Group V Chitin Deacetylases Influence the Structure and Composition of the Midgut of Beet Armyworm, Spodoptera exigua

Chitin deacetylase (CDA) can accelerate the conversion of chitin to chitosan, influencing the mechanical properties and permeability of the cuticle structures and the peritrophic membrane (PM) in insects. Putative Group V CDAs SeCDA6/7/8/9 (SeCDAs) were identified and characterized from beet armyworm Spodoptera exigua larvae. The cDNAs of SeCDAs contained open reading frames of 1164 bp, 1137 bp, 1158 bp and 1152 bp, respectively. The deduced protein sequences showed that SeCDAs are synthesized as preproteins of 387, 378, 385 and 383 amino acid residues, respectively. It was revealed via spatiotemporal expression analysis that SeCDAs were more abundant in the anterior region of the midgut. The SeCDAs were down-regulated after treatment with 20-hydroxyecdysone (20E). After treatment with a juvenile hormone analog (JHA), the expression of SeCDA6 and SeCDA8 was down-regulated; in contrast, the expression of SeCDA7 and SeCDA9 was up-regulated. After silencing SeCDAV (the conserved sequences of Group V CDAs) via RNA interference (RNAi), the layer of intestinal wall cells in the midgut became more compact and more evenly distributed. The vesicles in the midgut were small and more fragmented or disappeared after SeCDAs were silenced. Additionally, the PM structure was scarce, and the chitin microfilament structure was loose and chaotic. It was indicated in all of the above results that Group V CDAs are essential for the growth and structuring of the intestinal wall cell layer in the midgut of S. exigua. Additionally, the midgut tissue and the PM structure and composition were affected by Group V CDAs.


Introduction
Spodoptera exigua (Hübner) is a moth (Lepidoptera) of the Noctuidae family and a considerable polyphagous economic pest in the world [1,2], causing damage to a variety of crops. S. exigua is found in all provinces, regions and cities in China. Seventeen Chinese provinces have reported S. exigua outbreaks, resulting in significant losses in agricultural production [3,4]. Chemical pesticides are currently the primary means of controlling the reproduction of S. exigua. However, because of an over-reliance on chemical pesticides, S. exigua resistance has grown fast. In some regions, a high level of cross-resistance to various pesticides has been developed in field populations in recent years [5][6][7].

Group V SeCDAs Sequence Analysis
Nucleotide sequences of the cDNAs for Group V SeCDAs were obtained from the S. exigua genome (Table 1). It was shown by the NCBI Blastp comparisons that SeCDAs only have a CDA domain (Figure 1). It was demonstrated by the deduced protein sequences that SeCDAs proteins are synthesized as preproteins with signal peptides using the SignalP 5.0 Server. The secreted SeCDAs contain multiple N-glycosylation sites and five conserved motifs of the CDA domain (Supplementary Figure S1).

Group V SeCDAs Expression Analyses
It was shown via qRT-PCR analysis that Group V SeCDAs are expressed in different developmental stages of S. exigua. SeCDA6 gene expression was highest in eggs, followed

Group V SeCDAs Expression Analyses
It was shown via qRT-PCR analysis that Group V SeCDAs are expressed in different developmental stages of S. exigua. SeCDA6 gene expression was highest in eggs, followed by the fifth-and second-instar larvae. Expression of SeCDA7 and SeCDA9 genes was found in the fifth-instar larvae. The SeCDA8 gene was highly expressed in eggs and prepupae ( Figure 3).  The expression of SeCDAs in different tissues of the fifth-instar day-2 larvae was shown via qRT-PCR analysis. SeCDAs were expressed in all tissues of the fifth-instar larvae and were highly expressed in the midgut ( Figure 4A). The SeCDAs expression was more abundant in the anterior region of the midgut, followed by the middle and posterior regions of the midgut ( Figure 4B). The expression of SeCDAs in different tissues of the fifth-instar day-2 larvae was shown via qRT-PCR analysis. SeCDAs were expressed in all tissues of the fifth-instar larvae and were highly expressed in the midgut ( Figure 4A). The SeCDAs expression was more abundant in the anterior region of the midgut, followed by the middle and posterior regions of the midgut ( Figure 4B).

Expression of Group V SeCDAs in Response to 20E and JHA Signaling
Furthermore, qRT-PCR was used to analyze the effects of the injection of 20-hydroxyecdysone (20E) and juvenile hormone analog (JHA) on the SeCDAs of S. exigua. The injection resulted in the down-regulation of the transcriptional expression level of SeC-DAs when treated with 20E ( Figure 5A). The expression of SeCDA6 and SeCDA8 decreased after JHA treatment. However, the expression of SeCDA7 and SeCDA9 increased after JHA treatment ( Figure 5B).

Functional Analysis of Group V SeCDAs
qRT-PCR analysis showed that the injection of dsSeCDAs can cause the silencing efficiency of SeCDAs to be 79.28%, 75.55%, 73.23% and 87.57%, respectively. The relative expression level of SeCDAs was reduced by 99.89%, 90.45%, 84.46% and 96.60% after the injection of dsSeCDAV ( Figure 6). Group V SeCDAs were silenced in larvae injected with dsSeCDA6 and dsSeCDA8, respectively. SeCDA6 and SeCDA7 were silenced when dsSeCDA7 was injected into the fourth-instar larvae. Only the SeCDA9 gene was silenced when dsSeCDA9 was injected. This was indicated by the results that the silencing efficiency of SeCDAs can be up-regulated via dsSeCDAV injection. After the GFP and SeCDAs were silenced, the larvae were molted and pupated normally (Supplementary Figure S4).

Expression of Group V SeCDAs in Response to 20E and JHA Signaling
Furthermore, qRT-PCR was used to analyze the effects of the injection of 20-hydroxyecdysone (20E) and juvenile hormone analog (JHA) on the SeCDAs of S. exigua. The injection resulted in the down-regulation of the transcriptional expression level of SeCDAs

Functional Analysis of Group V SeCDAs
qRT-PCR analysis showed that the injection of dsSeCDAs can cause the silencing efficiency of SeCDAs to be 79.28%, 75.55%, 73.23% and 87.57%, respectively. The relative expression level of SeCDAs was reduced by 99.89%, 90.45%, 84.46% and 96.60% after the injection of dsSeCDAV ( Figure 6). Group V SeCDAs were silenced in larvae injected with dsSeCDA6 and dsSeCDA8, respectively. SeCDA6 and SeCDA7 were silenced when dsSeCDA7 was injected into the fourth-instar larvae. Only the SeCDA9 gene was silenced when dsSeCDA9 was injected. This was indicated by the results that the silencing efficiency of SeCDAs can be up-regulated via dsSeCDAV injection. After the GFP and SeCDAs were silenced, the larvae were molted and pupated normally (Supplementary Figure S4).

Immunohistochemistry
The specific signal of SeCDAs was weak in the midgut after dsSeCDAV injection, which confirmed the specificity and silencing effect of the SeCDAs antibody. The immunohistochemistry method was used to explore the biological role of SeCDAs. Group V SeCDA protein was localized on the intestinal wall cells of the midgut of S. exigua. After

Immunohistochemistry
The specific signal of SeCDAs was weak in the midgut after dsSeCDAV injection, which confirmed the specificity and silencing effect of the SeCDAs antibody. The immunohistochemistry method was used to explore the biological role of SeCDAs. Group V SeCDA protein was localized on the intestinal wall cells of the midgut of S. exigua. After silencing SeCDAs, the intestinal cell layer became thinner, and the intestinal cells were more closely arranged (Figure 7).

Scanning Electron Microscopy Analysis
Seventy-two hours after the injection of dsSeCDAV, the morphology of the midgut and PM were observed under a scanning electron microscope. The results showed that the vesicles on the midgut epithelia disappeared after the injection of dsSeCDAV compared with the control injected with dsGFP ( Figure 8A). The PM treated with dsSeCDAV became rough with large folds and pores; in contrast, the chitin microfilament structure was loose and disordered ( Figure 8B). Detection of SeCDAs using a specific antibody. The specific signal of SeCDAs was detected after dsGFP and dsSeCDAV injection into fourth-instar larvae. The negative control was with the pre-immune serum. The SeCDAs protein is represented by green and the cell nucleus by red. Cm: circular muscle; Pm: peritrophic membrane; Rc: regenerative cells; Dc: digestive cells; scale bar is 50 µm.

Scanning Electron Microscopy Analysis
Seventy-two hours after the injection of dsSeCDAV, the morphology of the midgut and PM were observed under a scanning electron microscope. The results showed that the vesicles on the midgut epithelia disappeared after the injection of dsSeCDAV compared with the control injected with dsGFP ( Figure 8A). The PM treated with dsSeCDAV became rough with large folds and pores; in contrast, the chitin microfilament structure was loose and disordered ( Figure 8B).

Discussion
According to the GenBank databases, five groups of CDA proteins have been classified according to the alignment of their predicted chitin deacetylase domain sequences. SeCDA6/7/8/9, which only contains a CDA catalytic domain, belongs to the Group V CDAs protein (Figure 1). It was indicated by recent studies on CDA in insects that the abundance of these proteins was different in the tissue, developmental stages and species. The first CDA protein, TnPM-P42, a midgut-specific protein, was identified from T. ni in 2005 [13]. In subsequent studies, Group I CDA was mainly found in ectodermal tissues, such as the foregut, hindgut, integument and trachea. Group III CDA occurred primarily in specialized epidermal tissues, and Group V CDAs were mainly expressed in the midgut-specific proteins. BmCDA6/7/8 were specifically described in the anterior part of the midgut depending on the larval instar. The expression of BmCDA6 was concentrated primarily in the middlemolting stage. BmCDA7 and BmCDA8 were only highly expressed in the feeding stages. [29]. TcCDA6/7/8/9 were expressed specifically in the midgut during the larval feeding stage [12]. SeCDAs identified from S. exigua were more abundant in the most anterior region of the midgut, which was similar to Group V CDAs in B. mori [29] and T. castaneum [12]. This indicated that SeCDAs were primarily synthesized in the anterior part of the midgut. We speculate that SeCDA7 and SeCDA9 may be involved in modifying and maintaining midgut characteristics to stabilize the metabolism. Moreover, SeCDA6 and SeCDA8 proteins were more frequently detected in the eggs. We speculate that SeCDA6 and SeCDA8 may be involved in midgut formation during egg development and the modification of their structures.
In insect, the bioactive hormone 20E and juvenile hormone (JH) play a crucial role in development, metamorphosis and reproduction. Insect molting may be induced by this process [30] and regulate specific gene expression [31][32][33], such as five chitin synthesis

Discussion
According to the GenBank databases, five groups of CDA proteins have been classified according to the alignment of their predicted chitin deacetylase domain sequences. SeCDA6/7/8/9, which only contains a CDA catalytic domain, belongs to the Group V CDAs protein (Figure 1). It was indicated by recent studies on CDA in insects that the abundance of these proteins was different in the tissue, developmental stages and species. The first CDA protein, TnPM-P42, a midgut-specific protein, was identified from T. ni in 2005 [13]. In subsequent studies, Group I CDA was mainly found in ectodermal tissues, such as the foregut, hindgut, integument and trachea. Group III CDA occurred primarily in specialized epidermal tissues, and Group V CDAs were mainly expressed in the midgut-specific proteins. BmCDA6/7/8 were specifically described in the anterior part of the midgut depending on the larval instar. The expression of BmCDA6 was concentrated primarily in the middle-molting stage. BmCDA7 and BmCDA8 were only highly expressed in the feeding stages. [29]. TcCDA6/7/8/9 were expressed specifically in the midgut during the larval feeding stage [12]. SeCDAs identified from S. exigua were more abundant in the most anterior region of the midgut, which was similar to Group V CDAs in B. mori [29] and T. castaneum [12]. This indicated that SeCDAs were primarily synthesized in the anterior part of the midgut. We speculate that SeCDA7 and SeCDA9 may be involved in modifying and maintaining midgut characteristics to stabilize the metabolism. Moreover, SeCDA6 and SeCDA8 proteins were more frequently detected in the eggs. We speculate that SeCDA6 and SeCDA8 may be involved in midgut formation during egg development and the modification of their structures.
In insect, the bioactive hormone 20E and juvenile hormone (JH) play a crucial role in development, metamorphosis and reproduction. Insect molting may be induced by this process [30] and regulate specific gene expression [31][32][33], such as five chitin synthesis pathway genes in S. exigua [34]. Insect egg maturation is triggered by JH, which also helps maintain the larval condition, adult sexual development regulation, and premature larval metamorphosis prevention [35]. The JHA was used instead of JH in the larvae. The treatment of 20E accounted for a significant increase in mRNA expression levels of BmCDA6; however, it decreased the expression levels of BmCDA7 and BmCDA8. The treatment of JHA up-regulated the expression of BmCDA7 and BmCDA8 [16]. Some results obtained after the treatment of 20E in other insects also regulated the CDA gene expression [16,36,37]. In this study, after the larvae of S. exigua were injected with 20E, the expression level of the SeCDA6/7/8/9 was decreased. After treatment with JHA, the expression of SeCDA6 and SeCDA8 decreased, while SeCDA7 and SeCDA9 expressions increased. These results indicate that both 20E and JHA can regulate the expression of CDA genes.
RNAi technology has been widely used to regulate post-transcriptional genes in fungi, nematodes, and insects. According to their functions, insect CDAs can be divided into intestinal-specific proteins (Group V CDA) and parenteral-specific proteins (Groups I, II, III and IV CDA). The TcCDA6/7/8/9 were only expressed in the midgut, which belongs to midgut-specific proteins. There was no aberrant phenotype in individuals treated with dsTcCDA6/7/8/9 [12]. RNAi was used to silence Mythimna seperata CDA1 to evaluate PM alterations [38]. It was shown in the investigation that in the treated specimens, the dsRNA concentration was raised, the PM holes were wide, and the chitin microfilament structure became looser [38]. In our study, the silencing effectiveness of the SeCDA6 gene following RNAi treatment was 79.28%. The SeCDA7 gene silencing effectiveness was 75.55%. SeCDA8 silencing efficacy was 73.23%, and SeCDA9 gene silencing effectiveness was 87.57%. After the injection of dsSeCDAV, the silencing efficacy of SeCDA6/7/8/9 was 99.89%, 90.45%, 84.46% and 96.60%, respectively. The silencing efficiency of all Group V CDAs was considerably better than that of any Group V CDA alone, and a new way to improve the silencing efficiency of Lepidoptera genes was provided. The expression of SeCDAs decreased in all larvae injected with dsSeCDA6 or dsSeCDA8. Additionally, the DNA sequence alignment identity of SeCDA6 and SeCDA8 was as high as 93.64%.
This result may be due to the high similarity of the gene sequences. The silence of the SeCDA7 gene led to decreased gene expression of SeCDA6 and SeCDA7. Only the SeCDA9 gene was silenced when dsSeCDA9 was injected into the larvae. The DNA sequence alignment similarity between SeCDA6 and SeCDA7 was 53.95%. The sequence alignment similarities between SeCDA9 and SeCDA6, SeCDA7, and SeCDA8 are 65.64%, 60.88%, and 65.49%, respectively. The sequence similarity between SeCDA7 and SeCDA8 was only 55.53%. It was suggested by these results that the desired precise targeting effect may not be produced by highly similar gene sequences or the dsRNA targeting the same domain. The reasons for these results are unclear, and more research will be studied in the future.
BmCDA8 was discovered in both the PM and the ectoperitrophic gap between the midgut epithelial cells and the PM. BmCDA7 was discovered in the PM of B. mori [27,39,40]. These midgut CDAs may have been involved in PM production in the anterior midgut, where PM synthesis begins [41,42]. The structure of the chitin matrix in many insects is impacted by the deacetylation process [24,43,44]. It has been noted that PM permeability in vitro may be boosted by recombinant HaCDA5a, a CDA from the moth H. armigera [45]. In this study, we examined the localization of SeCDAs using immunohistology to determine their biological function. SeCDAs are present in the intestinal wall cells of the midgut of S. exigua. When SeCDAs were silenced, the intestinal cell layer shrank, and the intestinal cells were more evenly distributed. Group V CDAs were demonstrated by these findings to be critical in developing and organizing the intestinal wall cell layer in the midgut of S. exigua.
The maintenance of PM mechanical strength in H. armigera is probably correlated with CDA [17]. A change in the peritrophic membrane might make insects more prone to infection, which may eventually be lethal [46]. It was shown via scanning electron microscopy (SEM) analysis of the surface structure of the S. exigua midgut after dsSeCDAV treatment that the microvilli on the epithelia of the larval midgut did not alter substantially. However, all vesicles disappeared after the injection of dsSeCDAV compared to the control. This demonstrates that CDAs do not damage the microvilli but are crucial to the development of vesicles. Compared with the control, the PM of the larvae injected with dsSeCDAV became coarse, with many folds, pores and gaps. The chitin microfilament structure became loose and disorderly, similar to the results of CDA gene silencing in M. separata [38]. Therefore, it is hypothesized that CDA is crucial for maintaining the structural integrity of the PM of S. exigua. Group V CDAs are essential in forming the midgut and PM of S. exigua. However, no phenotypic changes were observed during molting in S. exigua with silenced Group V CDAs. The silencing of Group V CDAs affected the structure of the midgut and the PM. It was demonstrated by all the above results that the PM structure was disrupted when the Group V CDAs were silenced.

Insect Larvae
Larvae of S. exigua were reared on an artificial diet by the Laboratory of Insect Molecular Biology at Hebei Agricultural University. The rearing temperature was 26.5 ± 1 • C, the relative humidity from the egg to the third-instar larvae was 75 ± 10%, the relative humidity from the third-instar larva to the pupal stage was 65 ± 15%, and they were reared under a 14 L:10 D photoperiod. Fifth-instar larvae were used for dissection to isolate the PMs and various tissues for analyses.

Sequence Analysis
The sequence of the gene encoding Group V CDA was analyzed using Clustal X 1.83 and GeneDoc. The gene structure was mapped using the Exon-Intron Graphic Maker (http://www.wormweb.org/exonintron (accessed on 17 May 2021)). The open reading frame, coded amino acid, isoelectric point (pI) and molecular protein weight were predicted using the DNAMAN 6.0 program (LynnonBiosoft, USA). The NetNGlyc 1.0 Server (http:// www.cbs.dtu.dk/services/NetNGlyc/ (accessed on 18 February 2022)) was used to predict the N-glycosylation sites. The TMHMM Server v.2.0 (https://services.healthtech.dtu.dk/ service.php?TMHMM-2.0 (accessed on 19 February 2022)) was used to determine whether it belonged to a transmembrane protein, and SignalP 5.0 (https://services.healthtech. dtu.dk/service.php?SignalP-5.0 (accessed on 19 February 2022)) was used to predict the presence or absence of signal peptides. The NCBI Conserved Domain Database (CCD, https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi (accessed on 23 June 2021)) was used to predict the conservative protein structure domain of SeCDA. Clustal X 1.83 was used for multi-sequence alignment, MEGA 5.0 software was used to build a neighborjoining (NJ) phylogenetic tree, and 1000 repetitions were calculated using the bootstrap method, and the model was the Poisson model.

Gene Expression Analyses
Five tissues (integument, midgut, hindgut, malpighian tubules and fat body) were dissected from fifth-instar larvae for total RNA extraction to determine the tissue-specific expression pattern of SeCDAs. The hindgut and malpighian tubules from 10 larvae and other tissues (integument, midgut and fat body) from three larvae were pooled as one treatment. Three biological replications were prepared (each with three technical replicates). Eggs, first-, second-, third-, fourth-, and fifth-instar larvae, prepupae, and pupae were collected to determine their expression profiles during development. Each sample contained 3-20 individuals. Samples from each developmental stage were analyzed in biological triplicate. Each replicate was given a sample of at least 100 mg of tissue or larvae. The samples were stored at −80 until all samples were collected for RNA extraction.
Total RNA from the frozen tissues was isolated using the RNAprep Pure Tissue Kit (TIANGEN, Co., Ltd., Beijing, China) and treated with DNase I. A spectrophotometer (BioSpectrometer fluorescence, Eppendorf, Germany) and 1.2% agarose gel electrophoresis were used to measure RNA concentration and quality. The entire experimental operation was conducted in an RNase-free environment. A total of 1 µg of total RNA per reaction was used to perform cDNA synthesis using the PrimeScript RT Reagent Kit with gDNA Eraser (Perfect Real Time) (TaKaRa, Beijing, China) with an Oligo dT Primer.
The primers for qRT-PCR were designed using DNAMAN 6.0 software and are listed in Table 2. The melting curve was determined for each sample to detect the gene-specific peak and verify the absence of primer dimers. qRT-PCR was carried out in a 20 µL reaction volume containing 1 µL of cDNA, 10 µM of each primer, 7 µL of H 2 O and 10 µL of TB Green ® Premix Ex Taq™ II (Tli RNaseH Plus), ROX plus (TaKaRa, Beijing, China). PCR was conducted on a CFX-96 real-time PCR detection system (Bio-Rad, Hercules, CA, USA). The amplifications were performed using an Eppendorf thermocycler. They comprised an initial denaturation step at 95 • C for 3 min, followed by 39 cycles at 95 • C for 15 s, 55 • C for 30 s and 72 • C for 30 s, and a final extension step of 5 min at 72 • C. Relative transcript levels of target genes were calculated with the 2 −∆∆Ct method. Three biological repeats, each with three technical replicates, were set for each cDNA, and β-actin was used as the internal reference gene to calibrate the total amount of RNA. β-actin proved to be a suitable housekeeping gene applied in the qRT-PCR of S. exigua in our previous study [46]. The amplification efficiencies of the target and reference were approximately equal (Supplementary Figure S5). The significance of differences between samples was assessed using SPSS 22.0 software (SPSS Inc., Chicago, IL, USA), and the data were presented as means and standard errors (SE). Table 2. Primers used in this study.

Preparation and Treatment with 20E and JHA
A total of 40 fourth-instar larvae of S. exigua with similar body weights and body shapes were selected for the experiment. 20E (Solarbio, Beijing, China) was dissolved in dimethyl sulfoxide (DMSO) at a working concentration. A total of 5 µL of a 100 ng/µL 20E solution was injected into the fourth-instar larvae on the second day, and DMSO was injected into the control group. JHA (Sigma-Aldrich, Shanghai, China) was dissolved in dimethyl ketone (DMK). JHA and DMK were injected into S. exigua as described for the 20E treatment and control. RNA was extracted from whole S. exigua larvae after 12 h of treatment. Reverse transcription was performed, and the expression of SeCDAs genes was analyzed using qRT-PCR.
A total of 150 fourth-instar 1-day larvae were collected and randomly divided into five groups. Aliquots of 10 µg of dsSeCDA6, dsSeCDA7, dsSeCDA8, dsSeCDA9, dsSeCDAV and dsGFP were injected into the hemocoel of the abdomen between the second and third segments with a microinjector. The silencing efficiency of dsSeCDA6, dsSeCDA7, dsSeCDA8, dsSeCDA9 and dsSeCDAV was detected 72 h after dsRNA injection via qRT-PCR. Three biological replications (each with three technical replicates) were used to determine the silencing efficiency. The remaining larvae were maintained for phenotypical observation. The visible phenotypical changes were recorded daily until the larvae molted into adults.

Immunohistochemistry
Fourth-instar larvae were injected with dsGFP or dsSeCDAV separately to determine the specificity of the SeCDAs antibody. At 72 h after injection, the larvae were collected to prepare paraffin sections. Immunohistology was conducted as follows: SeCDAs were detected with the Group V SeCDAs rabbit antiserum (1:200) as primary antibodies and HRP Rabbit IgG/FITC (1:200) (Servicebio, Wuhan, China) as the secondary antibody for fluorescence detection. Propidium iodide (PI) is an analog of ethidium bromide that binds strongly to DNA and emits red fluorescence upon intercalation in double-stranded DNA, which stains the nucleus red. PI (Servicebio, Wuhan, China) was used for nuclear staining. Images were captured using a transmission electron microscope (HT7800/HT7700, Nikon, Japan) with excitation at 465-495 nm (HRP Rabbit IgG/FITC) and 330-380 nm (PI).

Scanning Electron Microscopy
The midgut and the PM of S. exigua larvae injected with dsGFP and dsSeCDAV were rinsed gently with PBS (Servicebio, Wuhan, China). The midgut and PM were quickly fixed using an electron microscope fixative (Servicebio, Wuhan, China) at room temperature for 2 h and transferred to 4 • for storage. The midgut was washed, and PM was used three times for 15 min each in 0.1 M PBS. The tissues were placed in 1% OsO 4 for 2 h at room temperature (Ted Pella Inc., Redding, CA, USA) in 0.1 M PBS. The tissues were washed three times for 15 min each in 0.1 M PBS. The tissues were sequentially added to 30%, 50%, 70%, 80%, 90%, 95% and 100% alcohol and isoamyl acetate for 15 min each time. The samples were dried with a Critical Point Dryer (K850, Quorum, UK) and sputter-coated with gold for 30 s with Lon Sputtering Apparatus (MC1000, Nikon, Japan). The tissues