Antimicrobial peptides in Dendrobium officinale: Genomic parameters, peptide structures, and gene expression patterns

Abstract

A weak codon usage bias was found in Dendrobium catenatum (D. officiale) antimicrobial peptides (AMPs), after the analysis of relative synonymous codon usage, GC contents, and the effective number of codons. The codon usage preference was mainly influenced by natural selection pressure. The self‐optimized prediction method and SWISS‐MODEL were applied for peptide structural and domain analyses, and some typical antimicrobial domains were found in D. officinale AMP amino sequences, such as knot1 domain, gibberellins‐stimulated domain, cupin_1 domain, defensin_like domain, and SLR1‐BP (S locus‐related glycoprotein 1 binding pollen coat protein) domain. To investigate the AMPs gene expression pattern, abiotic stresses, such as salt stress, drought stress, salicylic acid (SA), and methyl jasmonate (JA), were applied and the gene expression levels were detected by the real‐time fluorescent quantitative polymerase chain reaction. Results showed that, even though the basic AMPs gene expressions were low, some AMPs can still be induced by salt dress, while the drought dress did not show the same impact. The SA and JA signaling pathways might be involved in most of the AMPs expressions. The natural selection of the D. officinale AMPs and thus forming diverse types of AMPs enhanced the plant's innate immunity and disease resistance capability, which would lead to a better understanding of the molecular mechanism for D. officinale adapting to the environment. The finding that salt stress, SA, and JA signaling pathways can induce AMP expression lays a foundation for the further development and functional verification of D. officinale AMPs.

Core Ideas

• Dendrobium officinale antimicrobial peptides (AMPs) genome had a weak codon usage bias and was mainly influenced by natural selection pressure.

• Diverse types of D. officinale AMPs enriched the innate immune system and disease resistance capability.

• AMPs can be induced by salt dress, and salicylic acid and jasmonic acid signaling pathways are involved.

• These results will help D. officinale AMPs engineering (biosynthesis) and functional verification.

Abbreviations

AMP

antimicrobial peptide

CDS

coding sequence

EMBOSS

European molecular biology open software suite

ENC

effective number of codons

GASA

gibberellins‐stimulated

GRAVY

grand average of hydropathicity

JA

jasmonate

PCR

polymerase chain reaction

PEG

polyethylene glycol

RSCU

relative synonymous codon usage

SA

salicylic acid

SLR1‐BP

S locus‐related glycoprotein 1 binding pollen coat protein

SOPMA

self‐optimized prediction method

1. INTRODUCTION

Antimicrobial peptides (AMPs) have raised an upsurge of research interest since their first report in the 1980s (Zasloff, 2002). In plants, AMPs are the most common and prominent defense barriers that help plants to deal with pathogenic assaults (Olga et al., 2020) and are part of the innate immune system with a broad spectrum of antimicrobial activity against the invasion of exogenous fungi, bacteria, and viruses. AMPs interact with the phospholipid in the cell membrane instead of a specific receptor, and resistance against AMPs in pathogens can hardly be induced. Therefore, AMPs were considered one of the most promising antibiotic candidates in plant genetic engineering and disease resistance breeding.

Dendrobium catenatum (D. officinale) is a precious traditional Chinese herb in the Orchidaceae family and is currently listed as endangered species. Dendrobium officinale has many medicinal functions, such as nourishing stomach conditions, boosting vital energy (or called nourishing Yin energy), heat‐clearing, anticancer, antioxidant, antifatigue, hepatoprotective, enhancing immunity, and others (Tang et al., 2017; Teixeira da Silva & Ng, 2017; Wang, 2021; Zhang et al., 2005). Reports have revealed that the extract from the Dendrobium genus plants has antibacterial and antifungal activities against Escherichia coli, Staphylococcus aureus, Bacillus subtilis, Klebsiella pneumonia, Vibrio cholera, etc. (Devi et al., 2009; Marasini & Joshi, 2012; Sandrasagaran et al., 2014). Transcriptome sequencing of D. officinale in our group showed that there were multiple genes annotated as antimicrobial peptides (Table S1). It implies that the plant's defense system could be activated and D. officinale AMPs were therefore produced against the pathogen invasions (Ali et al., 2018; Bolouri Moghaddam et al., 2016).

The antifungal activities of D. officinale AMPs are relatively investigated. For this study, we first focused on the genome codon usage, peptide structure, functional domains, and expression pattern of AMPs in D. officinale. Studies on the AMPs genome codon usage bias and its influence factors would lead to a better understanding of the molecular mechanism for D. officinale adapting to the environment (Kollmar & Mühlhausen, 2017). And for more fundamental information on the D. officinale AMPs, peptide properties, secondary structure, domain analysis, and tertiary structure modeling were then carried out. Finally, abiotic and biotic stresses were applied to investigate the expression patterns of D. officinale AMPs, which provided a deep insight into the correlation between the gene induction and regulation mechanism of AMP genes.

Core Ideas

• Dendrobium officinale antimicrobial peptides (AMPs) genome had a weak codon usage bias and was mainly influenced by natural selection pressure.

• Diverse types of D. officinale AMPs enriched the innate immune system and disease resistance capability.

• AMPs can be induced by salt dress, and salicylic acid and jasmonic acid signaling pathways are involved.

• These results will help D. officinale AMPs engineering (biosynthesis) and functional verification.

2. MATERIALS AND METHODS

2.1. Source of D. officinale AMP genome

Ten completely sequenced and annotated Dendrobium catenatum (D. officinale, wild type from Guangnan, Yunan, China) antimicrobial peptide genes and the coding sequence (CDS) were retrieved and downloaded from Uniprot.org (https://www.uniprot.org/), respectively, and are listed in Table S2. The D. officinale AMPs were then renamed as DoAMP1–DoAMP10 (Table 1).

TABLE 1.

Primers of Dendrobium officinale antimicrobial peptide (AMP) genes for real‐time fluorescent quantitative polymerase chain reaction.

Peptide	Gene accession	Forward primer (5′−3′)	Reverse primer (5′−3′)
DoAMP1	A0A2I0WLB7	TCATTGACTGCTCCTTCGTT	ATCACCACATGCTCCACCTT
DoAMP2	A0A2I0VR83	TCTTTTAGGGCTGCTTCTGG	CTCTGTAGCAAGGGCATTCA
DoAMP3	A0A2I0VUD1	TCCTCGCTTCTGGGATTG	TGGAAGCCCTCAGTCCTACA
DoAMP4	A0A2I0WLU1	GAAGGATTTCGGATCATACC	CCAGCTAATAGTTCCTCTGC
DoAMP5	A0A2I0VSB5	TTCTGATCTCCCTTCTCG	GCAGCCTATTTCTAATCTTG
DoAMP6	A0A2I0VT35	TTCTTTTCGGAAACATTGT	CATCAGCAGGACTCCAAAA
DoAMP7	A0A2I0VNA0	ATCGAGGTCAACGAACACG	AGGATAAACCACTTCCACCG
DoAMP8	A0A2I0VEZ2	TACAAACTTGGTGCTCAATGAGTC	GCATGGCTACTTTCATTTGGCTT
DoAMP9	A0A2I0VFV6	CACAAAAGAGATGCTTAATGATTC	ACAGTCGTAAACACAGAGA
DoAMP10	A0A2I0VT45	CCTATGGTAAAGGCAGACAG	TAGCAATCGTATGTGCAAGC
β‐Actin	XM_020842888.2	AGGAAGGCGGCTTTGAATC	CCATGCCAACCATGACACC

2.2. Base composition and codon bias indices

GC contents in each AMP gene were obtained using the European molecular biology open software suite (EMBOSS) explorer, “cusp” program (https://www.bioinformatics.nl/emboss‐explorer/). It calculates the contents of GC bases at position 1 (GC1), position 2 (GC2), position 3 (GC3), the average GC content (GC), and the average value of GC1 and GC2 (GC12). The effective number of codons (ENC) was calculated by the “chips” program in EMBOSS Explorer.

The relative synonymous codon usage (RSCU) was obtained using the CodonW 1.4.2 software. The Hemi 1.0.3.7 software was used to generate an RSCU heat map and hierarchical cluster analysis.

The neutrality analysis was used to estimate the varying role of codon usage preference between mutational pressure and natural selection. In the neutrality plot, the GC3 value was plotted against its GC12 value. A high correlation (close to 1) between GC12 and GC3 indicates that the codon usage preference is mainly affected by mutation pressure. In contrast, it was mainly affected by natural selection.

A correlation plot between ENC and GC3 was plotted using the ENC value versus the GC3 value. If the codon usage of one gene is completely random and unbiased, then the expected ENC curve follows the equation (Wright, 1990): $\mathrm{ENC}=2+\mathrm{GC}3+\frac{29}{\mathrm{GC}{3}^2+{(1-\mathrm{GC}3)}^2}$. Dots that fell on the curve are regarded as unbiased, whereas dots that are located away from the curve are considered high codon usage biased.

The Parity Rule 2 bias was presented by plotting the value of AT‐bias [A/(A+T)] as the ordinate and GC‐bias [G/(G+C)] as the abscissa, where A represented the adenine content, T represented the thymine content, G represented guanine content, and C represented cytosine content. In this plot, the center of the plot is the place where A  = T and G  =  C, values that are located in this central area indicated no codon usage bias.

2.3. Peptide properties, secondary structure, domains analysis, and tertiary structure modeling

Peptide properties, such as amino acids, molecular weight, isoelectric point (pI), instability index, grand average of hydropathicity (GRAVY), and a lipophilic index, were calculated using ProtParam (https://web.expasy.org/protparam). Subcellular localization prediction was performed by WoLFPSORT (https://wolfpsort.hgc.jp/).

Sequence analysis of secondary structures and domain predictions were performed by self‐optimized prediction method (SOPMA) (https://npsa‐prabi.ibcp.fr/cgi‐bin/npsa_automat.pl?page = npsa_sopma.html) and simple modular architecture research tool (SMART) (http://smart.embl‐heidelberg.de/), respectively. SignalP‐5.0 (https://services.healthtech.dtu.dk/service.php?SignalP‐5.0) was used for signal peptide prediction and analysis. SWISS‐MODEL (https://swissmodel.expasy.org/) was applied to perform the homology modeling for the tertiary structure of D. officinale AMPs.

2.4. Plant materials, abiotic, and biotic stresses processing

Dendrobium officinale (wild type from Guanzhai, Fujian, China) was grown in a greenhouse at the Institute of Subtropical Agriculture, Fujian Academy of Agricultural Sciences (Zhangzhou, Fujian, China). Dendrobium officinale seedlings were selected for the abiotic and biotic stress experiments and were put in an MGC‐450HP growth chamber (Bluepard). The culture condition was set as 13,000 lux light intensity, 16 h/8 h day/night light cycle, and 25°C with 75% humidity. Abiotic stresses were applied as 100 mmol·L⁻¹ sodium chloride (NaCl), 10% polyethylene glycol (PEG), 50 mg·L⁻¹ salicylic acid (SA), and 0.2 mmol·L⁻¹ methyl jasmonate (JA), respectively. Normal growth D. officinale seedlings were used as a control. After 24 and 48 h of soaking in the above‐mentioned abiotic stress solutions, seedlings were instantly collected and put into liquid nitrogen and then transferred to a freezer at −80°C for further gene expression analysis. A fungal pathogen of Colletotrichum gloeosporioides, which is the cause of anthracnose disease in D. officinale (Lan et al., 2016), was isolated and used as the biotic stress. Colletotrichum gloeosporioides were grown on potato dextrose agar plates for 7 days at 28°C. Spores were collected by spreading sterile distilled water (with 0.01% (v/v) tween‐20) onto the surface of the plate and filtered through two layers of sterilized cheesecloth. The spore suspension was then diluted into 1 × 10⁶ spores·mL⁻¹ and sprayed on the plant leaves for infection.

2.5. Real‐time fluorescent quantitative PCR (RT‐qPCR)

The total RNA from the roots, stems, and leaves of D. officinale was extracted using a SteadyPure Plant RNA Extraction Kit (Accurate, China. No. AG21019), and the cDNA template was reverse transcribed by an Evo M‐MLV RT Mix Kit with gDNA clean for qPCR (Accurate, China. No. AG11728). RT‐qPCR primers were designed by Primer Premier 5.00 software (Table 1) and were synthesized by Shanghai Biotechnology Co., Ltd.

The RT‐qPCR was performed using the SYBRGreen Premix Pro Taq HS qPCR Kit (Accurate, China. No. AG11701) in a Roche lightcycle96 real‐time PCR system (Roche), with three technical replicates. The qPCR volume was 20 μL in total, including 3 μL template cDNA, 1 μL forward primer and 1 μL reverse primer, 10 μL SYBRGreen Pro Taq HS Premix, and 5 μL RNase‐free water. A housekeeper gene, D. officinale β‐actin, was used as the internal reference (Wang et al., 2018; Yang et al., 2020), which provided a basis for the relative quantification of the AMP genes.

The amplification reaction cycle at which the fluorescence of a sample rises above the background fluorescence (positive call) is called the quantification cycle (Cq) of the sample, also well known as the threshold cycle (Ct). In this study, the fluorescence threshold value was set as 0.2. The relative quantification of each AMP gene was calculated by the ratio of the Ct value of the target gene to the Ct value of the β‐actin gene in each sample.

2.6. Statistical analysis

The correlation analysis was carried out by using the statistical software SPSS (version 25.0) with the Spearman's rank method.

3. RESULTS AND DISCUSSION

3.1. Relative synonymous codon usage

A total of 2189 codons (Table 2) were collected and used for relative synonymous codon usage (RSCU) value calculation. A heat map of the RSCU value for each synonymous codon in the genes is shown in Figure 1 and was clustered into two groups.

TABLE 2.

Relative synonymous codon usage (RSCU) values in Dendrobium officinale antimicrobial peptide (AMP) genes.

Amino acid	Codon	Number	RSCU	Amino acid	Codon	Number	RSCU
Phe	UUU	45	0.85	Ser	UCU	24	1
	UUC	61	1.15		UCC	21	0.88
Leu	UUA	23	0.82		UCA	22	0.92
	UUG	22	0.79		UCG	15	0.63
	CUU	39	1.39	Pro	CCU	32	1.28
	CUC	43	1.54		CCC	14	0.56
	CUA	11	0.39		CCA	32	1.28
	CUG	30	1.07		CCG	22	0.88
Ile	AUU	44	1.32	Thr	ACU	23	1.05
	AUC	35	1.05		ACC	14	0.64
	AUA	21	0.63		ACA	34	1.55
Met	AUG	64	1		ACG	17	0.77
Val	GUU	43	1.14	Ala	GCU	33	1.04
	GUC	22	0.58		GCC	30	0.94
	GUA	25	0.66		GCA	38	1.2
	GUG	61	1.62		GCG	26	0.82
Tyr	UAU	20	0.87	Cys	UGU	32	0.73
	UAC	26	1.13		UGC	56	1.27
TER	UAA	2	0.6	TER	UGA	8	2.4
	UAG	0	0	Trp	UGG	22	1
His	CAU	35	1.43	Arg	CGU	6	0.24
	CAC	14	0.57		CGC	13	0.52
Gln	CAA	36	1.14		CGA	15	0.6
	CAG	27	0.86		CGG	14	0.56
Asn	AAU	46	1.14	Ser	AGU	22	0.92
	AAC	35	0.86		AGC	40	1.67
Lys	AAA	55	0.67	Arg	AGA	49	1.97
	AAG	108	1.33		AGG	52	2.09
Asp	GAU	59	1.33	Gly	GGU	29	0.68
	GAC	30	0.67		GGC	41	0.96
Glu	GAA	108	1.03		GGA	61	1.43
	GAG	102	0.97		GGG	40	0.94

Note: Codons with preferred usage (RSCU values greater than 1) are given in italics.

FIGURE 1.

Heatmap of the relative synonymous codon usage (RSCU) for Dendrobium officinale antimicrobial peptide (AMP) genes.

The RSCU value of a codon is the ratio of the observed codon usage to the expected usage. It was performed to investigate the overall synonymous codon usage bias of the target genes. Generally, RSCU = 1 means no codon usage preference and an RSCU value less than 1 represents the codon is underused. When an RSCU value is greater than 1, it indicates that this codon usage is biased, and the RSCU value of a highly biased codon is greater than 1.5.

In this study, among the 64 synonymous codons within the 10 AMP genes, 53% of codons were underused than expected, and 42% of codons have preferred usage within which there were only 10.9% of codons were highly biased (RSCU > 1.5). The highest RSCU is AGG (arginine, Arg) with a value of 2.09, and codon UAG was not used (RSCU = 0). Generally, the overall codon usage preference is weak. Generally, species or genes have certain preferences on the usage of one or several synonymous codons to code their amino acids, which is so‐called the codon usage bias (Hershberg & Petrov, 2008). The stronger the codon bias is, the higher is the gene expression level in an organism (Quax et al., 2015; Zhou et al., 2016).

Our results revealed that the overall codon usage preference of the D. officinale AMP genes had a weak bias. The hierarchical cluster of RSCU showed that DoAMP1, DoAMP2, DoAMP3, and DoAMP7 genes had similar codon usage biases, while DoAMP4, DoAMP6, DoAMP5, DoAMP8, DoAMP9, and DoAMP10 had similar codon preference (Figure 1). Obviously, genes in the same cluster suggested a similar codon usage pattern and a similar influence derived from mutation pressure and natural selection.

3.2. Base composition and effective number of codons

GC contents for the first, second, and third codon positions and the entire gene (GC1, GC2, GC3, and GC) of the 10 AMP genes are shown in Table 3, with average values of 0.46, 0.45, 0.48, and 0.47 respectively. In the present study, D. officinale AMP genes’ GC3 values varied from 0.27 to 0.66, with an average of 0.48. Regarding that, there were 10 genes only, the GC3 value distribution is relatively scattered.

TABLE 3.

Summary of GC contents and effective number of codons (ENC) of Dendrobium officinale antimicrobial peptide (AMP) genomes.

AMP	GC1	GC2	GC3	GC	GC12	ENC
DoAMP1	0.56	0.46	0.55	0.52	0.51	57.24
DoAMP2	0.51	0.47	0.66	0.55	0.49	48.05
DoAMP3	0.51	0.52	0.66	0.56	0.51	48.14
DoAMP4	0.51	0.33	0.43	0.43	0.42	50.26
DoAMP5	0.42	0.47	0.38	0.43	0.45	44.07
DoAMP6	0.48	0.38	0.41	0.42	0.43	49.23
DoAMP7	0.57	0.41	0.63	0.54	0.49	54.95
DoAMP8	0.39	0.56	0.32	0.42	0.47	37.68
DoAMP9	0.36	0.43	0.32	0.37	0.39	43.37
DoAMP10	0.35	0.46	0.47	0.43	0.40	35.93
Average	0.47	0.45	0.48	0.47	0.46	46.89

Kawabe and Miyashita (2003) suggested that when the GC3 values are widely distributed, which means the amino acids can be coded by more than one synonymous codon, and it indicates that the codon usage bias is greatly affected by mutation pressure. If the GC3 value varied in a narrow range, it indicates that codon usage bias is greatly affected by selection pressure. Moreover, a strong selection pressure could cause no correlation between GC12 and GC3. In the present study, as shown in Table 4 the correlation analysis of GC contents and ENC. GC12 was positively correlated with GC and GC3, the correlations were 0.885 (p < 0.01) and 0.703 (p < 0.05) (also shown in Figure 2a), respectively. ENC was positively correlated with GC1 and the correlation was as high as 0.925 (p < 0.01). Hence, the GC content and ENC value study showed that the codon usage bias of D. officinale AMP genes was mainly affected by mutation pressure.

TABLE 4.

Correlation analysis of GC contents and effective number of codons (ENC) in Dendrobium officinale antimicrobial peptide (AMP) genomes.

Item	GC	GC1	GC2	GC3	GC12	ENC
GC1	0.766**
GC2	0.230	−0.308
GC3	0.953**	0.717*	0.049
GC12	0.885**	0.698*	0.466	0.703*
ENC	0.581	0.925**	−0.442	0.545	0.528
Gene length	0.221	0.721*	−0.667*	0.227	0.169	0.734*

^* p < 0.05

^** p < 0.01.

FIGURE 2.

The correlation between codon bias indices for Dendrobium officinale antimicrobial peptide (AMP) genes: (a) neutrality plot (GC12 against GC3); (b) relation between GC3 and ENCs (ENC plot); (c) Parity Rule 2 (PR2) bias plot [A3/(A3 + T3) against G3/(G3 + C3)].

The ENC plot shows the correlation between ENC and GC3 (Figure 2b). When a dot locates on the expected ENC curve, it represents the codon usage is completely random and unbiased (Wright, 1990), which indicates it is greatly affected by mutation pressure during the evolution process. In contrast, if the gene codon bias is mainly affected by selection pressure from the environment, its location will be far away from the expected ENC curve. In this study, 10 AMP genes are all located below the expected curve. There were two genes, DoAMP1 and DoAMP7, located close to the curve, suggesting that their codon usage preferences were affected by mutation pressure. The other eight genes were tilting away from the curve, indicating their codon usage preferences were mainly affected by the selection pressure.

Generally, the ENC value varies from 20 to 61, it describes the preference degree of the synonymous codon usage deviated from completely random selection. When an ENC value is close to 20, it was regarded as highly biased. In contrast, it indicates weak usage bias (Gun et al., 2018; Lal et al., 2016). In this study, the ENC values of the 10 D. officinale AMP genes ranged from 35.93 to 57.17 with an average value of 40.19, indicating that the gene's codon usage bias is weak.

Synonymous codons have a degeneration phenomenon, mutations in the third position of genes can be easily tolerated and this contributes to a stable protein amino sequence. It is generally believed that when the G and C (or A and T) bases are proportionately distributed at the third position, the codon usage preference is strongly affected by mutation pressure only. If the G and C (or A and T) bases are not proportionately distributed at the third position, it indicates that the codon usage preference is affected mainly by selection pressure. (Kawabe & Miyashita, 2003) In this study, the distribution of G and C (or A and T) at the third position in the synonymous codons is not proportional, as shown in Figure 2c, indicating that codon usage preference is mainly affected by selection pressure.

3.3. Physicochemical properties and subcellular localization

The physicochemical properties of peptides play a critical role in their biological functions and provide important hints for distinguishing them from non‐AMPs (Kang et al., 2012; Torrent et al., 2011). And the subcellular localization of these AMPs is highly associated with their antimicrobial functions and biological processes (Weerawanich et al., 2018). Therefore, the basic information such as amino acids, molecular weight, pI (isoelectric point), instability index, GRAVY, and aliphatic index calculated by ProtParam and the subcellular localization predicted by WoLFPSORT are shown in Table 5.

TABLE 5.

Physicochemical properties for Dendrobium officinale antimicrobial peptides (AMPs).

AMP	Amino acids	Molecular weight	pI	Instability index	GRAVY	Aliphatic index	Subcellular localization
DoAMP1	434	47417.48	8.92	56.06	−0.241	83.94	Extracellular
DoAMP2	97	10917.97	9.05	49.44	−0.276	66.39	Extracellular
DoAMP3	92	10065.81	9.06	42.99	−0.061	66.85	Extracellular
DoAMP4	629	73420.63	5.65	61.68	−1.037	61.80	Endomembrane
DoAMP5	75	8338.66	7.55	31.92	−0.173	68.93	Extracellular
DoAMP6	201	23343.09	6.80	54.66	−0.113	76.12	Nucleus
DoAMP7	484	53617.56	5.86	42.20	−0.141	86.80	Extracellular
DoAMP8	56	6033.06	7.51	61.31	0.282	66.25	Extracellular
DoAMP9	55	6279.55	9.28	88.67	−0.391	70.91	Extracellular
DoAMP10	56	6205.20	8.13	88.41	−0.166	54.11	Extracellular

Abbreviations: GRAVY, grand average of hydropathicity; pI, isoelectric point.

The number of amino acids ranged from 55 aa (DoAMP9) to 629 aa (DoAMP4) and molecular weights ranged from 6033.06 Da (DoAMP8) to 73420.63 Da (DoAMP4). As shown in Table 5, the pI of these AMPs varied from 5.65 (DoAMP4) to 9.28 (DoAMP9). The GRAVY is negative, except DoAMP8 (0.282), suggesting that the D. officinale AMPs are more hydrophobic.

Generally, the hydrophobicity property of an AMP affects its antimicrobial activity and selectivity (Edwards et al., 2016). It was previously reported that increasing the hydrophobicity on the positive charge side of an AMP can improve its antimicrobial activity, while reducing its hydrophobicity can reduce the antimicrobial activity (Bahar & Ren, 2013). Moreover, some studies have shown that the range of an AMP's targets can be changed by increasing the hydrophobicity of an AMP within an optimal window (Dathe et al., 1997). For instance, some AMPs, such as magainin and pro‐hinged peptides, were only effective against Gram‐negative bacteria, after artificial increasing their hydrophobicity, they were found to have anti‐Gram‐positive activity and can also kill eukaryotic cells (Dathe et al., 1997; Hyun et al., 2020). The present AMPs have both hydrophilicity and hydrophobicity, indicating their wide spectrum of antimicrobial activities.

Eight out of the 10 AMPs were predicted extracellular location while DoAMP4 was predicted in the endomembrane and DoAMP6 was in the nucleus (Table 5). The results indicated that most of the D. officinale AMPs were secreted outside the plant cell, and we conjectured that these AMPs may be involved in the extracellular defense barrier against exogenous antigens, while the intracellular AMPs (DoAMP4 and DoAMP6) mainly participated in the intracellular immune process.

3.4. Secondary structure analysis

To get a better understanding of the peptide secondary structure and to lay a foundation for the following analysis, the prediction of the domains and spatial structure of these AMPs were performed by SOPMA. The secondary structures analysis (Table 6) indicates that there are four types of secondary structures composing these peptides. They are α‐helix, extended strand, β‐turn, and random coil. Generally, AMPs with α‐helices and β‐turns tend to be more stable and with high cytotoxicity (Ruiz et al., 2014).

TABLE 6.

Components of secondary structure in Dendrobium officinale antimicrobial peptides (AMPs) peptides sequences.

AMP	α‐Helix (%)	Extended strand (%)	β‐turn (%)	Random coil (%)
DoAMP1	27.88	17.74	10.83	43.55
DoAMP2	27.84	8.25	2.06	61.86
DoAMP3	36.96	10.87	5.43	46.74
DoAMP4	44.04	20.03	6.68	29.25
DoAMP5	26.67	16.00	4.00	53.33
DoAMP6	48.26	11.44	5.47	34.83
DoAMP7	23.55	27.48	7.85	41.12
DoAMP8	5.36	25.00	3.57	66.07
DoAMP9	16.36	20.00	5.45	58.18
DoAMP10	19.64	26.79	8.93	44.64

3.5. Domain prediction and analysis

To have a more detailed study on structure analysis, a modular architecture tool, simple modular architecture research tool (SMART), was used to perform domain analysis and prediction of the 10 AMPs. As shown in Figure 3, the functional domains were signal peptide, transmembrane regions, kont_1, gibberellins‐stimulated (GASA), cupin_1, defensin_like, and SLR1‐BP, etc. The knot_1 domain is usually characterized by cystine‐knot, which is a small antimicrobial protein rich in cysteine (Emamifar et al., 2021; Postic et al., 2018). The GASA domain, also known as Snakin, is an important biomolecule in plant innate immunity (Oliveira‐Lima et al., 2017). The GASA protein family contains a conserved 12 cysteine residues, which have significant antimicrobial activity (Filiz & Kurt, 2020; Li et al., 2021). The cupin_1 domain represents a conserved β‐barrel fold of barrel structure, which is highly homology with vicilin (Dunwell, 1998). It was reported that vicilin‐derived peptides have antifungal activity activities (Vieira Bard et al., 2014; Xie et al., 2016). Defensin_like domain represents a cysteine‐rich protein, which shows insecticidal and antifungal activity (De Coninck et al., 2013; Terras et al., 1995). SLR1‐BP domain is a cysteine‐rich protein and is stabilized by two to six disulfide bridges (Takayama et al., 2000). SLR1‐BP was reported to interact with S locus‐related glycoproteins and pathogens, and was recognized as part of plant primitive “immune system” (Brugière et al., 2000).

FIGURE 3.

Domain‐analysis of Dendrobium officinale antimicrobial peptide (AMP) sequences by simple modular architecture research tool (SMART). The knot_1 domain was knottin characterized by a cystine‐knot; the gibberellins‐stimulated (GASA) domain represented the gibberellin regulated protein family, which was a conserved cysteine‐rich domain; cupin_1 represented a conserved barrel domain (“cupa” is the Latin term for a small barrel); SLR1‐BP domain was S locus‐related glyco‐protein 1 binding pollen coat protein also cysteine‐rich.

Consequently, the complexity of the structural conformation influences the mechanisms of AMPs’ action, including interactions with cell membranes and their specificity of targets. Based on annotations in the SMART database, D. officinale AMPs possess a range of antimicrobial activities including antibacterial, antifungal, and antiviral activity (De Coninck et al., 2013; Dunwell, 1998; Emamifar et al., 2021; Filiz & Kurt, 2020; Li et al., 2021; Oliveira‐Lima et al., 2017; Postic et al., 2018; Takayama et al., 2000; Terras et al., 1995; Vieira Bard et al., 2014; Xie et al., 2016).

In the present study, at least one or two above‐mentioned functional domains were found in all the D. officinale AMPs through domain analysis. The knot_1 domain was shown in DoAMP1 and DoAMP3. DoAMP2 contains the GASA domain. The cupin_1 was detected in DoAMP4 and DoAMP7. The defensin_like domain was in DoAMP5. The most common domain is the SLR1‐BP domain and was seen in DoAMP6, DoAMP8, DoAMP9, and DoAMP10.

A short segment of secretory signal peptide (Sec/SPI) (Garcion et al., 2021) was found in DoAMP1, DoAMP2, DoAMP3, DoAMP5, and DoAMP7 (Supplemental Table S3). These AMP peptides were first synthesized intracellular and then secreted and cleaved afterward under the direction of signal peptides (Auclair et al., 2012; Walker & Lively, 2013). There were two AMPs structurally without a signal peptide, DoAMP4 and DoAMP6; they were predicted to be located in the endomembrane and nucleus (Table 5). We speculated they may be involved in the intracellular immune process (Owji et al., 2018).

Previous studies revealed that plant defensins consisted of one α‐helix and a triple‐stranded antiparallel β‐sheet. This stable structural motif was formed through the disulfide bridges between cysteines (Kobayashi et al., 1991), whereas thionins contain two antiparallel α‐helices and one antiparallel double‐stranded β‐sheet, composing an L shape structure, with the two antiparallel α‐helices as the long arm and the double‐stranded β‐sheet as the short arm (Broekaert et al., 2010). Typically, in a vicilin‐like AMP, a unique structural motif is a conserved barrel where the N‐terminus and the C‐terminus form cupin folds. Reports have shown that vicilin‐like AMPs exhibit activities against bacteria and fungi (Vieira Bard et al., 2014).

Based on our findings, we conjectured that DoAMP1, DoAMP3, DoAMP5, DoAMP6, and DoAMP9 were plant defensins, while DoAMP2, DoAMP8, and DoAMP10 were thionins. DoAMP4 and DoAMP7 were vicilin‐like AMPs.

3.6. Homology modeling

To determine the tertiary structure of D. officinale AMPs, the SWISS‐MODEL homology modeling method was applied in this study (Waterhouse et al., 2018). As shown in Figure 4, d . officinale AMPs were modeled with at least one α‐helix structure, one β‐sheet structure, or some coiled‐coil structure. It should be noted that the α‐helix structure is important for an AMP, which was so‐called pore‐forming activity. α‐Helices can interact with the membrane and form a pore structure (Rahnamaeian, 2011), which damages the permeability of the cell membrane and kills cells. Generally, the homology modeling results are consistent with the secondary structure analysis shown in Table 6.

FIGURE 4.

Homology modeling of the tertiary structure for Dendrobium officinale antimicrobial peptides (AMPs) by SWISS‐MODEL. Briefly, the target antimicrobial peptide (AMP) amino acid sequences were input manually and evaluated under the automated mode. The suitable modeling templates of each AMPs were automatically identified by SWISS‐MODEL based on BLAST (Camacho et al., 2009) and HHblits (Steinegger et al., 2019). Top‐ranked templates and alignments were then selected and the models were built accordingly.

3.7. Gene expression pattern

In terms of gene expression pattern, which is devoted to determining where the AMP genes are expressed and analyzing the induce mechanism of these AMP, the basic expression levels of the 10 AMP genes in root, stem, and leaf were examined. As shown in Figure 5, the relative expression levels are low in all parts of the plant. The highest level is merely up to 0.04 (relative to β‐Actin), which indicates that AMPs are low expressed in healthy D. officinale. Moreover, some AMPs can be considered to have no expression, such as DoAMP4, DoAMP6, and DoAMP8, in particular. In general, the basic expression levels of the AMPs genes were highest in leaves, and were lowest in roots.

FIGURE 5.

Relative expression levels of Dendrobium officinale antimicrobial peptide (AMP) genes to β‐actin in root, stem, and leaf. The values of relative expression levels were ratios of (𝐶𝑡_{𝑡𝑎𝑟𝑔𝑒𝑡}/𝐶𝑡𝛽_{−𝐴𝑐𝑡𝑖𝑛}).

It is a complex systemic mechanism between pathogen attack and immune defense in plants. Studies on the response of AMPs to abiotic and biotic stresses are helpful strategies to understand the response mechanism or the related metabolic pathways in plants (Castro & Fontes, 2005). To mimic environmental stresses, NaCl was used to simulate salt stress, while PEG was used to simulate drought stress. We also studied two molecules with biological significance. SA is a signal molecule, that participates in the regulation of various physiological functions in plants and plays an important regulatory role in improving the resistance of plants to abiotic stress (An & Mou, 2011). JA is a hormone and signal molecule related to plant damage, which mediates plant resistance to insects and pathogens (Santino et al., 2013; Wang et al., 2020). Since most of the presented DoAMPs were annotated antifungal activity, the C. gloeosporioides fungus, the cause of anthracnose disease, was used as the biotic stress (Lan et al., 2016).

In this study, 100 mmol·L⁻¹ NaCl, 10% PEG, 50 mg·L⁻¹ SA, 0.2 mmol·L⁻¹ JA, and 1 × 10⁶ spores·mL⁻¹ C. gloeosporioides were applied to induce AMP genes expression to investigate the expression pattern under abiotic stress. The gene expression levels in D. officinale leaves are most abundant among all plant compartments (Figure 5), and therefore leaves were selected to examine the gene expression after treatment. As shown in Figure 6, the induced expression level varied among different treatments. Roughly, the genes’ expression levels in group A were all lower than 25‐fold relative to control, and that of group B were higher than 100‐fold and reached up to 650‐fold. As for the biotic stress in group C, most AMPs had similar expressions with group A AMPs.

FIGURE 6.

Relative expressions of Dendrobium officinale antimicrobial peptide (AMP) genes under (a and b) abiotic stresses, and (c) biotic stress. The relative quantifications were done relative to the expression of the housekeeping gene β‐actin (𝐶𝑡_{𝑡𝑎𝑟𝑔𝑒𝑡}⁄𝐶𝑡_{𝛽−𝐴𝑐𝑡𝑖𝑛}). To normalize the varied expression among different controls, gene expressions under sodium chloride (NaCl), polyethylene glycol (PEG), salicylic acid (SA), jasmonate (JA), and C. gloeosporioides treatments were normalized relative to their controls, respectively.

As shown in Figure 6, group A has six low‐expression genes, which were DoAMP1, DoAMP3, DoAMP4, DoAMP7, DoAMP8, and DoAMP9. Under the salt (NaCl) stress, genes DoAMP3 and DoAMP4 expressed sevenfold higher than control after 24 h of treatment and declined back to normal in 48 h. Gene DoAMP7 had 4.7‐fold of expression higher than control after 48 h. The other three genes (DoAMP1, DoAMP8, and DoAMP9) were not sensitive to salt stress. Under drought stress PEG, only gene DoAMP1 expressed sixfold than that of control in 24 h and fell back to normal in 48 h. The other five genes were not sensitive to PEG. Five genes in group A had higher expression after 48 h of SA treatment with folds varying from 2 to 13, except for DoAMP8 whose expression was relatively depressed. In the meantime, four genes were induced expression after 48 h of JA treatment the folds were up to 20.8, except for genes DoAMP3 and DoAMP4 whose expression was not influenced. Generally, the six genes in group A have low sensitive responses to NaCl, SA, and JA, and no response to PEG treatment.

There were four genes in high‐expression group B, including DoAMP2, DoAMP5, DoAMP6, and DoAMP10. Under the salt (NaCl) stress, gene DoAMP2 expression was 390‐fold higher than control after 24 h of NaCl treatment and reached up to 416‐fold in 48 h. DoAMP6 was highly expressed after 48 h of NaCl treatment (99‐fold). Under drought PEG stress, genes DoAMP5, DoAMP6, and DoAMP10 were not sensitive, while DoAMP2 expressed 184‐fold in 24 h and 54‐fold in 48 h. For the SA treatment, all four genes had the highest expression after 48 h treatment, with 16‐, 22‐, 67‐, and 151‐fold, respectively. The same results were found in JA treatment; we got the highest expression of these four genes after 48 h, with 378‐, 259‐, 115‐, and 636‐fold, respectively.

In terms of biotic stress, D. officinale AMP genes were all low expressed after 24 and 48 h of C. gloeosporioides treatment, as shown in Figure 6C. Two sensitive AMP genes were found higher expressions after C. gloeosporioides treatment, DoAMP2 and DoAMP10, which were also found in high‐expression group B. It is believed that AMPs can be produced immediately upon infection while some are constitutively expressed (Broekaert et al., 2010; Rahnamaeian, 2011). In this study, eight out of these 10 D. officinale AMP genes were not expressed, some genes were even depressed, including DoAMP4, DoAMP7, DoAMP8, and DoAMP9, which were also found in low‐expression group A.

The result revealed that most of the D. officinale AMP genes had very low responses to salt (NaCl) and drought (PEG) stresses, except for DoAMP2 to NaCl and PEG, DoAMP6 to NaCl, and DoAMP8 to PEG. Applying SA and JA treatment to D. officinale can induce most of the D. officinale AMP genes’ expression, indicating that the SA and JA signaling pathways may be involved in the D. officinale AMP genes expression pattern. This result will contribute to lateral research and a better understanding of the express mechanism of AMPs in D. officinale.

4. CONCLUSIONS

In this study, we had an overview of the D. officinale AMPs genome CDSs, and a weak codon usage bias was observed. During the long period of evolution, the D. officinale AMPs genome was influenced by both natural selection and mutation pressure, while the selection pressure against mutation pressure is much stronger.

Structural analysis and location prediction revealed that D. officinale AMPs had various types of antimicrobial domains (knot1, GASA, cupin_1, defensin_like, and SLR1‐BP) and distributed inter and extracellular. Diverse types of AMPs enriched the defensive system and improved the nonspecific immune and disease resistance capability in the growth process of D. officinale.

All these AMPs genes had low basic expressions in all parts of the plant. Genes DoAMP1, DoAMP3, DoAMP4, DoAMP7, DoAMP8, and DoAMP9 can be induced under salt stress, drought stress, SA, and JA treatment, while the expression level remained low. The expression levels of DoAMP2, DoAMP5, DoAMP6, and DoAMP10 have raised hundreds of folds under all the above abiotic stresses. It could imply that DoAMP2 (also sensitive to drought stress) and DoAMP6 were salt stress‐sensitive AMPs, while SA treatment‐sensitive AMPs were DoAMP6 and DoAMP10. Moreover, all AMPs were sensitive to JA treatment, which means the SA and JA signaling pathways are involved in the D. officinale AMPs gene expression. It is noteworthy that plant genes are expressed under certain conditions, while some may not be expressed at all. However, this does not hinder the value that we discover, study, and utilize the AMP genes. The investigation of expression patterns laid a solid foundation for the further development and functional verification of D. officinale AMPs.

Although chemicals like pesticides are efficient in plant pathogen controls, there are still environmental risks to be faced. Plant AMPs provide better disease management. It is much easier to get AMP from metabolite production in suspension culture than from plant material. This study gives hints on D. officinale AMPs engineering (biosynthesis) and their sequences and structures architecture.

AUTHOR CONTRIBUTIONS

Huiming Huang: Conceptualization; funding acquisition; investigation; methodology; software; writing–original draft; writing–review and editing. Wen‐Yee Lee: Writing–review and editing. Hui Zou: Resources. Haiming Li: Investigation; resources. Shuhe Zhang: Resources. Heping Li: Data curation. Jiangbo Lin: Conceptualization; funding acquisition; methodology; project administration; writing–original draft.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Supplemental Table S1. Genes that were identified and annotated as antimicrobial peptides in transcriptome sequencing of Dendrobium officinale in our group.

Supplemental Table S2. Coding sequence (CDS) of Dendrobium officinale AMP genes.

Supplemental Table S3. Signal peptide predictions for Dendrobium officinale AMPs sequences by SignalP‐5.0.

Huang, H. , Lee, W.‐Y. , Zou, H. , Li, H. , Zhang, S. , Li, H. , & Lin, J. (2023). Antimicrobial peptides in Dendrobium officinale: Genomic parameters, peptide structures, and gene expression patterns. The Plant Genome, 16, e20348. 10.1002/tpg2.20348

DATA AVAILABILITY STATEMENT

The data that supports the findings of this study are available in the Supporting Information of this article.