Quantitative trait loci (QTL) underlying phenotypic variation in bioethanol-related processes in Neurospora crassa

Bioethanol production from lignocellulosic biomass has received increasing attention over the past decade. Many attempts have been made to reduce the cost of bioethanol production by combining the separate steps of the process into a single-step process known as consolidated bioprocessing. This requires identification of organisms that can efficiently decompose lignocellulose to simple sugars and ferment the pentose and hexose sugars liberated to ethanol. There have been many attempts in engineering laboratory strains by adding new genes or modifying genes to expand the capacity of an industrial microorganism. There has been less attention in improving bioethanol-related processes utilizing natural variation existing in the natural ecotypes. In this study, we sought to identify genomic loci contributing to variation in saccharification of cellulose and fermentation of glucose in the fermenting cellulolytic fungus Neurospora crassa through quantitative trait loci (QTL) analysis. We identified one major QTL contributing to fermentation of glucose and multiple putative QTL’s underlying saccharification. Understanding the natural variation of the major QTL gene would provide new insights in developing industrial microbes for bioethanol production.

Introduction metabolism and identify superior combinations of alleles for trait enhancement for CBP. 83 To this end, we chose to perform QTL analysis on a laboratory generated population of 84 the model filamentous fungus Neurospora crassa. 111 offspring and two parental strains 85 were genotyped by sequencing, evaluated for their ability to decompose cellulose and 86 ferment glucose, and subjected to QTL analysis. A major QTL was identified for 87 fermentation and multiple putative QTL's were identified for saccharification of 88 cellulose. (HGLM) (1X Vogel's salts, 2% glucose, 0.5% L-Arginine, pH 5.8) were used to generate 97 mycelial mats in petri plates, which were used to generate replicate mycelial pads for 98 each experiment using a bore punch.

100
Enzyme activity assay 101 A modified FPA assay was performed using 96-well plates as described by Camassola 102 and Dillon, in which secreted protein extracts were taken from wells in 6-well plates containing 1% CMC broth (1X Vogel's salts, 1% CMC) 4 days after inoculation with 10 gauge mycelial pads [27]. Culture broth containing secreted enzyme was filtered and 105 centrifuged at 13.2k rpm to remove any fungal cells or debris. 50μL of supernatant was 106 added to 100μL of 50mM sodium acetate buffer pH 5.6 in a 96-well deep-well plate, 107 which was then equilibrated to 50°C for 5 min in a hot-water bath. A 5mg strip of 108 Whatman Grade No. 1 filter paper was submerged in the solution, and the plate was 109 incubated at 50°C for 60 min. After 60 min, 300μL of 3,5-dinitrosalicylic acid (DNS) 110 Reagent was added to stop the enzymatic reaction and visualize glucose equivalents. The  To characterize fermentation of glucose among the lab generated N6 first filial (F1) 123 generation and the parental strains (FGSC 2223 and FGSC4825), fermentation was 124 carried out in a 96-well format in deep-well plates. Replicates for each strain were 125 collected from mycelial mats grown from spore suspension in high glucose liquid media 126 (HGLM) with a 6 gauge punch and inoculated into 750µl of HGLM (2% glucose), sealed with aluminum ThermowellTM seals, and allowed to ferment for 7 days in 12:12 128 Light/Dark conditions at 25°C. All samples were performed in biological quadruplicate.

129
After fermentation, 600 µL of media was recovered and cell debris was removed by 130 sequential centrifugation at 13.2k rpm for 5 min. The recovered supernatant was  that were not polymorphic between parents were removed, followed by any markers in 147 which one of the two parents was missing data. The resulting 4900 markers were 148 formatted for Excel for further filtering. Chi-square tests were performed for each marker 149 and those markers with unequal segregation among progeny (>20% disparity) were removed, followed by markers with >10% missing data among progeny. The filtered 151 genotype data was combined with phenotype data from FPA and glucose fermentation 152 assays and formatted for R-QTL. The formatted data was imported into R-QTL as and 153 the create map function was used to generate a linkage map from a physical map based

Results
To test if there exist a natural variation of the bioethanol-related processes in one 173 mapping population N6, we performed a QTL analyses in N6. Significant variation was 174 observed among the N6 population for both saccharification of cellulose and 175 fermentation of glucose (Fig 1). Both traits demonstrated transgressive segregation, with   however, none of these were above the LOD threshold for 90% CI. Using the fitqtl function, it was observed that the 4 QTL markers account for 210~26% of the observed variation in fermentation of ethanol. The QTL on LG I accounts 211 for 13% of the observed variation, while QTL on LG IV, VI and VII account for 4, 7, and 212 6% of the variation, respectively. Interestingly, marker segregation analysis revealed that 213 the parental genotypes contributing to greater ethanol production varied between 214 markers, with parental genotype A accounting for higher production at marker Nc-M878 215 on LG I and genotype B accounting for higher production at marker Nc-M4092 on LG VI 216 (Fig 4A and B).  16 Unspecified), although none were above the LOD threshold for 90% CI (Fig 5).

234
Combined, the putative QTL account for 19.5% of the total variation observed in 235 saccharification of cellulose. The major QTL on LG II accounts for 11% of the observed 236 variation, while minor QTL on LG VI and VI account for 4.6% and 5.7%, respectively. Marker segregation analysis at QTL markers on LG II and LG IV revealed that 243 genotype B was the main contributor to increased potential for saccharification of 244 cellulose (Fig 6A and B). From this finding, we chose to investigate the potential for LG IV (Nc-M2984) (Fig 7A). Marker interaction plots demonstrated that the genotype B specific increases in FPA activity only held for the interactions between QTL markers on 251 LG II and LG IV, but not with the interactions between adjacent markers on LG II ( Fig   252   7B and C). sensing or degrading cellulose. However, two major regulatory genes of interest were 269 also identified, vib-1 and xlr-1, on LG II and LG IV respectively (Table 2). In current study, we have identified that there exists a substantial variation in bioethanol-