Desirable feedstock qualities are largely dependent on the nature of processing technologies (Carroll and Somerville 2009). Currently, the two major biomass processing technologies are: thermal conversion and biological conversion. In thermal conversion (e. g., direct combustion or pyrolysis), it is more desirable to have feedstock with a lower amount of mineral residues and a higher energy content, which often correlates with a high lignin content of the biomass (Boateng et al. 2008). In biological conversion for biofuel production, the feedstock with lower lignin content has higher saccharification efficiency through enzyme hydrolysis, resulting in an increase in enzymatic fermentation efficiency (Fu et al. 2011a). Other cell wall components, such as hemicellulose (Lee et al. 2009) and pectin (Lionetti et al. 2010), also have a negative impact on bioenergy production using biochemical conversion technologies.

About 80 percent of the dry plant biomass is comprised of plant cell walls, which stores most of the biomass energy (Vogel and Jung 2001). Cellulose, hemicellulose, and pectin are the polysaccharide components of plant cell walls, of which cellulose is the primary component for biofuel (ethanol) production via fermentation (Carroll and Somerville 2009). Cell walls, especially secondary cell walls, are strengthened by lignin, a phenolic polymer derived from hydroxycinnamyl alcohols and produced by means of combinatorial radical coupling reactions (Boudet 2007). Lignin deposition reinforces plant cell walls to enable water transport, provide mechanical support and a barrier to pathogens, and help convey abiotic tolerance (Halpin 2004; Boudet 2007). However, high lignin content is not desirable for bioconversion of the lignocellulosic feedstock to biofuel for three reasons: 1) it prevents access of the hydrolytic enzymes to the polysaccharides, 2) it absorbs the hydrolytic enzymes, and 3) it inhibits the activities of the hydrolytic and fermentable enzymes used in the biofuel conversion process (Halpin 2004; Endo et al. 2008; Abramson et al. 2010). Studies using different alfalfa transgenic lines, with variable reduced lignin content, proved the negative correlation between lignin content and fermentable sugar release efficiency (Chen and Dixon 2007). Therefore, there is a strong interest in developing low-lignin content switchgrass cultivars for biofuel production.

The grass lignin polymer is usually composed of three monolignols [hydroxyphenyl (H), guaiacyl (G), and syringyl (S)] (Hatfield et al. 1999). Monolignols are derived from the amino acid phenylalanine through the monolignol biosynthesis pathway. The pathway has about ten key enzymes that catalyze the reaction steps and the pathway is evolutionarily conserved across angiosperms (Rastogi and Dwivedi 2008). Gene families encoding these key enzymes went through a rapid expansion after the divergence of monocots and dicots (Xu et al. 2009). By BLASTing against the switchgrass EST database, and utilizing phylogenetic analysis, we can find switchgrass homologs of all monolignol biosynthesis genes in model plant species. Through gene-expression patterns, in vitro enzymatic assays, and the generation of stable RNAi transgenic plants, a few switchgrass genes [4 Coumarate:Coenzyme A Ligase (4CL1,2), Cinnamyl Alcohol Dehydrogenase (CAD1,2), Catechol-O-methyltransferase (COMT)] in the monolignol biosynthesis pathway have been identified. RNAi: PvCOMT, RNAi: PvCAD2 and RNAi: Pv4CL1 transgenic plants have significantly less lignin content than wild type plants (Fu et al. 2011a, b; Saathoff et al. 2011a, b; Xu et al. 2011b).

Fu et al. (2011a) cloned a COMT cDNA and down-regulated its expression in cv. Alamo. Up to 90% of the COMT transcript was reduced and over 70% reduction in COMT enzyme activity was observed. Lignin content was reduced by 6 to 15% while the S/G ratio of the lignin was reduced from 0.69-0.71 in control plants to 0.37-0.40 in transgenics, mainly by reduction of S lignin content. The growth and development of transgenic plants appeared normal and the height and fresh and dry weight were similar to the controls under greenhouse conditions. Transgenic lines increased the ethanol yield by up to 38% using conventional biomass fermentation processes. The down-regulated lines required less severe pretreatment and 300-400% lower cellulase dosages for equivalent product yields using simultaneous saccharification and fermentation with yeast. Furthermore, fermentation of diluted acid-pretreated transgenic switchgrass using Clostridium thermocellum with no added enzymes showed better product yields than obtained with unmodified switchgrass. Therefore, this apparent reduction in the recalcitrance of transgenic switchgrass has the potential to lower processing costs for biomass fermentation-derived fuels and chemicals significantly (Fu et al. 2011a).

Saathoff et al. (2011b) demonstrated that switchgrass cv. Kanlow has at least two functional CAD genes. RNAi approach using 575 bp of PviCAD2 coding fragment (96% identical to the same region of PviCAD1 gene) was employed in an attempt to silence both CAD genes (Saathoff et al. 2011a). The CAD transcripts, protein amount and enzyme activities were substantially reduced in most transgenic lines. Four transgenic lines with single transgene copy were further analyzed. The lowest CAD activity found in these lines was less than 10% of that in the vector controls. The total lignin and cutin amount in these lines were reduced by 23% in average compared to the vector controls. Two transgenic lines had significantly higher glucose release after alkaline pretreatment and enzymatic saccharification. In a similar approach, Fu et al. (2011b) also cloned a CAD (PvCAD) cDNA from swithcgrass, which has 98-99% identity at amino acid level with the predicted proteins encoded by the PviCAD1 and 2 genes. Phylogenetic analysis suggests the gene is involved in lignin biosynthesis. For the eight RNAi transgenic plants analyzed, the extractable CAD activities were only 17-39% of that in control plants (using coniferaldehyde as a substrate). The transgenic plants grew normally in the greenhouse. Total lignin content in transgenic plants was 14-22% lower than in controls as determined by the acetyl bromide method, and both S and G lignins were reduced. In addition, chlorogenic acid, a soluble phenolic compound, was substantially increased in transgenic plants. Without acid pretreatment, transgenic plants released 28-59% more glucose with enzymatic hydrolysis than did the controls while 15-35% more glucose release was observed with pretreatment. Similarly, saccharification efficiency (total sugar release) increased by 19-89% without pretreatment, and by 19-44% with pretreatment. Sugar release was negatively correlated to lignin content but not to the S/G ratio, indicating reduced lignin content is the main reason for the improved sugar release.

Xu et al. (2011b) identified two 4CL genes in switchgrass. Phylogenetic and gene-expression pattern and enzymatic activity analyses suggest that Pv4CL1, but not Pv4CL2, is involved in monolignol biosynthesis. The RNAi:Pv4CL1 T0 transgenic plants downregulated Pv4CL1 expression to 0.05-0.73 fold of the WT controls. The 4CL enzyme activity was reduced by 80% on average as measured in T transgenic plants. The above-ground biomass yield of the transgenic plants was comparable to WT controls grown in the greenhouse conditions while brown color was seen in midvein, internodes, and mature roots of some transgenic plants. Pooled Tx transgenic plants had 22% reduction of the acid-insoluble lignin as well as total lignin. Lignin composition was also changed. T1 transgenic plants had 47% less G lignin and 45% more H lignin, than non-transgenic T segregates. Dilute acid-pretreated samples enhanced enzymatic hydrolysis of glucan but not xylan. With pretreatment, transgenic plant materials yielded 57.2% more fermentable sugar than the WT plants. In a similar effort, two highly homologous 4CL cDNAs were isolated and an RNAi construct attempting to suppress both genes were introduced into switchgrass. Up to 90% of the transcripts of both genes were suppressed. Although the total lignin content was not changed or only modestly reduced (up to 5.8%), the structure and composition of lignin appeared to have altered. That was reflected by significant reduction of acid insoluble lignin (up to 8.5) and increase of the ratio of acid soluble lignin vs. acid insoluble lignin (ASL/AIL increase by 21.4-64.3%), and by the increase of S/G ratio (11.8-164.5% higher in transgenic plants). Consequently, with alkaline pretreatment, glucan and xylan conversion efficiency of the best transgenic plant was increased by 16% and 18%, respectively (Wang et al. 2012).

The lignin biosynthesis pathway is regulated by complex transcription networks that involve many transcriptional activators and repressors. The transcriptional repressors could simultaneously inhibit the expression of several genes in the monolignol pathway, which could be another way to reduce lignin synthesis. Very recently, a transcription factor gene involved in lignin biosynthesis, PvMYB4, was cloned and characterized (Shen et al. 2012). PvMYB4 is an R2R3-MYB transcriptional repressor. PvMYB4 binds to the AC-rich AC-I, AC-II, and AC-III elements of monolignol pathway genes in vitro in EMSA assays, and down-regulates these genes in vivo. Ectopic overexpression of PvMYB4 in transgenic switchgrass, under control of the ZmUbi1 promoter, reduces lignin content by at least 40 to 50 percent; however, the S/G monolignol ratio remains unchanged. Additionally, the ester-linked p-CA: FA ratio in these plants is reduced by approximately 50 percent. Monosaccharide release after enzymatic saccharification, without acid pretreatment, is threefold higher in these transgenic plants.

However, total sugar release from cell wall residues remained the same. The morphology of the transgenic switchgrass was affected. The plant height was reduced by an average of 40 percent, but tiller numbers could be increased as much as 2.5 fold. Whether or not the total biomass was affected remains to be investigated (Shen et al. 2012).

Maize Corngrass1 (Cg1) gene encodes a grass-specific tandem repeat of miR156 gene, which "promotes juvenile cell wall identities and morphology" (Chuck et al. 2011). Overexpression of the gene, like in the maize mutant, increases biomass due to continuous initiation of tillers and leaves, and had less lignin and more glucose and other sugars in the leaves. Chuck et al. (2011) overexpressed Cg1 cDNA in switchgrass in an attempt to improve its biomass yield and the feedstock quality. In a field test, high and moderate expressers were dwarfed and had smaller leaves and lower yield of biomass whereas yield of low expressers were comparable to WT controls while producing four time more branches. The transgene affects flowering: none of the transgenic plants ever flowered after being grown in the field for two summers and a winter. This may be a favorite trait to prevent transgene flow. Total lignin content was moderately reduced in transgenic plants. Interestingly, the low expressers accumulated more than 250% starch in their stems compared to the WT controls. Consequently, saccharification using a mix of amyloglucosidase and a-amylase without pretreatment released 3-4 times more glucose from stems of the low expressers, which was similar to the amount released by dilute acid pretreatment, indicating that pretreatment could be reduced or completely eliminated in saccharification. In a similar approach, Fu et al. (2012) overexpressed a rice OsmiR156b precursor gene in switchgrass. Low expressers flowered normally, moderate expressers had reduced height and did not flower, while high expressers’ growth was severely stunted. Both low and moderate expressers had improved biomass yield: 58-101% more than the controls in greenhouse condition, mainly attributing to increase in tiller number. Solubilized sugar production (g/plant) after acid pretreatment and enzymatic hydrolysis increased by 40-72%.