Improved CTAB method for RNA extraction of thick waxy leaf tissues from sago palm ( Metroxylon sagu Rottb.)

Background: There is a growing interest in transcriptomics studies parallel to the advancement of transcriptome databases and bioinformatics, which provided the opportunity to study responses to growths, stimuli and stresses. There is an increase in demand for excellent RNA extraction techniques. General RNA extraction protocols can be used in RNA extraction, but the quality and quantity vary in different types of tissues from different organisms. Hence, a specific RNA extraction method for each organism’s tissue type is required to obtain the desired RNA quality and quantity. Results: The improved CTAB RNA extraction method is superior to the PCI method and MRIP method for thick waxy leaves that were applied for mature sago palm ( Metroxylon sagu Rottb.) leaf tissue and produce total RNA extract with good purity (OD 260/280 ≥ 1.8, OD 260/230 ≥ 2.0) and integrity (RIN ~ 7). RNA sequencing was conducted with the extracted samples and showed good assembly results (Q20 ≥ 97, Q30 ≥ 91%, assembly mean length ≥ 700 bp). Conclusion: The improved CTAB RNA extraction method enables rapid, cost‑effective, and relatively simple RNA extraction from waxy, fibrous and high‑in‑polyphenol sago palm ( M. sagu Rottb.) leaf tissue with next‑generation RNA sequencing recommended quality.


Introduction
Sago palm (Metroxylon sagu) is one of the starchproducing crops and is the best starch-producing species within the Metroxylon spp. which could be found from southern Thailand to the Solomon Islands [1,2]. Among the Metroxylon spp. only M. sagu is both hapaxanthic (once-flowering) and soboliferous (produce suckers) [3]. M. sagu lifecycle is divided into four stages which is the rosette stage, bole forming stage, inflorescence stage and fruit ripening stage where the starch accumulation in the trunk starts at the bole formation stages and reaches its peak at the beginning of the inflorescence stage which is approximately nine years after plantation [3]. The advantages of sago palm are its abilitiy to adapt to acidic, saline soil with seasonal flooding such as peat swamps which other crops are unable to grow [4]. There is growing interest in this palm as one of the candidates to combat the world food shortage.
The major differences between the improved CTAB method based on Untergasser [42] with modified Zeng and Yang [30] and modified Gasic et al. [34] are the extraction buffer is not pre-heated to 55 °C-65 °C and the precipitation using ice-cooled isopropanol for 10 min instead of overnight precipitation with lithium chloride. This change minimized the damage to the RNA due to high temperature and the extraction complete In this paper, we describe a validated method for the extraction of RNA for sago palm (Metroxylon sagu Rottb) leaf tissue for RNA sequencing (RNA-seq) purpose modified from Untergasser [42] RNA Miniprep using the CTAB method [42]. These procedures demonstrated the successful extraction of RNA from sago palm leaf tissue for RNA sequencing. The procedures described here are rapid, cost-effective and relatively simple.

Sampling
Sago palm leaf samples were wiped with 70% ethanol to remove debris, stored in labeled containers and snapfrozen in liquid nitrogen on-site to preserve the RNA. The samples need to be kept in liquid nitrogen before long-term storing in a − 80 °C freezer. Selection of the leaf age influences the total RNA yield and purity, in which young leaf gives better RNA yield and purity. In this study, mature sago palm leaf was selected in compliance with the motif of the main study.

Total RNA extraction
Three different extraction methods were compared in this study, i.e., improved CTAB method, PCI method [23] and MRIP method [41], to determine the most suitable method for extraction of total RNA from Metroxylon sagu leaf tissue for RNA sequencing ( Table 1). The improved CTAB method is a modified method based on Untergasser [42] which focus on the simplicity, rapidity and reproducibility of RNA extraction from mature sago palm leaf     sample. PCI method is an RNA extraction method practice on sago palm young leaf sample while MRIP method is a published RNA extraction method for young coconut palm leaf which shows promising results. One gram of mature sago palm leaf samples was grounded finely in prechilled mortar and pestle with liquid nitrogen for all three methods. The ribonucleic acid pellet obtained, which contains both DNA and RNA, from the three protocols were further subjected to DNase treatment. The pellet was resuspended with 0.5 ml of Tris-HCl buffer. Next DNase treatment was conducted with RQ1 RNase-Free DNase (Promega) according to protocol. Then, the mixture was centrifuged at 16,100 g, 4 °C for 15 min. After that, the mixture was transferred into a new 1.5ml microcentrifuge tube. Then, an equal volume of icecooled isopropanol was added, mixed and incubated at room temperature for 10 min. After that, the mixture was centrifuged at 16,100 g, 4 °C for 15 min. Then, the supernatant was discarded and the pellet was washed with 1 ml of 70% ethanol twice. Lastly, the pellet was resuspended with 100 µl Tris-EDTA buffer (10 mM Tris-HCl, 1 mM disodium EDTA, pH 8.0) for further analysis.

Improved CTAB method
This method was modified based on RNA miniprep using the CTAB method [42] ( Table 2). One gram of the powdered sago palm leaf sample was transferred into a 50-ml polypropylene centrifuge tube. Then, 15 ml of CTAB buffer with 1% (v/v) β-mercaptoethanol was added, mixed and then incubated for 5 min. After that, 15 ml of chloroform was added into the mixture, mixed and then incubated for another 5 min. The mixture was centrifuged at 20,922 g, 4 °C for 5 min. Next, the top aqueous layer from the tube was transferred into a new polypropylene centrifuge tube. Then 15 ml of chloroform was added into the aqueous layer tube, mixed and then incubated for 5 min. The mixture was centrifuged

PCI method
This method was adapted based on the RNA extraction method performed by Hussain [23]. One gram of the powdered sago palm leaf sample was transferred into a 50-ml polypropylene centrifuge tube.

MRIP method
This method was adapted based on the RNA extraction method performed by Xiao et al. [41]. One gram of the powdered sago palm leaf sample was transferred into a 50-ml polypropylene centrifuge tube. Then, 15 ml MRIP buffer [3.05 g ammonium thiocyanate (% w/v), 9.44 g guanidine thiocyanate (% w/v), 3.33 ml 3 M sodium acetate (pH 5.2) (% v/v) and 38 ml phenol (% v/v), adjust pH to 5.0 with acetic acid] was added, mixed and incubated for 5 min. Then, 5 ml ice-cooled chloroform was added, mixed and further incubated for 5 min. After that, the top aqueous layer from tube was transferred into a new 50-ml polypropylene centrifuge tube. Next, equal volume of ice-cooled isopropanol were added, mixed and incubated for 10 min. Then, the mixture was centrifuged at 20,922 g, 4 °C for 15 min. The supernatant was discarded and the pellet was washed with 70% ethanol twice. Lastly, the pellet was resuspended with appropriate buffer for further analysis or storage.

RNA purity test
RNA purity test was performed using Eppendorf Biophotometer Plus for RNA extraction methods comparison and Nanodrop was used to test the RNA quality for RNA sequencing samples. Samples pellet were dissolved in 100 µl Tris-EDTA and transferred into Eppendorf UVette cuvette for the instrument measurement in Eppendorf Biophotometer Plus while Samples pellet were dissolved in nuclease-free water for nanodrop. The purity of the sample's RNA extract was measured by the OD260/OD230 and OD260/OD280 ratio where the higher the value of the ratio the purer the sample's RNA [43].

RNA quality test
RNA quality test was performed using Tris-borate-EDTA buffered Agarose Gel Electrophoresis (TBE-AGE) for RNA extraction methods comparison and Agilent 2100 Bioanalyzer was used to test the RNA quality for RNA sequencing samples. In TBE-AGE, sample pellets were dissolved in 100 µl Tris-HCl. 5 µl of the RNA samples was mixed with 2 µl loading dye. The samples mixtures and GeneRuler 1-kb DNA ladder were loaded into the gel to perform the electrophoresis. In TBE-AGE, good-quality RNA sample represented by intact 28 s and 18 s rRNA band without smearing. In Agilent 2100 Bioanalyzer, sample pellets were dissolved and analyzed. RNA Integrity Number (RIN) is use to determine the quality of the RNA sample scored from 1 (worst) to 10 (best) [44].

Statistical analysis
IBM SPSS Statistics 28.0.1.1 (15) was used to analyse the 260/280 ratio, 260/230 ratio and RNA weight data obtained for RNA extraction method comparison. All the data were assessed using normality test to determine its parametricity. Parametric data were further subjected to multivariate analysis of variance (MANOVA) coupled with Tukey's test while non-parametric data were subjected to Kruskal-Wallis test.

RNA sequencing
The improved CTAB method was used to extract RNA from sago palm leaf tissue and sent for sequencing. The RNA samples were pelletized in 100% absolute ethanol and shipped for sequencing service. BGISEQ-500 sequencing platform was used in this study. The RNA Fig. 6 Test results of sample N2 by Agilent 2100. Source: Adapted from BGI Pre-sequencing Report sequencing data were measured by Q20 and Q30 values. BGI NGS RNA sequencing guideline for goodquality data are Q20 ≥ 90%, Q30 ≥ 80 [45]. The assembly sequences median length (N50) and the mean length provides information on abundance of the sample RNA length, the closer the N50 value to the mean length the more normalized the distribution of the sample RNA length.

RNA extraction methods comparison
The improved CTAB method was compared with the PCI method [23] and MRIP methods [41] and showed the best result in RNA quality and quantity (Table 3 and Fig. 1). Due to low RNA yield of MRIP method, there are no visible bands in TBE-AGE. On the other hand, there are presence of visible smear at the 5 s rRNA region, but absent at 28 s and 18 s rRNA band in TBE-AGE of PCI method indicating the considerable degradation of the RNA. CTAB method gives the best RNA quality as shown by the intact 28 s and 18 s rRNA band with slightly visible smear at the 5 s rRNA region in TBE-AGE. This outcome shows that the PCI method is not suitable for extracting RNA from mature sago palm leaf tissue. The lower sample weight used in this test improved the RNA yield and purity of the extracted RNA compared to the reported result. This outcome also shows that MRIP method is not suitable for RNA extraction from mature sago palm leaf tissue although it shows promising result when applied on RNA extraction from young coconut leaf. The purity and the RNA yield are very much lower compared to the reported result on application of MRIP method on young coconut leaf.
The 260/280 ratio, 260/230 ratio and the RNA weight data were subjected to normality test using IBM SPSS Statistics 28.0.1.1 (15). The null hypothesis (H 0 ) for the normality test assumes the population is normally distributed and the standard significance level (α) is 0.05. The P-values of 260/280 ratio and 260/230 ratio data are higher than α so we accept H 0 , hence they are normally distributed (Table 4). On the other hand, P-value of RNA weight is lower than α so we reject H 0 , hence the data are not normally distributed. Since both 260/280 ratio and 260/230 ratio data are normally distributed, they were subjected to MANOVA coupled with Tukey's test while because of the RNA weight data are not normally distributed, they were subjected to independent-samples Kruskal-Wallis test using IBM SPSS Statistics 28.0.1.1 (15). The H 0 for the Tukey's test assumes all means being compared are from the same population and the standard significance level (α) is 0.05. The P-values between all the RNA extraction methods are lower than the α (0.05) value so we reject H 0 , hence the 260/280 ratio and 260/230 ratio data between all the RNA extraction methods are not the same ( Table 5). The H 0 for the Kruskal-Wallis test assumes the distribution of the weight is the same across the RNA extraction methods and the standard significance level (α) is 0.05. The P-values of all the Table 8 RNA sequencing clean reads quality metrics from sequencing report. Source: Reproduced from BGI RNA-Sequencing Report   RNA extraction methods are higher than the α (0.05) value so we accept H 0 , hence distribution of the weight is the same across the RNA extraction methods ( Table 6). The improved CTAB method gives the best RNA yield with an average of 177.5 µg RNA for 1 g mature sago palm leaf tissue while PCI method and MRIP method only gives 11.17 µg and 7.97 µg RNA, respectively, even though the difference is not significant between RNA extraction methods. The RNA purity of the CTAB method for protein contamination as shown by the 260/280 ratio was significantly higher with an average of 1.93 than the PCI method and MRIP method with an average of 1.635 and 1.18, respectively. The RNA purity of the CTAB method for other organic compounds contamination as shown by 260/230 ratio was also significantly higher with an average of 1.955 than the PCI method and MRIP method with an average of 0.86 and 0.2, respectively. Hence, the improved CTAB method is superior for RNA extraction on mature sago palm leaf samples compared with the PCI method and MRIP method.

RNA extraction of mature sago palm leaf tissue by improved CTAB method
Pre-sequencing RNA sample quality and quantity tests were performed on the samples with NanoDrop and Agilent 2100 (refer Table 7 and Figs. 2, 3, 4, 5, 6, 7). Note that the RNA yield by the sequencing service provider is underestimating the actual RNA yield because the sample pellet was not fully dissolved during pellet resuspension by the sequencing service provider (data on actual RNA yield before pelletized and shipped for RNA sequencing are provided as Additional File 1).
There is high intensity of 5 s and some consistent signals pattern at the 18 s and 5 s fast region across the samples which is not ideal compared to the RIN standard but the outcome of the sequencing results showed otherwise. RNA sequencing of the extracted sago palm leaf tissue RNA by improved CTAB method gives a good result as all the samples exceed good-quality data definition for next-generation sequencing where Q20 ≥ 90%, Q30 ≥ 80% (refer Table 8). The de novo assembly of the sequenced RNA transcripts shows mean length > 700 bps and N50 > 1300 bps (refer Table 9) can be used as a guideline for future RNA sequencing of sago palm RNA extracts. The high intensity of the 5 s and some consistent signals pattern at the 18 s and 5 s fast region could indicate the RNA profile of mature sago palm leaf tissue and do not represent the degradation of the RNA sample. The below-average sample T1 RIN and 28 s/18 s ratio do not affect the sequencing result much as the clean read ratio, clean read Q30, mean assembly length and N50 values are comparable with other samples.

Conclusion
In this study, three RNA extraction methods, i.e., improved CTAB, PCI and MRIP, were compared and evaluated based on the best RNA purity and quality extracted from sago palm leaf tissue. Overall, the best result of RNA purity, quantity and quality of RNA obtained are by the improved CTAB method. The total RNA extracts using the improved CTAB method were then sequenced and this method generates good sequencing data Q20 ≥ 97, Q30 ≥ 91. The mean length ≥ 700 bp and the N50 value ≥ 1.3 kbp of the assembly can be used as a guideline for future RNA sequencing in sago palm.