Characterization of Heat Shock Transcription Factor in Bemisia Tabaci MED and Regulation of Genes Encoding Heat Shock Proteins During Temperature Stress


 The primary function of heat shock transcription factor (HSF) in the heat shock response is to activate the transcription of genes encoding heat shock proteins (HSPs). The phloem-feeding insect Bemisia tabaci (Gennadius) is an important pest of cotton, vegetables and ornamentals that transmits several plant viruses and causes enormous agricultural losses. In this study, the gene encoding HSF (Bthsf1) was characterized in MED B. tabaci. The full-length cDNA encoded a protein of 652 amino acids with an isoelectric point of 5.55. The BtHSF1 deduced amino acid sequence showed strong similarity to HSF in other insects. Expression analyses using quantitative real-time PCR indicated that Bthsf1 was significantly up-regulated in B. tabaci adults and pupae during thermal stress. Although Bthsf1 was induced by both hot and cold stress, the amplitude of expression was greater in the former. Bthsf1 had distinct, significant differences in expression pattern during different duration of high but not low temperature stress. Oral ingestion of dsBthsf1 repressed the expression of Bthsf1 and four heat shock proteins (Bthsp90, Bthsp70-3, Bthsp20 and Bthsp19.5) in MED B. tabaci during hot and cold stress. In conclusion, our results show that Bthsf1 is differentially expressed during high and low temperature stress and regulates the transcription of multiple hsps in MED B. tabaci.


Introduction
Insects are continually stressed by various environmental factors, and thermal stress is perhaps the most common and direct of these stressors. In response to thermal stress, insects deploy innate resistance mechanisms to alleviate the damage caused by temperature stress [1][2][3] . Among these, heat shock proteins (HSPs) directly respond to temperature stress and have a pivotal role in protecting insects from thermal damage 4 . In insects, HSPs can be subdivided into HSP100, HSP90, HSP70, HSP60, HSP40 and small heat shock proteins (sHSPs) depending on their structure, function and molecular weight [5][6][7][8] . Studies have shown that HSPs interact with heat shock elements (HSE) in the promoter region of genes via heat shock transcription factors (HSFs); this interaction facilitates the recruitment of other transcription factors and the formation of a transcription complex that promotes hsp expression 9,10 .
Heat shock transcription factors are crucial regulatory factors of the heat shock response that are conserved in eukaryotes 10,11 . HSFs are commonly divided into four types, including HSF1, HSF2, HSF3 and HSF4; of these HSF1 is considered to be the main regulator of hsp expression 10,12 . HSF1 is highly conserved in Drosophila, yeast and vertebrates, and its function cannot be replaced by the other three HSF regulators. HSF1 is expressed in response to heat stress in most tissues and cells 13 and has conserved domains: DNA-binding domain (DBD) 14 . After DNA binding, oligomerization, and nuclear localization, HSF1 regulates the expression of stress-induced hsps to foster the organismal response to environmental stressors such as high temperature, heavy metals, and protease inhibitors 15 . The function of HSF1 has been well-studied in insects 16, 17 . In Drosophila, hsf is constitutively expressed in the cytoplasm and nucleus. In vitro studies have con rmed that Drosophila HSF can directly respond to high temperature and oxidative stress, thus indicating that HSF acts as an "thermometer" to regulate the stability of the intracellular environment when physiological tolerance is exceeded 18 .
The white y, Bemisia tabaci (Gennadius), is a species complex that contains 44 cryptic species 19 . It is polyphagous and colonizes over 600 known host plants 20,21 . B. tabaci feeds directly on plants, secretes honeydew, and disseminates plant viruses; it is an invasive pest that causes damage to host plants and serious economic losses to crop production worldwide 22,23 . The invasive species represented by MED cryptic species (B. tabaci Q) is the most serious form of this pest. It can spread rapidly and competes to replace indigenous species, including the MEAM1 cryptic species (B. tabaci B). The adaptability of the MED cryptic species is related to many external factors, including pesticide sensitivity, behavioral interactions and host range [24][25][26][27][28] . The malleability of the MED cryptic species is the primary reason it can quickly adapt to different habitats, including those with temperature extremes [29][30][31] .
Several studies have demonstrated that thermotolerance of the MED cryptic species correlates with hsp expression, especially hsp90, hsp70 and shsps 30,32,33 . However, the relationship between these three hsp gene families and the white y heat shock transcription factor is not clear. In the present study, we cloned and identi ed the full-length gene encoding B. tabaci heat shock transcription factor 1 (Bthsf1) and analyzed its expression during temperature stress. RNA interference (RNAi) was used to further understand the role of BtHSF1 in the regulation of hsps in B. tabaci, which may ultimately lead to improved control methods for this important pest.

Materials And Methods
Insects. B. tabaci were reared on tomato (Solanum lycopersicum) in controlled temperature chambers plants as described 33 . Identi cation of the B. tabaci MED cryptic species was determined using the mitochondrial cytochrome oxidase I (mtCOI) gene as described previously 53 .
Isolation of RNA, cloning and RACE. Total RNA was isolated from B. tabaci pupae and adults as described previously and stored at -80 °C until needed 33 . cDNA was synthesized using an oligo(dT) 18 primer (TaKaRa), and full-length cDNAs encoding HSF1 were obtained by 5'-and 3'-RACE (SMART RACE, Clontech) using the primers listed in Table 1. HSF sequences were con rmed by 5' RACE.
Isolation and characterization of Bthsf1. The fragment of HSF1 was isolated and identi ed based on analysis of the published transcriptome data 54 . The primers used for amplifying fragment are provided in Table 1. PCR products were puri ed, cloned, and sequenced as described 33 .
Established methods were used for identifying ORFs and aligning amino acid sequences 33,55 . Bthsf1 sequences were analyzed with tools available at the ExPASy Molecular Biology Server (https://www.expasy.org/) including Compute pI/MW, BLAST, and Translate. Phylogenetic analyses were conducted as described previously 33,56 . The three-dimensional (3D) structure of the DBD domain was predicted by the SWISS-MODEL website (https://swissmodel.expasy.org/) using the Drosophila melanogaster DBD domain (SMTL ID: 1hkt.1) as a template.
Synthesis of dsRNA. Full-length B. tabaci HSF1 gene was identi ed using the online website (http://sidirect2.rnai.jp/); the regions for RNA silencing were determined, and primers for RNAi were designed. Sense and antisense primers included a T7 promoter sequence (TAATACGACTCACTATAGGG) at the 5' ends to catalyze transcription from both cDNA strands (Table 1). dsRNA speci c to the gene encoding green uorescence protein (dsGFP) was used as a control (Table 1). PCR products were cloned in pGEM-T easy (Promega, Madison, WI, USA) and resulting constructs were used as template DNA in subsequent ampli cations. The PCR product was used for preparation of double-stranded RNA (dsRNA) using the MEGAscript® RNAi kit according to the manufacturer's instructions (Thermo, Waltham, MA, USA) 57 . The quality of dsRNA was evaluated by spectrophotometry and gel electrophoresis and then diluted into 30% (w/v) sucrose for use in experiments.
Oral ingestion of dsRNA. Feeding chambers for delivering dsRNA were constructed as described previously 58 with minor modi cations. Two pieces of Para lm membrane (2×2 cm 2 ) were stretched out by hand until they were each 2-fold their original length. A tube was sealed with 2 layers of membrane containing 30% (w/v) sucrose solution (300 μL) between them. One side of the tube was covered with a piece of meshed net to allow aeration. Adult white ies (aged less than 12 h) were released into the Para lm chamber before covering it with a meshed net. A Para lm sandwich was positioned into the top of the tube and the tube was incubated at 25 °C. For experiments, various amounts (500 ng/μL) of dsBthsf1 or dsGFP were diluted into 30% (w/v) sucrose solution. Experiments were conducted four times under identical conditions. Temperature exposure. B. tabaci adults and pupae were collected, placed in glass tubes, and exposed to each of the following temperatures for 1 h: -12, -10, -8, -6, 39, 41, 43, and 45 °C. Adults and pupae that were maintained at room temperature (26 °C) were used as controls. Treated adults and pupae were allowed to recover at 26 °C for 1 h and were then frozen in liquid nitrogen and stored at -80 °C (N=4).
In experiments with different duration of temperature, B. tabaci adults (n=60) were exposed to high temperatures (31, 37 and 43 °C) for 15 min, 30 min, 1 h, 1.5 h, and 2 h and low temperatures (-10, -4 and 2°C ) for 30 min, 1 h, 1.5 h, 2 h and 3 h. Insects were then allowed to recover at 26 °C for the same duration as the temperature treatment (N=4). Adults and pupae maintained at room temperature (26 °C) were used as controls.
For RNAi, newly emerged B. tabaci adults were supplied with dsBthsf or dsGFP for 1 day, exposed to -6 and 41 °C for 1 h, and then allowed to recover at 26 °C for 1 h. The mortality of B. tabaci was checked after temperature stress, and the surviving B. tabaci were frozen in liquid nitrogen and stored at -80 °C. Each treatment included four biological replications.
Quantitative real-time PCR. The cDNA template was transcribed from RNA with the HiScript III RT SuperMix for qPCR (Vazyme, Nanjing, China) as recommended, and primers were designed with Primer 5.0 software (Table 1). Quantitative real-time PCR (qRT-PCR) was performed in 20 μL total reaction volumes comprised of 10 μL of 2×ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China), 1 μL of each gene speci c primer (Table 1), and 2 μL of cDNA templates. It was carried out that reactions on a CFX-Connect real-time PCR system (Bio-Rad, Berkeley, CA, USA) using the following conditions: 3 min at 95 °C, 40 cycles of denaturation at 95 °C for 30 s, and annealing (30 s) at 60 °C for each gene. and gene expression was calculated using the 2 −ΔΔCt method and normalized to the abundance Elongation factor 1 alpha (EF-1α) and 60S ribosomal protein L29 (RPL29) 59 .
Data Analysis. One-way ANOVA, followed by Tukey's and Duncan's multiple comparison, was used to detect signi cant differences among temperatures using SPSS v. 16.0 60 . For ANOVA, data were transformed for homogeneity of variances, and differences were considered statistically signi cant when P < 0.05.
For RNAi, the relative abundance of target genes and survival rates were compared to the dsGFP control. Student's t-test was used to compare differences in gene expression and mortality with SPSS v. 16.0, and differences were considered signi cant at P < 0.05.

Data availability
Data of this study were included in the supplementary materials.

Results
Sequence analysis of Bthsf1. The full-length cDNA of Bthsf1 was 2500 bp and encoded a predicted protein containing 725 amino acids (GenBank accession no. MW478139) (Fig. S1). The predicted protein product of Bthsf1 was 80.23 kDa with an isoelectric point of 5.90. When the GenBank and PROSITE databases were compared, the BtHSF1 deduced protein showed high similarity to the HSF1 family; InterPro analysis indicated that BtHSF1 contained a conserved DNA-binding domain (DBD) at amino acid residues 10-114 (Fig. 1A). The 3D structure of HSF1 in B. tabaci was modeled using the DBD domain in D. melanogaster (SMTL ID: 1hkt.1) as a template (Fig. 1B); HSF1 showed 69.81% sequence identity to the D. melanogaster orthologue.
Phylogenetic analysis of BtHSF1. The BtHSF1 deduced amino acid sequence was compared with orthologous proteins in other insects. B. tabaci HSF1 showed high sequence identity with HSF in D. melanogaster, Apis cerana, Bombyx mori and Nilaparvata lugens (Fig. 1A). A phylogenetic tree was generated using the amino acid sequences of 15 HSF family members in orders Lepidoptera, Diptera, Coleoptera and Hemiptera (Table S1). BtHSF1 grouped in a well-supported cluster with other members of the Hemiptera and was well-separated from insects in other orders (Fig. 2).

Discussion
A variety of internal and external stimuli can activate HSF, including heat shock. There are three key steps in HSF function in heat stress including the following: polymerization of HSF from monomer to trimer; recognition and binding of HSF to the HSE in hsp promoter regions; and transcriptional activation of the hsps 34 (Fig. 7). Therefore, it is important to study how HSF regulates the expression of genes encoding HSPs when insects undergo thermal stress.
In this study, we cloned and identi ed the Bthsf1 in B. tabaci MED cryptic species. The deduced BtHSF1 contains the conserved motif (DNA-binding domain) of the HSF family. The predicted amino acid sequence of Bthsf1 shows considerable sequence similarity with HSF orthologues in D. melanogaster, B. mori, A. cerana and N. lugens. Phylogenetic analysis revealed that BtHSF1 resides within a phylogenetic group that includes HSF in other Hemiptera insects, including NlHSF in the brown planthopper (N. lugens), AgHSF in the cotton aphid (Aphis gossypii) and DnHSF in the Russian wheat aphid (Diuraphis noxia). The phylogenetic conservation of HSF within the Hemiptera indicates that BtHSF1 could be potentially useful in taxonomic studies. The deduced 3D structures of HSF1 share typical features of the HSF1 family, which indicates conservation in the HSF1 family.
Under normal conditions, HSF exists as an inactive monomer in the cytoplasm and is bound to HSPs 15 . When cells are subjected to thermal stress, the internal environment shifts, which relieves the inhibition of HSF activity. Interestingly, there are relatively few studies documenting hsf expression patterns in insects during temperature stress. Our results show that Bthsf1 can be signi cantly activated and expressed constitutively by high and low thermal stress, It shows that this transcription factor can interact with the HSEs when the white y resists external temperature stress, and promote the expression of HSPs 30,32,33 . In D. melanogaster, the transcription of Dmhsf, Dmhsfb and Dmhsfd were upregulated during temperature stress 17,35 , and genes encoding HSF in Mamestra brassicae and Agasicles hygrophila was also induced by thermal stress 36,37 , However, it is important to note that there obvious differences in hsf expression among insects; for example, Cchsf encoding HSF in Cotesia chilonis was induced by low but not high temperature stress 38 .
Several studies have shown that the duration of temperature stress impacts the growth and development of organisms [39][40][41] . In this study, we analyzed Bthsf1 expression during different duration of temperature stress. Bthsf1 had obvious peak expression levels during high temperature stress; e.g. prominent peaks at 15 min (31 °C) and 1 h (31 °C, 37 °C, 43 °C). However, these spikes in Bthsf1 transcript levels were not observed during low temperature stress. Our results indicate that B. tabaci is conditioned by heat stress to activate HSF1 and promote hsp expression. Furthermore, our ndings help explain why the fold increases in Bthsps transcription during high temperatures are so much higher than hsp expression levels during cold temperatures 30,33 . During prolonged periods of heat stress, Bthsf1 is gradually down-regulated; the protracted accumulation of HSPs becomes deleterious to the cell, which leads to the repression of HSF by HSP70 and other molecular chaperones [42][43][44] . Our results reveal the importance of studying the expression of hsf and hsp concurrently during thermal stress.
The feeding method of dsRNA delivery has been widely and successfully used to study gene function in hemipteran insects [45][46][47][48][49][50] . When B. tabaci was supplied with dsBthsf1 for one day, Bthsf1 expression was signi cantly downregulated after exposure to -6 °C and 41 °C, and mortality increased relative to the dsGFP control. The contribution of HSF to hsp expression, fecundity and survival during adverse conditions has been studied in other organisms. For example, in Artemia franciscana, hsf1 knockdown decreased hsp expression in diapausing embryos 51 , and RNAi-mediated suppression of hsf in Haliotis diversicolor downregulated several hsps 52 . In A. hygrophila, microinjection of dsAhHsf into newly-emerging adults reduced the expression of two different hsps and decreased egg production and survival 37 . In our study, RNAi with dsBthsf1 resulted in a signi cant down-regulation of Bthsp90, Bthsp70-3, Bthsp20, and Bthsp19.5 at -6 °C and 41 °C, indicating that Bthsf1 is involved in the regulation of multiple HSP genes in B. tabaci. Expression of Bthsp70-1 was not diminished in RNAi experiments, indicating that other regulatory pathways function to control hsp expression. Collectively these ndings indicate that Bthsf1 can regulate the expression of some but not all hsps, and further studies are warranted to con rm BtHSF1 interactions and regulatory functions.   Table  S1. (B) 3D predicted structure of HSF1.

Figure 2
Phylogenetic analysis of HSF1 in B. tabaci and other insect species. Numbers on the branches are bootstrap values obtained from 1,000 replicates. Accession numbers and abbreviations for the insect species are listed in Table S1.   The expression of hsf1 and hsps in B. tabaci after oral delivery of dsBthsf1 and dsGFP. (A) -6 °C, (B) 41°C . Asterisks represent signi cant differences between dsGFP and dsBthsf1-treated insects; ns indicates no signi cant difference.

Figure 6
Effects of thermal treatments on the mortality of dsRNA-ingested B. tabaci. Mortality of B. tabaci was determined after thermal treatments for 1 h. Asterisks represent signi cant differences between dsGFP and dsBthsf1.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.