A universal method for high-quality RNA extraction from plant tissues rich in starch, proteins and fiber

Using existing protocols, RNA extracted from seeds rich in starch often results in poor quality RNA, making it inappropriate for downstream applications. Though some methods are proposed for extracting RNA from plant tissue rich in starch and other polysaccharides, they invariably yield less and poor quality RNA. In order to obtain high yield and quality RNA from seeds and other plant tissues including roots a modified SDS-LiCl method was compared with existing methods, including TRIZOL kit (Invitrogen), Plant RNeasy mini kit (Qiagen), Furtado (2014) method, and CTAB-LiCl method. Modifications in the extraction buffer and solutions used for RNA precipitation resulted in a robust method for extracting RNA in seeds and roots, where extracting quality RNA is challenging. The modified SDS-LiCl method revealed intense RNA bands through gel electrophoresis and a nanodrop spectrophotometer detected ratios of ≥ 2 and 1.8 for A260/A230 and A260/A280, respectively. The absence of starch co-precipitation during RNA extraction resulted in enhanced yield and quality of RNA with RIN values of 7–9, quantified using a bioanalyzer. The high-quality RNA obtained was demonstrated to be suitable for downstream applications, such as cDNA synthesis, gene amplification, and RT-qPCR. The method was also effective in extracting RNA from seeds of other cereals including field-grown sorghum and corn. The modified SDS-LiCl method is a robust and highly reproducible RNA extraction method for plant tissues rich in starch and other secondary metabolites. The modified SDS-LiCl method successfully extracted high yield and quality RNA from mature, developing, and germinated seeds, leaves, and roots exposed to different abiotic stresses.


Materials and methods
Two winter wheat (Triticum aestivum L.) genotypes Tascosa and Tx86A5606 with contrasting responses to high night temperature (HNT) stress were used to isolate RNA from different plant tissue and growth conditions. RNA was isolated from physiologically mature wheat seeds collected from field grown plants 18 , flag leaves from controlled environment chamber grown plants exposed to HNT and control conditions 19 , germinated seeds under cold stress, roots of wheat seedlings grown on Murashige and Skoog (MS) media exposed to freezing stress. Our modified SDS-LiCl method was extensively compared to other currently available methods developed by public institutes and private companies. RNA was also isolated from plant samples exposed to cold, freezing and HNT stress to demonstrate that the efficiency of the new method can be extended to plants exposed to different abiotic stress conditions. In addition, the method was adopted for extracting RNA from developing and mature seeds of field-grown maize and sorghum. Details of the sample collection and stress imposition and associated references are detailed below.
Tissue sampling from control and high night temperature (HNT) stress exposed plants. Ma-ture seeds. Physiologically mature wheat seeds were collected from field grown plants, 21 days after 50% flowering. The field experiment was carried out at Kansas State University, Agronomy research farm at Manhattan, Kansas. High night temperature (HNT) stress was imposed by placing custom built field-based heat tents on already established wheat plants from 10 days after 50% flowering till maturity 18 . The heat tents maintained temperatures < 1 °C higher compared to ambient conditions during the day. Higher night-time (19:00 h to 6:00 h) temperature was achieved using a custom built semi-automated heating system operating independently in each tent during grain filling (For additional details on the construction, operation, and temperature imposition see Hein et al. 18 ). Physiologically mature seeds were collected from two winter wheat genotypes (Tascosa and Tx86A5606) grown under ambient night temperature (control; 22.7 °C) and an elevated night temperature of + 3.2 °C (HNT stress) during the grain filling, frozen in liquid nitrogen, and stored at − 80 °C for RNA extraction.
Developing seeds. Developing wheat seeds were collected from controlled environment chamber grown plants at 14 days after heading/HNT stress exposure, from control and HNT treatments. Two winter wheat genotypes, SY Monument (Tolerant) and KS07077M-1 (Sensitive) contrasting for HNT stress were exposed to control (28/15 °C) and a range of HNT treatments (28/18 °C; 28/23 °C; 28/25 °C and 28/27 °C) using the controlled environment chamber facility, at the Department of Agronomy, Kansas State University (For additional details on the high night-time temperature stress imposition see Impa et al. 20  Germinated wheat seeds exposed to cold stress. Seeds of the two wheat genotypes (Tascosa and Tx86A5606) were germinated in plastic Petri dishes (100 mm by 20 mm). Seeds were sterilized with 70% ethanol for 2 min, followed by 3% sodium hypochlorite for 5 min. Subsequently, seeds were rinsed 2 to 3 times with sterile distilled water before placing them on the germination paper. Ten seeds were placed in each petri dish containing the germination paper moistened with sterile distilled water. The Petri dishes were incubated in an incufridge (RevSci, Minnesota, USA) maintained at 15 °C to impose cold stress, and at 30 °C for control treatment 21 . The germinated seeds were collected five days after germination was initiated and immediately frozen in liquid nitrogen and stored at − 80 °C for RNA extraction. The sample collection was timed to ensure that each germinating seed had five days of continued cold stress after initiation of germination. Wheat roots exposed to freezing stress. Seeds  CTAB-LiCl method: RNA was extracted based on the method described by White et al. 3 . In this method, ionic detergent cetyltrimethylammonium bromide (CTAB) was used in the extraction buffer, and LiCl was used to precipitate RNA.
New Modified method (SDS-LiCl): This method follows the SDS-Phenol based extraction and RNA precipitation with sodium acetate and LiCl (Fig. 1). . The RNA quality of the sample was measured by monitoring the absorbance at A260/280 and A260/230 ratios ( Table 1). The purity and integrity of the total RNA bands were evaluated by loading 200 ng of RNA samples on 1% agarose gel stained with ethidium bromide and pictured under UV light in GEL DOC XR + and CHEMIDOC XRS + imaging Systems containing the IMAGE LAB software (Biorad, CA, USA). Further, the total RNA was analyzed with the Total RNA 6000 Nano Kit (Vers. II), precisely optimized for total RNA analysis with Agilent 2100 Bioanalyzer with RNA 6000 Nano Lab Chip (Agilent Technologies, CA, USA). The RIN (RNA integrity number) values were calculated using the Eukaryote Total RNA Nano Assay 2 . The electropherograms were analyzed using Agilent 2100 Expert B.02.06 SI648 software composed of data collection, presentation, and interpretation functions.  Relative gene expression was quantified using the 2 −ΔΔCT method as described in Livak and Schmittgen 22 , and BMY gene expression was normalized using Actin as a reference gene. A melt curve profile was incorporated in the thermal cycling method to check the specificity (with no gDNA contamination, no primer dimers, and no non-specific product) of the RT-qPCR reaction.

Results and discussion
SDS-LiCl method extracted higher yield and quality RNA from seeds. Previous methods using guanidine isothiocyanate based RNA extraction buffer results in starch solidification and co-precipitation of RNA with starch, reducing the quantity and quality of RNA 1, 2,23,24 . Modified SDS-LiCl method described here provides a robust method to obtain high-yield and high-quality RNA from wheat seeds containing high starch and other polysaccharides, and other tissues such as roots that are rich in fiber. In addition, this method has the potential to extract high yield and quality RNA from developing and mature seeds exposed to different abiotic stresses. Analysis of total RNA from mature wheat seed on agarose gel electrophoresis revealed that ribosomal RNA bands were intact and intense, demonstrating that the total RNA extracted by modified SDS-LiCl was not degraded (Fig. 2e). However, other methods produced faint bands of RNA, with higher impurities in the gel-well and a smeary background during the run (Fig. 2a-d). Absorbance spectra of RNA with a broad peak pattern at 260 nm ( Fig. 3) and ratios of A260/A280 and A260/A230 being close to or above 2 (Table 1) for all the RNA samples extracted using modified SDS-LiCl method. This demonstrates the RNA extracted to be free from protein contaminants, organic solvents, and other secondary products compared to other RNA isolation methods. Evaluation of total RNA quality using Agilent bioanalyzer 2100 revealed superior quality of RNA extracted using modified SDS-LiCl method with the electropherograms showing distinct, prominent peaks for ribosomal RNAs (rRNAs), whereas other methods exhibited either many small peaks and/or no protuberant peaks for ribosomal RNAs (Fig. 4 and Supplementary Figs. S1-S5). RIN values determined on Bioanalyzer (Agilent 2100) using Expert software for total RNA extracted using modified SDS-LiCl method ranged from 7.6 to 9.1 ( Table 2). The capillary gel electrophoresis virtual gel images (Agilent 2100 Expert B.02.06 SI648) revealed the intact rRNA bands of high quality RNA isolated by modified SDS-LiCl method compared to RNA extracted using other methods, except CTAB-LiCl method, which was equally effective in extracting high quality RNA (Supplementary Figs. S1-S5). Total RNA isolated with modified SDS-LiCl method produced significantly higher RNA yield (ranging from 19.41 to 37.24 μg/100 mg FW) compared to other methods including CTAB-LiCl method, which was on average 80% lower compared to modified SDS-LiCl method (Table 2). Similarly, lower RNA yield from wheat seeds using CTAB-LiCl has been reported 23 . The CTAB, as a cationic surfactant, favorably interacts with the anionic nucleic acids (DNA and RNA), including proteins and other biopolymers (polysaccharides and secondary metabolites). Therefore, due to the higher affinity of CTAB with both nucleic acids and other biopolymers, leads to reduced yield of RNA. On the other hand, anionic surfactant SDS does not interact with nucleic acids (anionic) but highly reactive with proteins and other polysaccharides and forms aggregates. Thus, SDS is more effective in separating RNA from other biopolymers during RNA extraction [25][26][27][28][29] . Moreover, CTAB-LiCl method has a lengthy procedure and is time consuming compared to the modified SDS-LiCl method. Higher RNA yield with the modified SDS-LiCl extraction method suggests the absence of starch interference during RNA precipitation. www.nature.com/scientificreports/ Components that make SDS-LiCl effective in extracting quality RNA from seeds and other tissue. Even though the extraction buffer used in the SDS-LiCl method contains NaCl, PVP, Tris/HCl, EDTA, β-mercaptoethanol, and SDS (see supplementary document), similar to existing methods, we optimized the concentrations of these chemicals for seeds and other tissues including roots. The chemical composition confirmed that 2.5 M NaCl, 2.5% PVP, 100 mM Tris, 25 mM EDTA, 2.5% β-mercaptoethanol, and 2% SDS, helps obtain high quality and yield of RNA without the interference of starch and other impurities. The enhanced yield and quality of RNA achieved in the modified SDS-LiCl method compared to CTAB-LiCl is due to the addition of SDS, and higher concentration of other chemicals (PVP and NaCl) in the extraction buffer. In agreement with previous studies, an increased concentration of NaCl, PVP, β-mercaptoethanol and SDS in the extraction buffer of the modified SDS-LiCl method, not only promotes cell lysis but also prevents RNAse and polyphenols in the isolated RNA 1,2,5,8,30-32 . In addition, post-grinding addition of SDS directly to the homogenate suppressed excessive foaming and facilitated proper lysis of seed samples and prevented starch solidification 1,2, 33 . Additionally, in the modified SDS-LiCl method, phase purification using acidic phenol (pH 4.3) saturated with citrate buffer helped to improve the RNA quality by separating RNA from bulk starch residues compared to CTAB-LiCl method. Efficient separation of bulk starch residues and degraded proteins during the centrifugation of RNA containing aqueous phase using acid phenol saturated with 0.1 M citrate buffer (pH 4.3) supports earlier findings 2 . A combination of 2 M Lithium chloride and 1/10 volumes of 3 M sodium acetate (pH 4.8) possibly facilitated better RNA precipitation. This is in agreement with previous studies wherein lithium chloride and sodium acetate were used to promote precipitation and resulted in high RNA yield 2,3,34 . Adding both lithium chloride and sodium acetate in the precipitation step enhanced the yield, whereas in CTAB-LiCl method, there was no phenol separation and LiCl alone was used for precipitation. Overall, the RNA obtained from modified SDS-LiCl method from field-grown wheat seeds under both control and HNT stress had high yield and quality, demonstrating a marked improvement compared to all other existing methods.
High quality RNA extracted using SDS-LiCl method suitable for downstream applications. PCR amplification of EPSPS gene full length CDS with 1.2 kb product size and a prominent band provides support that cDNA obtained from RNA using the modified SDS-LiCl method is high quality compared to other methods. In addition, it indicates that RNA integrity is sufficient to amplify large size gene fragments ( Supplementary Fig. S6). The RT-qPCR results of total RNA quantification show a tendency to obtain lower Cq values for the cDNA from RNA samples extracted by the modified SDS-LiCl method compared to other four different methods (Supplementary Table S1). When comparing the reproducibility expressed in terms of SEM values, the best results were obtained for samples extracted by the modified SDS-LiCl method. Comparatively, www.nature.com/scientificreports/ using other methods, high Cq values and SEM were obtained suggesting that RNA extracted using modified SDS-LiCl method resulted in high quality cDNA with lesser interference of contaminants. Quantitative real-time PCR (RT-qPCR) was performed with the cDNA synthesized from RNA isolated from mature seeds using the modified SDS-LiCl method. Amplification of all the samples with strait amplification curves and Ct values within the range permitted to perform meaningful analysis of expression data (Fig. 5, see Impa et al. 20 for published expression data on developing seeds using the modified SDS-LiCl method). The expression analysis of 21 genes encoding enzymes involved in the starch metabolism in developing wheat seeds exposed to a range of increasing night temperatures was carried out using the RNA extracted by the modified SDS-LiCl method (See Impa et al. 20 for published expression data). A single peak in the melt curve analysis and lower standard errors between the Ct values for a gene indicated the specificity of the primers to bind to the cDNA synthesized from the RNA, with no interference of PCR inhibitors (Fig. 5b).
Modified SDS-LiCl method equally effective on other tissues exposed to different abiotic stresses. Further, the modified SDS-LiCl method was used successfully to extract high-quality RNA from germinated wheat seeds, flag leaves, and root tissues that were exposed to cold, freezing and HNT stress. Agarose gel electrophoretic separation and Nanodrop UV spectrophotometer measurements of total RNA from different www.nature.com/scientificreports/ tissues of wheat genotypes exposed to different abiotic stresses were confirmed for RNA quality with A260/280 and A260/A230 ratios of ≥ 2 and a similar RNA yield to that of control samples ( Fig. 6; Table 3; Supplementary  Figs. S7-S9). Analysis of RNA samples by Bioanalyzer Agilent 2100 further confirmed high-quality RNA with prominent peaks of rRNA in electropherograms with a higher RIN values (> 7) for various wheat plant tissue RNA isolated by the modified SDS-LiCl method ( Fig. 7; Supplementary Figs. S10-S12). Our method reliably provided ample yield and high quality RNA from developing wheat seeds that were exposed to a range of increasing night temperatures (15-27 °C) under controlled environment chambers (Supplementary Figs. S13-S18; Supplementary Table S2; also see Impa et al. 20 for additional gene expression data). Previous studies have reported decreased concentration of RNA from the samples subjected to abiotic stress [12][13][14]35 . Similarly, in our study we observed less RNA concentration in the developing grain samples of KS07077M-1 subjected to HNT stress compared to control using the SDS-LiCl method (Supplementary Table S2). However, a similar trend was not observed with the RNA samples extracted using the SDS-LiCl method that were subjected to HNT or other abiotic stresses. Also, RNA extracted using other methods (Ambion TRIZOL, RNeasy Plant Mini Kit (Qiagen), Furtado 6 ) did not differ between control and HNT stress samples due to interference of starch. Similarly, an independent study recorded no difference in the RNA yield of wheat leaf tissue subjected to freezing stress 13 . Previous studies have reported that reduced concentration of RNA and interference of secondary metabolites in RNA extraction depends on the severity of stress and tissue age 12 also the type of tissue 13 . Hence, the lack of www.nature.com/scientificreports/  www.nature.com/scientificreports/ a significant impact on RNA concentration in this study could be due to the short duration of exposure under different stresses or due to moderate HNT stress 18 levels.
Modified SDS-LiCl method also yielded high quality RNA from other cereals seeds (mature and developing maize seeds and developing sorghum seeds) rich in starch, protein, and polyphenols (Supplementary Figs. S19, S20). RNA extraction protocols and commercial kits are limited to specific species and/or tissue 3,17,36 , whereas the  Fig. 3. The cropped images were auto corrected for the same brightness/contrasts using PowerPoint image tools. Table 3. Nano-spectrophotometric analysis of RNA yield, purity and Integrity Numbers (RIN) using Bioanalyzer (Agilent 2100) for the total RNA isolated using the modified SDS-LiCl method from germinated wheat seeds, flag leaf and root tissue exposed to different abiotic stresses. Yield and quality were assessed by absorption spectra and ratios of A260/A280 and A260/A230. Tx-Tx86A5606. a Values represent the mean ± SD from at least three technical samples.

Germinated seeds
Tx control ( www.nature.com/scientificreports/ modified SDS-LiCl method was suitable for extracting high quality RNA from various crop tissues differing in developmental stages and exposed to different abiotic stresses under controlled environment and field conditions. In summary, the modified protocol was exclusively developed for extraction of RNA from cereal seed tissue containing high starch, as it is one of the major hurdles challenging functional studies involving developing or mature seeds. This protocol is cost-effective compared to commercially available kits and has been demonstrated to be successful in obtaining quality RNA from mature wheat grains, while TRIZOL, CTAB, and other kits generally failed. The robustness of the modified SDS-LiCl method helped to extract significantly higher yield and quality of RNA from different wheat plant tissues, including, mature, developing and germinated seeds, leaves and roots, exposed to various abiotic stresses, whereas previously published protocols have limited the protocol to either seed tissues of wheat or other cereals 1,2,5,24 or only leaf tissues 3 . The protocol standardized here has been modified to be a rapid yet reliable, using basic laboratory chemicals that are easily accessible by researchers, whereas previously methods needed more chemicals in the extraction buffer, or expensive reagents kits 1,2,5 . Besides, the versatility of the universal protocol is demonstrated by extracting high-quality RNA from other cereal crop seeds. A comparative assessment including the advantages of the SDS-LiCl method compared to previous protocols is presented in Table 4.

Conclusion
In conclusion, we have optimized a robust method for extracting high yield and quality RNA from wheat seeds, free from starch and other contaminates. The obtained RNA quality was appropriate for downstream applications demonstrating the effectiveness of the protocol with seeds, that contain high starch and polysaccharides. The modified SDS-LiCl method extracted significantly higher yield and quality of RNA from different wheat tissues, including mature, developing and germinated seeds, leaves and roots exposed to different abiotic stresses. Besides, the versatility of the protocol was further strengthened with high quality RNA extracted from field grown maize and sorghum seeds.

Data availability
All related data not presented in the manuscript are made available in the supplementary documents.