Evaluation of Protein Extraction Methods for Vitis vinifera Leaf and Root Proteome Analysis by Two-Dimensional Electrophoresis

Abstract

An efficient protein extraction method is crucial to ensure successful separation by two-dimensional electrophoresis (2-DE) for recalcitrant plant species, in particular for grapevine (Vitis vinifera L.). Trichloroacetic acid-acetone (TCA-acetone) and phenol extraction methods were evaluated for proteome analysis of leaves and roots from the Tunisian cultivar ‘Razegui’. The phenol-based protocol proved to give a higher protein yield, a greater spot resolution, and a minimal streaking on 2-DE gels for both leaf and root tissues compared with the TCA-based protocol. Furthermore, the highest numbers of detected proteins on 2-DE gels were observed using the phenol extraction from leaves and roots as compared with TCA-acetone extraction.

Introduction

High-resolution two-dimensional electrophoresis (2-DE) allows for the quantitative and qualitative separation of complex proteins mixtures, usually extracted from cellular organisms. Sample preparation is an important step and is absolutely essential to allow reliable results. Most plant tissues do not provide a ready source of proteins and thus, need special treatment. Plant cells contain several substances that damage the yield and quality of the protein extraction and may be responsible for proteolytic breakdown, streaking, and charge heterogeneity. The most common interfering substances are phenolic compounds, proteolytic and oxidative enzymes, terpenes, pigments, organic acids, inhibitory ions, and carbohydrates (Carpentier et al. 2005).

In this regard, proteomic analysis of grape tissue involves a number of practical challenges since grape tissue contains lots of interfering elements, such as polysaccharides and phenolic compounds. Such contaminants are particularly problematic for 2-DE, resulting in horizontal and vertical streaking, smearing, and a reduction in the number of distinctly resolved protein spots.

Commonly, proteins are first isolated by precipitation and resolubilized by resuspending the dried pellet in an appropriate buffer. An incomplete precipitation and/or resolubilization will result in protein losses. In many cases, the extraction procedure must be optimized for one organism, tissue or cell compartment. Some reviews have indicated guidelines in order to perform an adequate protein extraction prior to 2-DE (Shaw and Riederer 2003; Gorg et al 2004).

Thus, Tesnières and Robin (1992) previously adapted Hsu and Heatherbell's protocol, based on phosphate buffer extraction, to study proteins from mature red berries and were able to visualize up to 52 spots on silver nitrate-stained 2-D gels. In the same year, Gianazza et al. (1992) isolated proteins from callus regenerated from Vitis leaves and petioles and visualized on silver nitrate-stained 2-D gels, on which they could resolve up to 900 spots. In 2004, Sarry et al. could detect up to 381 spots from the grape berry mesocarp using a trichloroacetic acid-acetone (TCA-acetone) based method. Recently, the phenol based method was used to perform a proteome analysis of grape berry clusters (Vincent et al. 2006), grape berry skin during ripening (Deytieux et al. 2007) and grape berry cell wall (Negri et al. 2008). For proteome analysis of grapevine leaves, Castro et al. (2005) used a lysis buffer containing urea, thiourea, 3-[(3-Cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) and sodium dodecyl sulfate (SDS) followed by TCA/acetone precipitation and Vincent et al. (2007) adapted TCA/acetone based method, according to Damerval et al. (1986), to study proteins from grapevine shoot tips (including the apex, the stem, four leaves, and tendrils).

To our knowledge, this is the first report on the evaluation of two protein extraction methods from leaves and roots of Vitis vinifera. The development of efficient protein extraction methods for such grapevine tissues is important since this species is recalcitrant regarding its polyphenols, polysaccharides and tannins contents. In these protocols, different extraction buffers (TCA/acetone, phenol) and precipitation steps (acetone, ammonium acetate) were evaluated.

Results and Discussion

In our study, TCA-acetone (Damerval et al. 1986) and phenol (Hurkman and Tanaka 1986) protocols were evaluated for protein extraction and precipitation. Comparison was done based on protein yield, spot focusing and resolution, number of resolved spots and intensity of spots. Both protocols were used to evaluate protein extraction from leaves and roots of a recalcitrant species, Vitis vinifera L.

Leaf and root extracts obtained using the two protocols were evaluated using the Bradford assay. As far as protein yields are concerned, the quantitative comparison showed that the phenol method provided highest yields compared with TCA-acetone extraction and that, for both leaf and root tissues. The concentration of leaf protein extracts was 2.23 ± 0.021 (Phenol), against 1.45 ± 0.033 mg/g fresh weight (FW) (TCA-acetone) while that of root protein extracts was 1.06 ± 0.068 (Phenol) against 0.045 ± 0.023 mg/g FW (TCA-acetone). Protein extracts from roots yielded a lower amount of protein than those obtained from leaves for each protocol (Table 1).

Table 1.  Protein yields and spot numbers from leaves and roots using phenol and trichloroacetic acid-acetone extraction protocols

	Leaves		Roots
	Protein yield (mg/g FW)	Number of spots	Protein yield (mg/g FW)	Number of spots
Values are the mean of three independent replicates ± standard deviation. TCA-aceton, trichloroacetic acid-acetone; Fw, Fresh Weight.
Phenol	2.23 ± 0.21	613 ± 23	1.060 ± 0.068	612 ± 29
TCA-acetone	1.45 ± 0.33	428 ± 18	0.045 ± 0.023	250 ± 24

These results are in accordance with (Nandakumar et al. 2003) who reported that TCA-acetone provides a pellet that is more difficult to redissolve in resuspended buffer, compared with the phenol extraction protocol. Therefore, the amounts of proteins extracted from different plant tissues depended on the organ type. Similar results were reported for tomato mature pollen, in which the amount of extracted proteins was greater using phenol than the TCA-acetone method (Saravanan et al. 2004). On the other hand, TCA-acetone proved to be better than phenol for Brassica seeds (Devouge et al. 2007). However, in Arabidopsis thaliana, no difference was obtained in protein yield with the TCA-acetone (1.39 mg/g FW) and the phenol protocol (1.40 mg/g FW) as well (Maldonado et al. 2008). Similar results were reported from banana leaves using both methods and comparable amounts of total proteins were obtained (Carpentier et al. 2005).

In the present study, equal amounts of leaf and root proteins (75 μg), extracted using both protocols, were separated by 2-DE under identical conditions in the 4 to 7 pH range. Representative silver-stained gels from each method are presented in Figure 1. The two methods resulted in good quality and well-resolved gels. However, quantitative and qualitative differences were observed in the spot patterns from both leaf and root extracts using both protocols, as highlighted in Figure 1.

Figure 1. Gels of protein extracts obtained using phenol (A) and trichloroacetic acid-acetone (TCA-acetone) (B) methods from leaves and roots of Razegui grapevine cultivar.

The difference in yield was visually noticeable on the 2-D patterns, with an increased number of spots using phenol extraction method as compared with TCA-acetone protocol, in both leaves and roots. On the other hand, a higher number of spots was detected in the electrophoretic profiles obtained from leaves as compared with roots by TCA-acetone extraction. In contrast, similar numbers of spots were observed in both samples by phenol extraction method (Table 1). Thus, about 613 and 428 spots from leaves could be resolved on 2-D gels using phenol and TCA-acetone protocols, respectively. From roots, up to 612 and 250 spots could be resolved on 2-D gels using phenol and TCA-acetone protocols, respectively. These results are similar to those reported from the leaves of Arabidopsis (Maldonado et al. 2008), tomato (Saravanan et al. 2004) and rice (Islam et al. 2004). However, a lower spot number was detected from olive leaves (Wang et al. 2003), and a higher number was reported for Arabidopsis cell suspensions (Sarry et al. 2006) and banana meristem (Carpentier et al. 2005).

In general, the phenol method produced better spots resolution and minimal streaking than the TCA-acetone method. The low number of spots observed in 2-D patterns obtained by the TCA-acetone extraction method is attributed to the low protein yield, which suggests an incomplete protein resolubilization. Surprisingly, when performed from leaves, TCA-acetone extraction allowed us to obtain 2-D patterns displaying a higher number of spots than those obtained from roots. However, their spots resolution was poor (Figure 1). Furthermore, the TCA-acetone protocol resulted in irreproducible gels with streaks, precipitation and an under-representation of proteins. According to Carpentier et al. (2005), this method does not remove the interfering compounds such as carbohydrates, causing precipitation and bad focusing, and polyphenols that generate charge heterogeneity and streaking. It is also totally dependent on the effectiveness of the protease inhibitor cocktails (Carpentier et al. 2005).

In spite of its low yield, phenol extraction protocol led to a high number of spots from roots. Carpentier et al. (2005) also reported a high number of spots using the extraction phenol method from several plant species and organs. Phenol protocol might allow extraction of a greater number of proteins compared with the TCA-acetone method, resulting in higher amounts of resolubilized proteins. Some of these spots matched in the different gels, while others were specific to the used protocol. A high percentage of spots that were specific to each protocol were observed in both leaves and roots. This demonstrates that the specificity of the extracted proteins is protocol-related.

The acetone precipitation-based approach resulted in a low number of unique spots (12% spots from roots and 29% spots from leaves). The percentage of common spots that were present in gels from both extraction methods was 22% at most and averaged in leaves and roots (Table 2).

Table 2.  Total number of spots detected from grapevine leaves and roots after phenol and trichloroacetic acid-acetone extractions and percentages of spots specifically related to each of the extraction protocols

Grapevine tissues	Total number of spots	(%)
		Phenol	TCA-acetone
TCA-aceton, trichloroacetic acid-acetone.
Leaves	858	50	29
Roots	694	64	12
Mean ±SD	776	57 ± 11.2	20.5 ± 13.6

Saravanan and Rose (2004) observed 12% and 24% of common spots further to acetone and phenol extractions from tomato roots and leaves, respectively. Using the phenol extraction method, 50% and 64% of spots were detected from roots and leaves, respectively. In tomato leaves and roots, 16% and 27% of spots were obtained from phenol extraction; however, lower percentages were detected in the leaves and roots (4% and 21%, respectively) when extraction was based on TCA-acetone (Saravanan and Rose 2004). A number of hypotheses were tested to determine the basis of the substantially different composition of proteins in extracts made using the different techniques. The most obvious was the differential solubility and extraction, due to the biochemical characteristics of the proteins themselves.

As mentioned by Maldonado et al. (2008), no single method can extract the full proteome. We are conscious that we are considering a minimal part of the grapevine proteome in our analysis; particularly those extracted and solubilized using the tested protocol. Chevalier et al. (2004) previously reported that for plant tissues, the total number of spots resolved by 2-DE varies from a few hundred to 3 000. In the case of A. thaliana, the most complete proteome map was obtained by assembling the proteins identified from different organs, developmental stages, and undifferentiated cultured cells (Baerenfaller et al. 2008).

The TCA-acetone method could lead to the resolubilization of a greater quantity of specific proteins, therefore not necessarily producing a high number of spots but an increased spot abundance, in particular areas of the 2-D gels. To verify this hypothesis for each extraction protocol, all of the spot intensities were summed to estimate the protein load of the 2-D patterns. The most abundant spots were also isolated to evaluate the quantity range (Figure 2). The results show differences in the number of spots, sum of spot intensity and maximal abundance (Figure 2) between the extraction methods. The highest values were obtained by using the phenol extraction protocol.

Figure 2. Comparison of phenol and trichloroacetic acid-acetone (TCA-acetone) extraction methods based on the number of spots, sum of spot quantity and maximal abundance in leaves and roots of Razegui grapevine cultivar.

On the other hand, the phenol protocol constantly gave the greatest number of spots along pI and M_r gradients (Figure 3). In A. thaliana, Maldonado et al. (2008) reported a significantly higher number of resolved spots further to TCA-acetone protein extraction from leaves as compared with the phenol extraction method. The evaluation of the highest number of spots within different categories of M_r and pI, allowed us to notice that spots generated further to the phenol extraction are significantly more abundant (P≤ 0.05) in the 10 kDa and 60 kDa M_r ranges for leaves and 20 kDa and 80 kDa M_r ranges for roots (Figure 3).

Figure 3. Distribution of phenol and trichloroacetic acid-acetone (TCA-acetone) extracted proteins from leaves and roots of Razegui grapevine cultivar, according to their M_r and pI.

Carpentier et al. (2005) reported that the number of significant spots from the phenol extraction method in banana meristems is statistically more abundant in M_r category spots above 25 kDa, while the TCA-A method has more significant spots in the category range below 25 kDa. In grapevine, Vincent et al. (2006) reported that the phenol based extraction from mature berries gave the greatest number of spots along both pI and M_r gradients. These authors also observed that most of the spots resolved between 5 and 7 along the pI range and between 10 and 50 kDa along M_r range (Vincent et al. 2006).

For both protocols, most of the spots were statistically (P≤ 0.05) resolved between 4 and 7 along the pI range. The enhanced superiority of phenol protocol indicates that this procedure generates 2-D gels resolving not only more numerous but also more prominent spots. The histograms representing the number of spots for both tissues (Figure 4) along the pI gradient, demonstrate that the protein distribution is not homogenous within the pH gradient. This would justify the choice of a nonlinear pH gradient. The maximal abundant spots along M_r and pI ranges presented in (Figure 4) show that the phenol extraction protocol generates the most prominent spots within pI or M_r units.

Figure 4. Maximal quantities of phenol and trichloroacetic acid-acetone (TCA-acetone) extracted proteins from leaves and roots of Razegui grapevine cultivar, according to their M_r and pI.

Conclusion

In the present study, phenol and TCA-acetone protein extraction methods were evaluated in grapevine leaves and roots by 2-DE electrophoresis. The results demonstrate the enhanced efficiency of the phenol based protocol as compared with TCA-acetone extraction, if protein yield, gel quality, spot numbers and quantities are considered. This could be attributed to the fact that phenol is one of the strongest dissociating agents known to decrease molecular interactions between proteins and other materials, in addition to its selectivity as a solvent (Carpentier et al. 2005). Saravanan and Rose (2004) previously noticed that the phenol extraction method gave the greatest protein yield and was particularly effective with recalcitrant tissues. Regarding the TCA-acetone extraction method, this precipitant is very effective (Devouge et al. 2007), and allows instant elimination of proteolytic and other modifying enzymes (Wu et al. 1984). However, a disadvantage of TCA extracted proteins is that they are difficult to redissolve (Loomis and Bataille 1966). Even though this protocol was successfully used on grapevine shoots and green immature berries in the past, it gave poor results on mature berry clusters (Vincent et al. 2006) and grapevine leaves and roots as demonstrated above.

Given that in a comparative proteomic analysis, a major objective is to maximize the number of proteins that can be resolved, both phenol and TCA-acetone extraction methods reveal themselves to be complementary since each protocol allows the extraction of specific proteins.

Materials and Methods

Plant material

During dormancy, cuttings from Tunisian grapevine (Vitis vinifera L.) cultivar ‘Razegui’ were rooted under controlled conditions. Young grapevine plantlets were further grown in pots containing sand in a climate controlled greenhouse (25 °C/20 °C temperature, 16/8 h photoperiod and 60% relative humidity). Vines were regularly irrigated using Long Ashton nutrient solution (Hewitt 1966), before be transferred on hydroponic culture system based on Long Ashton solution. After 20 d of stabilization under the hydroponic conditions and vegetative development, leaves and roots were harvested for protein analysis. Three biological replicates were used in each treatment.

Protein extraction

Two protein extraction methods were evaluated; namely those based on phenol and trichloroacetic acid-acetone (TCA-acetone).

Phenol extraction

The phenol extraction procedure was based on a previously published method by Schuster and Davies (1983) described in Hurkman and Tanaka (1986). Five grams of frozen plant tissue (leaves and roots) were finely powdered in liquid nitrogen using a pestle and mortar and resuspended in 10 mL of extraction buffer 1 (1% (w/v) polyvinylpolypyrrolidone (PVPP), 0.7 M sucrose, 0.1 M KCl, 0.5 M Tris-HCl pH 7.5, 500 mM ethylenediaminetetraacetic acid (EDTA), 1 mM (PMSF), 2% (v/v) β-mercaptoethanol. The mixture was extensively homogenized on ice for 30 min, then an equal volume of water-saturated Phe was added and the mixture was rehomogenized for 30 min on ice and centrifuged at 6 000 g for 30 min. The upper Phe phase was removed and re-extracted with extraction buffer 1 described above. Proteins were precipitated from the final Phe phase with five volumes of saturated ammonium acetate (100 mM) in methanol overnight at −20 °C and finally centrifuged at 6 000 g for 30 min. Protein pellets were washed five times with 100 mM ammonium acetate in methanol and dried with nitrogen gas. Proteins were solubilized in urea buffer (Urea 5 M, Thiourea 2 M, CHAPS 2% (w/v), SB3–10 2% (w/v), Tris 40 mM, ampholytes (3/10) 0.2% (v/v), TBP (tributylphosphine) 2 mM). The protein concentration was determined according to Bradford's (1976) method using bovine serum albumin as a standard.

TCA-acetone extraction

Total protein extraction was carried out using the procedure of Damerval et al. (1986). Five grams of frozen plant tissue was finely powdered as described above and transferred to a 50 mL centrifuge tube containing four volumes of ice-cold acetone containing 10% (w/v) TCA, 1% (w/v) PVPP and 2% (v/v) β-mercaptoethanol. The mixture was homogenized and proteins were precipitated at −20 °C overnight. After incubation, the suspension was centrifuged at 10 000 g (4 °C) for 15 min. Supernatant was discarded and the pellet was resuspended with 5 mL acetone containing 0.07%β-mercaptoethanol and kept at −20 °C for 1 h. Centrifugation was repeated as above and pellets were washed five times with 5 mL acetone containing 0.07%β-mercaptoethanol, supernatant was decanted. Pellets were dried under vacuum and kept at −20 °C until use. Proteins were solubilized in urea buffer. The protein concentration was determined according to Bradford's (1976) method using bovine serum albumin as a standard.

IEF and SDS-PAGE

The 2-DE of total proteins was performed via capillary isoelectrofocalization (IEF). Proteins were separated using the Mini Protean II system from Bio Rad Laboratories. In the first dimension, gel containing 4% acrylamide/N,N0-methylenebisacrylamide, 9.2 M urea, 20% (v/v) TritonX100, 10% (v/v) ampholytes, 0.4% (v/v) riboflavin phosphatase and 0.025% ammonium persulfate was cast in 1 mm (internal diameter)-70 mm capillary tubes. The pH gradient was carried out on the capillary gel by a pre-run according to the following program: 200 V for 10 min, 250 V for 15 min and 300 V for 15 min. Further, 75 μg of proteins were loaded at the cathode on the gel and electrofocalized according to the following program: 250 V for 15 min, 300 V for 15 min and 400 V during 18 h. After focusing, the proteins were equilibrated by incubating the gels with 2 mL of equilibration buffer (Tris HCl 0.5 M pH 6.8, SDS 10%, Glycérol 10%, dithiothreitol (DTT) 154.2 mg, Bleu de Bromophénol 25 mg). The second dimension (denaturant sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)) was a discontinuous gel system described by Laemmli (1970).

Gel staining, imaging and data analysis

The gels were visualized by silver staining (Laemmli 1970). The 2-D gels were imaged using a VersaDoc Imaging System (BioRad) and the images were analyzed using PDQuest 2-D Version 8.0.1 Analysis software. The apparent M_r and pI of each protein was determined by referencing to protein markers.

(Co-Editor: Roberto Bassi)