Effects of PHENYLALANINE AMMONIA LYASE (PAL) knockdown on cell wall composition, biomass digestibility, and biotic and abiotic stress responses in Brachypodium

Abstract

The phenylpropanoid pathway in plants synthesizes a variety of structural and defence compounds, and is an important target in efforts to reduce cell wall lignin for improved biomass conversion to biofuels. Little is known concerning the trade-offs in grasses when perturbing the function of the first gene family in the pathway, PHENYLALANINE AMMONIA LYASE (PAL). Therefore, PAL isoforms in the model grass Brachypodium distachyon were targeted, by RNA interference (RNAi), and large reductions (up to 85%) in stem tissue transcript abundance for two of the eight putative BdPAL genes were identified. The cell walls of stems of BdPAL-knockdown plants had reductions of 43% in lignin and 57% in cell wall-bound ferulate, and a nearly 2-fold increase in the amounts of polysaccharide-derived carbohydrates released by thermochemical and hydrolytic enzymic partial digestion. PAL-knockdown plants exhibited delayed development and reduced root growth, along with increased susceptibilities to the fungal pathogens Fusarium culmorum and Magnaporthe oryzae. Surprisingly, these plants generally had wild-type (WT) resistances to caterpillar herbivory, drought, and ultraviolet light. RNA sequencing analyses revealed that the expression of genes associated with stress responses including ethylene biosynthesis and signalling were significantly altered in PAL knocked-down plants under non-challenging conditions. These data reveal that, although an attenuation of the phenylpropanoid pathway increases carbohydrate availability for biofuel, it can adversely affect plant growth and disease resistance to fungal pathogens. The data identify notable differences between the stress responses of these monocot pal mutants versus Arabidopsis (a dicot) pal mutants and provide insights into the challenges that may arise when deploying phenylpropanoid pathway-altered bioenergy crops.

Keywords: Bioenergy; Fusarium; ferulic acid; grass; herbivory; lignin; phenylpropanoid; saccharification; tyrosine ammonia lyase; ultraviolet light..

Fig. 1.

Bradi3g49250 (BdPAL1) and Bradi3g49260 (BdPAL2) transcript levels in Brachypodium CTL and BdPAL RNAi mutant tissues, and related protein activity. (A) Relative amounts of BdPAL1 and BdPAL2 transcripts in various CTL plant tissues, as determined by qRT–PCR. (B) BdPAL1 and BdPAL2 transcript levels in BdPAL RNA1-1 and 1-3 third internode stem tissues compared with the CTL. (C) Amounts of PAL activity (left three columns) and TAL activity (right three columns) in culm first internode extracts. Note that PAL activity is ~25-fold higher than TAL activity. In (A), means (represented by columns) were normalized to the first internode mean transcript level value, which was set to 1, using expression of BdUBC18 (Bradi4g00660) as reference. Bars represent standard deviations (SDs). Different letters above columns represent significant differences. n=4 biological reps and n=3 technical reps for (A) and (B), and 2 and 2 for (C).

Fig. 2.

Digestibility of stem biomass, comparing BdPAL RNAi1-1 and 1-3 with empty vector control (CTL) and the WT. Shown are the average amounts of glucose (top chart) and pentose sugars (bottom chart) released per milligram of senesced ground biomass using the various pre-treatments along with partial enzyme hydrolysis. Vertical bars represent the SD. Different letters for a given treatment represent significant differences. n=5 biological reps, n=3 technical reps.

Fig. 3.

BdPAL RNAi cell walls have altered lignin composition and reduced FA and pCA content. (A) Phloroglucinol staining of stem sections. Lighter staining is consistent with less lignin. Note that vascular bundles (marked by an arrow) appear normal (e.g. no collapsed cells) in both RNAi lines. (B) Amounts of acid-insoluble and acid-soluble lignin, as well as lignin S:G units ratios, in senesced stems plus leaf sheaths. (C) The amounts of ferulate (FA) and p-coumarate (pCA) released from lignin (pCA) and hemicelluloses (FA and pCA) upon 2M NaOH treatment. Vertical bars represent the SD. For (B), each column represents a biological rep with three technical reps averaged; significant differences can be found in Supplementary Table S3 at JXB online. For (C), n=3 biological reps and n=2 technical reps. Asterisks denote Student’s t-test significant differences (P<0.01) between each PAL knocked-down line and its corresponding control grown concurrently (BdPAL RNAi1-1 versus WT; BdPAL RNAi1-3 versus CTL).

Fig. 4.

Growth of BdPAL RNAi1-1 and 1-3 plants compared with the WT or empty vector control (CTL). (A) Time course measurements of the longest leaf lengths and tallest culm heights of BdPAL RNAi1-1 plants. Arrows delineate the time points at which 50% of the spikelets could be seen emerging from the leaf whorl. (B) Average daily root growth of BdPAL RNAi1-1 and WT seedlings. n=48. (C, D) Average above-ground senesced biomass dry weight (C) and final culm heights (D). Columns grouped together represent plants grown concurrently. Bars denote the SD. Asterisks represent Student’s t-test significant differences (P<0.03). 5<n<25.

Fig. 5.

Effect of BdPAL silencing on Fusarium disease resistance. (A) Comparison of the foliar necrotic lesions that developed on wounded leaves of WT Bd21-3 and BdPAL RNAi1-1 following F. culmorum infection. The image shows necrotic lesions 7 days post-infection (dpi). n=32. (B) Comparison of the root necrotic lesion sizes developed on WT and BdPAL RNAi1-1 roots following F. culmorum infection. 22<n<24. Image shows seedlings 9 dpi. Asterisks in (A) and (B) represent Student’s t-test P-value <0.001, comparing BdPAL RNAi1-1 with the WT for a given time point. (C, D) Comparison of the effect of ACC (C) and AVG (D) treatments on the necrotic lesion sizes developed on WT and BdPAL RNAi1-1 roots, 7 d following F. culmorum infection. 20<n<24. For (C, D), asterisks represent Student’s t-test P-values <0.001, which relate to the comparison of each treatment with its corresponding untreated control. For all graphs, error bars represent the standard error (SE).