Acta Physiologiae Plantarum

, Volume 31, Issue 4, pp 871–875

Improvement of an RNA purification method for grapevine (Vitis vinifera L.) suitable for cDNA library construction


    • Laboratoire de Physiologie Moléculaire de la Vigne
    • RLP AgroScience GmbH, AlPlanta
  • Synda Chenenanoui
    • Laboratoire de Physiologie Moléculaire de la Vigne
    • RLP AgroScience GmbH, AlPlanta
  • Ahmed Mliki
    • Laboratoire de Physiologie Moléculaire de la Vigne
  • Michael Höfer
    • RLP AgroScience GmbH, AlPlanta
Short Communication

DOI: 10.1007/s11738-009-0309-0

Cite this article as:
Daldoul, S., Chenenanoui, S., Mliki, A. et al. Acta Physiol Plant (2009) 31: 871. doi:10.1007/s11738-009-0309-0


A method is described, which consistently yields high quality total RNA from grapevine. Dissolving of crude RNA pellets in borate-containing buffer, instead of normally used water before a selective lithium chloride precipitation, was found to be a critical step, leading to a 2.5-fold increase of yield. The resulting RNA preparations were suitable for standard downstream applications and also for cDNA library construction. The method worked efficiently and reproducibly and could easily be scaled from milligram to gram quantities of plant material grown in hydroponic culture, sandy soil or Perlite. It was applied to different kinds of grapevine tissues (leaves, stem) and, after additional adaptation of the protocol, to roots.


cDNA library constructionGrapeRNA extractionTris–borateVitis vinifera L.



Diethyl pyrocarbonate


Ethylene diamine tetra-acetic acid


Elongation factor one alpha


Fresh weight






Reverse transcription-polymerase chain reaction


Standard deviation


Sodium dodecyl sulphate


Grapevine (Vitis vinifera L.) is among the most widely cultivated perennial plant species in the world, and viticulture and winemaking are of cultural and high economical importance. Due to the recent unravelling of the grapevine genome (Jaillon et al. 2007), this plant may become a model for fruit tree genetics, and the foundation for the improvement of quality, yield, and biotic and abiotic stress tolerance by biotechnology approaches has been laid (Troggio et al. 2008). Consequently, grapevine research requires studies of gene expression and function, which are highly dependent on RNA quality. However, grapevine is well known for its high content of interfering substances, which prevent the application of standard RNA isolation protocols. During regular development and especially when exposed to abiotic stresses such as drought and salt stress, unfavourable secondary metabolites like polysaccharides and polyphenols, particularly flavonoids, accumulate significantly (Boss and Davies 2001; Winkel-Shirley 2002). High concentrations of polymeric carbohydrates greatly reduce the yield of RNA from plant tissues. Those contaminants could render the RNA unusable for many molecular techniques such as cDNA library construction or gene expression profiling. It is therefore a major challenge to obtain sufficient amounts of high-quality RNA from plants like grapevine, especially when cultivated under abiotic stress conditions. Consequently, an efficient protocol for RNA extraction is a necessary prerequisite for many downstream applications (Gasic et al. 2004; Iandolino et al. 2004; Fleige and Pfaffl 2006).

In this paper, we describe an improved RNA extraction protocol originally reported by Renault et al. (2000), which provides highly intact total RNA from grapevine grown in hydroponic culture, sandy soil or Perlite and subjected to different abiotic stresses.

Materials and methods

Plant material

Vitis vinifera L. cv. Razegui and Cardinal (drought and salt tolerant), Syrah (salt sensitive) and Galb Sardouk (drought sensitive) were selected, which exhibit contrasting phenotypic responses to abiotic stress. Plants were grown individually in pots either in sandy loam soil, hydroponics or Perlite under greenhouse conditions (16 h light period at a photosynthetic active radiation of at least 300 μE m−2 s−1, temperature varied between 25 and 28°C and relative humidity ranged from 50 to 70%). Plant material (leaves, stems and roots) was harvested after 6 months of culture and immediately frozen in liquid nitrogen.

Before harvest, plants used for RNA isolations were subjected either to 3 weeks of water deficiency in sandy soil or salt stress (100 mM NaCl) in hydroponic culture, or were unstressed. All plants were healthy and were without any visible sign of leaf senescence. Three independent repetitions of the growth experiments were performed for RNA isolation and gene expression analyses.

RNA purification

All glassware used for the isolation of RNA was baked for 2 h at 180°C. All solutions were subjected to DEPC treatment except for Tris-containing buffers. All chemicals used were of p.a. grade and supplied by Carl Roth GmbH (Karlsruhe, Germany) and Sigma-Aldrich Chemie GmbH (Taufkirchen, Germany) except otherwise indicated.

Plant tissue was ground to a fine powder at the temperatures of liquid nitrogen using a mixer mill (MM200, Retsch, Haan, Germany). As much as 10 g of the frozen powder was added to 100 ml of extraction buffer (200 mM Tris–HCl pH 8.5, 300 mM lithium chloride, 10 mM EDTA 1.5% (w/v) SDS, 2.0% (w/v) PVPP) and mixed with 80 ml of chloroform for at least 5 min. After centrifugation at 14,300g (Sorvall RC5C-Plus, FiberLite® F10) for 25 min at 15°C, the supernatant was extracted with one volume of P:C:I three times and recovered after centrifugation as described above. After addition of 0.5 volumes of cold ethanol (AppliChem GmbH, Darmstadt, Germany), polysaccharides were precipitated by centrifugation at 14,300g for 20 min at 4°C. The supernatant was mixed with sodium acetate, pH 5.2, to a final concentration of 150 mM and one volume of cold 2-propanol. After overnight incubation at −20°C, nucleic acids were collected by centrifugation at 14,000g for 35 min at 4°C. The air-dried pellet was dissolved in Tris–borate buffer (80 mM Tris, 10 mM EDTA, pH 8.0 adjusted with crystalline boric acid) according to the following rule: for leaves and stems 4 ml and for roots 2 ml per g plant material initially used for homogenisation. RNA was precipitated with 2.5 M lithium chloride and resuspended in DEPC-treated water after centrifugation at 20,000g (Sorvall RC5C Plus, FiberLite® F13) for 35 min at 4°C.

RNA analysis

Total RNA was analysed by 1.2% agarose gel electrophoresis and ethidium bromide staining. Statistical analyses of yield were computed using UNIVARIATE (to check normal distribution of data) and MIXED (for analysis of variance) of the SAS 9.2 program package (SAS Inc., Cary, NC 27513, USA).

One step reverse transcription-polymerase chain reaction

As much as 200 ng of total RNA template was reverse transcribed in a reaction volume of 25 μl by means of SuperScript II one step RT-PCR system with Platinum Taq DNA polymerase (Invitrogen, Karlsruhe, Germany) according to the manufacturer’s instructions. V. vinifera elongation factor one alpha (EF1α; TIGR plant transcript assembly TA36627_29760)-specific primers (0.2 μM final concentration each, forward primer 5′-CAGTGGGACGTGTGGAGACTGGTG-3′, reverse primer 5′-GGGCATAGCCATTTCCGATCTGAC-3′), which amplified a product of 288 bp derived from one exon, were used. cDNA synthesis was performed at 50°C for 30 min and subsequent PCR as follows: denaturation at 94°C for 2 min, amplification at 94°C for 15 s, 65°C for 30 s and 68°C for 45 s for 25, 30 and 35 cycles followed by a final extension step at 68°C for 5 min. PCR without RT reaction was performed by means of Platinum Taq DNA polymerase (Invitrogene, Karlsruhe, Germany). Amplification products and DNA size marker (λ PstI-ladder, MBI) were separated on 1.5% agarose gel and visualised under UV light after ethidium bromide staining.

Results and discussion

Optimisation of RNA purification from leaf and stem material

Recently, a comparison of several RNA extraction protocols revealed that Tris–lithium chloride extraction buffer-based methods are most suitable for grapevine (Tattersall et al. 2005). Based on this result, a protocol published by Renault et al. (2000) was tested with leaf material from grapevine. Initial results of moderate quality (A260/280 = 1.7 ± 0.1SD) and insufficient yield (118 ± 35SD μg/g FW) indicated that optimisation of the procedure was essential.

The ratio of volume of homogenisation buffer to fresh weight of tissue was raised from 5 to 10 ml/g, which decreased the viscosity of the homogenate and finally improved the purity of the RNA (A260/280 = 2.1 ± 0.01SD) by minimising the carry-over of impurities from the prominent interphase during phenol extraction. Hence, the number of P:C:I extraction steps could be reduced to three instead of four as described in the original protocol. However, since RNA yield was still not sufficient, the resulting pellet obtained after the 2-propanol precipitation step was resuspended at 4°C in Tris–borate buffer, pH 8.0, instead of using RNase-free water as a solvent. As a result, approximately a 2.5-fold increase in yield was observed (Fig. 1) when compared to the results obtained with the original protocol of Renault et al. (2000). This could be due to the borate, which is known to form complexes with polyphenols and carbohydrates, resulting in an enhanced solubility of these polymers to which nucleic acids are susceptible to bind (Manning 1991).
Fig. 1

Dependency of yield of total RNA on the solvent used in an intermediate step of the isolation protocol using leaves of V. vinifera cv. Cardinal. Values are means of four independent preparations. Error bars represent ± SD

After introduction of the modifications described above, the yield and quality improved reliably. Northern blots performed to assess the integrity of the different batches of total RNA using an EF1α-specific probe always showed well-defined bands without any visible sign of degradation (data not shown). Statistical analysis by ANOVA of yield data derived from leaves of V. vinifera cv. Razegui (370 ± 77SD μg/g FW) and Syrah (385 ± 101SD μg/g FW), used for salt-stressed experiments revealed no significant differences between salt-stressed and control plants (at a significance level of P ≤ 0.05), which could be due to the moderate salt stress treatment. On the other hand, under drought stress, a considerable decrease of RNA yield ranging from 1.4 to 2.6-fold was observed (Fig. 2), similar to observations made in senescing leaves of tomato and poplar (Bhalerao et al. 2003). RNA prepared from leaf material was successfully used for RT-PCR and cDNA library construction in order to detect differentially expressed genes in moderately salt-stressed grapevine (Daldoul et al. 2008). Application of our protocol to woody stem material from grapevine grown under control conditions yielded on average 190 μg/g FW of total RNA compared to 80–120 μg/g FW, as reported by Thomas and Schiefelbein (2002).
Fig. 2

Comparison of RNA yield. a Yield of total RNA obtained from leaves of control (C) and 3 weeks drought-stressed (S) grapevine cv. Razegui (Ra) and Galb Sardouk (Ga). Mean values were calculated when possible and the number of isolations is indicated in brackets. Error bars represent ± SD

Adaptation of the protocol to root tissue

Application of the modified protocol to roots initially failed to result in acceptable amounts of RNA. In order to improve yield, decreasing volumes of Tris–borate buffer of 3, 2 or 0.4 ml per g FW of plant material, initially used for homogenisation (corresponding to 75, 50 and 10% of the original volume), were used for solubilisation of the crude pellet before lithium chloride precipitation (Fig. 3a). The resulting total RNA was tested by semi-quantitative RT-PCR. The size of the PCR product for EF1α in the gel corresponded to the predicted size. RNA samples consistently showed a poor amplification rate when solubilised in the lowest Tris–borate buffer volume (Fig. 3c). In PCR control experiments without reverse transcriptase reaction, accumulation of the EF1α-specific amplicon significantly lagged behind compared to the RT-PCR reaction, which ensured that the RT-PCR amplification product indeed originated mainly from the RT reaction (Fig. 3d). Apparently, conditions of 0.4 ml/g FW of Tris–borate buffer favoured the carry-over of contaminating compounds, which had a negative impact on the RT-PCR amplification efficiency (Wong and Medrano 2005). From our results, a volume of 2 ml/g FW of Tris–borate buffer proved, therefore, to be a good compromise between RNA recovery and quality (Table 1).
Fig. 3

Optimisation of RNA extraction from roots of grapevine. a Agarose gel electrophoresis of total RNA isolated from roots of hydroponically cultivated grapevines (1 μg/lane). Lane 1: λ DNA:PstI. Lane 2, 3 and 4: total RNA derived from preparations where the crude pellet was solubilised in either 0.4 ml (lane 2), 2 ml (lane 3) or 3 ml (lane 4) of Tris–borate buffer per gram of fresh weight of plant material initially used for homogenisation. Arrow indicates contamination by genomic DNA. b. Agarose gel electrophoresis of total RNA isolated from roots of perlite-cultivated grapevine (1 μg/lane). c Agarose gel electrophoresis of one-step reverse transcription-polymerase chain reaction (RT-PCR) using aliquots of total RNA as shown in a and EF1α-specific primers. Lane 1: λ DNA:PstI. Lane 2, 3 and 4: RT-PCR product using 200 ng of RNA as template as shown in a (lane 2) after 25, 30 and 35 cycles, respectively. Lane 6, 7 and 8: RT-PCR product using 200 ng of RNA as template as shown in Fig 3a (lane 3) after 25, 30 and 35 cycles respectively. Lane 10, 11 and 12: RT-PCR product using 200 ng of RNA as template as shown in a (lane 4) and amplified after 25, 30 and 35 cycles, respectively. Lane 5 and 9 are empty slots. Amplicon size 288 bp. Repetition of the experiment gave identical results. d Agarose gel electrophoresis of one-step RT-PCR (upper level) and PCR (lower level) performed in parallel using aliquots of total RNA as prepared in Fig. 3a (lane 2) and EF1α specific primers. Lane 1, 2, 3 and 4: RT-PCR product using 250 ng of total RNA as template after 26, 30, 34 and 38 cycles, respectively. Lane 5: λ DNA:PstI. Lane 6: negative RT-PCR control (H2O instead of RNA template). Lane 7, 8, 9 and 10: PCR product using 250 ng of total RNA as template after 30, 34, 38 and 42 cycles, respectively. Lane 11: λ DNA:PstI; Lane 12: negative PCR control (H2O instead of RNA template); Amplicon size 288 bp. Representative results are shown

Table 1

Optimisation of yield and quality of total RNA from roots of grapevine

Volume of Tris–borate buffer (ml/g FW)

Absorbance ratios

Yield (μg/g FW)















The average RNA yield from Perlite-grown roots was 100 μg/g FW and significantly higher than from roots of hydroponic cultures with an average yield of 55 ± 43SD μg/g FW. In addition, the integrity of RNA from Perlite-grown roots was always higher than of RNA from hydroponic cultures. This observation was confirmed by agarose gel electrophoresis, indicating that the type of culture dramatically influences the quality and yield of RNA extractions (Fig. 3a, b).


A protocol has been developed for the extraction of functional RNA from leaf, stem and root tissue of grapevine. The procedure works reliably, requires only standard laboratory equipment and is suitable for many downstream applications such as Northern hybridisation, mRNA isolation, cDNA library construction and RT-PCR. The protocol can also be easily adjusted linearly to varying amounts (0.1–20 g) of plant material according to the experimental requirements.


The authors thank Claudia Linhard (AlPlanta) for the excellent technical assistance and Manfred Jutzi (DLR Rheinpfalz, Neustadt a.d. Weinstrasse, Germany) for performing the statistical analyses. This work was supported in part by an ICGEB research grant (CRP/TUN06-01). S. Daldoul was also supported by a research grant of the German Academic Exchange Service (DAAD).

Copyright information

© Franciszek Górski Institute of Plant Physiology, Polish Academy of Sciences, Kraków 2009