**5. Humanization of moss glycosylation and an odd interspecies confusion**

All plants, including mosses, such as *P. patens,* are surprisingly apathetic towards switching off core fucosylation and xylosylation. Briefly, after the first cloning of the relevant enzymes fucosyl transferase and xylosyl transferase [15, 16], knock-down *Arabidopsis* and *Nicotiana* plants were generated [17, 18] and knock-out lines of *P. patens* were established for the production of biopharmaceuticals [19, 20]. The mainly resulting paucimannosidic structures, that is, N-glycans with two to five (a few = *pauci* in contrast to many = *oligo*) mannose residues and possibly one GlcNAc, xylose, and/or fucose residue were considered suboptimal for most purposes, and hence, efforts were undertaken to increase the levels of GnGn and to

*Protein Glycosylation in Bryophytes Differs Subtly from That in Vascular Plants DOI: http://dx.doi.org/10.5772/intechopen.107035*

### **Figure 3.**

*Sections of mass spectra of two mosses representative of mosses with plant-typical Lewis A antennae (panel A) and those with larger glycans (panel B). Mass increments indicate the addition of a 160.07 Da element, which could be explained as the addition of a pentose and a methyl group. Panel A shows the high mass section of N-glycans from Physcomitrium sphaericum (Ludw) Fuernr and panel B that of Amblystegium serpens (Hedw.) Schimp.*

eventually introduce the mammalian type core α1,6-fucose [21]. A big step toward humanized N-glycans was the introduction of β1,4-galactosylation [22]. The most stunning success, however, was the introduction of the entire mammalian sialylation pathway into plants—no less than six mammalian foreign genes that have to be actively expressed and correctly localized without harshly compromising the physiology of the host plant. This was at first realized with *Nicotiana* [23, 24] and then later with *P. patens* [3, 25]. As planned, these cloning steps led to the biosynthesis of substantial percentages of sialylated N-glycans. This established *P. patens* as in fact being very "patient", that is, tolerant and enduring towards this enormous intervention into the cellular machinery. Notably, certain otherwise difficult-to-express glycoproteins can be produced in suitable quality and quantity by glycoengineered *P. patens* [26].

However, next to some incomplete intermediate products, peaks with hitherto unknown mass levels occurred in mass spectra of moss lines expressing human β1,4-galactosyl transferase [27]. The mass increments of 132 Da indicated the addition of—against all rules—pentose. Being sensitive to α-arabinofuranosidase, this pentose was identified as furanosidic L-arabinose in α-linkage [27]. Its exact location was not known at the time of writing this chapter.
