ChIP-seq and RNA-seq for complex and low-abundance tree buds reveal chromatin and expression co-dynamics during sweet cherry bud dormancy

Chromatin immunoprecipitation-sequencing (ChIP-seq) is a robust technique to study interactions between proteins, such as histones or transcription factors, and DNA. This technique in combination with RNA-sequencing (RNA-seq) is a powerful tool to better understand biological processes in eukaryotes. We developed a combined ChIP-seq and RNA-seq protocol for tree buds (Prunus avium L., Prunus persica L Batch, Malus x domestica Borkh.) that has also been successfully tested on Arabidopsis thaliana and Saccharomyces cerevisiae. Tree buds contain phenolic compounds that negatively interfere with ChIP and RNA extraction. In addition to solving this problem, our protocol is optimised to work on small amounts of material. Furthermore, one of the advantages of this protocol is that samples for ChIP-seq are cross-linked after flash freezing, making it possible to work on trees growing in the field and to perform ChIP-seq and RNA-seq on the same starting material. Focusing on dormant buds in sweet cherry, we explored the link between expression level and H3K4me3 enrichment for all genes, including a strong correlation between H3K4me3 enrichment at the DORMANCY-ASSOCIATED MADS-box 5 (PavDAM5) loci and its expression pattern. This protocol will allow analysis of chromatin and transcriptomic dynamics in tree buds, notably during its development and response to the environment.

seq on the same starting material. Focusing on dormant buds in sweet cherry, we explored the link between expression 1 level and H3K4me3 enrichment for all genes, including a strong correlation between H3K4me3 enrichment at the 2 DORMANCY-ASSOCIATED MADS-box 5 (PavDAM5) loci and its expression pattern. This protocol will allow 3 analysis of chromatin and transcriptomic dynamics in tree buds, notably during its development and response to the The term 'epigenetics' has traditionally been used to refer to heritable changes in gene expression that take place without 8 altering DNA sequence (Wolffe and Matzke, 1999), but it is also used, in a broader sense, to refer to modifications of the 9 chromatin environment (Miozzo et al., 2015). Epigenetic modifications are important for a wide range of processes in    Leida et al., 2012), this is the first step-by-step detailed ChIP-seq 1 protocol in trees that includes all of these improvements: 2 • Our ChIP-seq/RNA-seq protocol can be carried out on complex plant tissues that contain interfering compounds 3 (phenolic complexes, scales, protective layers), by adding chelators of these compounds in the extraction buffers.

4
• The cross-linking step is performed on frozen, pulverised material, thus allowing sample collection in the field, where 5 cross-linking equipment is not available. It also allows studying fast responses by flash freezing material immediately 6 after a stimulus (e.g. transient temperature stress) rather than cross-linking directly on fresh tissue or cells. In previous 7 protocols, the cross-linking step was performed using a vacuum and lasted at least 10 minutes and up to 1 hour

11
• By using frozen, pulverized material, ChIP-seq and RNA-seq can be performed on the same starting material for a 12 direct and robust comparison of epigenetic regulation and gene expression.

13
• Our protocol can be used to perform ChIP-seq and RNA-seq on a small amount of biological material. We optimised 14 this protocol to start from 0.2 to 0.5 g of buds, which is considerably lower than the usual amount of 0.8 to 5 g of starting

18
We have used this protocol to analyse histone modification profiles in several tree species. In a first instance, we 19 analysed H3K27me3 in buds of Prunus persica L Batch (peach) and could replicate previously published results (de la 20 Fuente et al., 2015). To demonstrate the versatility of this protocol, we also successfully performed ChIP-seq for 21 H3K27me3 and H3K4me3 in sweet cherry (Prunus avium L.) and apple (Malus x domestica Borkh.). Furthermore, we 22 directly compared expression level and enrichment for H3K4me3 by performing ChIP-seq for H3K4me3 and RNA-seq 23 on the same sweet cherry bud samples. We demonstrated the correlation between chromatin status and gene expression 24 for AGAMOUS (AG) and ELONGATION FACTOR 1 (EF1) that are known to be under control of H3K27me3 and 25 H3K4me3, respectively . We expended our analysis of chromatin and expression in cherry buds 26 harvested at different stages of dormancy, first to DORMANCY-ASSOCIATED MADS-box 6 and 5 (PavDAM6 and 27 PavDAM5) genes, which are key regulators of dormancy in trees, and then to the entire genome. Dormancy is an 28 important developmental stage of fruit trees and is characterised by a period of repressed growth that allows trees to 29 persist under low winter temperature and short photoperiod (Faust et al., 1997). A proper regulation of the timing of the 30 onset and release of bud dormancy is crucial to ensure optimal flowering and fruit production in trees. Consequently, 1 unravelling the associated molecular mechanisms is essential and numerous studies have been conducted in trees to . We find that dormancy-associated PavDAM6 and PavDAM5 genes are more expressed in 7 dormant buds than in non-dormant buds and that H3K4me3 occupancy is associated with PavDAM5 expression level.

8
We also find significant changes in H3K4me3 level during dormancy for 671 genes, and that these changes are positively 9 associated with transcriptional changes during dormancy. Our results show the potential for future exploration of the link 10 between chromatin dynamics and expression at a genome-wide level during tree bud dormancy. Moreover, our combined 11 ChIP-seq and RNA-seq protocol, which is working on many tree species, will allow a better understanding of 12 transcriptional regulatory events and epigenomic mechanisms in tree buds.

15
Validation of the protocol robustness in peach

16
In order to validate our protocol, we analysed H3K27me3 enrichment in the gene body of a DAM gene cluster on peach

23
at the DAM genes confirms that our improved ChIP-seq protocol is working on tree buds.

25
Successful application of this protocol on different species

26
To demonstrate the versatility of this protocol, we performed ChIP-seq for two histone marks (H3K27me3 and 27 H3K4me3) on buds of three tree species: peach, apple (Malus x domestica Borkh.) and sweet cherry (Prunus avium L.)

28
( Figure 3). We analysed the signal at the genes ELONGATION FACTOR 1 (EF1), known to be enriched in H3K4me3, 29 and AGAMOUS (AG), known to be enriched in H3K27me3 . While H3K27me3 is associated with a 30 repressive transcriptional state, H3K4me3 is on the contrary associated with transcriptional activation. We observed a 1 strong H3K4me3 signal at EF1 and enrichment for H3K27me3 at AG locus for the three species ( Figure 3). This result 2 confirms that this ChIP-seq protocol works on buds for several tree species. To test the adaptability of this protocol for 3 other biological system, we also performed ChIP-qPCR in Arabidopsis thaliana and in Saccharomyces cerevisiae. We  Figure 1). This protocol is thus versatile and can be used on buds for several tree species as well as 7 in other biological systems such as Arabidopsis thaliana and yeast.  To test if chromatin state and gene expression can be directly compared using this protocol, we carried out RNA-seq and

11
ChIP-seq on the same starting material of sweet cherry floral buds. To start with, we analysed expression together with 12 enrichment for H3K27me and H3K4me3 for PavEF1 and PavAG genes. We observe that PavEF1, marked by H3K4me3, 13 is highly expressed, while PavAG, marked by H3K27me3, is very lowly expressed (Figure 3c, Suppl. Figure 4).

14
We then compared the abundance of H3K4me3 histone mark to the expression patterns of PavDAM5 and PavDAM6  The time of dormancy release, after which the buds are in ecodormancy, is defined when the percentage of bud break 18 reaches 50% at BBCH stage 53 (Meier, 2001). We observe that samples harvested in October and December are in 19 endodormancy and the ones harvested in January are in ecodormancy (Figure 4a). PavDAM5 and PavDAM6 are two key 20 genes involved in sweet cherry dormancy corresponding to the peach PpeDAM5 and PpeDAM6, respectively. We find 21 that PavDAM6 is highly expressed in October at the beginning of endodormancy and that its expression decreases in    H3K4me3 at these genes is relevant. We observe that H3K4me3 enrichment at PavDAM5 is higher in December 28 compared to October and January (Figure 4c). This is in agreement with the higher PavDAM5 expression in December.

29
We observe that H3K4me3 enrichment at PavDAM6 is also higher in December compared to October and January, while 1 its peak of expression is in October (Figure 4c).

3
Direct comparison of ChIP-seq and RNA-seq data 4 To further investigate the link between gene expression and H3K4me3 enrichments during bud dormancy, we analysed 5 genome-wide changes between time-points in expression level as well as H3K4me3 signal. For this, we employed a gene 6 centric approach by measuring the strength of the H3K4me3 signal at each gene, and identifying genes showing 7 significant changes between at least two of the three time-points (see material and methods for more detail). We

19
We explore signalling pathways that were represented in the different clusters (Figure 5c, Suppl. figure 7b, Suppl. Table   20 2). Among the genes classed in the green, gold and orange clusters, that increase expression over time, we identified the 21 GLUTATHION S-TRANSFERASE19 (PavGSTU19), LATE EMBRYOGENESIS ABUNDANT14 (PavLEA14) and    In this study, we described a combined ChIP/RNA-seq protocol for low abundance and complex plant tissues such as tree 8 buds. This method allows a robust comparison of epigenetic regulation and gene expression as we use the same starting 9 material. More notably, this protocol could permit to perform ChIP/RNA-seq for kinetic experiment with short intervals 10 (every minute or less) and to collect samples in the field.

11
Several studies have led to the identification of molecular mechanisms involved in dormancy, including a cluster of  Previous studies highlighted the importance of DAM genes as key components of dormancy in perennials. We have 3 shown that, in sweet cherry, PavDAM6 is highly expressed in October, at the beginning of dormancy, and then down-4 regulated over time ( Figure 4). Conversely, we found that PavDAM5 is highly expressed at the end of endodormancy 5 (December), and then down-regulated during ecodormancy ( Figure 4). Both PavDAM5 and PavDAM6 genes were

11
We conducted a ChIP-seq in sweet cherry for H3K4me3, which is associated with gene activation, in order to link the 12 abundance of histone marks to expression patterns at three different dates along dormancy (October, December and 13 January; Figure 4). We found an H3K4me3 enrichment around the translation start site of both PavDAM6 and

4
Genes involved in bud dormancy 5 In addition to DAM genes, we identified genes that showed differential H3K4me3 enrichment between the different bud 6 dormancy stages. We found genes involved in the response to drought, cold and oxidative stresses that were either highly     Three branches bearing floral buds were randomly chosen from the sweet cherry cultivar 'Burlat' trees at different dates.

5
Branches were incubated in water pots placed in forcing conditions in a growth chamber (25°C, 16h light/ 8h dark, 60-6 70% humidity). The water was replaced every 3-4 days. After ten days under forcing conditions, the total number of 7 flower buds that reached the BBCH stage 53 (Meier, 2001) was recorded. We estimate that endodormancy is released 8 when the percentage of buds at BBCH stage 53 is above 50% after ten days under forcing conditions.          then thaw rapidly by warming the tube in your hand. Repeat once more. This freeze-thaw step is not essential, but could improve the chromatin yield. Centrifuge the tube at 15,800 × g for 3 minutes at 4°C to pellet debris, and carefully recover 1 the supernatant into a new tube. Complete the tube to 300 µl with the Sonication buffer. Set aside a 10 µl aliquot of 2 chromatin in a PCR tube to serve as the non-sonicated control when assessing sonication efficiency by gel 3 electrophoresis and keep on ice. Shear the chromatin into ~300 bp (100-500 bp) fragments by sonication (e.g. using   13 and incubate at 45°C for 1 hour. During this step, take out the SPRI beads (e.g AMPure beads; Beckman Coulter, cat# 14 A63880) from the fridge and allow them to equilibrate at room temperature (for at least 30 minutes before use).

15
To extract DNA using SPRI beads, vortex the beads until they are well dispersed, add 126 µl of beads to 60 µl of sample

29
iii. MNase digestion analysis Complete the digested sample (50 µl) and non-digested (20 µl) aliquots to 55.5 µl with TE buffer, add 4.5 µl of 5M NaCl 1 and incubate in a PCR machine or thermocycler at 65°C for 8 hours to reverse cross-link. Add 2 µl of 10 mg/ml RNase A 2 and incubate at 37°C for 30 minutes. Add 2 µl of 20 mg/ml proteinase K and incubate at 45°C for 1 hour. During this 3 step, take out the SPRI beads from the fridge and allow them to equilibrate at room temperature (for at least 30 minutes 4 before use). Proceed to the DNA extraction, using SPRI beads as explained before in the sonication analysis section (ii).   14 Transfer 50 µl of protein A-and/or protein G-coupled magnetic beads (Invitrogen cat# 10-002D and cat# 10-004D,

10
After the second wash in TE buffer, resuspend the beads in 100 µl of TE buffer and transfer the beads to a PCR tube.

11
Place the tube on a magnetic rack, remove the TE buffer and resuspend the beads in 60 µl of Elution buffer [10 mM Tris-

1
Keep the rest of the DNA for sequencing (continue to "ChIP library preparation and size selection section").

3
ChIP library preparation and size selection: TIMING 2-3 days 4 Use 5-10 ng from the input fraction for the preparation of sequencing libraries. Quantify the input fraction using a DNA 5 high sensitivity Qubit kit. For the IP fraction, as yield is often too low to be able to quantify DNA, we recommend to use 6 the entire volume from the IP for the preparation of sequencing libraries. Using this protocol we extracted around 500 ng 7 of DNA from 300 to 500 mg of powder of sweet cherry buds. We recommend carrying out ChIP-seq library preparation      The raw reads obtained from the sequencing were analysed using several publicly available software and in-house 3 scripts. Firstly, we determined the quality of reads using FastQC (www.bioinformatics.babraham.ac.uk/projects/fastqc/).   reads for all H3K4me3 ChIP-seq compared to H3 ChIP-seq, suggesting that the H3K4me3 ChIP worked. Also, around 19 30% of the genome is not covered by reads, which can be explained by the fact that the ChIP-seq has been performed on 20 cherry tree buds, but mapped on the peach genome.

21
We used a gene-centric approach to identify genes with significant changes in the strength of H3K4me3 signal between  H3K4me4 signal for any particular gene. Next, we identified a subset of genes that exhibit significant differential binding 27 between any two time-points (October vs December; October vs January and December vs January). For this step, we 28 used the H3 ChIP as a control, instead of the INPUT, because the number of mapped reads for one of the INPUT is much 29 lower than other samples (Suppl. Table 1) and this might have created some bias in detecting genes with significant differences in H3K4me3 enrichment between time-points. The quality of biological replicates was assessed by 1 performing a correlation heatmap, and hierarchical clustering of samples (Suppl. Figure 7b), based on the H3K4me3 2 signal around TSS for all genes, normalised by H3. It shows that H3K4me3 ChIP-seq replicates are of high quality.

3
To identify groups of genes with similar H3K4me3 dynamics, hierarchical clustering was performed on the Z-score of 4 the H3K4me3 signal normalised by H3 using the function hclust on 1-Pearson correlation in the statistical programme R 5 (R Core Team 2014). The Z-score has the formula (signal for a time-point -average over all time-points/ STD across all 6 time-points), which allows the changes in H3K4me3 between time-points to be compared on a gene-to-gene basis, after 7 normalising for differences that exist between genes.  ChIP-seq and RNA-seq raw data will be made available on GEO upon acceptation of the manuscript.            ChIP results in different biological systems a) ChIP-qPCR on Arabidopsis thaliana seedlings. 7-day-old Col-0 seedlings grown at 22°C were harvested aRer 15-minute incubaIon at 17°C and flash frozen. ChIP was performed as outlined in the protocol, with an anI-HTA9 anIbody (44). QuanItaIve PCR was carried out using primers specifically amplifying the genomic region occupied by the +1 nucleosome of the HSP70/ AT3G12580 or the upstream nucleosome-free region (NFR). Occupancy is normalised to the input fracIon. b) ChIP-qPCR on Saccharomyces cerevisiae (budding yeast). YEF473a cells expressing FLAG×3tagged Hsf1 were grown in YPD medium unIll mid log phase at 28˚C and subjected to 5-minute heat treatment at 37°C, aRer which they were harvested by centrifugaIon and flash frozen. ChIP was performed as outlined in the protocol, with an anI-FLAG anIbody. QuanItaIve PCR was carried out using primers specifically amplifying disInct regions of the promoter of Hsf1-target gene SSA4 (denoted as distance from the start of the coding sequence). Occupancy is normalised to the input fracIon.

a b Supplemental figure 2
Example of ChIP-seq library profile with (a) and without (b) adapter contaminaIon. The peak at 125 bp corresponds to adapter contaminaIons. A size selecIon using SPRI beads is carried out to remove the adapter contaminaIon prior sequencing. DNA profiles were obtained using TapeStaIon 4200 (Agilent Genomics). Example of RNA-seq library profile with (a) and without (b) adapter contaminaIon. The peak at 120 bp corresponds to adapter contaminaIons. A size selecIon using SPRI beads is carried out to remove the adapter contaminaIon prior sequencing. DNA profiles were obtained using TapeStaIon 4200 (Agilent Genomics).