Overexpression of AhHDA1 results in retarded growth of hairy roots and altered cell metabolism in Arachis hypogaea

Author email addresses: Liangchen Su: sksky163@163.com, Shuai Liu: 408152101@qq.com, Xing Liu: 10150926@qq.com, Baihong Zhang: 1311369961@qq.com, Meijuan Li: 929722415@qq.com, Lidan Zeng: 2036428216@qq.com #These authors have contributed equally to this work. *Correspondence: Ling Li, Guangdong Provincial Key Laboratory of Biotechnology for Plant Development, School of Life Sciences, South China Normal University, No. 55, Zhongshan Avenue West, Tianhe District, Guangzhou 510631, China; liling502@126.com Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 September 2018 doi:10.20944/preprints201809.0486.v1


Introduction
During the various stages of development and in response to abiotic stresses, plant cells regulate the expression of many genes (Gan, et al., 2013;Lee, et al., 2015). Chromatin structure plays a key role in transcriptional regulation and is controlled by multiprotein complexes that recognize and instigate biochemical modifications of the histone proteins (Chen and Wu, 2015;Ganai, et al., 2016). Together, these multiprotein complexes, of which the core catalyzing enzymes are histone acetyltransferase proteins (HATs) and histone deacetylases (HDACs), maintain chromatin structure in a state of dynamic equilibrium (Mikkelsen et al., 2007). Thus, gene transcription can be activated by HATs through acetylation of lysine residues in the N-termini of histones H3 and H4, which switches the associated chromatin from a compacted to a looser state. In contrast, HDACs function as antagonists of HATs by removing acetyl groups from histone acetylation sites, resulting in transcriptional repression and gene silencing (Xing and Poirier, 2012;Ma et al., 2013). The fluctuating balance between HAT and HDAC activity comprises a large part of the epigenetic regulation of the genome; for example, recent evidence has revealed that regulation of histone acetylation plays a pivotal role in plant responses to salinity, drought and other stresses (Kim 4 et al., 2015).
HDACs have been isolated or identified in several plants and are well conserved across species, including Arabidopsis, rice, maize, grapevine and peanut (Arachis hypogaea) (Pandey et al., 2002;Ding et al., 2012;Yang et al., 2016;Aquea et al., 2010;Su et al.;. Phylogenetic analysis has classified HDACs into three distinct families, namely RPD3/HDA1-like HDACs, SIR2-like HDACs, and HD2 proteins, based on sequence similarity, substrate specificity, and cofactor requirement (Ma et al., 2013;Bourque et al., 2016). The majority of the HDACs in a given plant species are of the RPD3/HDA1-like category; for example, Arabidopsis thaliana has 12 RPD3/HDA1 subfamily genes among 18 putative HDAC family genes (Ma et al., 2013). It has been reported that HDACs function in plant growth, development, stress responses and gene silencing, and are also involved in other cellular processes such as cell death and the cell cycle. HDA6, HDA9 and HDA19 are the most-studied HDACs. Thus, HDA6 features in the responses to various plant hormones: it not only participates in jasmonic acid (JA)-and abscisic acid (ABA)-mediated plant defense responses, but is also involved in transgene silencing-related auxin responses; furthermore, recently the hda6 mutant was found to display a brassinosteroid (BR)-repressed phenotype in the dark and was less sensitive to BR biosynthesis inhibitors (Murfett et al., 2001;Devoto et al., 2002;Tanaka et al., 2008;Luo et al., 2012;Hao et al., 2016). HDA19 is involved in JA and ethylene signaling during the response to pathogens, and redundantly with HDA6 regulates the repression of embryonic properties during germination (Zhou et al., 2005;Tanaka et al., 2008). HDA9, which acts to oppose the effect of its homologs HDA6 and HDA19, is a negative regulator of germination in seedlings: hda9 mutant strains display 5 reduced seed dormancy and faster germination than wild-type plants (van Zanten et al., 2014).
Other HDACs have mostly been reported to have a role in plant growth and development. For instance, HDA5 interacts with FVE, FLD (FLOWERING LOCUS D) and HDA6, forming a protein complex involved in the regulation of flowering time ; both overexpression and silencing of HDA7 lead to a delay in growth at postgermination and later developmental stages, while downregulation of HDA7 decreases silique fertility in Arabidopsis (Cigliano et al., 2013); by regulating kinase genes, HDA18 is involved in positional information signaling during cellular patterning of Arabidopsis root epidermis .
From the above, it is clear that HDAs are important in plant growth and development, and are key players in signal transduction responses to plant hormones. Accordingly, the regulation of histone acetylation provides a new perspective in crop breeding research.
However, in many important crop plants, such as peanut, research on the epigenetic regulation of development and growth has rarely been reported and the mechanisms of these processes are not well understood. Others have reported that the peanut allergy gene Ara h 3 undergoes an increase in histone H3 acetylation and a decrease in histone H3K9 dimethylation in embryos at an early stage of maturation (Fu et al., 2010), while histone deacetylation modification participates in the repression of peanut seed storage protein gene Ara h 2.02 during germination . In our previous study, we isolated and characterized AhHDA1 from A. hypogaea, showing that AhHDA1 is very similar in structure to an Arabidopsis HDAC (AtHDA6) and, in recombinant form, possesses HDAC activity (Su et al., 2015). We also found that AhHDA1 is significantly upregulated by water deficit and 6 100 µM ABA. However, whether this HDAC mediates cell growth and development in peanut is not clear.
In this paper, we show that modifications in AhHDA1 expression, particularly overexpression, have a significant morphological effect on hairy root cells. Furthermore, tsranscriptome sequencing suggests that AhHDA1 regulates the biosynthesis of various carbon-mebolism-related biological molecules. These results provide insight into the role of AhHDA1 in peanut hairy roots, and lay the foundations for a wider understanding of histone deacetylase function in peanut, with potential impact on the breeding of new varieties.

Overexpression of AhHDA1 in transgenic hairy roots retards growth
35S::AhHDA1-eGFP (denoted 35S::AhHDA1 in the main text) and 35S::AhHDA1-RNAi transgenic hairy roots were obtained by transforming four-leaf peanut seedlings. We observed that 35S::AhHDA1-eGFP transgenic hairy roots became denser, with a dark coloration, after 20 d in vitro cultivation (Fig 1A and 1B). As the growth period progressed beyond 20 d, the 35S::AhHDA1-eGFP transgenic hairy roots grew more slowly and eventually stopped growing entirely. In contrast, both 35S::AhHDA1-RNAi and the control group maintained a relatively high growth rate, with the growth rate of AhHDA1-RNAi transgenic hairy roots being the highest (Fig 1C). Under the light microscope, the cells at the surface of 15 d-cultivated overexpressing transgenic hairy roots appeared thick and indistinct, with less root hair in the hairy zone than the control group and the RNA interference group. Treatment with 20% PEG seemed to exacerbate this phenotype further (Fig 2A). Although there was no significant difference in cell size among the different groups in the meristem and near elongation regions, the hairy root cells overexpressing AhHDA1 were significantly smaller than controls and the 35S::AhHDA1-RNAi group in the root hair and far elongation regions ( Fig 2B). Furthermore, the cell width-to-length ratios in 35S::AhHDA1 and 35S::AhHDA1-RNAi transgenic hairy roots were smaller than controls, presenting as a long, narrow cell form. Thus, altered expression of AhHDA1 had a clear effect on the morphology of cells near the root hair region ( Fig 2B and Fig 2C). The intracellular ROS concentration reflects the extent to which cells are experiencing environmental stress. We measured the relative concentration of superoxide anions and H2O2 in transgenic hairy roots using specific fluorescent probes. The superoxide anion concentration in 15 d overexpressed AhHDA1 hairy roots was 4.06 times that of the blank control, while in the 35S::AhHDA1-RNAi group it was 0.75 times that of controls. Similarly, the H2O2 concentration in overexpressed AhHDA1 hairy roots was 2.81 times blank control levels, while in the 35S::AhHDA1-RNAi group it was 0.77 times that of controls (Fig 3).
Taken together with their abnormal phenotype and morphology, the increase in ROS level of 35S::AhHDA1 transgenic hairy roots suggests they are experiencing cell damage during this early growth stage.

AhHDA1 overexpression in hairy roots influences expression of biomolecular synthesis and metabolism genes
To investigate how AhHDA1 affects cell morphology in hairy roots, we obtained transcriptome data for both 35S::AhHDA1 and 35S::AhHDA1-RNAi hairy roots (www.ncbi.nlm.nih.gov/Traces/study/?acc=SRP113568) and compared these separately to the control group to identify differentially expressed genes (DEGs). Using criteria of >2-fold difference in expression level and p<0.05, we found 2157 DEGs by comparing the RPKM in different groups, and these were first subjected to GO enrichment classification (Fig 4). Most 10 DEGs influenced by AhHDA1 could be grouped under the categories of cell part, membrane and intracellular in cellular component (166, 154 and 148 genes respectively), binding (824 genes) and catalytic activity (574 genes) in molecular function, metabolic process (637 genes), oxidation-reduction process (179 genes) in biological process. It is worth noting that in the category of biological process, relatively large numbers of genes could also be grouped under response to stimulus, protein modification process and transport, i.e. respectively, 144, 131 and 130 genes. To obtain a deeper understanding of the role of the DEGs identified, we examined the pathways in which their gene products are thought to operate. Unfortunately, most DEGs could not be found in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database because relatively few peanut genes and proteins have been annotated, but 468 out of the total 12 of 2157 DEGs could be assigned to a KEGG pathway (Fig 5). We found the starch and sucrose metabolism (ko00500), carbon metablism (ko01200) and photosynthesis pathways (ko00195) to be the most gene-enriched pathways with 41, 33 and 31 genes, respectively.
Other pathways, including phenylpropanoid biosynthesis (ko00940), oxidative phosphorylation (ko00190), plant hormone signal transduction (ko04075) and isoflavonoid biosynthesis (ko00943) pathways, were also indicated as modulated by AhHDA1 through its effect on the expression of downstream genes (Table S1). Of those annotated, 31 DEGs were identified as the most significantly differentially expressed, and 25 of these were found to be associated with biosynthesis and metabolism (Table 1). This result suggests that AhHDA1 influences several biosynthetic and metabolic pathways. In the starch and sucrose metabolism pathway, 12 beta-glucosidases and 5 trehalose-6-phosphate phosphatases were found to be differentially expressed: in 35S::AhHDA1 transgenic hairy roots, most of these genes were downregulated, while in the 35S::AhHDA1-RNAi group expression of these genes remained unchanged or was upregulated ( Fig S1). In the biomolecular synthesis pathways identified (which include phenylpropanoid biosynthesis, isoflavonoid and flavonoid biosynthesis, diterpenoid biosynthesis, and plant hormone biosynthesis, among others), 43 DEGs were upregulated and 78 DEGs were downregulated in 35S::AhHDA1 transgenic hairy roots ( Fig   S2). At the same time, six DEGs were downregulated and nine DEGs were upregulated in the 35S::AhHDA1 group in 19 DEGs, which were related with auxin (IAA), abscisic acid (ABA) and methyl jasmonate (MeJA) (Fig S3). In the photosynthesis pathway, 49 DEGs were downregulated in hairy roots overexpressing AhHDA1, with most DEGs remaining unchanged or being upregulated (Fig S4). These results imply that overexpression of AhHDA1 in peanut hairy roots directs a slowdown of the cell's primary metabolic activity, while promoting secondary metabolic activity.   Araip.S5EJ7 (isoflavone 2'-hydroxylase) in the isoflavonoid biosynthesis pathway. The transcriptional levels of these six DEGs in the 35S::AhHDA1 transgenic group were found to be significantly higher than in the control and 35S::AhHDA1-RNAi groups (Fig 6). This is consistent with the transcriptome data. Together, these results indicate that overexpression of AhHDA1 enhances the transcriptional level of phenylpropanoid, flavonoid and flavonoid biosynthesis genes, all of which are involved in secondary metabolism. The higher expression levels of these genes might be related to the slower growth rates in these transgenic plants.

Overexpression of AhHDA1 increases expression of phenylpropanoid secondary metabolic pathway genes
In earlier studies, peanut AhHDA1 was isolated and was found to be induced by ABA and PEG (Su et al., 2015). In the current study, we investigated the role of AhHDA1 in more detail and observed that overexpression of causes retarded growth during the early growth of hairy roots (Fig 1). Under the light microscope, significant changes in cell morphology were observed near the root hair region in these overexpressing plants, with the cell size being significantly smaller than in other groups. Moreover, the width-to-length ratios of hairy root cells in both the overexpressing and RNA interference groups were smaller than in controls, with cells near the root hair region being long and narrow in shape (Fig 2). Transcriptome sequencing was carried out to improve our understanding of how AhHDA1 might mediate this alteration in cell morphology. We found that pathways involved with various aspects of metabolism, biomolecular biosynthesis and photosynthesis were significantly influenced by AhHDA1 (Fig S1). Thus, 131 DEGs were downregulated and 58 DEGs were upregulated in various metabolic pathways, with the starch and sucrose metabolism, and carbon metabolism pathways containing the most DEGs (Fig S1). In particular, nine DEGs in the isoflavonoid biosynthesis pathway (five cytochrome P450 superfamily proteins and four O-methyltransferase family proteins) were upregulated and seven DEGs in the isoflavonoid biosynthesis pathway (six O-methyltransferase family proteins and one cytochrome P450 superfamily protein) were downregulated. In the phenylpropanoid biosynthesis pathway, 13 DEGs (including four cinnamyl alcohol dehydrogenase and four peroxidase superfamily proteins) were downregulated, while 11 DEGs in the phenylpropanoid biosynthesis pathway (including nine peroxidase superfamily proteins) were upregulated (Fig S2).  (Besseau et al., 2007). Furthermore, coordinated regulation of genes involved in flavonoid metabolism can redirect flux into the isoflavonoid branch of the phenylpropanoid pathway, by reducing competition for the flavanone substrate. The experimental results, which showed good reproducibility, were also validated by principal component analysis of the gene transcript data (Fig. S5). Based on this, our experiments demonstrate that AhHDA1 overexpression causes retarded growth of transgenic hairy roots, and is associated with increased expression of phenylpropanoid secondary metabolic pathway genes (Fig 6). It is clear that the expression of these secondary metabolite synthesis genes is relevant to AhHDA1 differential expression. is mediated by recruiting SAP18, a member of the SIN3 HDAC complex, which influences floral patterning (Liu et al., 2009). In another example, HDA15 was found to interact directly with PIF3 and NF-YCs in vivo and in vitro. PIF3 targets the genes involved in photosynthesis and chlorophyll biosynthesis in the dark and represses gene expression, while NF-YCs function as transcriptional co-repressors by interacting with HDA15 to inhibit hypocotyl elongation in photomorphogenesis during the early seedling stage Tang et al., 2017). Other researchers also reported that HDA-interacting proteins are involved in stress responses. Thus, the cold signaling attenuator High Expression of Osmotically Responsive Gene1 (HOS1) negatively regulates cold responses by interacting with its interacting partner histone deacetylase 6 (HDA6) under short-term cold stress (Jung et al., 2013). In wheat, FVE/MSI4 is a WD-40 repeat-containing protein that can form the FAE-HAD6-FLD complex, which plays important roles in co-regulating gene expression in multiple developmental processes including flowering time and abiotic stress Zheng et al., 2017). From the above, it is clear that RPD3/HDA1 family proteins are involved in cell development and stress responses.

AhHDA1 interacting proteins and their potential biological functions
Using the yeast two-hybrid technique, 37 AhHDA1-interacting protein candidates were identified (Table S2). Some of the most interesting proteins that interacted with AhHDA1 are APRR2-like, GLK1-like, bHLH77-like, NAC2 and MYB-like transcription factors. Three molecular chaperones or their regulators were also found, and these may function in nucleosome assembly. In addition to these examples, 18 proteins might be involved in cellular synthetic metabolic pathways, while nine proteins are related to stress responses and six proteins to photosynthesis and oxidative phosphorylation. Although these candidate interacting proteins require further experimental analysis, they potentially provide insight into how AhHDA1 might regulate synthetic metabolic pathways, cellular defense responses, and photosynthesis and oxidative phosphorylation via the proteins it interacts with.
In this study, differential expression of AhHDA6, which is a RPD3/HDA1 subfamily member, was shown to modify cell morphogenesis and developmental in peanut hairy roots and to retard cell growth (Fig 1 and Fig 2). Transcriptome analysis implicated certain synthetic metabolic pathways in the cause of the observed phenotype. Few similar results have been reported previously in other plants, perhaps because AhHDA1 possesses different features to RPD3/HDA1 genes in other species, or because the phenomenon of HDAC-inhibited cell growth is stage-and/or tissue-specific. In future studies, AhHDA1-interacting proteins will be identified and studied in detail, which will provide further insight into the molecular mechanisms of AhHDA1 function and its role in cell growth and development in hairy roots, thereby possibly laying the foundations for improvements in crop plant varieties.

Plant materials and growth conditions
Seeds of peanut (Arachis hypogaea L. cv Yueyou 7) (Fang et al., 2007) were sown in pots with a potting mixture of vermiculite, perlite and soil (1: 1: 1), and grown in an illuminated incubator with 16 h of light from fluorescent and incandescent lamps (200 µmol m −2 s −1 ) at 26˚C followed by 8 h of darkness at 22˚C. Plants were watered with half-strength Murashige and Skoog nutrient solution every other day (Su et al., 2015).

Agrobacterium strain and binary vector
The cucumopine-type A. rhizogenes strain K599 was used to induce transgenic hairy roots in peanut. The AhHDA1 cassette was released from pCanG-AhHDA1, which was previously

Yeast two-hybrid analysis
After treated by 30% PEG6000 (W/V) for 5 h, total RNA in peanut roots, stems and leaves were extracted to construct yeast cDNA library. The cDNA fragments were ligated to pGADT7 plasmid and stored in yeast cells. Full length AhHDA1 cDNA was amplified by PCR using the following primer pairs: AhHDA1-EcoR1-F: 5'-TACGAATTCATGGGGATA GAAGAAGAGAG-3', AhHDA1-Sac1-R: GAGCTCTCAGCAGCATCCATGTGG-3'. The PCR fragment was ligated into the vector pGBKT7 (Clontech, USA) after cleavage by restriction endonucleases EcoRI and SacI (New England Biolabs, USA). pGBKT7-AhHDA1 and the yeast library plasmid were co-transformed to AH109 competent cells. After screened on the SD/Trp -/Leu -/Hisselective medium, all fragments cloned were checked by PCR and sequenced. The two-hybrid assays were performed using the Matchmaker TM Gold two-hybrid system (Clontech) following the manufacturer's protocols (Liu et al., 2016a).

Measurements of intracellular reactive oxygen species
A superoxide anion chemiluminescent probe (Wako, Cat No. 075-05111) and a BES-H2O2-Ac H2O2 probe (Wako, Cat No. 028-17811) were used to measure superoxide anion content and H2O2 content, respectively, which reflect the levels of intracellular reactive oxygen species (ROS) (Kawase et al., 2004;Teranishi et al., 2007). After washing with phosphate-buffered saline, hairy root tips were soaked with BES-H2O2-Ac fluorescence probe and superoxide anion chemiluminescent probe for 20 min and imaged using a laser scanning confocal microscope (Leica TCS SP8) (Yamamoto et al., 2017). Then the samples were ground and homogenized in phosphate buffer solution (pH7.4), and the supernatant was collected after centrifugation. Plate-reader-based luminescence measurements were performed in 96-well plates using a multimode reader (Molecular Devices, SpectraMax M5; wavelengths 488 nm and 610 nm, respectively), with temperature set at 37˚C.

RNAseq and data analysis
After 20 d growth, 35S::eGFP, 35S::AhHDA1-eGFP and 35S::AhHDA1-RNAi transgenic hairy roots were collected from at least 10 different hairy roots that comes from different sites of hairy roots, as independent biological replicates for transcriptome analysis. Total RNA in every sample was no less than 5 μg. The total RNA samples were first treated with DNase I (NEB, Cat. No. M0303) to degrade any contaminating DNA. Then the samples were enriched for mRNA using oligo(dT) magnetic beads (NEB, Cat. No. S1419). After mixing with fragmentation buffer, the mRNA was broken into short fragments (about 200 bp). Then first-strand cDNA was synthesized using random hexamer primers (NEB, Cat. No. E7771).
Buffer, dNTPs, RNase H and DNA polymerase I were added to synthesize the second strand and double-stranded cDNA was purified with AMPure XP Beads (NEB, Cat. No. E7330).
End repair and 3'-end single nucleotide A addition were then performed using the Next ® Ultra TM II End Repair/dA-Tailing Module (NEB, Cat. No. E7546). Finally, sequencing adaptors were ligated to the fragments. The library products were sequenced on the Illumina Hiseq 2000 platform. Clean reads were obtained by removing reads containing adapter, reads containing polyN and low-quality reads from raw data, and were aligned to the peanut genome (http://peanutbase.org/home) using Soap (2.21) software (http://soap.genomics.org.cn/). The gene expression level was quantified using the RPKM method; the RPKM of each gene was calculated based on the length of the gene and the read count mapped to this gene. The genes' RPKM values were calculated based on all uniquely mapped reads. Differential expression analysis was implemented using the Poisson distribution model, and genes with a false discovery rate ≤ 0.001 and more than twice the difference between either the overexpressing AhHDA1 or the AhHDA1 interference group and the control group were assigned as differentially expressed (criteria of >2-fold difference in expression level and p<0.05).

Reverse transcription and real-time PCR
RNA extraction was carried out as described (Su et al., 2015). Three biological replicate RNA samples for each time point and treatment were used for downstream applications. Relative expression for each well was calculated . Expression data for A.