TrendsBecause many of the roles assumedby protein glycosylation in eukaryotesare not applicable to Bacteria orArchaea, this postmodification likelyserves distinct roles in prokaryotes.

In Bacteria, protein glycosylation sys-tems are found in nonpathogenic spe-cies, pointing to roles beyondvirulence.

In Bacteria and Archaea, protein gly-

ReviewSweet New Roles for ProteinGlycosylation in ProkaryotesJerry Eichler1,* and Michael Koomey2

Long-held to be a post-translational modification unique to Eukarya, it is nowclear that both Bacteria and Archaea also perform protein glycosylation,namely the covalent attachment of mono- to polysaccharides to specific pro-tein targets. At the same time, many of the roles assigned to this protein-processing event in eukaryotes, such as guiding protein folding/quality control,intracellular trafficking, dictating cellular recognition events and others, do notapply or are even irrelevant to prokaryotes. As such, protein glycosylation mustserve novel functions in Bacteria and Archaea. Recent efforts have begun toelucidate some of these prokaryote-specific roles, which are addressed in thisreview.

cosylation contributes to the integrityand proper architecture of glycopro-tein-containing assemblies.

Changes in protein glycosylation offersprokaryotes a rapid and reversiblemanner in which to respond to envir-onmental changes.

1Department of Life Sciences, BenGurion University of the Negev,Beersheva 84105, Israel2Department of Biosciences,University of Oslo, 0316 Oslo, Norway

*Correspondence:jeichler@bgu.ac.il (J. Eichler).

Prokaryotic Protein Glycosylation – Understanding the How but Not the WhyAs the list of completed genome sequences keeps growing, it is becoming increasingly clearthat the number of protein-coding genes cannot alone account for the size of an organism’sproteome. Sources of proteomic expansion include the various post-translational modifi-cations (see Glossary) a given protein can undergo. Of the various protein-processing eventsthat have been described, glycosylation, namely the covalent linkage of mono- to polysac-charides, is one of the most prevalent and probably the most complex (Box 1) [1,2]. Longthought to be restricted to Eukarya, it is now accepted that both Bacteria and Archaea also arecapable of N-glycosylation, where glycans are amide-bonded to select Asn residues of atarget protein, as well as O-glycosylation, where glycans are added to hydroxyl-presentingamino acids, particularly Ser and Thr [3–6] (Table 1). Despite the fact that numerous glyco-proteins have been identified in Bacteria and Archaea [7,8], that the structures of many of theglycans decorating prokaryal glycoproteins have been solved [9–13], and that considerableprogress has been made in delineating pathways of protein glycosylation in several bacterialand archaeal species [6,14,15], the roles served by the bacterial and archaeal versions of thisuniversal post-translational modification remain poorly defined.

In Eukarya, numerous functions have been assigned to protein glycosylation. The glycosylationprocess begins in the endoplasmic reticulum, the first stop on the secretory pathway, where alipid-bound polysaccharide core is transferred to target protein Asn residues. The N-linkedglycan is then augmented by individual sugars, also transferred from lipid carriers, to yield acomplex branched oligosaccharide [16–18]. The composition of the N-linked glycan dictatesinteractions of the modified protein with molecular chaperones, such as calnexin and calre-ticulin, and other enzymes that accommodate proper protein folding [19,20]. Indeed, thermo-dynamics-based studies have demonstrated the importance of protein glycosylation for proteinfolding [21,22]. At the same time, the same N-linked glycan structure is monitored by thequality-control system responsible for identifying aberrantly folded proteins and targeting themfor degradation, if necessary [23–25]. Once an N-glycosylated protein has successfully navi-gated the coordinated protein folding and quality control steps, it may be delivered to the Golgi,the next station along the secretory pathway, via a sorting process that can also rely on N-linked

662 Trends in Microbiology, August 2017, Vol. 25, No. 8 http://dx.doi.org/10.1016/j.tim.2017.03.001

© 2017 Elsevier Ltd. All rights reserved.

GlossaryArchaellum: the motility structure ofArchaea, functionally equivalent tothe bacterial flagellum.Autotransporters: found in a broadrange of Gram-negative bacteria,autotransporters comprise a family ofouter membrane or secreted proteinsthat facilitate their own transport tothe cell surface. In such proteins, theautotransporter domain, comprisingthe C-terminal portion of the protein,forms a beta-barrel structure in theouter membrane through which theN-terminal domain is presented onthe cell surface. Autotransporters areassociated with virulence,contributing to adhesion,aggregation, invasion, biofilmformation, and toxicity.N-glycosylation: the covalentlinkage of glycans to select Asnresidues in a target protein throughan amide bond.O-glycosylation: the covalentlinkage of glycans to hydroxyl-presenting amino acids, particularlySer and Thr.Oligosaccharyltransferase:oligosaccharyltransferases catalyzethe transfer of glycans from the lipidcarriers upon which they areassembled onto selected residues inglycoproteins. In Bacteria, Archaea,and lower eukaryotes, theoligosaccharyltransferase acts alone,whereas in higher Eukarya, theenzyme exists as a multimericcomplex.Post-translation modification: anevent that follows translation,designed to create variants of agiven protein through the covalentattachment of one or more of severalclasses of molecules (e.g., sugars,lipids, or small chemical groups, likeacetyl or methyl groups), theformation of intra- or intermolecularlinkages (e.g., disulfide bonds),proteolytic cleavage (e.g., signalpeptide removal), and/or anycombination thereof.Sequon: sequence motifs in apolypeptide denoting sites whereglycans are attached. In N-glycosylation, the Asn of a sequon,namely an Asn-Xaa-Ser/Thrsequence, where Xaa is any residuebut Pro, is modified. Variations to thecanonical sequon have beenobserved in prokaryotes. Thesequons processed in O-glycosylation are less well defined.

Box 1. Glycosylation As a Source of Protein Diversity

Of the different post-translation modifications to which a given protein can be subjected, glycosylation introduces themost diversity. Several factors are responsible for the enormous variability associated with protein glycosylation. Inaddition to the variability derived from how a glycan is linked to a protein (e.g., N-linked or O-linked), considerablediversity is generated at the level of individual sugars comprising protein-linked glycans. For instance, the incorporationof sugars that differ in the number of backbone carbons (e.g., pentoses and hexoses), that can exist in different epimericforms (e.g., glucose, mannose and galactose), and that can be distinguished via the addition of different chemicalgroups (e.g., amino or methyl groups) all contribute to glycan diversity. Further variablity arises when sugars start tooligomerize into a glycan due to the many possible linkages between any two sugars (in terms of both the position andstereochemistry of the connection), the possibility for branching, and the heterogeneity possible in a given oligosac-charide. Indeed, the variability of protein-linked glycans may be infinite because of the fact that no template limiting thesize of an oligosaccharide seems to exist. Together, these considerations result in a plethora of protein-linked glycansunique in composition and/or architecture.

glycan composition [26,27]. Once in the Golgi, the N-linked glycan is subjected to furtherprocessing through the addition and/or removal of constituent sugars to yield a range of N-linked oligosaccharides [28–30]. The Golgi is also the site of O-glycosylation, a second majorprotein glycosylation event that can also introduce considerable diversity into the glycosylationprofile of a glycoprotein [31,32]. In functional terms, the heterogeneity in glycan contentgenerated in the Golgi can be exploited for targeting different glycoproteins to distinct subcel-lular compartments [33,34] or, in the case of cell-surface-exposed glycoproteins, can contrib-ute to various cell–cell or other recognition events important for the development,differentiation, or physiology of a particular cell, tissue, or organism [35–38]. Such hetereo-geneity can, moreover, reflect different diseased states [39–42].

At the same time, it would appear that many of the roles assumed by protein glycosylation ineukaryotes are not relevant for prokaryotes. For instance, whereas protein folding and N-glycosylation are linked in the eukaryal secretory pathway, these processes occur on either sideof the plasma membrane in the case of bacterial and archaeal proteins secreted by the twinarginine translocation pathway. Here, such proteins fold in the cytoplasm [43,44], whileoligosaccharyltransferase-based N-glycosylation transpires on the outer surface of thebacterial and archaeal cell [45,46] (as do some versions of bacterial O-glycosylation [47]).Likewise, the need to sort proteins to distinct subcellular compartments is extremely limitedin prokaryotes. Finally, the number of recognition events required by a prokaryotic cell is likely tobe far less than its eukaryal counterpart. Therefore, Bacteria and Archaea must rely on proteinglycosylation for other purposes than how this post-translational modification is used in Eukarya(Figure 1, Key Figure). In this review, recent works addressing these roles are discussed.

Bacterial Protein Glycosylation – Not for Virulence AloneThe number of bacterial protein glycosylation systems recognized continues to grow. Thisprocess has been fueled in part by a few serendipitous discoveries followed by comparativegenomics that allows one to immediately see how broadly distributed protein glycosylationsystems (and genes) truly are. The take-home messages here are that (i) bacterial proteinglycosylation is much more prevalent than one could have imagined, and (ii) it is not strictlyassociated with pathogenic species. There are thus a number of outstanding questions relatingto the biological significance of bacterial protein glycosylation that are emphasized here.

In studying protein glycosylation in Bacteria, the predominant emphasis has been placed onpathogens. However, systems related to those found in pathogens abound in commensal andenvironmental isolates. Thus, bacterial protein glycosylation is not a canonical virulence factoras defined by the criteria established by Falkow [48]. Nonetheless, clear defects in colonizationand virulence in mammalian, insect, and plant model systems are seen for glycosylation nullmutants. This is particularly true of so-called dedicated systems in which the glycosylation

Trends in Microbiology, August 2017, Vol. 25, No. 8 663

S-layer: part of the cell envelope,the surface layer is a self-assemblingtwo-dimensional pseudocrystallinearray that covers the entire outersurface of the cell. Comprising asingle or a very small number ofproteins or glycoproteins, the S-layeris ubiquitous in Archaea and iscommon in Bacteria.

Table 1. N- and O-glycosylation in Eukarya, Bacteria, and Archaea

Eukarya Bacteria Archaea

N-glycosylation

Essential Yes No Some species

Distribution Universal Limited Almost universal

Glycan diversity Common core Limited Extensive

Glycan lipid carrier Dolichol phosphate,dolichol pyrophosphate

Undecaprenolpyrophosphate

Dolichol phosphate, dolicholpyrophosphate

Oligosaccharyltranferase

Number of subunits Single ormultiple

Single Single

Catalytic subunit STT3 PglB AglB

Processing of protein-bound glycan

Yes No Yes

Different glycans onsame protein

No No Yes

O-glycosylation

Distribution Universal Some species Unknown

Glycan diversity Extensive Extensive Limited

Transfer of assembledglycan from lipid carrier

No Possible Unknown

Processing of protein-bound glycan

Yes No Unknown

machineries target a specific protein. These most often involve major surface-localized entities,such as flagella, autotransporters, and type IV pili.

Bacterial flagella come in both O-glycosylated and non-glycosylated forms [49]. In those caseswhere flagellin subunits undergo modification, null glycosylation mutants show a remarkablearray of disparate phenotypes. In Helicobacter, Campylobacter, and Aeromonas, as well asGram-positive Clostridia and Listeria, glycosylation-null mutants are nonflagellated and displayfla� phenotypes [50–54]. In contrast, such mutants in Pseudomonas and Burkholderia retainflagella and motility [55,56]. A breakthrough in understanding how flagellin glycosylation mightexert its influence on filament assembly came from studies in Aeromonas caviae. Using amutant lacking the Maf1 glycosyltransferase required for the transfer of pseudaminic acid toflagellin, Parker and colleagues established that glycosylation is dispensable for subunit exportbut is essential for filament assembly, since nonglycosylated flagellin is still secreted [57].Flagellin glycosylation in a chaperone mutant (flaJ) showed that glycosylation takes place in thecytoplasm (as long surmised but never previously documented) and before chaperone bindingand protein secretion. It was further shown that FlaJ preferentially bound glycosylated flagellin.It will be interesting and important to see if such findings can be extended to encompass theother flagellar glycosylation systems.

Bacterial autotransporters, adhesins of Gram-negative bacteria defined by shared structuralelements and surface localization mechanisms, also come in both glycosylated and non-glycosylated flavors. O-glycosylated autotransporters are modified with heptoses at multiplesites within the passenger domain via the action of cytoplasmic heptosyltransferases usingADP-heptose [58]. Three such adhesins have been characterized in Escherichia coli, namelyAIDA-1, Ag43, and TibA. Surprisingly, glycosylation null mutants have different effects onotherwise shared phenotypes but no major quantitative effects on secretion or surface

664 Trends in Microbiology, August 2017, Vol. 25, No. 8

Key Figure

Protein Glycosylation Serves Distinct Roles in Eukarya and in Bacteriaand Archaea

BacteriaArchaea

Eukarya

Protein foldingandquality control

Differen�alsubcellulartarge�ng

Cell–celland other

recogni�onevents

Assembly and strength of glycoprotein-based structures

Remodeling in response toenvironmental

changes

Virulence,host–microbe

interac�ons

Protein glycosyla�onProtein glycosyla�on

Figure 1. Given differences in the temporal relationship and location of protein glycosylation, in the structure of the cell, ininteractions with biological molecules or other cells, and in lifestyle, eukaryotes and prokaryotes (Bacteria and Archaea) relyon protein glycosylation for distinct purposes.

localization. Such mutants in the Ag43 system were found to be defective in associatedfunctions, such as adherence and autoagglutination, yet showed enhanced binding to ele-ments of the extracellular matrix [59]. In contrast, heptosyltransferase null mutants in the TibAsystem were altered solely in adherence phenotype [60]. These findings are consistant withothers, revealing how poorly understood the structure–function relationships of autotransport-ers are with regard to their different phenotypes. The recent, highly resolved structure of aglycosylated 300 residue-long TibA peptide represents a major advance in this arena [61]. Thatsame study also defined a unique family of structural autotransporter heptosyltransferases anddetermined their structure. Another breakthrough in this field was the identification of an entirelydifferent class of heptosyltransferases mediating N-linked glycosylation of a subset of auto-transporters in species including Actinobacillus, Kingella, and Aggregatibacter [62,63]. Here, thecytosolic N-glycosyltransferase utilizes nucleotide-activated monosaccharides to modify aspar-agine residues with a relaxed consensus sequon. These enzymes are related to HMW1C that N-glycosylates adhesins in Haemophilus influenzae [64,65]. Null mutants for the glycosyltransferaseare functionally defective for autotransporter function, as are the equivalent mutants in the O-linked systems. However, it remains unclear whether these phenotypes result from reducedfunction per se, as opposed to a reduction in surface localization. It is remarkable that despite thefact that the Aggregatibacter aphrophilus autotransporter is almost identical to that in Aggre-gatibacter actinomycetemcomitans, the former is N-glycosylated while the latter is O-glycosy-lated in a process involving enzymes used in the biosynthesis of the O-polysaccharide oflipopolysaccharide, including the WaaL-type O-antigen ligase [66]. Moreover, an HMW1C-likeN-glycosyltransferase is not readily identifiable in the A. actinomycetemcomitans genome.

N- and O-linked broad-spectrum (aka general) protein glycosylation systems encompass thosethat modify a large number of extracytoplasmically targeted substrates and are exemplified bythe N-linked system of Campylobacter jejuni and the O-linked systems of pathogenic Neisseriasp. and Mycobacterium tuberculosis, respectively. In C. jejuni, substrates are modified with en

Trends in Microbiology, August 2017, Vol. 25, No. 8 665

bloc transferred oligosaccharides by an oligosaccharyltransferase structurally related to theSTT3 subunit of the eukaryotic oligosaccharyltransferase complex [67,68]. The Neisseria sp.system involves the en bloc transfer of oligosaccharides by an oligosaccharyltransferasestructurally related to the WaaL family of O-antigen ligases [15,69], while the mycobacterialsystems rely on the transfer of mannose from polyprenol carriers by protein mannosyltransfer-ases related to their eukaryotic counterparts [70,71]. Despite their differences, a number ofshared themes connect these systems. First, membrane-linked protein substrates (lipoproteinsor proteins bearing membrane-spanning domains) are over-represented in all three. Likewise,the vast majority of glycoproteins in each are not predicted to be surface-exposed. In the C. jejuniand Neisseria systems, each utilizes a di-N-acetyl-bacillosamine sugar at the reducing end of theglycan that is synthesized and transferred to undecaprenylphosphate by a conserved andevolutionarily related set of components [72,73]. Although the Neisseria and mycobacterialsystems do not utilize classical targeting sequons, the sites of modified Ser and Thr residuesin each lie within interdomain regions of reduced complexity enriched in proline, as well asthreonine and serine residues. Taken together, these systems exemplify how amalgamationsof related components connect, which otherwise are perceived as being highly unrelated.

Broad-spectrum systems have also been identified in Acinetobacter and Burkholderia viabioinformatics, and each of these also encompass glycosylation of the pilin subunit proteinof the type IV pilus colonization factor [74,75]. Again, mutants defective in glycosylation areattenuated in infection models for both. It should be noted as a matter of curiosity that an activeprotein-targeting oligosaccharyltransferase (VC0393) has been identified in Vibrio cholerae thatO-glycosylates target proteins from Neisseria when coexpressed in Escherichia coli [76].However, protein glycosylation has yet to be observed in V. cholerae. In addition, a forwardgenetic screen using Tn-seq showed no fitness alterations during infection, dissemination, orsurvival in the aquatic environment for the VC0393 null mutant [77]. Future work needs toresolve the status of a potential V. cholerae protein O-glycosylation system.

Nearly all of these systems are associated with pathogens, and glycosylation null mutants havebeen seen to be colonization-defective or of reduced virulence in host model systems (oftenwhile showing few, if any, in vitro phenotypes) [74,78–81]. Understanding the mechanismsbehind these attenuated phenotypes is complicated by the large number of proteins lackingglycosylation. Thus, it remains unclear whether these phenotypes are the result of the non-glycosylation of a single protein or the cumulative effects of multiple protein perturbations.

As noted earlier, similar glycosylation systems are found in related commensal and environ-mental species in each of the above cases. The unique general O-linked system documented inBacteroides fragilis is particularly interesting as this species is considered to be an importantmember of the healthy gut microbiota [82]. Mutants with altered glycan structures were unableto compete with wild-type organisms in gut colonization. In addition, this seems to be uniqueamong broad-spectrum O-glycosylation systems in its use of a sequon-like element. Althoughthe oligosaccharyltransferase has yet to be defined in this system, the basic structure of theglycoforms (and some of the biosynthetic machinery), as well as a defined glycoproteome, hasbeen established.

Little attention is placed here on the detailed synthetic pathways and precise structures of theglycans/oligosaccharides themselves. Needless to say, there is an enormous amount ofdiversity to be found. That said, one can ask why specific glycoforms are associated withspecific systems. Might they be driven by forces of the innate and/or adaptive immunesystems? Or are they shaped by mere metabolic considerations relating to the costs ofsynthesizing certain oligosaccharides or conflicts with coexpressed glycoconjugate synthesispathways? In some cases, the systems merely tap into or co-opt a pre-existing pathway for

666 Trends in Microbiology, August 2017, Vol. 25, No. 8

lipopolysaccharide or capsular polysaccharide biosynthesis. It is conceivable that such glycansmight mimic host glycan structure and thus perturb or ameliorate immune recognition or, moregenerally, merely mask elements that might otherwise elicit immune responses [83]. In somebacteria, glycoform structure is very stable, while in others there are significant degrees of intra-and interstrain diversity. This situation is perhaps exemplified by the related C. jejuni andpathogenic Neisseria broad-spectrum systems. The former utilizes an incredibly conservedheptasaccharide, while the latter employs an array of glycoforms that can be expressed by asingle strain. Why is the system so constrained in the case of the former and so deconstrainedin the latter? Likewise, some flagellin systems are very static while in others, such as those ofHelicobacter, Campylobacter, and Clostridium species, large genomic islands associated withdramatic glycoform variability are found. With regard to the adaptive immune response, it isnotable that only few studies have examined antibody responses to protein-associated glycansor the effects of glycan-specific antibodies. This is surprising, first as it is well established thatconjugating oligosaccharides to a protein results in transforming such oligosaccharides into T-cell-dependent antigens, underlying a robust antibody response. Second, many glycoengin-eering endeavors are predicated on exploiting bacterial protein glycosylation systems toproduce vaccines targeting bacterial glycoconjugates (Box 2). Two recent works suggest thatglycan-specific antibodies may have significant biological activities. Of note, Pluschke andcolleagues identified a Bacillus anthracis lineage lacking expression of the spore surface-linkedoligosaccharide and suggested that this could be the result of prior immunization of cattle with avaccine based on oligosaccharide-positive spores [84]. More directly, Szymanski et al. reportedthat immunization with glycoconjugates bearing the C. jejuni N-linked heptasaccharide led to a10-log reduction in colonization of chickens, in association with the elicitation of glycan-directed antibodies [85]. Thus, while considerable progress has been made, there is still muchrelated to the roles played by protein glycosylation in Bacteria that remains to be elucidated.

Archaeal N-Glycosylation – Reasons to Sweeten at ExtremesLike Bacteria, Archaea are also capable of both N- and O-glycosylation [6,86]. Yet, whereas N-glycosylation is currently restricted to delta/epsilon proteobacteria [5], it seems to be an almostuniversal process in Archaea [87]. Because of this, and since only three examples of archaealO-glycosylation have been reported [88–90], the bulk of research into protein glycosylation inArchaea has focused on N-glycosylation. Such efforts have not only revealed novel aspects ofthis universal post-translational modification but have also provided insight into what purpose itserves in Archaea.

Having provided the first example of a non-eukaryal glycoprotein [88], halophilic (salt-loving)archaea have long served as models of choice for understanding N-glycosylation in Archaea.As such, not only are halophiles central to research aimed at defining archaeal N-glycosylation

Box 2. Glycoengineering with Bacteria: Escherichia coli As a Glycofactory

With the description of a pathway for N-glycosylation in Campylobacter jejuni [14], efforts subsequently focused onintroducing the ability to glycosylate proteins into E. coli, a bacterial workhorse of biotechnology. Accordingly, E. colitransformed with the gene cluster encoding components of the C. jejuni N-glycosylation pathway was able toglycosylate recombinant C. jejuni glycoproteins, as well as proteins into which a sequence recognized by the N-glycosylation machinery was inserted [68,117]. It was also reported that similar N-glycosylation was performed by E. coliexpressing the C. jejuni oligosaccharyltransferase PglB and enzymes involved in the assembly of different O-poly-saccharides, components of the lipopolysaccharide surrounding Gram-negative bacteria, demonstrating that variabilityof the glycan added to target proteins in engineered E. coli was possible [118]. More recently, efforts have focused ongenerating E. coli strains capable of producing glycoproteins bearing human N-linked glycans. In such efforts, genesencoding yeast enzymes common to the yeast and human N-glycosylation pathways have been introduced into E. colicells expressing C. jejuni PglB, yielding proteins bearing the core N-linked glycan of eukaryotic glycoproteins [119].Although much remains to be done before E. coli can serve as host cell for economically viable glycoprotein production,considerable efforts are being directed at optimizing glycosylation efficiency and glycoprotein yield, as well ashumanizing the added glycans.

Trends in Microbiology, August 2017, Vol. 25, No. 8 667

pathways [6], they have also helped elucidate the reasons for this post-translational modifica-tion. At present, these efforts have largely focused on three haloarchaeal glycoproteins, namelysurface (S)-layer glycoproteins that comprises the S-layer surrounding such cells, archaellins,the building blocks of the archaellum (corresponding to the archaeal counterparts of bacteriaflagellins and flagella, respectively), and pilins, the basic subunit of pili, another surfaceappendage found in haloarchaea [91–95]. Such studies have revealed that N-glycosylationis important for the integrity and strength of the structures assembled from these glycoproteins.In some of the earliest studies on archaeal N-glycosylation, it was shown that treatingHalobacterium salinarum with bacitracin, a compound known to interfere with other glycosyla-tion systems, not only affected S-layer glycoprotein mass but also converted these normallyrod-like cells into spheres [96]. More recently, it was shown that interfering with properHaloferax volcanii N-glycosylation through the deletion of genes encoding proteins involvedin this post-translational event rendered the S-layer more susceptible to proteolytic degradationor more readily released to the surrounding medium, apparently due to under-glycosylation ofthe S-layer glycoprotein [97–102]. The importance of N-glycosylation for Hfx. volcanii S-layerintegrity was further shown with membrane vesicles prepared from cells deleted of N-glyco-sylation pathway genes; such vesicles presented only partially intact S-layers [102]. Moreover, itwas reported that truncation of the pentasaccharide N-linked to the Hfx. volcanii S-layerglycoprotein resulted in reduced secretion of a reporter protein to the growth medium,presumably due to effects on the assembly or architecture of the S-layer [102] (Figure 2).

N-g

lyco

syla

�on

mut

ants

Plasma membranePlasma membrane

S-layerS-layer

Enhancedsensi�vity toadded protease

Increasedrelease intothe growthmedium

Loss of integrityupon vesicleforma�on

Decreased protein exportto the medium

Pare

nt

Figure 2. Compromised N-Glycosylation Has a Detrimental Effect on the Hfx. volcanii S-Layer. The Hfx.volcanii S-layer surrounding the cell comprises a single protein, the S-layer glycoprotein (schematically depicted in the toppanel, Parent strain). In mutants where N-glycosylation is perturbed by deletion of genes involved in the process, the S-layer shows increased sensitivity to added protease (second panel; [98–102]), is more readily released into the growthmedium (third panel; [97]), loses its integrity when right-side-out membrane vesicles are prepared (fourth panel; [102]), andinterferes with the export of protein to the growth medium (bottom panel; [102]).

668 Trends in Microbiology, August 2017, Vol. 25, No. 8

Outstanding QuestionsWhat additional roles does protein gly-cosylation serve in prokaryotes? Doeukaryotes also rely on protein glyco-sylation for purposes now seeminglyunique to prokaryotes?

Why are there so many different bac-terial protein glycosylation systems?Do the innate and/or adaptive immunesystems provide the driving force orare metabolic considerations relatingto the costs of synthesizing certainoligosaccharides responsible?

How widespread is the use of modifiedN-glycosylation as a response to envi-ronmental change in Archaea? Howdoes modified N-glycosylation contrib-ute to the ability of Archaea to adapt toenvironmental changes? Do Bacteriaalso modify protein glycosylation inresponse to changing surroundings?Does O-glycosylation change in theface of shifting environments?

How can prokaryotic protein glycosyl-ation be exploited for appliedpurposes?

Compromised N-glycosylation also led to defects in the assembly of archaella and pili in Hfx.volcanii, as well as losses of motility and adhesion, demonstrating the functional importance ofproper archaellin and pillin N-glycosylation, respectively [95,103].

The importance of archaellin N-glycosylation for archaellum assembly and motility was likewisedemonstrated in the methanogenic archaea Methanococcus voltae [104] and Methanococcusmaripaludis [105]. In the thermoacidophile Sulfolobus acidocaldarius, compromised N-glyco-sylation did not affect archaellum assembly, structure or stability, yet did compromise motility[106,107]. Disrupted N-glycosylation also interfered with proper S. acidocaldarius S-layerstructure [107]. Moreover, perturbed N-glycosylation substantially decreased S. acidocaldariusgrowth rates as the salinity of the surrounding medium increased [106,108]. It is noteworthythat in S. acidocaldarius, a species belonging to the major archaeal phylum Crenarchaeota, N-glycosylation is essential for survival [109], whereas this is not the case in Euryarchaeota, asecond major archaeal phylum that includes the halophiles and methanogens discussed above[104,110].

More striking than the contribution of N-glycosylation to the architecture, integrity, and functionof glycoprotein-based structures is the apparent involvement of this protein-processing event inarchaeal responses to changing environments. Although many Archaea are denizens of‘extreme’ surroundings, they too can experience environment-related stress, as conditionsin such habitats may not remain static. Post-translational modifications, including proteinglycosylation, provide a rapid and reversible response to such environmental changes. InHfx. volcanii, the S-layer glycoprotein is modified by an N-linked pentasaccharide [13]. How-ever, when the salinity of the growth medium dropped below a certain threshold, both dolicholphosphate, the lipid upon which N-linked glycans are assembled in Hfx. volcanii, and S-layerglycoprotein Asn-498, a position not modified when the cells are grown at higher salinity, weremodified by a tetrasaccharide of novel composition [111–113]. Moreover, if assembly of thepentasaccharide was compromised, the so-called ‘low salt’ tetrasaccharide was added toAsn-498 even under high-salt conditions, indicative of the coordinated behavior of the two N-glycosylation pathways in this organism. In other studies, it was reported that the N-glycosyla-tion profile of M. maripaludis archaellins was modified when the growth temperature exceededa certain threshold [114]. Whereas the archaellin FlaB is normally modified by a tetrasaccharideat its four N-glycosylation sites, in cells grown at temperatures �38�C, this protein was insteadmodified by only the first three sugars, with sugar three lacking the attached Thr residue andacetamidino group seen at this position at lower temperatures. It was further confirmed that theappearance of the truncated N-linked glycan was not the result of tetrasaccharide degradationat the elevated growth temperature. These results suggest that modulating the N-glycosylationpattern of proteins that undergo such post-translational modification somehow helps Archaeaadapt to changes in the environment. It remains, however, to be determined just how modifiedN-glycosylation contributes to such adaptability. With this in mind, it is worth noting that in Hbt.salinarum, the transcription factor TrmB regulates the expression of metabolic enzymes,including those involved in the production of sugars used in protein glycosylation, in responseto glucose levels [115]. In summary, novel uses for protein glycosylation by Archaea may reflectyet another creative solution these organisms employ to cope with the extreme environments inwhich they are often found.

Concluding RemarksProtein glycosylation is of undoubtable biological significance in a variety of prokaryotes.However, current studies into protein glycosylation clearly represent only the tip of an icebergand can be considered as having addressed the ‘low hanging fruit'. As seen by mutant analysesin Bacteria, the phenotypes are all over the map and in a number of cases, lack coherence. InArchaea, the limited number of species amenable to experimental manipulation represents an

Trends in Microbiology, August 2017, Vol. 25, No. 8 669

obstacle to the drawing of general conclusions. As such, although evidence for new andseemingly prokaryote-specific roles for protein glycosylation is starting to accumulate, itremains a challenge to define precise functions for this post-translational modification in suchorganisms (see Outstanding Questions). This is not new. In 1993, Varki summarized thepotential biological significance of oligosaccharides with the conclusion that “all of the theoriesare correct” [116]. When one includes prokaryotic protein glycosylation into the equation, thisconclusion may be more true now than ever.

AcknowledgmentsJ.E. was supported by grants from the Israel Science Foundation (ISF) (grant 109/16), the ISF within the ISF-UGC joint

research program framework (grant 2253/15), the ISF-NSFC joint research program (grant 2193/16) and the German-

Israeli Foundation for Scientific Research and Development (grant I-1290-416.13/2015). M.K. was supported in part by

Research Council of Norway (RCN) project 214442, as well as the Centre for Integrative Microbial Evolution at the

Department of Biosciences, University of Oslo.

References

1. Apweiler, R. et al. (1999) On the frequency of protein glycosyla-

tion, as deduced from analysis of the SWISS-PROT database.Biochim. Biophys. Acta 1473, 4–8

2. Spiro, R. (2002) Protein glycosylation: nature, distribution, enzy-matic formation, and disease implications of glycopeptidebonds. Glycobiology 2, 43R–56R

3. Zarschler, K. et al. (2010) Protein tyrosine O-glycosylation – arather unexplored prokaryotic glycosylation system. Glycobiol-ogy 20, 787–798

4. Iwashkiw, J.A. et al. (2013) Pour some sugar on it: the expandingworld of bacterial protein O-linked glycosylation. Mol. Microbiol.89, 14–28

5. Nothaft, H. and Szymanski, C.M. (2013) Bacterial protein N-glycosylation: new perspectives and applications. J. Biol. Chem.288, 6912–6920

6. Jarrell, K.F. et al. (2014) N-linked glycosylation in Archaea: astructural, functional and genetic analysis. Microbiol. Mol. Biol.Rev. 78, 304–341

7. Bhat, A.H. et al. (2012) ProGlycProt: a repository of experimen-tally characterized prokaryotic glycoproteins. Nucleic Acids Res.40, D388–D393

8. Schäffer, C. and Messner, P. (2017) Emerging facets of pro-karyotic glycosylation. FEMS Microbiol. Rev. 41, 49–91

9. Young, N.M. et al. (2002) Structure of the N-linked glycanpresent on multiple glycoproteins in the Gram-negative bacte-rium, Campylobacter jejuni. J. Biol. Chem. 277, 42530–42539

10. Kelly, J. et al. (2009) A novel N-linked flagellar glycan fromMethanococcus maripaludis. Carbohydr. Res. 344, 648–653

11. Nothaft, H. et al. (2012) Diversity in the protein N-glycosylationpathways within the Campylobacter genus. Mol. Cell Proteom.11, 1203–1219

12. Messner, P. et al. (2013) Bacterial cell-envelope glycoconju-gates. Adv Carbohydr. Chem. Biochem. 69, 209–272

13. Kandiba, L. et al. (2016) Structural characterization of theN-linked pentasaccharide decorating glycoproteins of the halo-philic archaeon Haloferax volcanii. Glycobiology 26, 745–756

14. Linton, D. et al. (2005) Functional analysis of the Campylobacterjejuni N-linked protein glycosylation pathway. Mol. Microbiol. 55,1695–1703

15. Aas, F.E. et al. (2007) Neisseria gonorrhoeae O-linked pilinglycosylation: functional analyses define both the biosyntheticpathway and glycan structure. Mol. Microbiol. 65, 607–624

16. Larkin, A. and Imperiali, B. (2011) The expanding horizons ofasparagine-linked glycosylation. Biochemistry 50, 4411–4426

17. Aebi, M. (2013) N-linked protein glycosylation in the ER.Biochim. Biophys. Acta 1833, 2430–2437

18. Bieberich, E. (2014) Synthesis, processing, and function of N-glycans in N-glycoproteins. Adv. Neurobiol. 9, 47–70

19. Caramelo, J.J. and Parodi, A.J. (2015) A sweet code for glyco-protein folding. FEBS Lett. 589, 3379–3387

670 Trends in Microbiology, August 2017, Vol. 25, No. 8

20. Lamriben, L. et al. (2016) N-Glycan-based ER molecular chap-erone and protein quality control system: the calnexin bindingcycle. Traffic 17, 308–326

21. Hanson, S.R. et al. (2009) The core trisaccharide of an N-linkedglycoprotein intrinsically accelerates folding and enhancesstability. Proc. Natl. Acad. Sci. U. S. A. 106, 3131–3136

22. Shental-Bechor, D. and Levy, Y. (2009) Folding of glycoproteins:toward understanding the biophysics of the glycosylation code.Curr. Opin. Struct. Biol. 19, 524–533

23. Hebert, D.N. and Molinari, M. (2012) Flagging and docking: dualroles for N-glycans in protein quality control and cellular proteo-stasis. Trends Biochem. Sci. 37, 404–410

24. Hebert, D.N. et al. (2014) The intrinsic and extrinsic effectsof N-linked glycans on glycoproteostasis. Nat. Chem. Biol.10, 902–910

25. Benyair, R. (2015) Glycan regulation of ER-associated degrada-tion through compartmentalization. Semin. Cell Dev. Biol. 41,99–109

26. Kamiya, Y. et al. (2012) Molecular and structural basis forN-glycan-dependent determination of glycoprotein fates in cells.Biochim. Biophys. Acta 1820, 1327–1337

27. Satoh, T. et al. (2014) Structural basis for disparate sugar-binding specificities in the homologous cargo receptorsERGIC-53 and VIP36. PLoS One 9, e87963

28. Stanley, P. (2011) Golgi glycosylation. Cold Spring Harb.Perspect. Biol. 3, a005199

29. Moremen, K.W. et al. (2012) Vertebrate protein glycosylation:diversity, synthesis and function. Nat. Rev. Mol. Cell Biol. 13,448–462

30. Medzihradszky, K.F. et al. (2015) Tissue-specific glycosylation atthe glycopeptide level. Mol. Cell. Proteom. 14, 2103–2110

31. Cylwik, B. et al. (2013) Congenital disorders of glycosylation:part II. Defects of protein O-glycosylation. Acta Biochim. Pol. 60,361–368

32. Bard, F. and Chia, J. (2016) Cracking the glycome encoder:signaling, trafficking, and glycosylation. Trends Cell Biol. 26,379–388

33. Coutinho, M.F. et al. (2012) Mannose-6-phosphate pathway:a review on its role in lysosomal function and dysfunction.Mol. Gen. Metabol. 105, 542–550

34. Karabasheva, D. et al. (2014) Roles for trafficking and O-linkedglycosylation in the turnover of model cell surface proteins.J. Biol. Chem. 289, 19477–19490

35. Corfield, A.P. (2015) Mucins: a biologically relevant glycanbarrier in mucosal protection. Biochim. Biophys. Acta 1850,236–252

36. Kavanaugh, D. et al. (2015) The intestinal glycome and itsmodulation by diet and nutrition. Nutr. Rev. 73, 359–375

37. Clerc, F. et al. (2016) Human plasma protein N-glycosylation.Glycoconj. J. 33, 309–343

38. Furukawa, J. et al. (2016) Glycomics of human embryonic stemcells and human induced pluripotent stem cells. Glycoconj. J.33, 707–715

39. Drake, R.R. (2015) Altered glycosylation in prostate cancer. Adv.Cancer Res. 126, 345–382

40. Kudelka, M.R. et al. (2015) Simple sugars to complex disease –

mucin-type O-glycans in cancer. Adv. Cancer Res. 126,53–135

41. Pinho, S.S. and Reis, C.A. (2015) Glycosylation in cancer:mechanisms and clinical implications. Nat. Rev. Cancer 15,540–555

42. Lauc, G. et al. (2016) Mechanisms of disease: the human N-glycome. Biochim. Biophys. Acta 1860, 1574–1582

43. Palmer, T. and Berks, B.C. (2012) The twin-arginine transloca-tion (Tat) protein export pathway. Nat. Rev. Microbiol. 10, 483–496

44. Berks, B.C. (2015) The twin-arginine protein translocation path-way. Annu. Rev. Biochem. 84, 843–864

45. Lechner, J. et al. (1985) Transient methylation of dolichyl oligo-saccharides is an obligatory step in halobacterial sulfatedglycoprotein biosynthesis. J. Biol. Chem. 260, 8984–8989

46. Nita-Lazar, M. et al. (2005) The N-X-S/T consensus sequence isrequired but not sufficient for bacterial N-linked protein glyco-sylation. Glycobiology 15, 361–367

47. Faridmoayer, A. et al. (2007) Functional characterization ofbacterial oligosaccharyltransferases involved in O-linked proteinglycosylation. J. Bacteriol. 189, 8088–8098

48. Falkow, S. (1988) Molecular Koch’s postulates applied to micro-bial pathogenicity. Rev. Infect. Dis. 10, S274–S276

49. Logan, S.M. (2006) Flagellar glycosylation – a new component ofthe motility repertoire? Microbiology 152, 1249–1262

50. Canals, R. et al. (2007) Non-structural flagella genes affectingboth polar and lateral flagella-mediated motility in Aeromonashydrophila. Microbiology 153, 1165–1175

51. Schirm, M. et al. (2004) Flagellin from Listeria monocytogenes isglycosylated with beta-O-linked N-acetylglucosamine. J. Bac-teriol. 186, 6721–6727

52. Schirm, M. et al. (2003) Structural, genetic and functional char-acterization of the flagellin glycosylation process in Helicobacterpylori. Mol. Microbiol. 48, 1579–1592

53. Shen, A. et al. (2006) A bifunctional O-GlcNAc transferasegoverns flagellar motility through anti-repression. Genes Dev.20, 3283–3295

54. Twine, S.M. et al. (2009) Motility and flagellar glycosylation inClostridium difficile. J. Bacteriol. 191, 7050–7062

55. Arora, S.K. et al. (2001) A genomic island in Pseudomonasaeruginosa carries the determinants of flagellin glycosylation.Proc. Natl. Acad. Sci. U. S. A. 98, 9342–9347

56. Hanuszkiewicz, A. et al. (2014) Identification of the flagellinglycosylation system in Burkholderia cenocepacia and the con-tribution of glycosylated flagellin to evasion of human innateimmune responses. J. Biol. Chem. 289, 19231–19244

57. Parker, J.L. et al. (2014) Maf-dependent bacterial flagellin gly-cosylation occurs before chaperone binding and flagellar T3SSexport. Mol. Microbiol. 92, 258–272

58. Lu, Q. et al. (2014) An iron-containing dodecameric heptosyl-transferase family modifies bacterial autotransporters in patho-genesis. Cell Host Microbe 16, 351–363

59. Reidl, S. et al. (2009) Impact of O-glycosylation on the molecularand cellular adhesion properties of the Escherichia coli auto-transporter protein Ag43. Int. J. Med. Microbiol. 299, 389–401

60. Cote, J.P. et al. (2013) Glycosylation of the Escherichia coli TibAself-associating autotransporter influences the conformationand the functionality of the protein. PLoS One 8, e80739

61. Yao, Q. et al. (2014) A structural mechanism for bacterial auto-transporter glycosylation by a dodecameric heptosyltransferasefamily. eLife 3, e03714

62. Naegeli, A. et al. (2014) Molecular analysis of an alternative N-glycosylation machinery by functional transfer from Actinobacil-lus pleuropneumoniae to Escherichia coli. J. Biol. Chem. 289,2170–2179

63. Rempe, K.A. et al. (2015) Unconventional N-Linkedglycosylation promotes trimeric autotransporter function inKingella kingae and Aggregatibacter aphrophilus. mBio 6,e01206

64. Grass, S. et al. (2010) The Haemophilus influenzae HMW1Cprotein is a glycosyltransferase that transfers hexose residues toasparagine sites in the HMW1 adhesin. PLoS Pathog. 6,e1000919

65. Gross, J. et al. (2008) The Haemophilus influenzae HMW1adhesin is a glycoprotein with an unusual N-linked carbohydratemodification. J. Biol. Chem. 283, 26010–26015

66. Tang, G. and Mintz, K.P. (2010) Glycosylation of the collagenadhesin EmaA of Aggregatibacter actinomycetemcomitans isdependent upon the lipopolysaccharide biosynthetic pathway.J. Bacteriol. 192, 1395–1404

67. Lizak, C. et al. (2011) X-ray structure of a bacterial oligosacchar-yltransferase. Nature 474, 350–355

68. Wacker, M. et al. (2002) N-linked glycosylation in Campylobac-ter jejuni and its functional transfer into E. coli. Science 298,1790–1793

69. Vik, A. et al. (2009) Broad spectrum O-linked protein glycosyla-tion in the human pathogen Neisseria gonorrhoeae. Proc. Natl.Acad. Sci. U. S. A. 106, 4447–4452

70. Dobos, K.M. et al. (1996) Definition of the full extent of glycosyl-ation of the 45-kilodalton glycoprotein of Mycobacterium tuber-culosis. J. Bacteriol. 178, 2498–2506

71. VanderVen, B.C. et al. (2005) Export-mediated assembly ofmycobacterial glycoproteins parallels eukaryotic pathways.Science 309, 941–943

72. Hartley, M.D. et al. (2011) Biochemical characterization of theO-linked glycosylation pathway in Neisseria gonorrhoeaeresponsible for biosynthesis of protein glycans containing N,N0-diacetylbacillosamine. Biochemistry 50, 4936–4948

73. Kelly, J. et al. (2006) Biosynthesis of the N-linked glycan inCampylobacter jejuni and addition onto protein through blocktransfer. J. Bacteriol. 188, 2427–2434

74. Lithgow, K.V. et al. (2014) A general protein O-glycosylationsystem within the Burkholderia cepacia complex is involved inmotility and virulence. Mol. Microbiol. 92, 116–137

75. Scott, N.E. et al. (2014) Diversity within the O-linked proteinglycosylation systems of acinetobacter species. Mol. Cell.Proteom. 13, 2354–2370

76. Gebhart, C. et al. (2012) Characterization of exogenous bacterialoligosaccharyltransferases in Escherichia coli reveals the poten-tial for O-linked protein glycosylation in Vibrio cholerae andBurkholderia thailandensis. Glycobiology 22, 962–974

77. Kamp, H.D. et al. (2013) Gene fitness landscapes of Vibriocholerae at important stages of its life cycle. PLoS Pathog. 9,e1003800

78. Iwashkiw, J.A. et al. (2012) Identification of a general O-linkedprotein glycosylation system in Acinetobacter baumannii and itsrole in virulence and biofilm formation. PLoS Pathog. 8,e1002758

79. Liu, C.F. et al. (2013) Bacterial protein-O-mannosylating enzymeis crucial for virulence of Mycobacterium tuberculosis. Proc.Natl. Acad. Sci. U. S. A. 110, 6560–6565

80. Szymanski, C.M. et al. (2002) Campylobacter protein glycosyl-ation affects host cell interactions. Infect. Immun. 70, 2242–2244

81. Vik, A. et al. (2012) Insights into type IV pilus biogenesis anddynamics from genetic analysis of a C-terminally tagged pilin: arole for O-linked glycosylation. Mol. Microbiol. 85, 1166–1178

82. Fletcher, C.M. et al. (2009) A general O-glycosylation systemimportant to the physiology of a major human intestinal symbi-ont. Cell 137, 321–331

83. Cress, B.F. et al. (2014) Masquerading microbial pathogens:capsular polysaccharides mimic host-tissue molecules. FEMSMicrobiol. Rev. 38, 660–697

84. Tamborrini, M. et al. (2011) Identification of an African Bacillusanthracis lineage that lacks expression of the spore surface-associated anthrose-containing oligosaccharide. J. Bacteriol.193, 3506–3511

Trends in Microbiology, August 2017, Vol. 25, No. 8 671

85. Nothaft, H. et al. (2016) Engineering the Campylobacter jejuni N-glycan to create an effective chicken vaccine. Sci. Rep. 6, 26511

86. Tytgat, H.L.P. and Lebeer, S. (2014) The sweet tooth of bacte-ria: common themes in bacterial glycoconjugates. Microbiol.Mol. Biol. Rev. 78, 372–417

87. Kaminski, L. et al. (2013) Phylogenetic- and genome-derivedinsight into the evolution of N-glycosylation in Archaea. Mol.Phylogenet. Evol. 68, 327–339

88. Mescher, M.F. and Strominger, J.L. (1976) Purification andcharacterization of a prokaryotic glucoprotein from the cellenvelope of Halobacterium salinarium. J. Biol. Chem. 251,2005–2014

89. Sumper, M. et al. (1990) Primary structure and glycosylation of theS-layer protein of Haloferax volcanii. J. Bacteriol. 172, 7111–7118

90. Lu, H. et al. (2015) Identification of the S-layer glycoproteins andtheir covalently linked glycans in the halophilic archaeon Hal-oarcula hispanica. Glycobiology 25, 1150–1162

91. Jarrell, K.F. et al. (2010) Biosynthesis and role of N-linkedglycosylation in cell surface structures of archaea with a focuson flagella and s layers. Int. J. Microbiol. 2010, 470138

92. Jarrell, K.F. et al. (2010) S-layer glycoproteins and flagellins:reporters of archaeal posttranslational modifications. Archaea2010, 612948

93. Albers, S.V. and Meyer, B.H. (2011) The archaeal cell envelope.Nat. Rev. Microbiol. 9, 414–426

94. Jarrell, K.F. et al. (2013) Surface appendages of archaea: struc-ture, function, genetics and assembly. Life (Basel) 3, 86–117

95. Esquivel, R.N. et al. (2016) Identification of Haloferax volcaniipilin N-glycans with diverse roles in pilus biosynthesis, adhesion,and microcolony formation. J. Biol. Chem. 291, 10602–10614

96. Mescher, M.F. and Strominger, J.L. (1976) Structural (shape-maintaining) role of the cell surface glycoprotein of Halobacte-rium salinarium. Proc. Natl. Acad. Sci. U. S. A. 73, 2687–2691

97. Abu-Qarn, M. et al. (2007) Haloferax volcanii AglB and AglD areinvolved in N-glycosylation of the S-layer glycoprotein andproper assembly of the surface layer. J. Mol. Biol. 374,1224–1236

98. Yurist-Doutsch, S. et al. (2008) aglF, aglG and aglI, novel mem-bers of a gene cluster involved in the N-glycosylation of theHaloferax volcanii S-layer glycoprotein. Mol. Microbiol. 69,1234–1245

99. Kaminski, L. et al. (2010) AglJ participates in adding the linkingsaccharide in the Haloferax volcanii N-glycosylation pathway. J.Bacteriol. 192, 5572–5579

100. Yurist-Doutsch, S. et al. (2010) N-glycosylation in Archaea: onthe coordinated actions of Haloferax volcanii. AglF and AglM.Mol. Microbiol. 75, 1047–1058

101. Arbiv, A. et al. (2013) AglQ is a novel component of the Haloferaxvolcanii N-glycosylation pathway. PLoS One 8, e81782

102. Tamir, A. and Eichler, J. (2017) N-glycosylation is important forproper Haloferax volcanii S-layer stability and function. Appl.Environ. Microbiol. 83, e03152-16

103. Tripepi, M. et al. (2012) N-glycosylation of Haloferax volcaniiflagellins requires known Agl proteins and is essential for bio-synthesis of stable flagella. J. Bacteriol. 194, 4876–4887

672 Trends in Microbiology, August 2017, Vol. 25, No. 8

104. Chaban, B. et al. (2006) Identification of genes involved in thebiosynthesis and attachment of Methanococcus voltae N-linkedglycans: insight into N-linked glycosylation pathways in Archaea.Mol. Microbiol. 61, 259–268

105. VanDyke, D.J. et al. (2009) Identification of genes involved in theassembly and attachment of a novel flagellin N-linked tetrasac-charide important for motility in the archaeon Methanococcusmaripaludis. Mol. Microbiol. 72, 633–644

106. Meyer, B.H. et al. (2013) Agl16, a thermophilic glycosyltransfer-ase, mediating the last step of the N-glycan biosynthesis in thethermoacidophilic crenarchaeon Sulfolobus acidocaldarius. J.Bacteriol. 195, 2177–2186

107. Meyer, B.H. et al. (2015) N-Glycosylation of the archaellumfilament is not important for archaella assembly and motility,although N-glycosylation is essential for motility in Sulfolobusacidocaldarius. Biochimie 118, 294–301

108. Meyer, B.H. et al. (2011) Sulfoquinovose synthase – an impor-tant enzyme in the N-glycosylation pathway of Sulfolobusacidocaldarius. Mol. Microbiol. 82, 1150–1163

109. Meyer, B.H. and Albers, S.V. (2014) AglB, catalyzing the oligo-saccharyl transferase step of the archaeal N-glycosylation pro-cess, is essential in the thermoacidophilic crenarchaeonSulfolobus acidocaldarius. Microbiologyopen 3, 531–543

110. Abu-Qarn, M. and Eichler, J. (2006) Protein N-glycosylation inArchaea: defining Haloferax volcanii genes involved in S-layerglycoprotein glycosylation. Mol. Microbiol. 61, 511–525

111. Guan, Z. et al. (2010) Distinct glycan-charged phosphodolicholcarriers are required for the assembly of the pentasaccharideN-linked to the Haloferax volcanii S-layer glycoprotein. Mol.Microbiol. 78, 1294–1303

112. Guan, Z. et al. (2012) Protein glycosylation as an adaptiveresponse in Archaea: growth at different salt concentrationsleads to alterations in Haloferax volcanii S-layer glycoproteinN-glycosylation. Environ. Microbiol. 14, 743–753

113. Kaminski, L. et al. (2013) Two distinct N-glycosylation pathwaysprocess the Haloferax volcanii S-layer glycoprotein uponchanges in environmental salinity. mBio 4, e00716–e00713

114. Ding, Y. et al. (2016) Effects of growth conditions on archaella-tion and N-glycosylation in Methanococcus maripaludis. Micro-biology 162, 339–350

115. Todor, H. et al. (2014) A transcription factor links growth rate andmetabolism in the hypersaline adapted archaeon Halobacteriumsalinarum. Mol. Microbiol. 93, 1172–1182

116. Varki, A. (1993) Biological roles of oligosaccharides: all of thetheories are correct. Glycobiology 3, 97–130

117. Kowarik, M. et al. (2006) N-linked glycosylation of folded pro-teins by the bacterial oligosaccharyltransferase. Science 314,1148–1150

118. Feldman, M.F. et al. (2005) Engineering N-linked proteinglycosylation with diverse O antigen lipopolysaccharide struc-tures in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 102,3016–3021

119. Valderrama-Rincon, J.D. et al. (2012) An engineered eukaryoticprotein glycosylation pathway in Escherichia coli. Nat. Chem.Biol. 8, 434–436