Heterologous glycosyl hydrolase expression and cellular reprogramming resembling sucrose-induction enable Zymomonas mobilis growth on cellobiose

Plant derived fuels and chemicals from renewable biomass have significant potential to replace reliance on petroleum and improve global carbon balance. However, plant biomass contains significant fractions of oligosaccharides that are not usable natively by many industrial microorganisms, including Escherichia coli, Saccharomyces cerevisiae, and Zymomonas mobilis. Even after chemical or enzymatic hydrolysis, some carbohydrate remains as non-metabolizable oligosaccharides (e.g., cellobiose or longer cellulose-derived oligomers), thus reducing the efficiency of conversion to useful products. To begin to address this problem for Z. mobilis, we engineered a strain (Z. mobilis GH3) that expresses a glycosyl hydrolase (GH) with β-glucosidase activity from Caulobacter crescentus and subjected it to an adaptation in cellobiose medium. Growth on cellobiose was achieved after a prolonged lag phase in cellobiose medium that induced changes in gene expression and cell composition, including increased expression and secretion of GH. These changes were reversible upon growth in glucose-containing medium, meaning they did not result from genetic mutation but could be retained upon transfer of cells to fresh cellobiose medium. After adaptation to cellobiose, our GH-expressing strain was able to convert about 50% of cellobiose to glucose within 24 hours and use it for growth and ethanol production. Alternatively, pre-growth of Z. mobilis GH3 in sucrose medium enabled immediate growth on cellobiose. Proteomic analysis of cellobiose- and sucrose-adapted strains revealed upregulation of secretion-, transport-, and outer membrane-related proteins, which may aid secretion or surface display of GHs, entry of cellobiose into the periplasm, or both. Our two key findings are that Z. mobilis can be reprogrammed to grow on cellobiose as a sole carbon source and that this reprogramming is related to a natural response of Z. mobilis to sucrose that enables sucrose secretion.

We also tested cellobiose adaptation of Z. mobilis GH3 by serial passage using Z. mobilis 262 pVector as a control (Fig 3). The first passage was performed from RMG to RMC with 0.4 mM 263 IPTG and all subsequent passages performed in RMC+IPTG (Fig 3A). To prevent carryover of 264 extracellular GH between passages, cells were pelleted, washed, and resuspended in fresh After the first passage to RMC, little to no growth was observed over the course of 3 days 278 for both Z. mobilis pVector and Z. mobilis GH3 (Fig 3B). After the second passage, Z. mobilis 279 GH3 cell density more than doubled after two days whereas control cells did not grow (Fig 3C).
280 Z. mobilis GH3 continued to grow robustly after the third passage with a doubling time of <24 281 hours (Fig 3D) 292 sucrose (RMCS) and growth was monitored (Fig 4A). We found that sucrose-grown cells 293 resumed growth efficiently on cellobiose without a long lag phase, similar to RMC-adapted cells 294 (Fig 2B), but neither RMG-grown Z. mobilis GH3 nor pVector control cells grew significantly.
295 However, RMG-grown Z. mobilis GH3 did eventually resume growth after a lag, as seen 296 previously (data not shown). This finding suggests that sucrose can induce changes in Z. mobilis 297 that enable GH3-mediated growth on cellobiose. To investigate the effects of sucrose on GH 298 activity, we assayed and compared GH activity of cellobiose-, sucrose-, and glucose-grown Z.
299 mobilis GH3. We found that both cellobiose-and sucrose-grown cells exhibited higher 300 extracellular and whole cell GH activity than glucose-grown cells (S5 Fig) consistent (Fig 4B). pVector 315 control cells were only able to grow in RMC+0.8% sucrose, suggesting that 0.8% sucrose is the 316 minimum amount of sucrose that will support growth of Z. mobilis in RMC medium (Fig 4B). 324 mobilis GH3, here defined as cells grown in RMS for 48 hours, were also inoculated into fresh 325 RMC with increasing amounts of sucrose (i.e., RMC+0.2-0.8% sucrose). In each culture, the 326 sucrose-adapted cells consumed almost all sucrose in the medium by 24 hours and after which 327 cells continued to grow on cellobiose (Fig 4B and S5C Fig). These results suggest that 0.2% 16 328 sucrose is sufficient to remodel Z. mobilis GH3 for growth on cellobiose, but that adaptation in 329 higher sucrose concentrations will support more cell growth and greater rates of cellobiose

346
We observed greater similarity in intracellular protein levels (R 2 = 0.37) between cellobiose-347 and sucrose-adapted cells when normalized to glucose-grown cells whereas extracellular protein 348 levels were less similar across conditions (R 2 = 0.17) (S7 Fig). Given our observations that 349 extracellular GH activity increases during adaptation to cellobiose and that sucrase and 350 levansucrase are known to be secreted from Z. mobilis (23) in response to sucrose, we looked at 17 351 levels of secretion-related proteins in both sucrose-and cellobiose-adapted cells. Notably, the 352 levels of a majority of annotated transport and secretion-related proteins increased in both 353 sucrose-and cellobiose-adapted cells (Fig 5, S7 Fig). Interestingly, GH3 CC_0968 was also 354 upregulated in both extracellular and intracellular fractions in both cellobiose and sucrose media 355 (Fig 5, S8 Fig)  To understand the nature of protein remodeling during sucrose and cellobiose adaptation 367 more completely, we performed K-means clustering on normalized log2-fold change values 368 (normalized to glucose-grown cells) for proteins that were measured in both the intracellular and 369 extracellular fractions (1199 proteins in total). In total, 50 clusters were produced revealing 370 similar remodeling patterns between sucrose-and cellobiose-adapted cells (Fig 6).  (Figs 5 and 6). These results are consistent with GO-term enrichment 378 analysis of differential protein levels in each sample compared to glucose-grown cells where GO 379 terms related to stress (oxidoreductase), secretion, and transport (periplasmic space, OM, 380 transport, and protein secretion) were enriched in both intra-and extracellular fractions of 381 cellobiose-and sucrose-adapted strains (S4 Table). Within the large cluster of primarily 382 downregulated proteins several growth-related pathways were enriched including the ATP 383 synthase, ribosome, translation, and purine and pyrimidine biosynthesis pathways (Fig 6). A 384 decrease in growth-related proteins can be attributed to the reduced growth rate of cells 385 metabolizing cellobiose or sucrose relative to cells growing on glucose.