]. Numerous research demonstrated that the structure and diversity of insect gut bacteria can be significantly affected by host diets [15,20,21]. Our study compared the changes of gut bacterial communities of S. frugiperda fed on two traditional Chinese herbal medicines and three major food crops. The diversity and abundance of the microbial community of S. frugiperda fed on various host plants were significantly different. The abundance and diversity of gut bacterial community of S. frugiperda larvae fed on rice were higher than that on other host plants. It was similar to the research that the gut bacterial diversity in G. molesta and P. xylostella was significantly changed after the host plants shifted [20,21].
At the phylum level, Firmicutes, Actinobacteriota, and Proteobacteria dominated microbial communities of S. frugiperda larvae. Among them, Firmicutes was predominant in all samples (relative abundance >80%). This result was similar to the intestinal bacterial community of many Lepidoptera insects, such as S. exigua, S. litura, and P. xylostella [42-

Discussion
Insect gut microbes are important to the expansion of host populations and they enhance the adaptability to external adverse factors [13,15]. Numerous research demonstrated that the structure and diversity of insect gut bacteria can be significantly affected by host diets [15,20,21]. Our study compared the changes of gut bacterial communities of S. frugiperda fed on two traditional Chinese herbal medicines and three major food crops. The diversity and abundance of the microbial community of S. frugiperda fed on various host plants were significantly different. The abundance and diversity of gut bacterial community of S. frugiperda larvae fed on rice were higher than that on other host plants. It was similar to the research that the gut bacterial diversity in G. molesta and P. xylostella was significantly changed after the host plants shifted [20,21].
At the phylum level, Firmicutes, Actinobacteriota, and Proteobacteria dominated microbial communities of S. frugiperda larvae. Among them, Firmicutes was predominant in all samples (relative abundance >80%). This result was similar to the intestinal bacterial community of many Lepidoptera insects, such as S. exigua, S. litura, and P. xylostella [42][43][44]. Lv et al. also reported that Firmicutes were dominant bacterial phyla in gut bacterial communities of S. frugiperda reared on different host plants [33]. Numerous studies indicated that Firmicutes bacteria are indispensable in the material and energy metabolism of the host [13,45,46]. Firmicutes can synthesize a variety of carbohydrate-degrading enzymes and convert glycogen into nutrients available to the host. Firmicutes can also produce sugar transporters, especially for mannose, which can reduce the toxicity of hazardous substances for insects [47,48]. Besides, Ptaszyńska et al. found that Firmicutes may participate in the infection of honeybees by pathogenic bacteria [49]. The high relative abundance of Firmicutes in all samples in our study supports their importance in the digestion and metabolism processes of S. frugiperda. The high level abundance of Actinobacteriota occurred in S. frugiperda larvae fed on wheat and rice. It has been reported that Actinobacteria is involved in various metabolic and physiological activities, including extracellular enzyme production and antimicrobial activity [50].
Enterococcaceae was the most common gut bacteria in Lepidoptera, such as G. molesta, Manduca sexta, Spodoptera littoralis, and Helicoverpa armigera [15,[51][52][53]. In our study, Enterococcaceae was also a dominant family of microorganisms with the highest abundance in all treatments. Researchers have confirmed that Enterococcaceae play a positive role in host adaptability, which was related to the degradation of the plant cell walls into available nutrients and the metabolization of toxic secondary metabolites [53][54][55]. Recent studies have found that the S. frugiperda invading China has a higher fitness level when fed on maize than when fed on rice [19]. He et al. reported that the survival rate of S. frugiperda larvae fed on rice was only 0.4%, and that they were unable to pupate successfully [29]. We found that the relative abundances of Enterococcaceae in S. frugiperda larvae fed on rice was significantly lower than that on maize. This phenomenon provided a new idea to reveal that the S. frugiperda damage to maize was more serious than to rice in China. Bacillus and Enterbacter were detected in our study with high relative abundance in the S. frugiperda fed on honeysuckle leaves. Bacillus is one of the main symbiotic bacteria that produces digestive enzymes in insects, which can produce cellulase, protease, and lipase [56,57]. Enterbacter plays an important role in glucose metabolism, which can also degrade cellulose [21,58]. These two genera both had the highest abundance in S. frugiperda fed on honeysuckle leaves, which may be related to the high content of cellulose in honeysuckle leaves.
Klebsiella within Enterobacteriaceae is a beneficial microbe, which wildly exists in the guts of insects and other herbivores [59,60]. Klebsiella could increase the fitness of hosts and promote the development and reproduction of insects [21]. Acevedo et al. found that Klebsiella was able to inhibit the defense response of plants, which could improve the survival rate and the environmental adaptability of S. frugiperda larvae [61]. Niyazi et al. reported that Klebsiella could promote the reproductive capacity of Ceratitis capitata [62]. In our study, Klebsiella was detected with a high relative abundance in S. frugiperda fed on maize, wheat, and honeysuckle leaves than that on other diets. Furthermore, our previous study showed that the S. frugiperda in these three groups had shorter larval development duration, higher larval survival rate, and higher reproductive ability than other groups [30]. These results suggested a potential relationship between the abundance of Klebsiella and the development and reproduction of S. frugiperda.

Conclusions
In this study, high-throughput sequencing was used to compare the gut bacterial communities of S. frugiperda reared on different host plants. The results confirmed that the host plants can significantly affect the gut bacterial diversity and structure of S. frugiperda. Spodoptera frugiperda fed on rice had higher bacterial richness and diversity. Firmicutes, Actinobacteriota, and Proteobacteria were the most abundant bacterial phyla. PICRUSt2 analysis indicated that most predicted functions were related to metabolism. This investigation of gut microbiota can supply a theoretical basis for clarifying the host adaptation mechanism of S. frugiperda.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/insects14030264/s1, Figure S1: Alpha diversity index rarefaction curves. (A) Sobs index (B) Shannon index; Table S1: Statistical analysis of V3-V4 hypervariable region of 16S rDNA of the bacteria in the gut of S. frugiperda larvae; Table S2: Analysis of alpha-diversity of midgut bacteria in different treatments of S. frugiperda larvae.

Data Availability Statement:
The data presented in this study are available in the article.