Tailoring poplar lignin without yield penalty by combining a null and haploinsufficient CINNAMOYL-CoA REDUCTASE2 allele

Lignin causes lignocellulosic biomass recalcitrance to enzymatic hydrolysis. Engineered low-lignin plants have reduced recalcitrance but often exhibit yield penalties, offsetting their gains in fermentable sugar yield. Here, CRISPR/Cas9-generated CCR2(−/*) line 12 poplars have one knockout CCR2 allele while the other contains a 3-bp deletion, resulting in a 114I115A-to-114T conversion in the corresponding protein. Despite having 10% less lignin, CCR2(−/*) line 12 grows normally. On a plant basis, the saccharification efficiency of CCR2(−/*) line 12 is increased by 25–41%, depending on the pretreatment. Analysis of monoallelic CCR2 knockout lines shows that the reduced lignin amount in CCR2(−/*) line 12 is due to the combination of a null and the specific haploinsufficient CCR2 allele. Analysis of another CCR2(−/*) line shows that depending on the specific CCR2 amino-acid change, lignin amount and growth can be affected to different extents. Our findings open up new possibilities for stably fine-tuning residual gene function in planta.

correlation (HSQC) NMR spectra (aromatic regions) of extract-free, ball-milled, whole-cell-wall material swelled in DMSO-d 6 :pyridine-d 5 (4:1 v/v). Volume-integrals (means ± standard deviation) are given for the lignin aromatic and substructure units that are labeled and color-coded to match their assignments in the spectrum. The aromatic percentages are on a S+G+H=100% basis and the sidechain (not shown) percentages are on an A+B+C=100% basis. pBA (p-hydroxybenzoate) and FM (ferulate marker) levels are given on an S+G+H=100% basis; note that pBA levels are over-represented because they are slowly-relaxing pendent groups on the lignin sidechain. Source data are provided as a Source Data file. Sequence information of the targeted CCR2 locus in poplar. For each poplar line, the target sequences of the CRISPR/Cas9 construct are shown. Both alleles (the upper sequence is the P. tremula CCR2 allele and the lower sequence is the P. alba CCR2 allele) and the indel patterns are shown. gRNA2 (underlined) and protospacer adjacent motif (PAM; bold text) sequences are highlighted for the wild type. In CCR2(+/-) line 114, also a substitution or insertion (in red) occurred. All lines had a monoallelic mutation in the CCR2 locus. The status of the CCR2 alleles present in P. tremula x P. alba is denoted between the parentheses; the first one represents that of the P. tremula allele, the second one that of the P. alba allele; +, wild type; -, knock-out; *, protein-modified. All other lines had monoallelic mutations in the CCR2 locus. The status of the CCR2 alleles present in P. tremula x P. alba is denoted between the parentheses; the first one represents that of the P. tremula allele, the second one that of the P. alba allele; +, wild type; -, knock-out; *, protein-modified.

Supplementary Table 1. Weekly height of CCR2(-/*) line 12.
Height was measured on a weekly basis during a growth period between 8 and 20 weeks in the greenhouse. Values in cm are given as means ± standard deviation. P value of the two-tailed Student's t-test is given. Wild type, n = 10 biologically independent samples; CCR2(-/*) line 12, n = 11 biologically independent samples. Source data are provided as a Source Data file.

Supplementary Table 3. Saccharification results of CCR2(-/*) line 12.
Plants were grown for 20 weeks in the greenhouse. Means ± standard deviation of glucose yield expressed as percentage of cell wall residue (CWR), cellulose-to-glucose conversion efficiency (%), or the glucose yield per plant (based on the dry weight of the stem). Samples were saccharified for 72 h using no pretreatment, acidic pretreatment (1 M HCl), or alkaline pretreatment (62.5 mM NaOH). For all pretreatments tested, CCR2(-/*) line 12 had an increased glucose yield compared to the wild type (two-tailed Student's t-test, the exact P values are shown in the table; wild type, n = 10 biologically independent samples; CCR2(-/*) line 12, n = 11 biologically independent samples). Source data are provided as a Source Data file.

Supplementary Note 1. The phenotypes of CCR2(-/*) line 12 can be attributed to the specific mutations present in the CCR2 alleles
While reviewing our manuscript, the referees were concerned that the observed phenotypes in CCR2(-/*) line 12 (lignin alterations, red xylem phenotype, normal growth) might be the consequence of offtarget mutations. Our reply (and corresponding changes made to the manuscript) convinced the referees of our claims and is shown below.
(A) In silico prediction of off-targets. The theoretical possibility of off-targets is extremely low. (i) Offtarget mutation in another CCR gene family member which might explain the typical CCR-deficient phenotypes such as lower lignin amount, red xylem phenotype and increased incorporation of ferulic acid in the lignin, is excluded by our in silico analysis: the eight closest homologs of CCR2 in P. tremula x P. alba (Potri.001G046100, Potri.001G046400, Potri.001G045500, Potri.001G045100, Potri.001G045900, Potri.001G045800, Potri.001G045000 and Potri.T134100) have 3 or 4 mismatches with gRNA1 (5'-GACCAAAAATGTGATCATTG-3') and, more importantly, also lack the subsequent and necessary NGG PAM sequence to allow genome editing in their respective homologous target regions. (ii) The chance of having an off-target mutation in other regions of the Populus tremula x P. alba genome is also very unlikely by itself. As off-target effects decrease when the number of mismatches increases, most genome editors consider 4 or more mismatches between the gRNA and any genomic sequence too much to be considered as potential off-targets 5,6 . Following this rule, in silico analysis predicted gRNA1 to have only one region in the genome of P. tremula x P. alba that could potentially be considered as an off-target (i.e., < 4 mismatches and presence of PAM sequence). However, this region still contains 3 mismatches with gRNA1 and, moreover, because it is intergenic (between Potri.001G282700 and Potri.001G282800), even a possible off-target mutation of this region would be very unlikely to lead to the typical CCR-deficient phenotypes described above. We conclude that, based on our in silico analysis, it is highly unlikely that the phenotypes observed in CCR2(-/*) line 12 are caused by an off-target mutation.
phenomenon is described elaborately in literature. We discovered that CCR2(-/*) line 12 had a mildly reduced lignin amount (i.e., not below the critical level). Biomass yield was therefore not affected in CCR2(-/*) line 12. This is the most likely explanation for the observed phenotype. By contrast, the lignin amount in CCR2(-/*) line 206, which is higher than that of CCR2(-/-), but still significantly reduced when compared to that of wild type, has dropped below the critical level and is therefore insufficient to maintain a normal growth phenotype. To better frame our finding, we added a paragraph to the discussion describing the link between CCR2 activity (and the amino acid changes), lignin amount and growth, and discuss what is known about this issue in literature.