**3. GHs identified in sugarcane and** *B. distachyon* **cell wall proteomes**

In sugarcane, 49 GHs have been identified in cell wall proteomes [28, 29]. They are distributed in 16 GH families. The GH3 family is the best represented (~20%), followed by GH17 (~16%), GH18 (~12%), and GH1 (~8%) (**Figure 1**). This distribution varies according to organ and developmental stage. In cell suspension cultures, only 4 GH families were identified among which GH3 was the most populated [28]. In 2-month-old stems, 7 GH families were found, GH3 also being the most represented [29]. Leaves recovered few GHs, from families 19, 27 (young leaves only), 28, and 31 (young leaves only). Apical internodes mainly contained GH3 members, whereas mostly GH17 members were found in basal ones [30]. Noteworthy, it should be mentioned that the absence of some GH families in a given cell wall proteomes could be due to technical limitations or differential accessibility as a consequence of differences in cell wall structure.

In *B. distachyon*, 114 CW GHs were identified in cell wall proteomes [31–34]. They are distributed into 21 families. The most populated one was GH17 (~17%), followed by GH28 (~13%), GH1 (~9%), GH3 (~8%), and GH35 (~8%) (**Figure 1**). GH28, followed by GH1, GH3, and GH16, had the highest number of members in young leaves. In mature leaves, GH17, GH18, and GH28 were the most represented. In internodes, they were GH28 and GH17. In seedlings and seeds, GH17 was the most populated family.

**Figure 1.** GH family distribution of presently known cell wall proteomes in sugarcane (a) and *B. distachyon* (b). The number of family members identified in each of them is indicated. Same colors indicate same GH families in both species.

The large size of the GH1, GH17, and GH28 families is probably linked to their roles in the assembly and in the rearrangement of cell wall polysaccharides [17]. Usually the GH1, GH16, GH17, and GH35 families are less represented in dicots than in monocots [31]. GH17 display glucan-1,3-β-glucosidase activity and possible substrates could be mixed (1,3)(1,4)-β-D-glucans [8]. This is consistent with the fact that only type II grass cell walls present this kind of hemicellulose.

After a survey of the cell wall proteomes described so far and collecting information regarding microorganism enzymes used for biomass deconstruction, we decided to focus this review on the GH1, GH3, GH17, GH27, GH35, and GH51 families. We have predicted functional and structural domains in newly identified CWPs using the PredictProtein bioinformatic software and grouping them in families [35]. Since plant cells perform cell expansion themselves by involving cell wall polysaccharide rearrangements, the plant mechanisms could be mimicked by the enzymes used in cocktails. The comparison of plant and microorganisms enzymes presently used for biomass hydrolysis could contribute to determining their common characteristics and which specificities of plant enzymes could be copied in order to improve industrial cell wall deconstruction processes. Conversely, this comparative study could help in identifying which of the characteristics of microorganism enzymes could be engineered in plant species in order to obtain biomass with less recalcitrant cell walls.
