DNA Preparation In Vitis Vinifera L. for Third Generation Sequencing


 Background: There have been several attempts to sequence the genome of Vitis vinifera L. (grapevine), utilizing low-resolution second-generation platforms. Nevertheless, the characterization of the grapevine genetic resources and its adaptation to vulnerable conditions could be better addressed through extensive and high-resolution genome sequencing.MinION is a third-generation sequencer preferred by many laboratories due to its relatively low cost, ease of use and small size. Even though this long-read technology has been rapidly improving, to reach its full potential requires high-quality DNA.Results: Here we establish a workflow for DNA extraction suitable for MinION sequencing long reads from grapevine. This protocol was tested with leaf samples from different positions on annual growing branches of grapevine, Purified nuclei from fresh young leaves that led to high quality, long DNA fragments, suitable for long-read sequencing were successfully generated. It is evident that longer reads in grapevine associate with both fresh tissue and adjusted conditions used for nuclei purification.Conclusions: We propose that this workflow presents a significant advancement for long-read quality DNA isolation for grapevine and likely other plant species.


77
Here we establish a workflow for DNA extraction compatible with MinION sequencing for long reads 78 from grapevine leaves. To fully benefit from nanopore sequencing platforms by obtaining longer sequencing read 79 lengths, DNA fragmentation needs to be minimized. Currently, sample preparation still requires optimization 80 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.435003 doi: bioRxiv preprint before sequencing, which can be challenging. In grapevine, the yield and quality of DNA can be significantly 81 compromised by polyphenols, polysaccharides, and proteins, which are abundant during different stages of berry 82 and leaf development (Iandolino et al., 2004). ONT proposed a method to obtain long-read sequencing by 83 integrating nuclei isolation with Nanobind DNA isolation (Circulomics Inc.) (Workman et al., 2018). Nanobind 84 is a magnetic disk covered with a high-density micro and nanostructured silica. This technology yields high 85 molecular weight genomic DNA (50-300 kb) from nuclei; it was successfully used in Sequoia sempervirens and 86 Zea mays (Workman et al., 2018). Our main objective here was to optimize the workflow for grapevine, to get 87 high quality, long DNA fragments, suitable for long-read sequencing. We foresee that the pipelines we suggest 88 herein will apply to a large number of species.  centrifugal forces (herein 2500g, 4000g, and 5000g). The supernatants were discarded and 1 mL cold NIB was 104 added to each pellet, and then resuspended with a paintbrush presoaked in NIB; suspensions were transferred in 105 a new tube. Volumes were adjusted to 15ml by ice-cold NIB and the samples were centrifuged at 5,000g, 2,500g 106 or 4,000g for 15 min at 4 °C. This step was repeated 3-4 times until the supernatants were colourless. Next, pellets 107 were resuspended in 1 mL 1X Homogenization Buffer (HB: prepared according to Workman et al. 2018) and 108 transferred to 1.5ml microcentrifuge tubes. Samples were either snap frozen or used directly. For long term 109 storage, samples were spun down at 2,500 g for 5 min, supernatants were discarded and pellets were snap-frozen 110 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

123
The supernatants were transferred to 1.5 mL Protein LoBind microcentrifuge tubes using a wide bore pipette.

124
Nanobind disks were added to the supernatants followed by addition of a 1X volume of isopropanol. The

135
According to the manufacturer, 3 to 20 fmoles of DNA are required for Flonge Cells. Libraries were prepared 136 using the 1D ligation sequencing kit (SQK-LSK109). Briefly, the workflow consisted of first an end-repair/dA-137 tailing step to repair blunt ends and to add an "Adenine" to the 3' end of the amplicon, followed by adapter and 138 motor protein ligation onto the prepared ends using NEB ligase (New England Biolabs). Subsequently, a bead-139 based purification step was used to enrich the adapter-ligated fragments and remove excessive nucleotides and 140 enzymes. Finally, the library mixture was loaded into the Flongle flow cell and the MinKNOW GUI (version 141 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.435003 doi: bioRxiv preprint 6 19.06.8) was used for data acquisition as per recommendation from ONT. Each sequencing experiment was run 142 for 24 hours (>60 active pores) or after a minimum of 500 reads were acquired for downstream data analysis.

143
Promega Biomath Calculator (https://worldwide.promega.com/resources/tools/biomath/) was used to determine 144 the amount of DNA in fmoles, using the starting yield of nucleic acids and the estimation of N50 of the reads ca.

145
at 30-40 kb. The final yield of the libraries varied between 3-5 fmoles, which is the minimum amount required.

146
Therefore, for the construction of the second library (referred to as "Vitis 2", 4,000g), the starting yield of DNA 147 was increased 4x at 2.5μg. After the ligation with the reagents used in double quantities, the final yield was 0.63μg, 148 with 25% recovery, and was loaded in the Flongle.  To establish an efficient protocol for grapevine DNA purification compatible with 3 rd generation sequencing, we 160 tested different protocols for nuclei isolation. As initial attempts to purify nuclei at standard low (2,000g) or high 161 (6,000g) relative centrifugal forces (g=timesxEarth's gravitational force) did not yield nuclei, we assumed that the 162 parameter g may be a limiting factor when it comes to nuclei purification and should be experimentally 163 determined. Nuclei from different species, organs, tissues and cell ages show significant differences in their size 164 and ploidy and thus mass. We thus further tried three different nuclei purification schemes, with variable 165 centrifugal force (i.e. 2,500, 4,000 ("Vitis 2") and 5,000 g ("Vitis 1"). We determined that the optimal scheme for 166 nuclei purification was Vitis 2, while Vitis 1 yielded nuclei that were mechanically perturbed (Fig. 1)

178
Next, we evaluated the performance of both sequencing runs. To this end, we performed read counts, 179 read length-average base-call quality score comparison, and total Gbs sequenced for comparing the two runs, Vitis 180 1 and Vitis 2 (Fig.4). From these results, it is evident that longer reads in grapevine associate with both fresh tissue               (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.