Proteomic Analysis of the Low Mutation Rate of Diploid Male Gametes Induced by Colchicine in Ginkgo biloba L.

Colchicine treatment of G. biloba microsporocytes results in a low mutation rate in the diploid (2n) male gamete. The mutation rate is significantly lower as compared to other tree species and impedes the breeding of new economic varieties. Proteomic analysis was done to identify the proteins that influence the process of 2n gamete formation in G. biloba. The microsporangia of G. biloba were treated with colchicine solution for 48 h and the proteins were analyzed using 2-D gel electrophoresis and compared to protein profiles of untreated microsporangia. A total of 66 proteins showed difference in expression levels. Twenty-seven of these proteins were identified by mass spectrometry. Among the 27 proteins, 14 were found to be up-regulated and the rest 13 were down-regulated. The identified proteins belonged to five different functional classes: ATP generation, transport and carbohydrate metabolism; protein metabolism; ROS scavenging and detoxifying enzymes; cell wall remodeling and metabolism; transcription, cell cycle and signal transduction. The identification of these differentially expressed proteins and their function could help in analysing the mechanism of lower mutation rate of diploid male gamete when the microsporangium of G. biloba was induced by colchicine.


Introduction
Ginkgo biloba L. is a living fossil tree with no close living relatives. It is cultivated because of its importance in traditional medicine and food value; and as an ornamental tree for its attractive shape and foliage. Recent studies have also implicated a use for Gingko extract in prevention of Alzheimer's disease [1] and dementia [2], thus creating a high market demand for this tree species. However, it has a tedious and inefficient growth process, and although effective breeding methods are emerging, it is still a challenge to meet the increased production requirement. Gingko extract contains secondary metabolites that have medicinal value; therefore, breeding methods that help in enhancing the quantity of these secondary metabolites are desirable. Since triploid (3n) plants have fast growth and a higher content of secondary metabolites as compared to 1n and 2n, ploidy breeding can be a useful tool for breeding new economic varieties of G. biloba.
Triploid plants are produced when diploid male gametes are induced and fertilized with normal female gametes. Colchicine is a wide-ranging and effective triploid-inducer used in studies of polyploidy breeding. It hinders the formation of spindle fibers in the metaphase of cell division and plays a minor role in affecting the structure of chromosomes [3].
Previous studies have shown that diploid pollen could be produced from the microsporocytes treated with colchicine, however, the rate of mutation was only 7 per cent [4] as compared to other trees such as, white Populus -88% [5], Eucommia-49.5% [6], and black Populus 290.25% [7]. The low mutation rate in diploid male gametes of G. biloba is insufficient for its application in ploidy breeding.
Sun et al. [8] showed that the microsporangium and microsporocyte wall are not the major factors hindering the process of colchicine induction of diploid male gametes in G. biloba. Several other studies have indicated that the proteome is an influencing factor in the formation process of male gametes. The events of meiosis are controlled by a protein enzyme complex known as the ''maturation promoting factor.'' These enzymes interact with one another as well as with other cell organelles to cause the breakdown and reconstruction of the nuclear membrane, the formation of the spindle fibers, and the cell division process [9][10][11][12].
We hypothesize that perhaps there are some proteins that do not destroy the chemical structure of colchicine; however, they can interfere with the process in which tubulin combines with colchicine. Therefore, it is necessary to determine differences between protein function in colchicine-treated and untreated microsporocytes during meiosis. Since colchicine stress has not yet been extensively examined at the proteome level in plant breeding studies, no data on this topic are currently available in the literature. Therefore, this study provided a framework to investigate the cause of low-level diploid male gametes formation in G. biloba treated with colchicine.

Plant materials
The male floral branches of G. biloba were collected in the spring season from the Teaching and Experimental Forestry Center of Beijing Forestry University (40.04060 N, 116.05256E). The branches were placed in cold storage (4uC) for 24 h before the microsporangiums of G. biloba began to develop in meiosis. The branches were cultured in the greenhouse (25uC) in order to induce meiosis. The microsporangia of G. biloba were treated with colchicine solution as soon as most of them turned green, were full in granule and the top bud scales began to open. Colchicine was applied using cotton balls soaked in 0.8% (8 mg/mL) colchicine solution [4]; the control samples were treated with cotton balls soaked distilled water only. Three independent biological replicates of each sample were taken. The samples were collected 48 h post treatment.

Protein Extraction
Total protein extracts were prepared according to the method described by Wang et al. [13] with minor modifications. The powder of the microsporangia of G. biloba was suspended in 1 mL cold extraction solution (10% TCA in acetone containing 1 mM PMSF and 0.07% b-mercaptoethanol) and then incubated at -20uC for 12 h. The samples were then centrifuged at 12,0006g for 30 min at 4uC. The pellets were washed thrice with prechilled acetone containing 0.07% b-mercaptoethanol, and kept at -20uC for 1 h before centrifugation at 12 0006g for 30 min at 4uC, and then dried under vacuum for10 min at 4uC.

2-D Electrophoresis
Two-dimensional gel electrophoresis was performed according to the Bio-Rad handbook and the procedure developed by Gorg et al. [14]. 700 mg of the sample was loaded onto analytical as well as preparative gels. For IEF, the Bio-Rad Mini-PROTEAN II 2-D system and pH 4-7 IPG strips (17 cm) were used according to the manufacturer's recommendations. The IPG strips were rehydrated for 14 h in 350 mL rehydration buffer containing protein samples. The images were taken as described by Wang et al. [13].
The gel strips were equilibrated for 15 min with two equilibration buffers: (i) reducing buffer for 15 min (50 mM Tris-HCl buffer, pH 8.8, 6 M urea, 20%v/v glycerol, 2% w/v SDS, and 1% w/v DTT). (ii) alkylating buffer for 15 min (50 mM Tris-HCl buffer, pH 8.8, 6 M urea, 20%v/v glycerol, 2% w/v SDS, and 2.5%w/v iodacetamide). SDS-PAGE was performed with 12% gels using the PROTEAN II xi Cell system (Bio-Rad, Hercules, CA, USA). The gels were run at 0.5 h at 1 W per gel, then at 8 W per gel until the dye front reached the gel bottom. The separated proteins were visualized by CBB G250 staining.

Image Acquisition and Data Analysis
The 2-DE images were acquired with an Image Scanner GS-800 (Bio-Rad, Hercules, CA, USA) in transmission mode. The image analysis was carried out by a combination of manual visualization and software calculations using PDQuest 8.0.1 Gel Analysis Software (Bio-Rad, Hercules, CA, USA). In order to correct the variability due to CBB-staining and to reflect the quantitative variations in intensity of protein spots, the spot volumes were normalized as a percentage of the total volume in all of the spots present in the gel. All 2-DE images were globally analyzed by the software and the identified spots were manually rechecked. Significantly different spots were drawn from a comparison of any two gels.

Tryptic Digestion
The CBB stained spots were excised from the 2-DE gels and destained for 30 min 63-4 times until the gel was transparent with no color, using a destaining solution consisting of 50% acetonitrile and 25 mM ammonium bicarbonate and then they were immersed in acetonitrile (100%) for 10 min. The gels were dried for 30 min using a Speed-Vac system. 2.5 mL of 25 mM (NH 4 )HCO 3 was added to 25 mg of trypsin (final concentration 10 ng/mL); 10 mL of this trypsin solution was pipetted on each dried protein spot and the sample was incubated at 4uC for 60 min. The supernatant was discarded to minimize autodigestion of trypsin. Then the Ep tubes were placed upside down and incubated at 37uC for 14 h. To extract the peptide fragments from the tryptic digest, 20 mL of 5% (v/v) TFA was added to the digest and the sample was incubated at 37uC for 60 min, then the supernatant was transferred into another Ep tube. Thereafter, 20 mL of 50% (v/v) acetonitrile [containing 2.5% (v/v) TFA] was added to the gel and the sample was incubated at 30uC for 60 min. The supernatants were pooled together and dried for 2 h heating in a Speed-Vac system.

Protein identification by MALDI-TOF/TOF-MS
Before obtaining the mass spectra of the peptide mixture, the digested peptides were desalted and cleaned with ZipTip C18 pipette tips (Millipore Corporation, Bedford, MA, USA) according to the manufacturer's instructions. All analyses were performed using a Bruker Daltonics Autoflex (Bruker Daltonics Billerica, MA, USA) operated in the delayed extraction of 190 ns and reflector mode with an accelerating voltage of 20 kV. The peptide mixture was analyzed using a saturated solution of R-cyano-4-hydroxycinnamic acid (CHCA Bruker Daltonics Billerica, MA, USA) in 50% acetonitrile/0.1% trifluoroacetic acid. External calibration was performed with a peptide calibration standard (Bruker Daltonics Billerica, MA, USA, Part No.: 206195) and internal calibration with trypsin autoproteolytic fragments. The samples were analyzed on a MALDI-TOF/TOF-MS 4800 proteomics analyzer (Aglient, USA) and data were analyzed using GPS explorer software (Applied Biosystems, Foster City, Calif.) and MASCOT software (Matrix Science, London, UK).

Comparative protein profiles of colchicine treated and control microsporangia
Approximately, 1000 microsporangia proteins were detected on coomassie brilliant blue-stained gels ( Figure 1B). To accurately and quantitatively analyze proteomic changes, spot volume differences of more than 2-fold between two identical spots were defined as significant. The spots that revealed changing their abundance significantly after quantitative image analysis were randomly selected to check again. A total of sixty-six proteins were differentially expressed in response to colchicine ( Figure 1A, 1B). It was noted that some protein spots demonstrated qualitative changes in intensity. The spots 7704, 8104 and 3109 were absent in the colchicine-treated gels (Figure 1).

MOLDI-TOF/TOF-MS identification and classification of colchicines-responsive proteins
Thirty-nine of these differentially expressed spots were arbitrarily selected and excised from the gels and subjected to MOLDI-TOF/TOF-MS. Twenty-seven proteins were successfully identified and are listed in Table 1.
The identified proteins were associated with a wide variety of cellular processes. They were classified into several categories according to their function (Figure 2), including the ATP generation, transport and carbohydrate metabolism, protein metabolism, ROS scavenging and detoxifying enzymes, cell wall remodeling and metabolism, transcription, cell cycle, and signal transduction.
The identified spots corresponded to proteins that were related to energy, carbohydrate metabolism and photosynthesis. In our study, nine proteins were in this category. Four of them were found to be up-regulated, and three of them were changed in the opposite way. The rest of them evenly has been disappeared after colchicine treated. Several chaperones involved in protein processing, folding and degrading were also identified as being responsive to colchicine treatment. These included two HSP (spot3605 and spot3611) and two EF-Tu (spot4312 and spot5313) proteins. In this study, the levels of protein (3605), which was analogous to HSP-70 (3611) and heat shock protein 70, were found to be elevated ( Figure 3).
The SE-wap41 (spot 2207) was found to be down-regulated. 2dehydro-3-deoxyphosphooctonate aldolase (spot 7110), a protein involved in cell wall remodeling and metabolism, was found to be up regulated. Both of them indicating that colchicines might perhaps have interfered with cell wall formation. The three spots, 4607, 4801 and 8505 were related to transcription, cell cycle, and signal transduction. Spot 4607 was found to be a response factor of a plant growth hormone and was up-regulated by colchicine treatment. Spot 4801 was found to be the homolog of CDC48 and spot 8505 was classified as a putative polyprotein; and expression levels of both proteins were elevated in this study.
APX (spot 6008) and GST (spot 5004) were found to be a part of the antioxidant system employed by plants [15]. Previous microarray results demonstrated that these proteins were responsive to various stresses including osmotic, drought, and cold stresses [16][17][18]. We found that the abundance of GST was three times higher post colchicine treatment. However, the abundance of APX was drastically down regulated (half of the level in control).

Discussion
The protein profiles of microsporangia treated with colchicine enabled us to classify the individual response proteins on the basis of their function.

Cell wall remodeling and metabolism
The decline in the abundance of se-wap41 after 48 h of colchicine treatment corroborated the results of Chen et al. [19]. Se-wap41, a salt-extractable 41-kD wall-associated protein that has been reported to label plasmodesmata and the Golgi was found to be a class 1 reversibly glycosylated polypeptide (C1 RGP) [9]. Using immunogold labeling, Dhugga et al. [20] showed that RGP1 was specifically localized to Golgi stacks, where it was likely to be involved in xyloglucan biosynthesis [21]. This suggests that colchicine can interfere with the cell wall remodeling. Hogetsu and Shibaoka [22] also demonstrated that colchicine could inhibit cell  membrane for organic molecules, increases the negative charge of the cell wall and stabilizes the overall membrane structure. Wang et al. [23] found that 2-dehydro-3-deoxyphosphooctonate aldolase was upregulated in boron deficient Brassica napus roots. It is speculated that plants synthesize LPS to resist injury when cell membranes suffer the colchicine stress.

Proteins involved in ATP generation, transport and carbohydrate metabolism
Respiration is critical to metabolism in higher plants. In addition to its link with carbon metabolism, respiration releases energy stored in carbon-based compounds in a controlled manner for cellular use. It also generates many carbon precursors for biosynthesis. Carbon metabolism enzymes appear to be major targets for oxidative modification and breakdown in vitro and in vivo [24]. In our study, 48% of the proteins were found to be involved in ATP generation, transport and carbohydrate metabolism. Most of them responded in a manner similar to other abiotic stress conditions. For example, it was noted that enolase (spot5305), which catalyzed the formation of high-energy phosphoenol pyruvate from 2-phosphoglycerate in the glycolytic pathway, decreased after 48 h of colchicine stress. Previous reports showed that a mutation in the enolase locus results in the repression of cold-responsive genes and therefore, it acts as a positive regulator of cold-responsive genes [25]. The decreased abundance of enolase protein was similar to the response of Arabidopsis roots to NaCl stress [26]. Other energy-related proteins, including the two ATP synthases (spot 8408 and 3407) were found to be elevated in response to colchicine treatment. This response was similar to that observed in other stress conditions such as salt or cold stress [27][28][29].
Thiazole biosynthetic enzyme (spot 4018) also showed higher abundance in response to colchicine treatmenttreated. Tunc-Ozdemir et al. [30] indicated that when Arabidopsis (Arabidopsis thaliana) plants showed an elevated expression of transcripts encoding thiamin biosynthetic enzymes in response to abiotic stress conditions, such as high light, cold, osmotic stress, salinity, and oxidative treatments.

ROS scavenging and detoxifying enzymes
Abiotic stresses induce the production of ROS, which can cause damage to cellular components and act as a signaling molecule for stress responses [15]. In recent years, additional reports have indicated that GSTs in plants play major roles in herbicide detoxification [33][34][35][36][37][38][39][40]. GSTs also conjugate natural products such as anthocyanins, serving as binding proteins or ligandins for plant hormones, or catalyze GSH-dependent peroxidase and isomerase reactions [34,41]. Herbicide safeners are chemicals used to minimize the effect of the herbicides on crop plants. They protect cereal crops from herbicide injury by increasing the expression of GSTs as well as other herbicide-detoxifying enzymes and proteins such as cytochrome P-450 s, glucosyl transferases, and tonoplast transporters that can detoxify herbicides and transport them into the vacuole [39]. In this study, colchicine treatment resulted in elevated levels of GSTs, perhaps to reduce colchicine toxicity. An increased detoxification of compounds mediated by high levels of glutathione (GSH) and glutathione transferase (GST) has been found in cells resistant to cochicine [42]. Ruiz-Gomez et al. [42] suggested that GSTs may play an important role in main multidrug resistance (MDR) pumps in cells that express high levels of P-gly-coprotein (P-gly).
APX in plants utilizes ascorbate as an electron donor to catalyze H 2 O 2 conversion into H 2 O. It is a multigenic enzyme localized in different cellular compartments, which complements and coordinates antioxidant defenses [43]. In rice it was induced under conditions like ozone stress, drought stress, chitosan, and in blast lesion mimic mutants [44][45][46][47]. Studies on the possible interference of colchicine and H 2 O 2 in Arabidopsis thaliana cv.Col-0 suggested that APX was the most sensitive enzyme to colchicine and H 2 O 2 [48].
FNR (spot 9111) is a widespread family of FAD-containing enzymes. In different tissues and organisms these flavoproteins catalyze the reversible electron transfer between two molecules of the obligatory one-electron carrier Fd and a single molecule of NADP (H) [49]. The expression of a cloned plant FNR gene in a mutant E. coli strain did restore the oxidative tolerance to wild-type levels, indicating that the eukaryotic flavoenzyme behaves as a toxic radical scavenger in the bacterial host [50]. In our study, the FNR (spot 9111) was found to increase under colchicines stress. This indicates the toxic radical scavenger function of FNR in gymnosperms.

Protein metabolism
The proteolytic system, consuming metabolic energy, plays an essential role in the regulation of cellular functions by catalyzing rapid and irreversible reactions for various important biological processes, such as the cell cycle, apoptosis, signal transduction, protein processing, and immune and stress responses [51]. We detected three proteins that promote the proper folding of proteins and proteolysis and found that their abundance increased following colchicine treatment. One of them was the 26 s proteasome ATPase subunit. G. N. DeMartino and his colleagues [52] attempted the molecular cloning of cDNAs encoding the human 26S proteasome regulatory subunits. These subunits are divided into two distinct sub-groups, ATPase and non-ATPase subunits. The 26S proteasomal regulatory complex has intrinsic ATPase activity that seems to play an essential role in its function [52,53]. The 26S proteasome attacks poly-or multi-Ub chain to act as a degradation signal for proteolysis. Furthermore, research has shown that different kinds of stresses can result in degradation of the Ub chain. Treatment with HgCl 2 increases the expression of ubiquitin genes in Nicotiana; and a strong accumulation of a ubiquitin conjugating enzyme (Ubc1) was observed in tomato after exposure to CdCl 2 [54,55]. The expression of polyubiquitin gene is affected by darkness, UV radiation, starvation, and an enhanced level of ozone [56][57][58][59]. A significant increase in the level of de novo biosynthesis of ubiquinone and plastoquinone (PQ) in plant tissues was observed as a result of colchicine treatment [60].
One of the first described chaperone regulators of Tau is the Hsp70 chaperone family. Increased levels of Hsp70 promote tau solubility and tau binding to microtubules; the tau protein is normally expressed in the cytoplasm of cell bodies and axons, where it binds to and stabilizes microtubules [61]. The greatly increased abundance of HSP 70 s in our study suggested that the G. biloba microsporangia were building resistance to colchicine. This powerful resistance helped in maintaining the stability of microtubule when the microtubule of G. biloba was treated by colchicine.
Elongation factor-1a (EF-1a), a highly conserved protein named for its role in protein translation, is also a microtubule-associated protein (MAP) [62] and binds to the microtubule lattice [63][64][65]. This protein can also modulate microtubule dynamics in vitro and in vivo [66][67][68][69][70][71]. In our study this protein was found to be down regulated. Therefore, G. biloba might have a strong self-protection system to defend the colchicine invasion, which hinders the induction of diploid gametes.

Transcription, cell cycle and signal transduction
Auxin plays a crucial role in diverse aspects of plant growth and development, including tropic responses, apical dominance in the shoot, lateral root formation and differentiation of the vascular system. In this study, the increased ARF-L1protein indicated that colchicine may stimulate the differentiation of microsporangiums.

Conclusion
In this study, G. biloba microsporangia were subjected to 48 h of colchicine treatment followed by proteomic analysis of treated and control samples. A total of 27 colchicine-responsive proteins were identified (Table 1) by 2-D gel electrophoresis. Complementary strategies to protein electrophoresis can be employed at the transcript and metabolite levels in future to gain insight into the intricate network of plant responses to colchicine. Proteins are the most direct reflection of coclhicine response. And genes are the most fundamental evidences. Through above analysis, we can see the possible roles the twenty-seven proteins played in defending colchicine invasion. So a new dimension to our work is pointed out. The differently expressed proteins should be the most important part of future study. The real reason for the low mutation rate in diploid male gametes of G. biloba might be found out among them. Much work still to be done about the twentyseven proteins, such as comparing the twenty-seven genes with colchicine responsive genes in other species to see if any of them are specific to G. biloba. If we can confirm the significant proteins and find the mechanism of lower mutation rate of diploid male gamete when the microsporangium of G. biloba was induced by colchicine, we can try to optimize the ploidy breeding method in G. biloba.