Optimization of RT-PCR reactions in studies with genes of lignin biosynthetic route in Saccharum spontaneum

Saccharum spontaneum has been used for the development of energy cane a crop aimed to be used for the production of second-generation ethanol, or lignocellulosic ethanol. Lignin is a main challenge in the conversion of cell wall sugars into ethanol. In our studies to isolate the genes the lignin biosynthesis in S. spontaneum we have had great difficulty in RT-PCR reactions. Thus, we evaluated the effectiveness of different additives in the amplification of these genes. While COMT and CCoAOMT genes did not need any additives for other genes there was no amplification (HCT, F5H, 4CL and CCR) or the yield was very low (CAD and C4H). The application of supplementary cDNA was enough to overcome the non-specificity and low yield for C4H and C3H, while the addition of 0.04% BSA + 2% formamide was effective to amplify 4CL, CCR, F5H and CCR. HCT was amplified only by addition of 0.04% BSA + 2% formamide + 0.1 M trehalose and amplification of PAL was possible with addition of 2% of DMSO. Besides optimization of expression assays, the results show that additives can act independently or synergistically.


INTRODUCTION
Due to the increasing demand for clean energy, there has been a growing interest in biofuels obtained from biomass.In the production of the second-generation ethanol (E2G) or lignocellulosic ethanol the cell wall polymeric sugars cellulose and hemicellulose are hydrolyzed to simpler units which become fermentable (Ruth and Thomas 2003, Aden and Foust 2009, Huang et al. 2009).
Lignin is a bottleneck for E2G production in several plants.This phenolic polymer chemically binds to cellulose and hemicellulose in a systematic arrangement, increasing the recalcitrance of the biomass (Boerjan et al. 2003, Vanholme et al. 2010).As a consequence, gene manipulation has been tried aiming to decrease or change lignin composition (Chapple and Carpita 1998).The degree of JUAN P.P. LLERENA, PEDRO ARAÚJO and PAULO MAZZAFERA reduction or modification of lignin depends on the position of the enzyme in the biosynthetic pathway and the pattern of expression (Vanholme et al. 2012, Poovaiah et al. 2014).Change in the expression of genes encoding enzymes operating in the earlier part of the route usually changes the total lignin content (Bonawitz and Chapple 2010), while genes positioned later lead to changes in the lignin type (Li et al. 2003, Ralph et al. 2006, Chen and Dixon 2007).
Recently, breeders are transferring and increasing the number of Saccharum spontaneum alleles in commercial sugarcane hybrids, aiming to increase fiber content in the bagasse and, consequently, the potential for E2G generation (Carvalho-Netto et al. 2014).S. spontaneum culms are rich in lignin and other polyphenolic compounds (Devi and Muthu 2014) which can negatively interfere in RT-PCR reactions during analysis of gene expression.Additionally, the genomes of the Saccharum genus are polyploid, redundant, and present a high GC content as verified in other grasses (Šmarda et al. 2014), which often leads to the observation of non-specific bands or smears in the agarose gel as a result of RT-PCR analysis.In our case, the difficulty to amplify cDNA of genes of the lignin biosynthesis route in S. spontaneum internodes, using good quality (non-degraded) RNA, led us to test different additives in RT-PCR reactions.

RT-PCR is an essential technique in gene
expression and other biotechnological studies (Farell and Alexandre 2012) but it is prone to several problems caused by inhibitors in the samples (Samarakoon et al. 2013).Additionally, some nucleotide regions in primers with high melting temperature (Tm) or with high GC content (>60%) can form secondary structures "interrupting" the reaction (Zeng et al. 2006, Mamedov et al. 2008, Farell and Alexandre 2012).Several alternatives have been proposed to minimize these problems, such as the addition of substances of varied molecular nature (Simonović et al. 2012) that can act synergistically (Ralser et al. 2006, Farell andAlexandre 2012) or independently (Chakrabarti and Schutt 2001) in increasing the amplification process and the specificity of the reaction.Formamide (Sarkar et al. 1990, Chakrabarti andSchutt 2001), dimethyl sulfoxide -DMSO (Winship 1989, Varadaraj and Skinner 1994, Chakrabarti and Schutt 2002), non-ionic detergents (Demeke and Adams 1992), betaine (Baskaran et al. 1996, Weissensteiner and Lanchbury 1996, Henke et al. 1997), trehalose (Spiess et al. 2004, Samarakoon et al. 2013) and other compatible solutes have been used to solve the problems of formation of secondary structures in the template of cDNA (Ralser et al. 2006, Farell andAlexandre 2012).BSA (bovine serum albumin) has a high content of lysine which binds to phenolic compounds (Katcher and Schwartz 1994) and polysaccharides (Demeke and Adams 1992), avoiding polymerase inactivation (Kreader 1996).The beneficial effects of these additives on PCR reactions are often specific to the type of cDNA template and primer and they have to be empirically determined (Simonović et al. 2012).
Commercially available ready-to-use solutions, containing additives, have led to good results, but they possess the disadvantage of having unknown composition and the lack of flexibility to adjust the concentration of individual components (Ralser et al. 2006).
RNA extraction was made with Trizol (Reagent Tri-Phasis -Bio Agency).Approximately 1/3 of the volume of an Eppendorf of 2 mL was filled with a mixture 1/1 (v/v) of young (2°+3°) and mature (8°) internode powders (in order to identify as many isoforms possible) from the five replicates and 1.5 mL of Tri-Phasis reagent was added.Further on the extraction followed the manufacturer recommendations.The quantification of total RNA was made in Nanodrop spectrophotometer (Thermo Scientific) at 260 nm and the purity evaluated by 260 nm/280 nm ratio.The RNA quality was checked by 1% agarose electrophoresis with ethidium bromide and further visualization under UV light (Gel Doc 2000, BioRad).RNA samples were treated with DNAse (Turbo DNAse-free kit, Ambion) and submitted to reverse transcription using the RevertedAid First Strand cDNA Synthesis kit, according to manufacturer's instructions (Thermo Scientific).

In silico ANALYSIS OF DATABASE AND DESIGN OF PRIMERS
The sequences of the genes coding for the enzymes: 4-hydroxicinnamoyl CoA: ligase (4CL), cinnamoyl CoA reductase (CCR), ferulate were carried out in the NCBI, Gene Index and Phytozome databases.Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) = positive control.We used as bait the sequences more expressed in hybrid sugarcane tissues identified by Bottcher et Aiming to isolate genes of the lignin biosynthesis in S. spontaneum for further studies on gene expression and sugarcane transformation, we used formamide, BSA, trehalose, betaine, DMSO and PEG-8000 as additives in RT-PCR reactions.
We were able to amplify all genes of the lignin biosynthesis route.

PLANT MATERIAL
The plants used in the study were clones obtained  2013).Among the sequences returned from the databases we selected those from sorghum (Sorghum bicolor), rice (Oryza sativa), corn (Zea mays), wheat (Triticum aestivum), rye grass (Lolium perenne) and Arabidopsis thaliana.Only sequences with full CDS, low e-value (10-6) and verified by BLASTx were used (Altschul et al. 1997).These sequences were aligned in the BioEDIT software (Hall 1999) and from the alignment, primers were designed in conserved regions with Primer 3 (Rozen and Skaletsky 2000), using as parameters Tm in the range of 57°C -60°C, 2°C of maximum difference between forward and reverse primers and % GC between 55 and 60.When this was not possible, degenerate primers were designed.The avoid formation of secondary internal structures and dimers the primers were verified by the OligoAnalizer 3.1 software.The primer sequences, GC content and the expected size of the amplicon are in Table I.

TABLE I
Primers used in the RT-PCR reactions, expected amplicon size (base pairs) and % of G+C in the amplicon.Non-specific bases (degenerate primers) are indicated in parenthesis.

RESULTS AND DISCUSSION
Preliminarily, RT-PCR reactions were carried out without additives (Figure 1).Eight of the 10 genes analyzed did not have good amplification or they did not amplify.The amplicons were nonspecific to PAL and C3H; they were not observed with HCT, F5H, 4CL and CCR, or showed low yield with CAD and C4H.
Only COMT and CCoAOMT genes presented good amplification without additives, observing intense and specific bands at the expected size of 756 and 649 bp, respectively.C4H and C3H reactions produced bands with the expected size of 1281 and 1343 bp, respectively, but with low specificity and low yield.The genes coding for these two enzymes could be simply and satisfactorily  amplified when 2 µL of cDNA was used in the RT-PCR reactions, increasing yield and specificity of the bands (Figures 2a and 2b).
For CAD gene amplification, all PCR reactions were initially set up with 1 µL of cDNA and, preliminarily, additives were tested solely and at fixed concentrations, according to the literature (Lorenz 2011, Simonović et al. 2012).Only treatments 0.04% BSA and 0.04% BSA + 2% formamide allowed CAD gene amplification, and the combination of both additives showed better results in relation to the standard reaction, without additives (see Figure 1), generating a specific band with 943 bp, as expected (Figure 3a).The low efficiency with BSA may be due to its sensitivity to high temperatures used in the RT-PCR reaction, suggesting that its role would be more like a co-additive in the amplification of "templates" containing high GC % and PCR inhibitors (Farell and Alexandre 2012).The combination of BSA with formamide seemed to proportionate a synergistic effect, stabilizing Taq polymerase by inhibitors binding (mainly phenols) to BSA (Woide et al. 2010, Samarakoon et al. 2013), and formamidecaused destabilization of GC rich regions, thereby decreasing Tm and facilitating the extension step in the reaction.In other reactions using BSA and formamide, but varying template amount, 1.5µL or 2 µL of cDNA proved most efficient for CAD gene amplification (Figure 3b).No amplification products were observed for F5H, 4CL, HCT and CCR genes without additives (see Figure 1).Taking as reference the results with CAD, reactions were carried out using 0.04% BSA + 2% formamide (Figure 4a).Although specific bands were observed for F5H, CCR, and 4CL genes (625, 823 and 1340 bp, respectively), the yield was low.When more cDNA was added to the reactions the combination 3 µl of cDNA + 0.04% BSA + 2% formamide gave good results for these genes (Figure 4b).
The need to employ additives in reactions with CAD, F5H, CCR, and 4CL may be related to the fact that we used degenerate primers.Thus, the primers would be in fact a "pool" of oligomers and could be leading to the formation of secondary structures, "hairpin loops" or dimers, which would be competing with the annealing process in the RT-PCR reaction, decreasing amplification.(Singh et al. 2000, Linhart and Shamir 2002, Strien et al. 2013).Formamide decreases the Tm of cDNA template and consequently optimizes the annealing process of the primer (Simonović et al. 2012).Formamide is much more efficient as an additive when used at a concentrations up to 5%, its effectiveness being abolished at 10% (Chakrabarti and Schutt 2010), due to inhibition of Taq polymerase activity (Varadaraj and Skinner 1994).For this reason, we used an intermediate concentration (2%), but the yield was improved in the reactions of CAD, CCR, 4CL and F5H genes only with cDNA supplementation (Figure 4b).
BSA + formamide combination was not effective for amplification of HCT gene, which has the highest % GC among genes herein studied (Table I).No amplification band was observed in the expected size of 1317 bp (Figure 4a).Others additives with unknown or not well established function are being successfully tested in various experiments of gene amplification by RT-PCR (Nordstrom et al. 2006, Spink et al. 2007).Trehalose (0.1 M) and 8% PEG-8000 were chosen as additives to overcome the amplification problem of HCT in reactions containing 0.04% BSA + 2% formamide.BSA + formamide + trehalose allowed the amplification of HCT, resulting in a band with 1317 bp, as expected (Figure 5).PEG-8000 did not generate positive results.Additional cDNA was not needed to obtain a good yield and specificity.
It has been reported that besides decreasing the Tm of cDNA, trehalose also has a thermal stabilizer effect of Taq polymerase.(Weissensteiner and Lanchbury 1996, Carninci et al. 1998, Spiess et al. 2004, Samarakoon et al. 2013).The fact that trehalose is an innocuous chemical compound and does not inhibit Taq polymerase suggests that in the reactions with HCT it would be acting in synergism with formamide in the annealing process, resulting in minimal competition between the annealing process and the formation of hairpinloops or dimers in the pool of primers used in this experiment.
Little success was observed in the amplification of PAL gene using the basic conditions of RT-PCR (see Figure 1), with 3 diffuse and non-specific bands, below the expected size of 2033 bp.The addition of BSA and formamide also did not result in success (Figure 4a).PAL was amplified with the addition of 2% DMSO, presenting intense bands on expected 2033 bp position (Figures 6a  and 6b).DMSO can stimulate RT-PCR reactions within the final concentration range of 1% to 10% (Jung et al. 2001, Kitade et al. 2003, Simonović et al. 2012).However, other studies have shown that above 2% this solvent can inhibit the activity of Taq polymerase (Lorenz 2011), which defined the concentration used in our study.
Taken together, our data shows that the appropriate use of single or combined additives in RT-PCR reaction mixtures can significantly improve the expression effectiveness of genes of lignin biosynthesis in sugarcane.The guidelines for optimizing amplification may be applied to other recalcitrant plant genes.
from tissue culture material provided by the Agronomic Institute of Campinas (IAC), Ribeirão Preto, São Paulo-Brazil.Culms were cut in segments containing two nodes and were planted in plastic trays containing vermiculite and kept in the greenhouse, with daily irrigation.After 40 days, the seedlings obtained were transplanted to plastic vessels of 50 L (3 plants per vessel) containing organic substrate (nitrogen= 0.5%, moisture = 50%, organic carbon = 15%, pH = 6, C/N = 130 mmol/ Kg, CIC/C ratio= 8%, Genefertil-Genesol) and kept in the greenhouse under sprinkler irrigation (30 minutes, twice a day) for a period of approximately one year.After this period, the vessels were transferred to full sun, out of the greenhouse and plants were left to grow for another four months, under sprinkler irrigation (30 minutes, three times a day).Plant material was harvested in June 2014, with 5 replicates (5 plants).Healthy culms were collected from internodes 2 nd +3 rd (young) and 8 th (mature), not including the node.With the aid of a blade the culms were cut into small pieces, frozen in liquid nitrogen and freeze-dried.The lyophilized material was ground to a powder in sterile mortars and cooled with liquid nitrogen and then stored in freezer -80°C.JUAN P.P. LLERENA, PEDRO ARAÚJO and PAULO MAZZAFERA al. ( /C)GA(G/C)GAGGTGAAG GT(C/T)CCACTCCTTGAGGCACT 2033 (67%) GADPH TTGGTTTCCACTGACTTCGTT CTGTAGCCCCACTCGTTGT --------------

Figure 1 -
Figure 1 -RT-PCR amplification of lignin biosynthesis genes using standard reaction conditions as described in the text.cDNA was produced from a pool of young (2°+3°) and mature (8°) internodes from S. spontaneum.The numbers indicate the base pairs of the expected amplicon and arrows indicate weak bands.bp = base pairs (DNA 1 kb ladder New England Biolabs).GAPDH = positive control.Negative control reactions (template omitted) did not amplify any bands (not shown).

Figure 2 -
Figure 2 -RT-PCR amplification of C4H (a) and C3H (b) using standard reaction conditions as described in the text but 1 or 2 µl of cDNA were included in the reaction mixture.cDNA was produced from a pool of young (2°+3°) and mature (8°) internodes from S. spontaneum.The numbers indicate the base pairs of the expected amplicon and arrows indicate weak bands.bp = base pairs [DNA 1 kb ladder from Thermo Scientific (a) and from New England Biolabs (b)].Negative control reactions (template omitted) did not amplify any bands (not shown).

Figure 4 -
Figure 4 -RT-PCR amplification of HCT, F5H, PAL, CCR and 4CL using 0.04% BSA + 2% formamide and 1 µl cDNA in the reaction mixture (a) and F5H, 4CL and CCR using BSA and formamide but 3 µl cDNA in the reaction mixture (b).cDNA was produced from a pool of young (2°+3°) and mature (8°) internodes from S. spontaneum.The numbers indicate the base pairs of the expected amplicon and arrows indicate weak bands.bp = base pairs (DNA 1 kb ladder from New England Biolabs).Negative control reactions (template omitted) did not amplify any bands (not shown).

Figure 5 -
Figure5-RT-PCR amplification of HCT and C4H using 0.04% BSA + 2% formamide and 1 µl cDNA in the reaction mixture, supplemented either with 8% PEG-8000 or 0.1 M trehalose.GAPDH was used as a control with PEG-8000 and trehalose.The expected amplicons of HCT and C4H were 1317 and 1281 base pairs.cDNA was produced from a pool of young (2°+3°) and mature (8°) internodes from S. spontaneum.bp = base pairs (DNA 1 kb ladder from New England Biolabs).Negative control reactions (template omitted) did not amplify any bands (not shown).

Figure 6 -
Figure 6 -RT-PCR amplification of PAL using standard reaction conditions as described in the text (a) or including 2% DMSO in the reaction mixture (b).cDNA was produced from a pool of young (2°+3°) and mature (8°) internodes from S. spontaneum.Arrows indicate weak bands.bp = base pairs (DNA 1 kb ladder from New England Biolabs).Negative control reactions (template omitted) did not amplify any bands (not shown).