An Ecient Chromatin Immunoprecipitation (ChIP) Protocol for Studying Histone Modications in Peach Reproductive Tissues

Perennial fruit trees display a perennial growth behaviour characterized by an annual cycling between growth and dormancy, with complex physiological features. Rosaceae fruit trees represent excellent models for studying not only the fruit growth/patterning, but also the progression of the reproductive cycle depending upon the impact of climate conditions. In addition, the current development of high ‐ throughput technologies is starting to have an important impact on Rosaceae tree research for investigating genome structure and function as well as (epi)genetic mechanisms involved in important developmental and environmental response processes during fruit tree growth. Among the epigenetic mechanisms, chromatin remodelling mediated by both histone modications and other chromatin-related processes play a crucial role in gene modulation, controlling gene expression process. A very useful technique to investigate the chromatin states in plants and their dynamics is chromatin immunoprecipitation (ChIP), generally applied for studies on chromatin states and enrichment in post transcriptional modications (PTMs) of histone proteins.


Page 3/21
ower bud initiation to fruit/seed maturation, is completed in two years after experiencing cold temperature to overcome the winter rest period, called dormancy. To date the reproductive cycle has been largely investigated both in physiological and molecular terms, but not in a comprehensive manner. As an initial broad observation, these studies established that the reproductive cycle resulted from coordinated changes in expression of hundreds to thousands of genes involved in the regulation of structural (cells/tissues differentiation of ower and fruit) and metabolic (fruit metabolites and hormones) traits of reproductive organs [2,3]. However, the attention was mainly paid to the fruit and to its ripening phase: in fact, in literature, there is an impressive number of studies focused on understanding the main processes that take place during the progression from an immature to ripe fruit, whereas the earlier phases, corresponding to the pre-pollination stage are not well clari ed [4]. In particular, the comprehension of bud dormancy phase is strategic when considering the impact of climate change on plant physiology.
Dormancy is an evolutionary process that is entrusted to temperature resilient structures such as buds and can be interpreted as "a state of self-arrest of the shoot apical meristem (SAM) which is maintained under growth-promoting conditions" [5,6]. Therefore, the dormancy mainly ensures survival under extremely low temperatures during winter and it can also in uence developmental functions, including fruit setting and patterning, when inadequate temperature compromises overcoming dormancy [7].
The emerging progress of high-throughput methods and bioinformatics technologies for analysing genome structure and function has had an important impact on research in fruit trees and has signi cantly contributed towards accelerating the discovery of speci c DNA regulatory elements that interact with transcription factors (TFs) responsible for plant growth and responses to plant-environment interaction [8]. For instance, whole genome sequencing projects of two fruit crop species, namely peach and apple, has fostered in-depth molecular studies in Prunus and Malus species over recent years [9][10][11], allowing the identi cation of factors which interact in a multilevel process and triggers the coordinated action of master regulators, including hormone signalling, microRNAs, and epigenetic mechanisms [12].
Among the latter, chromatin-remodelling mechanisms, mediated by both histone modi cations and other chromatin-related processes, play a crucial role in gene modulation, by in uencing the ability of transcription factors (TFs) to bind DNA regulatory elements and thereby controlling gene expression [13].
However, the vast majority of plant cis-elements in gene promoters are unknown [14], because optimised experimental protocols for recovering nucleic acids to be utilized for genome-wide analyses are still lacking.
A very useful technique to investigate DNA-protein interaction and chromatin states and their dynamics is chromatin immunoprecipitation (ChIP). It rely on the use of a speci c antibody raised against the target transcription factor or the histone modi cation under investigation and it is widely used for a few model systems, including Arabidopsis, though the establishment of this technique is still remarkably challenging in other plant systems [15,16]. In plants, ChIP is generally applied for studies on chromatin states and enrichment of post transcriptional modi cations (PTMs) in histone during development or stress response mechanisms proteins [17]. The integration of expression data, at either single gene or genomewide level, with histone PTM enrichments in speci c gene contexts can unveil possible direct correlations between gene transcriptional variations and histone modi cation dynamics, during plant tissue differentiation and development. Recent studies on fruit tissues revealed that a native chromatin immunoprecipitation protocol (N-ChIP), performed without cross-linking, is better for pro ling histones and histone modi cations studies for improved antibody speci city, higher pull-down e ciency, lower background and less bias when compared to an X-ChIP (cross-linked chromatin followed by immunoprecipitation) procedure [18].
In this work, we describe speci c protocols for ChIP procedure in different reproductive tissues ( ower buds and fruit) of Prunus persica with the aim to investigate histone marks distributions. Interestingly, histone modi cations have been implicated in regulating both bud dormancy and fruit development progression [19,20] and peach represents a model in the Rosaceae family for studies of biological processes like bud formation/dormancy and differentiation/development of climacteric fruits [10]. Here we describe in detail a method for chromatin extraction and immunoprecipitation to systematically optimize an X-ChIP protocol, suitable for subsequent gene target (ChIP-qPCR) and genome wide analyses (ChIP-seq). Our aim is to gure out a possible correlation between gene expression and presence of speci c chromatin marks. We focused our investigations on the role of two chromatin marks (described in detail in the Results section), namely trimethylation of histone H3 at lysine in position 4 (H3K4me3) and trimethylation of histone H3 at lysine 27(H3K27me3), in modulating speci c gene expression during these fundamental biological processes in the nectarine cultivar Fantasia.

ASSESSMENT OF CHROMATIN QUALITY: EFFICIENCY CHECK OF CHROMATIN FIXATION/EXTRACTION AND FRAGMENTATION
We have developed and optimized an immunoprecipitation protocol on crosslinked chromatin (X-ChIP) suitable for analysing either genome wide distributions and/or speci c distribution at target loci of single post-transcriptional histone modi cations, in peach reproductive tissues. One of the main advantages of the procedure proposed here is its applicability to hard plant tissues (i.e. FB and FM) stored at -80°C after freezing in liquid nitrogen.
Previous studies have described how the success of a ChIP procedure depends upon both the nature of the starting material and the initial processing steps. In these studies, the use of fresh and unfrozen plant tissue as starting material for the chromatin extraction and following analyses has often been strongly recommended [21]. Moreover, the use of vacuum mediated in ltration has also suggested to ensure an e cient penetration of the xative for the crosslinking reaction into the plant tissue; the crucial step which distinguishes a N-ChIP from X-ChIP procedure [16,18]. However, in both methods preserving chromatin structure during the isolation and subsequent steps is the main aim of the procedure. For this reason, the cross reaction with the xation agent (in our case formaldehyde) in a X-ChIP strategy was the initial step to be evaluated and optimized in our experiments. Insu cient crosslinking will not preserve the chromatin structure, while over crosslinking will hamper the ChIP procedure as reported by [22].
In Fig. 1A, the e ciency check conducted to determine the optimal crosslinking conditions for FB and FM tissues is reported. In FB, the developed protocol for chromatin xation and extraction was very e cient in all three stages (0, 475 and 770CU) during endodormancy. With the progression of endodormancy buds have an increase in number of scales and maturity degree while also accumulating soluble carbohydrates which hinder chromatin extraction. Bud scales act as protection for newly formed leaves and branch outgrowth. They form at the end of the growth season once the leaves have fallen off the branches. By removing the scales, the procedure was more e cient for all time points (Fig. 1A). Despite the absence of scale removal from buds in other published protocols [16], this process allowed to shorten the number of chromatin extraction steps and avoids the use of β-mercaptoethanol in our experiments.
For FM, the e ciency of chromatin xation and following extraction steps depended on fruit developmental stage. In fruit pre-ripening phase, from the onset to the end of the S3 phase, the chromatin extraction was reasonable: the extraction was satisfactory in FM collected at 83, 104 and 111 DAFB, when the fruit expansion and endoreduplication processes occur [20]. Indeed, during these stages, the accumulation of sugars and other metabolites is reduced due to the high energy requirement for the expansion of FM cells [23]. On the contrary, in samples collected at 118 and 125 DAFB the high level of polysaccharides and other secondary metabolites stored showed an inhibitory effect on the extraction phase, causing a dramatic reduction in terms of quality and yield (data not shown) as also reported in other works [19,24,25]. However, considering both the cellular uniformity of the mesocarp tissue and the general technical limits of the procedure, the chromatin extraction procedure was performed after pooling three biological replicates and the results are reported in the following sections.
According to other works conducted on different plant tissues [22], our results with FB and FM indicated that the addition of formaldehyde 1% (v/v) into the xation starting buffer is e cient for the following ChIP steps, since the chromatin is neither over-or under-crosslinked, and it allows to recover a substantial amount of DNA from the reverse crosslinking step (Fig. 1A).
The size of the chromatin fragments used as input material is the second determinant factor for the resolution and success of a ChIP procedure since a proper size distribution of DNA fragments is crucial for speci c ChIP applications. Ideally, the bulk of the chromatin for the following application includes a length between 250 and 750 bp but, depending on the intended use of ChIP applications (gene target vs whole sequencing), an appropriate shearing step must be determined for each chromatin preparation. In Fig. 1B we report the best results obtained after testing different sonication conditions, in terms of number, timing and amplitude of rounds. We decided to use a sonication procedure vs an enzymatic MNase-mediated fragmentation since when using formaldehyde crosslinking the access to chromatin is restricted [26]. After testing various sonication conditions, we used 60% amplitude with a number of rounds ranging between 15 (for FM) and 25 (for FB) for 10 sec each. These conditions allowed to obtain a physical shearing of chromatin compatible with the following purposes. For our samples a smear between 200 and 500 bp was observed for FB and between 200 and 800 bp for FM. In both cases, the fragmentation was suitable for the following molecular analyses and, in particular, the higher fragmentation of chromatin obtained from buds allowed for a successful library preparation for ChIP-Seq analysis accordingly [16].

ASSESSMENT OF CRHOMATIN IMMUNOPRECIPITATION ASSAY BY QPCR AND ChIP BY SEQUENCING IN REPRODUCTIVE PEACH TISSUES
After the application of speci c precautions in the chromatin xation/extraction protocol the quality and the pull-down e ciency of the immunoprecipitation was tested. The aim was to de ne whether our operative changes speci c for each peach plant material, affected the results of the following molecular procedures to be performed.
To monitor chromatin states and nd out potential correlations between gene transcript levels and the enrichment in speci c histone modi cation, two euchromatin regulative histone marks of interest were examined: trimethylation of lysine in position 4 of histone H3 (H3K4me3), which represents an active mark typically enriched around TSS of transcribed genes, and H3K27me3, a silencing mark generally distributed over the whole gene coding region [27][28][29].
For the mesocarp tissue, as target gene, we focused the attention on PpFLESHY (PRUPE_6G159200), also known as HECATE3 (HEC3)-like, a TF with a putative key role in eshy fruit mesocarp tissue identity [30]. In the FAN fruits, PpFLESHY did not exhibit relevant variations in its expression level during S1 and S2, while its transcription increased at S3 and highly accumulated at early S4 [30]. This expression pattern was recon rmed by our investigations on FLESHY transcript levels during different fruit developmental phases (Fig. S1). After chromatin immunoprecipitation with the two speci c Abs against H3K4me3 and H3K27me3 modi ed histones, qPCR Real Time assays were performed to verify and semi-quantify the enrichment of FLESHY in ChIPed DNA populations, following the indication reported by [21] (Fig. 2). Based on the different typical distribution patterns of H3K4me3 and H3K27me3 histone marks along the gene sequences, we analysed the FLESHY sequence by considering its predicted 'TSS around-' and 'gene body-' subregions and using four couples of primers (A, B couples for 'TSS around-' and C, D couples for 'gene body-' subregions. For details see table S1). qPCR results demonstrated a signi cant enrichment of H3K4me3 activation mark at the level of FLESHY 'TSS around-', in comparison to that measured in the 'gene body-'subregion (dark vs white bars). Additionally, whereas no enrichment was observed at the 'gene body' during 83, 104, and 111 DAFB, a higher and signi cant increase (relative to input), ranging from 4,3% up to 7%, was measured at the 'TSS around' region ( Fig. 2A). The H3K4me3 preferential enrichment at TSS level is in agreement with data reported in literature in other species and tissues [29,31,32].
An opposite trend was observed for the silencing mark H3K27me3: a lower enrichment, in terms of % IP of this mark was observed in all time points in comparison to that observed for the activating mark H3K4me3 in both the investigated gene regions. However, at TSS level a small reduction of H3K27me3 was measured during the progression of fruit growth, with values ranging from 1.4-0.5%, while at gene body level a comparatively lower and constant signal was measured with a less pronounced fold change (Fig. 2B). To exclude a reduced Ab e ciency, we analysed in the same mesocarp samples, the putative enrichment of H3K27me3 into a Polygalacturonase family member (ppePG22, PRUPE_4G262200), known to be a representative silenced locus in these same developmental stages (Fig. S2), con rming the enrichment of this histone mark along the whole gene sequence.
In parallel, we investigated the enrichment of the same histone modi cation marks H3K4me3 and H3K27me3 in FB tissues by performing chromatin immunoprecipitation followed by Illumina sequencing (ChIP-Seq). In brief, The CHIP-Seq raw reads were processed for adapter clipping and quality score trimming using Trimmomatic v 0.39 [33]. Clean reads were mapped to the P. persica genome v.2.0 [9] obtained from Ensembl (http://plants.ensembl.org/index.html) using the spliced aligner HISAT2 [34]. ChIP-Seq enrichment was calculated using Model-based Analysis of ChIP-Seq (MACS) [35].
For each histone modi cation, we calculated the enriched peaks by comparing the time points in pairs encompassing six different comparisons (0CU vs 475CU; 475CU vs 0CU; 475CU vs 770CU; 770CU vs 475 CU; 0CU vs 770CU and 770CU vs 0CU). The nearest gene was called after peak calling. Differentially expressed genes (DEGs; p adj < 0.05) identi ed in a previous RNA-Seq analysis (see Availability of data and materials section) data were associated to the called genes that were H3K4me3 and H3K27me3 enriched and the putative function for each gene was deduced from the Arabidopsis thaliana homolog. Transcript sequences were scanned by blastx against UniProt/Swiss-Prot and UniProt/ TrEMBL to homology search. Joined biological replicates were loaded into the Integrative Genomics Viewer (IGV) genome browser to visualize both H3K4me3 and H3K27me3 peaks along with gene expression (RNA-Seq) peaks in the peach genome (Fig. 3) in which we reported the expression pro le of some epigenetic regulator genes identi ed as marker genes [3] associated to their histone modi cation pro le during endodormancy and endo-ecodormancy transition.
As far as it concerns genes enriched in H3K4me3 (Fig. 3A), with the progression of endodormancy, an active transcription of the CHROMOMETHYLASE 3 (CMT3-like, PRUPE_6G011600) and a HDA19-like gene (PRUPE_8G183700) was observed. The DNA methyltransferase CMT3 maintains CHG (H = A, C, or T) methylation at constitutive heterochromatin in plants and thus it is important for maintaining genome stability [36]. Histone deacetylase 1of Arabidopsis (AtHD1 or AtHDA19), a homolog of yeast RPD3, is known to be a global regulator of many physiological and developmental processes [37]. The expression pattern and H3K4me3 distribution at loci of the reported genes indicate that bud dormancy overcoming is accompanied by the chromatin control of transcriptionally active regions through the enrichment in the H3K4me3 mark.
Regarding H3K27me3, our results indicate that there is not a good correlation between gene expression and chromatin enrichment in H3K27me3 (Fig. 3B). Although this histone mark is responsible for PCR2 mediated gene silencing in euchromatic regions, the enrichment in H3K27me3 does not always correlate with gene silencing or with a low gene expression. This is also the case of some selected chromatin modi ers, such as ATXR7-like (PRUPE_2G42400 a putative ARABIDOPSIS TRITHORAX-RELATED7 a putative Set1 class H3K4 methylase) [39], SUVR2-like (PRUPE_8G216300 a putative SU(VAR)3-9-like histone methyltransferase) [40] and JMJ24-like (PRUPE_6G322900 a jmjC histone demethylase possibly involved in gene silencing) [41]. All of them present a signi cant decrease in H3K27me3 mark in their chromatin, during dormancy progression; however, this does not correlate to their transcript level. In the case of ower bud, this missing correlation might be due to the strong tissue speci city of H3K27me3 chromatin mark [42] while ower buds are composed of different specialized tissues. However, this lack of correlation was observed by other authors and in other plant tissues.

Conclusions
ChIP represents a powerful tool in the study of histone modi cation status in plant tissues. Many examples in literature strongly support the indispensably important role of the ChIP approach in the eld of chromatin research. In this report, we describe an optimized X-ChIP procedure for recalcitrant plant tissues because of their structural properties and composition, such as peach reproductive tissues. To overcome some methodological obstacles for investigating speci c chromatin modi cations occurring in the reproductive organs during their development, we developed an e cient and reproducible protocol for the chromatin xation and extraction, applying changes that affect positively the following molecular investigations (Fig. 4). The general aim was to provide an affordable and repeatable procedure for studying the distribution of modi ed histones in buds and eshy fruits or other plant organs/tissues with high levels of polysaccharides, secondary metabolites (like phenols), high content of water and large vacuoles.
One advantage of the developed ChIP protocol is to avoid the step of clean nuclei isolation and the possibility to use frozen tissues for the extraction of chromatin, without the addition of βmercaptoethanol, as recommended by other procedures [16,18,43].
In addition, our procedure was optimized for the following molecular investigations and allowed to analyse the distribution of histone marks at single gene level by qPCR or at genome wide level by NGS sequencing. By integrating the information on gene expression and the enrichment/depletion of speci c modi ed histone at selected loci, we obtained information on the effect of a histone modi cation on gene expression and correlated the expression variation with variations in chromatin marks distribution and enrichment during fruit growth/ripening and bud dormancy in peach.
Results pointed out that, similar to what was observed for other species [12,44], in the reproductive cycle of peach, from bud to fruit, epigenetic regulation plays a fundamental role through the coordination of the response to endogenous and environmental signals.
Flower buds (FB) were collected at three time points during the winter 2018 and 2019 (05/11/2018, 10/12/2018, 07/01/19), corresponding to 0, 475, and 770 chilling units (CU) respectively, calculated as described by [45]. Daily temperature readings were retrieved from Agenzia Regionale per la Prevenzione e Protezione Ambientale del Veneto (ARPAV; https://www.arpa.veneto.it/). At each time point, buds were collected from groups of 3-4 plants each, corresponding to two (for 0CU and 770CU) and three (for 475CU) biological replicates, their scales were removed and immediately frozen in liquid nitrogen for their storage at -80° C until subsequent molecular analyses. Fruits were collected following the fruit growth by measuring the equatorial diameter. The fruit double sigmoidal growth kinetics have been divided in four phenological stages, named S1, S2 S3 and S4, as described by [46]. Phases corresponding to the rst exponential growth phase due to cell division, pit hardening phase, second exponential growth phase mainly owing to cell expansion, and ripening processes respectively, were described in detail in [20]. replicates, each composed of 5 fruits), was collected by taking, from each fruit, a radial section from the epicarp to the rst cell layer of endocarp tissue, with high cellular type homogeneity [20]. This portion was immediately frozen in liquid nitrogen and stored at -80°C.

Chromatin Extraction And Immunoprecipitation Assay
The general procedure described is valid for both FB and FM tissues. All buffers used were prepared using autoclaved stock solutions on the day of use and kept on ice until required. Phenylmethylsulfonyl uoride (PMSF, Sigma), Na-butyrate and protease inhibitor cocktail (PI, Sigma) should be added into the solutions just before use. The detailed explanation of all steps are reported in Supplementary M&M, with indicated "Notes" describing speci c points for each tissue sample.
In summary, for chromatin extraction/puri cation the frozen FB and FM tissues were nely powdered with liquid nitrogen and the chromatin was extracted by transferring the ne powder into a 50 mL tube and xing with 1% formaldehyde in cold NIB with a ratio 0,5 g FB or 1,5 g FM tissue in 25 ml NIB. Fixation was performed for 15' at RT and then blocked for 5' with 3.4 mL of 1M Glycine. Subsequently, the lysates were ltered through a single layer of Miracloth (Millipore) into a new tube and centrifuged at 5000 g for 15' at 4°C. The white pellet was respectively washed twice with 3 ml of cold Wash Buffer 1_FB/Wash Buffer 2_FB for Floral Buds or 5 ml Wash Buffer_FM for Fruit Mesocarp samples. The latter pellet was then additionally suspended in 5 ml of 15% Percoll solution and after centrifugation at 5000 g for 15' at 4°C. Finally, FB and FM pellets were suspended in 500 ul of Lysis Buffer. The extracted chromatin was sonicated by Microson Ultrasonic Cell Disruptor (Heat Systems, Germany) at power 30%-60%, for a variable number of pulses (10-15 s each, followed by 30s in ice) to 200-700 bp fragments and centrifuged at 16,000 g for 10' at 4°C. A fraction of the supernatant was saved and reverse cross-linking was performed with 0.2 M NaCl for 16 h at 65°C. DNA was puri ed by Phenol/Chloroform/isoamyl alcohol (25:24:1) as input in the following PCR reactions. Moreover, a quality check and yield estimation was conducted.
The subsequent immunoprecipitation procedure was performed using the Dynabeads protein G (ThermoFisher Scienti c) adapting the steps to our tissues. The Dynabeads protein G were -pre-washed and, after their recovery, suspended with 100 ul Lysis Buffer (for each biological sample and Ab used). The suspension was subdivided into two aliquots of 50 ul and one aliquot was incubated with 5 ug or 10 ug of FB or FM chromatin sample respectively, for 4-5 hours at 4-10°C ("pre-clearing of chromatin step"), while the other one was incubated with an appropriate quantity of speci c Ab and incubated on a rotating incubator at 4-10° for two hours (5 to 7 µg of antibody against H3K4me3 -Active Motif, 39159-or 10 µg of antibody H3K27me3 -Millipore, 07-449-) ("antibody preparation step"). The following immunoprecipitation step was performed by combining the pre-cleared chromatin samples with the Ab-Dynabeads protein G complexes, obtained from previous steps, and incubating them O/N at 4-10°C with gentle rotation. A chromatin aliquot processed similarly but without the addition of any antibody (No Ab sample) was used as background control.
The beads linked to histone-DNA immunoprecipitated complexes were then washed sequentially, once with Low Salt, High Salt, LNDET, and twice with TE Buffers. The histone-DNA complexes were eluted with 0.1M NaHCO 3 and 1% SDS and the cross-linking was reversed with 0.2 M NaCl for 16 h at 65°C. The DNA samples were precipitated with Glycogen (Sigma) and 2 volumes of absolute ethanol for 2h at -20°C.
After pellet collection followed by washing it with 70% ethanol, DNA was suspended in TE Buffer for proteinase K and treated with RNAse I for 30 min at 37°C and with proteinase K for 1 h at 42°C, extracted once with phenol-chloroform and with QIAquick PCR puri cation kit (QIAGEN). One microliter of this ChIPed DNA and an appropriate dilution of input were used for the following qPCR analyses.
One microliter of ChIPed DNA and an appropriate dilution of input were used for the following Real time qPCR analyses. For every enrichment of histone marks in each target gene investigation, at least two pairs of primers were considered for PCR reactions at speci c regions, and the sequences of best working primers (in terms of dimer formation and speci city), with relative positions compared to predict TSS, are reported in Table S1. qPCR data analyses were performed as reported in Rossi et al. 2007  FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) was used to assess the quality of the reads. The CHIP-Seq raw reads were processed for adapter clipping and quality score trimming using Trimmomatic v 0.39 [33]. Clean reads were mapped to the P. persica genome v.2.0 [9] obtained from Ensembl (http://plants.ensembl.org/index.html) using bowtie2 [47]. ChIP-Seq peaks calling and differential analysis between the three analysed samples were performed using Model-based Analysis of ChIP-Seq (MACS) [35].

Rna Extraction And Gene Expression Analysis
For FB, total RNA was extracted from 70-80 mg of frozen and ground sample using RNeasy Plant Mini kit (Qiagen) with minor modi cations: 1.5% PVP-40 was added in the extraction buffer RLT in a total volume of 750 µl instead 450 µl. On the contrary for FM, total RNA extraction was performed as reported in [30]. ktglcoyehjupzup), respectively.

Competing interests
The authors declare that they have no competing interests.

Funding
This work was funded by BIRD 2019-University of Padova to a grant to BC and SV.
Authors' Contributions SV and CB: conceptualization; MC and SF: methodology and writing draft and nal manuscript preparation; CF and JJ: review and editing. All authors read and approved the nal manuscript.
Aknowledgements Figure 1 Crosslinking e ciency and physical shearing chromatin analyses.
(a) Bud and mesocarp tissues were crosslinked in buffers containing increasing amounts of formaldehyde. Samples were subjected or not to a reverse crosslinking phase (decrosslinked sample + and -, respectively), and DNA isolated using phenol/chloroform extraction as described in MM section. While DNA is e ciently isolated from samples that were not crosslinked (lanes indicated with 0%), a decrosslinking procedure is required for the isolation of DNA from cross-linked samples (with 1% indicating the relative concentration of formaldehyde used in testing analyses, which resulted in a better yield of signal).
(b) Chromatin shearing check after the application of 60% amplitude with several 10 sec shearing rounds ranging between 25 (for FB) and 15 (for FM) followed by a reverse crosslinking phase and a DNA isolated using phenol/chloroform extraction.

Figure 2
Chromatin marks analysis by X-ChIP method for peach mesocarp tissue.   Schematic work ow of ChIP protocol.
The protocol phases are reported on the left of the picture and the main relative improvements applied are described on the right.