Symbiotic microbial studies in diverse populations of Aphis gossypii, existing on altered host plants in different localities during different times

Abstract Complex interactions between symbiotic bacteria and insects ultimately result in equilibrium in all aspects of life in natural insect populations. In this study, abundance of principal symbiotic bacteria was estimated using qPCR in 1553 individuals of aphids, Aphis gossypii. Aphids were sampled from primary and secondary host plants—hibiscus and cotton. Hibiscus aphids were collected from 24 different locations in April, September, and November, whereas cotton aphids were collected between 2015 and 2017 from areas with wide variations in climatic conditions. About 30%–45% aphids were recorded with the most dominant symbiont, Arsenophonus. The other symbionts were in low frequency, and about 7% of aphids were noted with Hamiltonella, Acinetobacter, and Microbacterium, and 3% of aphids were verified with Serratia and Pseudomonas. Aphids infected with Hamiltonella, Arsenophonus, and Serratia can influence Buchnera densities. Hamiltonella has positive interaction with densities of Arsenophonus and Serratia. Almost 100% coinfection of Hamiltonella and Arsenophonus was detected in Xinxiang aphids and 50% coinfection was reported in aphids from North China, while no coinfection was detected in Hainan aphids. These findings describe the prevalence pattern and richness of core community of symbiotic bacteria in naturally occurring populations of A. gossypii and provide new insights for the study of symbiotic bacteria.

geographical scale with a wide range of host plants, which provides a basis for a highly variable life cycle, with a holocyclic pattern in cold winters and anholocyclic forms in warm regions (Margaritopoulos et al., 2006).
Aphis gossypii usually migrates between the primary and the secondary hosts throughout the year (Kwon & Kim, 2017;Xia et al., 1999). In many areas, A. gossypii uses hibiscus plants as the primary host, between April and mid-May; then migrate to secondary host plants; and again return to feed on hibiscus from October to November, whereas some aphids eat hibiscus all the year long (Charaabi et al., 2008;Zhang & Zhong, 1990).
Symbiotic bacteria exist in many insects, known as primary symbionts (obligate symbionts), secondary symbionts, and facultative symbionts (Bright & Bulgheresi, 2010). Symbiotic bacteria have conditionally positive effects on the physical condition of the host insects. During host adaptation, bacterial symbionts suppress plant resistance (Su et al., 2015) or increases the expression of detoxifying enzymes (Singh et al., 2020) that efficiently enhance the host adaptation. Recent studies showed that symbiotic bacteria contribute to host variation, heat tolerance (Zhang, Leonard, et al., 2019), and temperature preference (Hague et al., 2020).
Meanwhile the host should also bear the metabolic and fitnessdependent cost to symbiont bacterial presence (Engl et al., 2020;Oliver et al., 2008); bacterial symbionts even can shape their host evolution (Coffman & Burke, 2020). Therefore, it is very important and necessary for insects to maintain a balanced population of symbiotic bacteria.
In primary symbionts, Buchnera is essential for sap-sucking aphids-reported in almost all aphid species, and responsible for essential amino acids and other nutrients for growth, reproduction (van Ham et al., 2003), and to improve heat tolerance (Zhang, Leonard, et al., 2019). The interaction of many facultative symbionts within aphids were well-studied, such as the bacterium Hamiltonella increases resistance to its parasitoid via toxin protein (Brandt et al., 2017) and Rickettsiella changes body color from red to green in its natural populations (Tsuchida et al., 2010).
Due to the difference in the life cycle, host type, pesticide selection pressure, dispersal, and migration, the synergy between symbiotic bacteria and aphids is quite different within aphid species or different in populations of same species. Some studies have been carried out in bacterial communities of A. gossypii based on 16S rRNA gene sequencing on Illumina platforms (Gallo-Franco et al., 2019;Xu et al., 2020;Zhao et al., 2016), quantitative PCR, and normalized host genes (Ayoubi et al., 2020;Chong & Moran, 2016;Zhang, Cao, et al., 2016). These findings reflect that symbiotic bacteria play an important role in A. gossypii, but the information about absolute quantity of symbiotic bacteria in natural populations and their changing trends within seasons, geographic areas, and hosts is still lacking. In this study, seven bacterial abundance in 1553 A. gossypii were estimated using quantitative PCR (qPCR), providing novel information about bacterial community related to biotypes, host plants, prevalence time, and geography.

| Field sampling and DNA extracting
For Illumina MiSeq DNA sequence analysis, wingless A. gossypii were collected from cotton fields located in Anyang (Henan Province) and Shihezi (Xinjiang Province) in late June 2016. Different instars of wingless aphids were picked up from cotton and then put in nuclease-free Eppendorf tubes. In order to make the sample data representative, only one aphid per cotton plant in each sampling field was collected. In this way, more than 50 aphids from 50 different fields were collected, mixed, and considered as one sample; 10 samples from each city were collected and immediately immersed in liquid nitrogen and stored at −80°C for further study.
For qPCR analysis, A. gossypii were collected from cotton and hibiscus plants, each adult wingless aphid was put in a nuclease-free Eppendorf tube. (a) From cotton, A. gossypii were collected from Xinjiang (2015) and Henan (2015Henan ( -2017 in June and from Hainan (2015) Province in February. Only one wingless adult aphid per plant from 8 to 48 cotton plants per site was collected to avoid sampling from the offspring of a single female (Zhang et al., 2018); aphids were immersed in liquid nitrogen and stored at −80°C for further study. (b) From hibiscus, A. gossypii were collected from ornamental hibiscus located in urban areas from each location; the locations are shown in Appendix S1. Aphids were collected in April, September, and November 2016. In order to avoid the confounding of the area F I G U R E 1 The mutualistic interaction between Aphis gossypii and ants on cotton. The cotton aphid, A. gossypii Glover, is an important global pest that sucks sap and transmits viral diseases to host plants, causing serious economic losses in agriculture. Due to the difference in the life cycle and other factors, such as the mutualistic interaction between cotton aphids and ants, the synergy between symbiotic bacteria and aphids is quite different within aphid species or different in populations of same species. In this study, seven bacterial abundances in 1553 A. gossypii were estimated using quantitative PCR, providing novel information about bacterial community related to biotypes, host plants, prevalence time, and geography caused by the transplanting of seedlings, we choose hibiscus which has been transplanted more than two years. Only one aphid per plant was sampled, and the next sample was collected from plants more than 10 meters away from the previous plants from where the sample was drawn. Sampled aphids were placed in 95% ethanol and stored in room temperature for further study.
Each sample was washed with 70% ethanol and rinsed three times with nuclease-free water, and total DNA from individual aphids or mixed samples was extracted using the TIANamp Genomic DNA Kit (TIANGEN Biotech (Beijing) LTD., China) according to the manufacturer's instruction. In order to break gram-positive bacterial cells, additional lysozyme (50 mg/ml) was added at the incubation step. Elution buffer, 30 μl, was added at the last step, and re-elution was done for one more time. Negative DNA extraction (control) includes DNA extractions of the nuclease-free water. The quantity and quality of the DNA were measured with a NanoDrop 2000C spectrophotometer (Thermo Scientific). The purified DNA samples were stored at −20°C, and the samples having lower concentration, <25 ng/μl, were excluded from the study.

| 16S rRNA gene amplification and sequencing
The V3-V4 hypervariable regions of the bacterial 16S rRNA gene were amplified using the primers 338F (5′-ACTCCTACGGGAGGCAGCAG -3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′). DNA from mixed aphid samples was used as template DNA. Amplicon generation of PCR products, quantification and qualification, PCR product mixing and purification, library preparation, and sequencing were carried out on an Illumina MiSeq platform at Shanghai Major Biopharm Technology Co., Ltd. Bioinformatics. Sequences with ≥97% similarity were assigned to the same OTUs.

| Quantification of symbiotic bacteria
qPCR was used to determine copies of 16S rRNA genes of dominant bacteria, and entire DNA from individual aphids-diluted 10 times-was used as template DNA. Bacterial special primers (Appendix S2) were designed according 16S rRNA genes using BEACON DESIGNER 7.6 (PREMIER Biosoft International, CA, USA).
Primer PCR efficiencies were tested using series-diluted templates (Appendix S2). The standard template was prepared, as previously reported , and the general step was cloning the target sequence in a plasmid of pEASY-T3 cloning vector (TransGen Biotech, China). Escherichia coli DH5α was used as a host for plasmid propagation, and the target sequence in the plasmid vectors was confirmed by sequencing. Each PCR reaction contained 5 μl 2×TransStart Green qPCR SuperMix (TransGen Biotech, China), 0.2 μl each of 10 mM forward and reverse primers, 2.0 μl template DNA (2.0 μl negative DNA considered as negative controls), and 0.2 μl 50 × ROX; nuclease-free water was added to make up to 10 μl.

The
Step OnePlus™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) was used to perform PCR, according to cycling conditions of 95°C for 3 min followed by 40 cycles of a two-step PCR (95°C for 5 s, 60°C for 30 s). All qPCR reactions were done in triplicate for each individual aphid DNA, and each reaction plate generated a corresponding standard curve. To make sure the primer specificity validated for each primer, checking steps were carried out, as previously reported .

| Statistical analysis
Bioinformatics of 250-bp paired-end reads were conducted, as previously reported . Principal component analysis (PCA) based on the Bray-Curtis method was performed to study alpha diversity (ACE and Chao1 estimators, Good's coverage estimates, and Shannon and Simpson diversity indices) among observed species. Alpha diversity and beta diversity analyses were executed based on normalized output data by random selected sequences per sample according to the sequence number of the sample with the least sequences, and the statistical analyses were carried out by using the independent two-sample t test to identify the differences between two groups, at p < .05, which is considered significant.
All analyses and estimates were carried out on the freely accessible Group comparisons of symbiotic bacteria and abundance were evaluated with the Mann-Whitney U test (group number > 2) and Kruskal-Wallis test (n > 2). Correlation between different bacteria was carried out using the Pearson correlation of the means of these parameters.

| Overview of the bacterial diversity from Henan and Xinjiang
The V3-V4 region of 16S rRNA gene was amplified from A. gossypii samples collected from Henan and Xinjiang provinces, and the Illumina MiSeq PE300 platform was used to generate raw reads. There were 78206-140328 raw reads from Henan samples, and 86988-145634 raw reads from Xinjiang-collected samples. Good's coverage estimates of sequencing data were noted with maximum coverage in all samples, remained more than 99% (Appendix S3). The quality filtering sequences for aphids from Henan and Xinjiang was assigned 274 and 453 OTUs containing 170 co-existed OTUs, respectively.
Xinjiang-collected aphids were found with higher bacterial biodiversity estimates (ACE and Chao1) than the aphids collected from Henan. Aphids from Xinjiang were also found with higher diversity of bacterial species (Figure 2a The most abundant symbiotic bacterial phylum was Proteobacteria, which accounted for 99.76% and 96.04% in Henan and Xinjiang samples, respectively. The most abundant genus was the primary symbiont Buchnera, having a relative abundance of 83.78% in aphids from Henan and 74.93% in aphids from Xinjiang. Generally, Arsenophonus, Acinetobacter, Serratia, Brevundimonas, and Pseudoxanthomonas were the top 5 most abundant facultative symbionts at the genus level in aphids from Henan, which accounted for 15.65% of the total. The bacterial genera Arsenophonus, Hamiltonella, Exiguobacterium, and Kosakonia including unclassified genus of the family Enterobacteriaceae were the 5 most abundant facultative symbionts in aphids from Xinjiang and accounted for 20.87% of the total (Figure 2d).

| The relationship between symbiotic bacteria
The infection frequency and abundance of 7 main symbiotic bacteria in single A. gossypii were analyzed by qPCR. From 1533 A. gossypii collected from North China-Xinjiang and Hainan provinces-100% F I G U R E 2 Bacterial diversity of Aphis gossypii from Henan (HN) and Xinjiang (XJ) provinces by high-throughput sequencing approaches. Community richness and diversity measured by Ace (a) and Chao1 (b), Student's t test was used for analysis significant differences of group mean value, *p < .05 and **p < .01. Principal component analysis based on the 20 samples (c) and symbiotic bacteria relative abundance of the bacteria in A. gossypii on genus level (d) individuals were found with primary symbiont Buchnera, 31.51% individuals were noted by at least one facultative symbiont, and 11.87% individuals were observed with two or more symbionts.
The density of Buchnera was influenced by some facultative symbionts. Both Hamiltonella-and Arsenophonus-infected aphids have higher density of Buchnera than the uninfected ones (Mann-Whitney-U tests, p < .001) (Figure 3a Coinfection with facultative symbionts was common in A. gossypii. It was found that 10.44% and 1.37% aphids were recorded with two or three facultative symbionts simultaneously. Only one aphid was coinfected with four facultative symbionts. The maximum coinfection type was of Arsenophonus and Hamiltonella, accounted for 4.44%, followed by coinfection of Arsenophonus and Microbacterium Buchnera has highest abundance in Xinjiang populations, which was 1.70-fold and 5.04-fold more than Henan (Mann-Whitney-U tests, p < .001) and Hainan populations (Mann-Whitney-U tests, p < .001), respectively. Hamiltonella has a higher titer level in the Hainan population (n = 8), which was 23.84-fold greater than in Xinjiang (n = 68).
The titer level of Serratia and Pseudomonas was maximum in Hainan (n = 11 and n = 12, respectively) than that in Henan populations (n = 6 and n = 8, respectively), which were 6.59-fold and 1.79-fold larger, respectively. The abundance of Arsenophonus was 1.96fold greater in the Xinjiang population than that in Henan (Mann-Whitney-U tests, p < .001) ( Figure 5). The aphids collected from Handan city (Hebei Province) had a higher abundance of Buchnera, 9.47-fold higher than Baoding city population, whereas the Baoding city population has the lowest abundance of Buchnera among 24 geographic populations.

F I G U R E 4 The relationship between facultative symbionts. The correlation between Arsenophonus and Hamiltonella copy numbers (a), Arsenophonus copy numbers in aphid co-infection with Hamiltonella (Ham.Co-infect) and aphids no co-infection with Hamiltonella (Non) (b), Hamiltonella copy numbers in aphid co-infection with Arsenophonus (Arse.Co-infect) and aphids no co-infection with Arsenophonus (Non) (c). The correlation between Serratia and Hamiltonella copy numbers (d), Hamiltonella copy numbers in aphid co-infection with Serratia (Serr.
Co-infect) and aphids no co-infection with (Non) (e). Scatter dot plot with mean ± standard deviation. **p < .01, data were analysis by Mann-Whitney U test  The abundance of Buchnera in aphid populations feeding on cotton was 1.48-fold higher than aphid populations feeding on hibiscus (Mann-Whitney-U tests, p < .001). Microbacterium abundance was 1.12-fold higher in the population feeding on cotton than aphid populations feeding on hibiscus (Mann-Whitney-U tests, p = .002).

| Assessment of infection frequency among aphid populations collected from various host plants
Aphids feeding on hibiscus have 1.47-fold higher Hamiltonella abundance than aphids feeding on cotton (Mann-Whitney-U tests, p = .029). There was no significant difference between Arsenophonus abundance between aphids feeding on hibiscus and aphids feeding on cotton (Mann-Whitney-U tests, p = .277) (Figure 7c).

| D ISCUSS I ON
Symbiotic bacteria are necessary for the survival, reproduction, host adaption, and resistance to biotic or abiotic stresses of most insects, especially for the sap-sucking insect-like aphids (Hosokawa et al., 2010;Ley et al., 2008;Moran, 2007;Teixeira et al., 2008). Henan province S handong province Hebei province ** ** ** ** ** ** ** ** ** population from both locations Henan and Xinjiang; however, the linear distance between the two places exceeds 2,500 km. Aphids from Henan Province have a higher proportion of facultative symbionts than aphids collected from Xinjiang Province, and both richness estimates and biodiversity of bacteria were higher in Xinjiang. PCA studies showed some aphid population in Xinjiang has large variations. Henan is warm temperate-subtropical area, whereas Xinjiang has wide range of temperature differences between days and nights.
Varying environmental conditions may be the cause of these differences during the spread of cotton aphid populations (Dunbar et al., 2007).
Symbiont abundance were previously measured in whole insects as the number of symbiotic bacteria per aphid genome using qPCR and normalized by single-copy gene abundance (Chong & Moran, 2016). In this study, adult aphids were used, during the early stage of adult aphid, and the number of symbiotic bacteria was basically stable (Ayoubi et al., 2020), so the age of adult aphids has little effect on the community of symbiotic bacteria. When aphids reproduce via viviparous parthenogenesis, ovaries occupy a large proportion of the female body, such as the last stage maturing embryo length was longer than 0.8 mm compare with the mother's body size 4.0-5.0 mm of Acyrthosiphon pisum (Rabatel et al., 2013). During development, the embryo of A. pisum just receives a small proportion of Buchnera from the mother at the beginning, immediately following rapid multiplication of the bacterium (Wilkinson et al., 2003). This causes the uneven distribution of symbiotic bacteria among the adult aphids, so we used the abundance of symbiotic bacteria in a single aphid to measure its richness. The DNA of each adult aphid for qPCR analysis was extracted according to a consistent method to ensure that the same DNA yields for each individual. the infection by avoiding the highly virulent strain of Pseudomonas (Hendry et al., 2018). Microbacterium belongs to Actinobacteria, which is widely present in air, soil, water, and plants, as well exists in guts of larvae and adult insects (Dantur et al., 2015), and has not influenced by spirotetramat insecticide used against cotton aphids (Zhang, Pan, et al., 2016). In the current study, there was a group of symbiotic bacteria, which accounts for a large proportion in cotton aphids, identified to family level, Enterobacteriaceae, and excluded due to ineffective detection by specific primers, utilized against Exiguobacterium.

Recent research shows
Hamiltonella also can have an effect on insecticide resistance of aphids (Li et al., 2020) or can alter the dynamics of host metabolic interactions with co-occurring microorganisms (Blow et al., 2020).
Costs and benefits of coinfection have influenced within-host interactions between these symbionts, resulting in coinfection dynamic changes within natural populations (Russell et al., 2013). Hamiltonella can confer tolerance to high temperatures in A. pisum (Russell & Moran, 2006), and some Serratia are known to be involved in the detoxification of insecticides (van den Bosch & Welte, 2017), or it could increase insect susceptibility to exposed insecticides (Skaljac et al., 2018). Although A. gossypii were collected from hibiscus, infection frequencies and abundance remained greatly different among various geographic populations. The symbiotic bacteria may help the host organism to improve adaptability, affected by environmental and/or historical factors (Tsuchida et al., 2002).
In this study, A. gossypii were collected from hibiscus in April, September, and November from North China, where eggs of the aphid hatched on the host plant-hibiscus-and later, after the month of April, migrated to the host plant in summer. There were some cotton aphids still hosting hibiscus plants in September, and alate adults return to the hibiscus to mate and oviposit in November (Xia, 1997).
The natural population of cotton aphids was suppressed to a small number several times a year because of the natural environment, especially due to rain and or absence of food plants (Hu et al., 2017;Liu et al., 2017), so migration of aphids among host plants occurs several times a year. There were differences in infection frequencies of bacterial symbionts among geographic populations, and most of the facultative symbionts spread more likely through horizontal transfer and can pass through plants (Henry et al., 2013) Hamiltonella, and Arsenophonus can affect the density of Buchnera with different modes, and there was also obvious interaction between facultative symbionts. By analysis of the symbiotic bacteria from different geographic areas, we speculate that A. gossypii migrate in a small area.

ACK N OWLED G M ENTS
This study was supported by a grant from the National Natural Science Foundation of China (31972355).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interests.

E TH I C S A PPROVA L
This study does not contain any studies with human participants or vertebrate performed by any of the authors. Cotton aphids are invertebrate insects and, according to the IUCN criteria, are not considered as endangered or protected species. Cotton aphid samples were collected from farmer's fields and ornamental hibiscus after obtaining their permission.

I N FO R M E D CO N S E NT
Informed consent was obtained from all individual participants included in the study.

DATA AVA I L A B I L I T Y S TAT E M E N T
All raw sequences were deposited in the NCBI Sequence Read Archive under accession number SRA Accession no. PRJNA725955.