Altered N-glycan composition impacts flagella-mediated adhesion in Chlamydomonas reinhardtii

For the unicellular alga Chlamydomonas reinhardtii, the presence of N-glycosylated proteins on the surface of two flagella is crucial for both cell-cell interaction during mating and flagellar surface adhesion. However, it is not known whether only the presence or also the composition of N-glycans attached to respective proteins is important for these processes. To this end, we tested several C. reinhardtii insertional mutants and a CRISPR/Cas9 knockout mutant of xylosyltransferase 1A, all possessing altered N-glycan compositions. Taking advantage of atomic force microscopy and micropipette force measurements, our data revealed that reduction in N-glycan complexity impedes the adhesion force required for binding the flagella to surfaces. This results in impaired polystyrene bead binding and transport but not gliding of cells on solid surfaces. Notably, assembly, intraflagellar transport, and protein import into flagella are not affected by altered N-glycosylation. Thus, we conclude that proper N-glycosylation of flagellar proteins is crucial for adhering C. reinhardtii cells onto surfaces, indicating that N-glycans mediate surface adhesion via direct surface contact.

Introduction N-glycosylation, as one of the major post-translational modifications, takes place along the ER/Golgi secretion route and consequently most N-linked glycans are found on proteins facing the extracellular space. Initial steps of N-glycosylation in the ER are highly conserved among most eukaryotes and consist of the synthesis of a common prebuilt N-glycan precursor onto a dolichol phosphate. Following the transfer of the glycan precursor onto the asparagine of the consensus sequence N-X-S/T of a nascent protein (where X can be any amino acid except proline), the glycoprotein is folded by the glycan recognizing chaperones Calnexin and Calreticulin (Stanley and Taniguchi, 2017). Subsequent N-glycan maturation steps in the Golgi are species dependent and give rise to a high variety of Nglycan structures. In land plants, Golgi maturation leads to N-glycans modified with b1,2-core xylose and a1,3-core fucose (Strasser, 2016). In Chlamydomonas reinhardtii, a unicellular biflagellate green alga, N-glycans can be decorated with core xylose and -fucose (Lucas et al., 2020;Oltmanns et al., 2019;Schulze et al., 2018). Additionally, 6O-methylation of mannose and addition of a terminally linked b1,4-xylose were reported (Mathieu-Rivet et al., 2013). While the functional advantage of Nlinked glycans to mature proteins is hardly understood, it is known that blocking the synthesis of a full N-glycan precursor results in hypoglycosylated proteins that cannot be folded properly (Gardner et al., 2013). Instead, they are degraded via the ER-associated degradation pathway (ERAD) and are consequently not targeted correctly (Adams et al., 2019;Cherepanova et al., 2016). Therefore, impairment of glycosylation at such early stage is lethal in both uni-and multi-cellular organisms and only when inhibition of glycosylation is carefully dosed (e.g. by tunicamycin), immediate physiological effects caused by hypoglycosylation can be observed (Kukuruzinska et al., 1987). Looking at C. reinhardtii, treatment of vegetative cells with tunicamycin lead to an impaired flagellar adhesiveness, indicating that glycoproteins are crucial for adhesion and the subsequent onset of gliding (Bloodgood et al., 1987). However, whether this phenotype is linked to mistargeting of proteins due to hypoglycosylation and/or the lack of N-glycans on the flagella surface is unclear.
Whole-cell gliding is one of two flagella-based motilities in C. reinhardtii besides swimming (Ishikawa and Marshall, 2011;Kozminski et al., 1993;Snell et al., 2004). In principle, the cell adheres to a surface via its flagella, positioning them in a 180˚angle and initiates gliding along the solid or semisolid surface into the direction in which one flagellum is pointing (designating it as leading flagellum) (Bloodgood, 2009). Interestingly, flagella not only bind to large solid surfaces, but they also bind to small, inert objects (e.g. polystyrene microbeads) that are moved along the flagellar membrane. While the two events, summarized as flagellar membrane motility, are believed to underly the same molecular machinery, it is assumed that they start with an adhesion of flagella membrane components to the surface (Bloodgood and Salomonsky, 1998). A micropipette force measurement approach recently showed that the flagella adhesion forces on different model surfaces with tailored properties lie in the range of 1 to 4 nN and that only positive surface charge diminished the adhesion force significantly (Backholm and Bäumchen, 2019;Kreis et al., 2019;Kreis et al., 2018). These findings imply that the adhesion system of C. reinhardtii has developed toward great flexibility instead of high specificity, in line with the high diversity of solid surfaces dwelled by the microalga in nature, ranging from soil and sand to wet leaves, moss, and bark (Harris, 2009). Remarkably, surface iodination experiments in the early 1980s revealed a single protein called flagellar membrane glycoprotein 1B (FMG-1B) as the main player mediating surface contact (Bloodgood and Workman, 1984). FMG-1B is exclusively located in the flagellar membrane and has a remarkable size of around 350 kDa (4389 amino acids) with a large extra-flagellar part (4340 amino acids) anchored in the membrane via a single predicted trans membrane helix of 22 amino acids (Bloodgood et al., 2019). As the name indicates, it is heavily N-and O-glycosylated. A recent knock down study showed, that it is the main constituent of the glycocalyx surrounding the flagellum. Additionally, a fmg-1B mutant showed a drastically reduced ability to glide (Bloodgood et al., 2019). Strikingly, FMG-1B is present at a high copy number and turns over rapidly within approximately 1 hr (Bloodgood, 2009). The rapid turnover is probably attributed to the fact that flagellar membrane components are constantly shed into the medium as flagellar ectosomes (Bloodgood, 2009;Wood et al., 2013). FMG-1B and another N-glycosylated membrane component, FAP113, have been shown to be eventually torn out of the membrane once bound to microbeads (Kamiya et al., 2018). Whether FAP113 is only involved in microbead binding or also in whole-cell gliding is unknown.
Recently, it was found that flagellar adhesion to surfaces is switchable by light, indicating that a blue-light photoreceptor signal is governing this process (Kreis et al., 2018). Following adhesion to the surface, a transmembrane signal mediates translation of the adhesion event into a calcium transient and protein phosphorylation cascade (Bloodgood, 2009;Collingridge et al., 2013;Kreis et al., 2018). According to the current model, an interaction of the short cytoplasmic part of FMG-1B with intraflagellar transport (IFT) may occur (Laib et al., 2009;Shih et al., 2013). IFT moves bidirectionally along the flagellar microtubules, the anterograde transport is driven by kinesin 2 and retrograde transport is driven by cytoplasmic dynein-1b (Cole et al., 1998;Huangfu et al., 2003;Kozminski et al., 1995;Lechtreck, 2015;Pedersen and Rosenbaum JLBT-CT in DB, 2008;Porter et al., 1999;Rosenbaum and Witman, 2002). Since retrograde IFT trains pause relative to the adhesion site while FMG-1B tethers to the solid surface through its large extracellular carbohydrate domain (Bloodgood, 2009), the force generated by retrograde motor protein dynein-1b will push the microtubule into the opposite direction, dragging the cell body and the second flagellum behind; the gliding process is initiated (Shih et al., 2013).
Due to the high N-glycosylation level of the extraflagellar domain of FMG1-B, interacting with the solid surface, it was suggested that N-glycosylation could be crucial for adhesion, beyond proper glycoprotein folding. Therefore, we compared flagellar membrane motility of different mutant strains impaired in N-glycan maturation (characterized in Schulze et al., 2018). We found that N-glycan maturation indeed impacts the interaction of flagellum and surface in all mutants analyzed.

Results
Altered N-linked glycans do not change the flagellar localization of FMG-1B To test whether N-glycan maturation in Golgi is important for flagellar surface motility in C. reinhardtii, two insertional mutants (IM) such as IM Man1A , IM XylT1A and their double mutant IM Man1A xIM XylT1A were studied. Initially, these mutants had been described in Schulze et al., 2018, where N-glycan patterns of supernatant proteins were analyzed and compared ( Figure 1A). The insertional mutagenesis giving rise to these mutants was performed in the parental strain CC-4375 (a ift46 mutant backcrossed with CC-124) complemented with IFT-46::YFP, referred to as WT-Ins throughout the current study. The first mutant, deficient in xylosyltransferase 1A (IM XylT1A ), produces N-glycans devoid of core xylose while simultaneously having a reduced length. Second mutant IM Man1A , a knock down mutant of mannosidase 1A, is mainly characterized by a lack of 6O-methylation of mannose residues while the N-glycan length is slightly greater than in WT-Ins. Furthermore, N-glycans of IM Man1A are slightly reduced in terminal xylose and core fucose. Finally, a double mutant of the above two single mutants (IM Man1A xIM XylT1A , obtained by genetic crossing) produces N-glycans devoid of 6O-methylation but of WT-Ins length and carrying core xylose and fucose residues ( Figure 1A). It is of note that in none of the mutants the flagellar length is altered as compared to WT-Ins (Figure 1-figure supplement 1). To confirm that flagellar N-glycan patterns of these mutants deviate from WT-Ins, whole-cell extracts and isolated flagella were probed with anti-HRP, binding to b1,2-xylose and a1,3-fucose attached to the N-glycan core (Kaulfürst-Soboll et al., 2011). In line with previous publications, the antibody showed a higher affinity toward N-glycoproteins synthesized by IM Man1A and IM Man1A xIM XylT1A while it showed a decreased affinity toward probes of IM XylT1A (Figure 1-figure supplement 2; Schulze et al., 2018). Further lectin-affino blotting with concanavalin A (ConA) was performed on whole-cell extracts, revealing increased ConA-affinity in all three N-glycosylation mutants compared to WT-ins ( Figure 1-figure supplement 3).
Since FMG-1B is the major constituent of the flagellar glycoproteome and to date the only protein proven to be involved in flagellar surface motility, different monoclonal antibodies raised against FMG-1B were employed to analyze FMG-1B localization (Bloodgood et al., 1986;Long et al., 2016). The whole-cell extracts or isolated flagella from IM Man1A , IM XylT1A , and their double mutant IM Man1A xIM XylT1A were probed separately with the glycan epitope recognizing antibody or the antibody against the protein backbone of FMG-1B. FMG-1B was found in whole cells and flagella of WT-Ins and all three mutants. Hereby, the protein amount was similar in all four strains as indicated by the use of the FMG-1B protein-specific antibody ( Figure 1B). Contrarily, the antibody raised against FMG-1B glycan epitopes barely detected whole-cell or flagella probes of IM Man1A and IM Ma-n1A xIM XylT1A . Also, the FMG-1B glycan signal decreased in the mutant IM XylT1A , particularly in the whole-cell sample. Taken together, these immuno-blots confirm that the N-glycan pattern of FMG-1B is altered in the mutants, while the protein localization is not affected by this alteration. In addition, label-free mass spectrometric quantification confirmed that FMG-1B is correctly targeted to the flagella in all mutants analyzed ( Figure 1C). The same is true for FAP113, another protein shown to be involved in flagella surface motility ( Figure 1D).
To compare the localization of glycan and protein of FMG1-B in WT-Ins and mutants, the cells were immuno-stained with the two FMG-1B-specific antibodies described above. A uniform signal of     the glycan-specific antibody was found in the flagella and the cell wall of WT-Ins ( Figure 1E, left panel). It should be noted, that such cross reaction with cell wall localized glycosylated proteins has been reported previously (Bloodgood et al., 1986). In line with the immuno-blotting experiment, no glycan signal was observed in the flagella or cell wall of IM Man1A and the double mutant, while signal intensity was low in the flagella of mutant IM XylT1A ( Figure 1E, left panel and). The faint FMG-1B signal in IM XylT1A compared to the WT-Ins was clearly observed when WT-Ins and IM XylT1A were mixed prior to immuno-staining ( Figure 1-figure supplement 6). When the FMG-1B peptide antibody was used, a uniform signal is found in the flagella of WT-ins, M Man1A , IM XylT1A , and the mutant IM Man1A xIM XylT1A ( Figure 1E, right panel). In summary, these data show that altered N-glycan maturation did not affect the flagellar localization of FMG-1B. This is in line with early findings by Bloodgood et al., reporting proper FMG-1B targeting to the flagella in a N-glycosylation mutant termed L23 showing increased ConAaffinity (Bloodgood et al., 1987). Although FMG-1B is the most prominent and best studied flagella membrane glycoprotein, there might be other glycoproteins involved in flagellar adhesion. Nevertheless, no protein was found consistently and significantly changed in abundance in flagella of the IM strains analyzed when compared to WT-Ins (Figure 1-figure supplement 7), also indicating that flagellar assembly is not considerably altered in the mutants versus WT.
Altered N-linked glycans attenuate bead attachment to and movement along the flagellar membrane In C. reinhardtii, polystyrene microspheres adhere to and move bidirectionally along the flagellar surface (Bloodgood, 1981). To check the effect of altered N-glycans on these processes, beads with a diameter of 0.7 mm were added to cell suspensions of IM Man1A , IM XylT1A , and the double mutant and the number of beads attached to-or moved along flagella were quantified ( Figure 2A   Source data 1. Raw data of attachment and movement of microbeads to and along flagella. found in Figure 2-source data 1. These data suggested that interaction of flagellar membrane and surface is altered due to altered N-glycan composition in these mutants.

Quantification of the flagella-mediated adhesion using atomic force microscopy
Flagella-mediated surface adhesion force was measured via atomic force microscopy (AFM) ( Figure 3A). Here, cells adhered to a cover slide were attached to an AFM cantilever via physical contact (Liu et al., 2011). Subsequently, the AFM cantilever was pulled upwards and the force required to pull the cells was recorded ( Figure 3B). To inhibit whole-cell gliding during the measurement, ciliobrevin D was used to inhibit dynein-1b activity and consequently the cell gliding (Firestone et al., 2012). Remarkably, the forces necessary to overcome the adhesion of C. reinhardtii flagella to the surface were significantly reduced in these three mutants analyzed as compared to WT-Ins ( Figure 3B and C). Especially in the double mutant the adhesion force was reduced from 8 nN in WT-Ins to 1 nN, while the average energy was reduced from 4 to 0.5 J nm À1 ( Figure 3C and D). Source data can be seen in Figure 3-source data 1. This result indicates that an altered N-glycan composition impacts the flagellar adhesion force onto a solid substrate.

Quantification of flagellar adhesion using a micropipette force measurement approach
To validate the AFM adhesion force measurements, an independent in vivo force measurement approach (Backholm and Bäumchen, 2019;Kreis et al., 2018) was used with another genetic background mutant, xylosyltransferase 1A (CRISPR XylT1A_1 ), generated in parental wildtype SAG11-32b (WT-SAG) by employing CRISPR/Cas9 (Figure 4-figure supplement 1). As a wavelength dependency of flagellar adhesion had been revealed using this approach (Backholm and Bäumchen, 2019), adhesion forces of the same cells were measured under precisely controlled blue-and redlight conditions using a micropipette adhesion force measurement approach. Adhesion forces were measured in presence and absence of dynein-1b inhibitor ciliobrevin D. Importantly, adhesion forces in CRISPR XylT1A_1 were significantly diminished in comparison to respective WT under both ciliobrevin D conditions confirming AFM results ( Figure 4). Illuminating cells with red light dramatically decreased the adhesion force in both WT-SAG and CRISPR XylT1A_1 in presence as well as in absence of ciliobrevin D. Notably, removal of ciliobrevin D resulted in a significant decrease in adhesion force for both WT-SAG and CRISPR XylT1A_1 from~2.6 nN to~1.3 nN in WT-SAG and from~1.8 nN to~1 nN in CRISPR XylT1A_1 . Source data can be seen in Figure 4-source data 1.

The effect of altered N-glycosylation on IFT and gliding
As presented in Figure 3, the strongest effect on adhesion forces assessed was observed in the double mutant IM Man1A xIM XylT1A when compared to WT-Ins. In the absence of ciliobrevin D, the adhesion force measured by AFM in mutant IM Man1A xIM XylT1A is still significantly lower than in WT-Ins ( Figure 5A). This is in line with the WT-SAG and CRISPR XylT1A_1 micropipette adhesion force measurement performed in the absence of ciliobrevin D, where also CRISPR XylT1A_1 had a lower adhesion force ( Figure 4). Interestingly, addition of ciliobrevin D resulted in significantly increased adhesion forces as seen for WT-SAG or WT-Ins ( Figure 4 and Figure 5A). Taken together, these results suggested that active dynein-1b might reduce surface adhesion forces. On the other hand, it implied that IFT might be hampered via altered N-glycosylation as surface adhesion forces were smaller in the N-glycosylation mutants. Therefore, IFT velocity and gliding ability of IFT46::YFP expressing WT-Ins and IM Man1A xIM XylT1A in absence of ciliobrevin D were assessed by using total internal reflection fluorescence (TIRF) microscopy. Videos of adhered cells generated by TIRF microscopy were evaluated manually with help of kymographs in Fiji software ( Figure 5B). Obtained data revealed that neither the proportion of gliding events (gliding velocity higher 0.3 mm*s À1 ), nor gliding speed distribution was significantly diminished when comparing WT-Ins and the double mutant ( Figure 5C). Likewise, anterograde and retrograde IFT velocities were found at WT-Ins values when comparing WT-Ins and the double mutant of adherent cells ( Figure 5C), implying no significant impact of altered N-glycan maturation on IFT. Source data for Figure 5 can be seen in Figure 5source data 1. Lastly, to rule out the possibility that ciliobrevin D might result in elevated adhesion forces due to a toxic side effect, we generated the triple mutant dynein-1b ts X IM Man1A X IM XylT1A-4-13# by crossing IM Man1A xIM XylT1A with CC-4423 ( Figure 5-figure supplement 1A-C). CC-4423 is characterized by the expression of a temperature sensitive transcript of dynein-1b leading to a depletion of dynein-1b at restrictive temperatures followed by an attenuation of retrograde IFT and flagella disassembly (Engel et al., 2012). Subsequently, adhesion forces were measured via AFM at restrictive temperatures, that is, under conditions mimicking the ciliobrevin D dependent inactivity of dynein-1b. As it cannot be excluded that flagella shortening, as induced upon temperature shift ( Figure 5-figure supplement 1D), might impact flagellar adhesion forces, the triple mutant treated with 20 mM NaPPi was assessed by AFM as control. Indeed, it was found that adhesion forces differed when measured at a defined flagella length of about 6.5 mm: while control cells showed an average adhesion force of 0.48 nN, cells depleted from dynein-1b showed a significantly elevated adhesion force of 1.64 N. Importantly, NaPPi only induces flagellar shortening ( Figure 5-figure supplement 1E) but does not affect dynein-1b (Dentler, 2005), therefore, differences observed are directly correlated to the action of dynein-1b. Thus we suggest, that the action of dynein-1b reduces flagellar adhesion forces due to a destabilization of surface adhered protein clusters.

Discussion
Our data revealed that the maturation of N-glycans has an impact on flagella-mediated cell adhesion in C. reinhardtii. At the same time, IFT and gliding velocity were not changed due to altered Nglycosylation. Microbead binding was found diminished in IM strains, implying that the flagellar surface has an altered affinity toward microbeads. In line with this, their surface adhesion forces were significantly reduced compared to WT-Ins. The AFM data were confirmed by assessing another XylT1A mutant created via CRISPR/Cas9, using micropipette force measurements. It should be noted that N-glycan patterns of IM XylT1A and CRISPR XylT1A were comparable and thereby strengthen the proposed role of XylT1A as core xylosyltransferase (Lucas et al., 2020;Schulze et al., 2018). Forces measured for WT strains and N-glycosylation mutants with AFM and micropipette force measurement confirmed that differential N-glycan maturation, i.e. altered N-glycan structures attached to mature proteins, lowers the adhesion force of flagella to a surface. These changes in adhesion forces were not accompanied by consistent drastic changes in the flagellar proteomes. For example, FMG-1B and FAP113, to date the only two known proteins involved in surface adhesion, were found in comparable amounts in WT-Ins and mutants (Figure 1 and  Kamiya et al., 2018). Of note, also FMG-1A is localized in flagella and its abundance was unaltered between WT and mutants in vegetative cells (Figure 1-figure supplement 4). This contrasts the current assumption that FMG-1A is solely expressed in reproductive cells and opens the question whether it might have a similar role as FMG-1B, given the high similarity of the two proteins (Bloodgood, 2009). Interestingly, gliding of mutant strains on solid surface was not affected. The current model for flagella-mediated cell adhesion and subsequent gliding proposes that the   extracellular part of certain glycoproteins such as FMG-1B adheres to the surface, cytoplasmic moieties of these proteins are bound to ongoing retrograde IFT directly or indirectly upon calcium-and light dependent stimulus which is followed by an onset of gliding (Kreis et al., 2018;Shih et al., 2013). Assuming that altered N-glycan maturation does not impact initial protein folding in the ER (as those steps are spatially and temporally separated), our data revealed that changed N-glycosylation, did not alter targeting of respective glycoproteins to flagella nor the velocity of IFT. Thus, changes in surface adhesion are likely linked to N-glycoprotein epitopes and their direct interaction with the solid or semisolid surface. How specific N-glycan moieties modulate adhesion force is subject of future research, particularly considering the finding that flagella-mediated adhesion of C. reinhardtii has been shown to be largely unaffected by different substrate surface properties (Kreis et al., 2019).
Notably, IFT and gliding were not changed between double mutant and WT-Ins ( Figure 5). The fact that altered N-glycosylation diminished the force of cells to adhere to surfaces but did not affect IFT, strongly suggests that adhesion to surfaces and IFT are not necessarily coupled. As discussed below, adhesion probably evolved independently of the necessity to enable cell gliding.   Moreover, it can be concluded that N-glycosylation does not significantly impact differences in light perception, as the adhesion forces for WT-SAG and CRISPR XylT1A were significantly stronger under blue than under red light (Figure 4).
In summary, taking advantage of single-cell adhesion force measurements, our data revealed that cell adhesion was significantly impaired in C. reinhardtii N-glycosylation mutant strains. Our data further suggested that flagellar assembly, IFT and FMG-1B transport into flagella were not affected by altered N-glycosylation implicating no role of N-glycosylation in these processes. Instead, proper N-glycosylation of flagellar proteins is crucial for adhering C. reinhardtii cells onto surfaces. Our observations further suggest that the remaining adhesion force, although diminished in N-glycan mutants, is still sufficient for gliding. Given the response of flagellar adhesion to blue light, it could potentially link adhesion to photo-protection which is also blue-light mediated, as adhesion might result in photoprotection via biofilm formation, which in turn would enable mutual cell shading (Kreis et al., 2018;Petroutsos et al., 2016).

Culture growth
Cells were grown photoheterotrophically in tris-acetate-phosphate (TAP) medium under constant illumination at 50 mmol photons*s À1 *cm À2 unless stated otherwise.

Measurement of flagellar length and flagellated cells
The cells were fixed with 0.5% Lugol's solution for 1 hr at room temperature. Flagellar length measurements were performed using a phase microscope (Nikon Eclipse Ti) equipped with an electron multiplying charged-coupled device. For each sample, at least 50 flagella were measured. For the measurement of flagellated cells, at least 100 cells were counted for each strain in biological triplicates.

Generation and analysis of a CRISPR/Cas9 mutant strain
Mutagenesis was performed on the WT strain SAG11-32b following the protocol described in Greiner et al., 2017 employing the transformation of a pre-built Cas9:guideRNA complex (Cas9 target seuence in the XylT1A gene: ACGAACACCCCAACACCAAT) simultaneously with a plasmid encoding for a paromomycin resistance via electroporation. Following selection with paromomycin, putative mutants were screened by PCR using the primer pairs short_fw: TACAAAGAACGGGACG-CAGG, short_rev: CATTGAAGCTCATCCAGACAC and long_fw: AAGGGTCACGGCACGGTATG, long_rev: CCTGAAGCACCCATGATGCACG. Genomic XylT1A regions of candidate strains showing not-WT like band patterns were sequenced. In total, two mutant strains differing in the DNA inserted following the Cas9 cutting site were identified (CRISPR XylT1A_1 and CRISPR XylT1A_2 ). Next, XylT1A protein levels were quantified by parallel reaction monitoring (PRM) and supernatant N-glycan compositions were assessed by IS-CID mass spectrometry. Additionally, flagella were isolated, separated by SDS-PAGE and, after transferred to a nitrocellulose membrane, probed with the protein backbone FMG-1B-specific antibody.

Flagella isolation
Flagella isolation from cultures in the mid-log growth phase was performed as described elsewhere by the pH shock method (Witman et al., 1972). Pellets containing flagella samples were stored at À80˚C until further use for immunoblotting or sample preparation for mass spectrometric measurements.

Immunoblotting
Frozen, dry flagella and whole-cell samples were resuspended in lysis buffer (10 mM Tris/HCl, pH = 7.4, 2% SDS, 1 mM Benzamidine and 1 mM PMSF) and subjected to sonication for 10 min. After pelleting cell debris, the protein concentration was determined using the bicinchoninic acid assay (BCA Protein Assay Kit by Thermo Scientific Pierce). Volumes corresponding to 30 mg of protein were separated by SDS-PAGE, transferred to nitrocellulose membranes and incubated with antibodies as indicated.

Lectin-affino blotting with concanavalin A and HRP
Frozen, dry whole-cell samples were resuspended in lysis buffer (10 mM Tris/HCl, pH = 7.4, 2% SDS, 1 mM Benzamidine and 1 mM PMSF) and subjected to sonication for 10 min. After pelleting the notsoluble cell debris, the protein concentration was determined using the bicinchoninic acid assay (BCA Protein Assay Kit by Thermo Scientific Pierce). Volumes corresponding to 50 mg of protein were separated by SDS-PAGE and transferred to nitrocellulose membrane. Membrane was incubated with ConA (1 mg/mL in TBST + 1 mM CaCl 2 + 1 mM MnCl 2 ) for 1.5 hr at room temperature. Subsequently membrane was washed and incubated 1 hr with HRP (5 mg/mL in TBST + 1 mM CaCl 2 + 1 mM MnCl 2 ). Excess HRP as well as Ca 2+ and Mn 2+ were removed by three washing steps with TBST, before affino blot was development via ECL.

Sample preparation for mass spectrometric measurements
Frozen, dry flagella and whole-cell samples were treated as described in Immunoblotting. Volumes corresponding to 60 mg of protein were tryptically digested and desalted as described elsewhere (Rappsilber et al., 2007).

Mass spectrometry measurements
Tryptic peptides were reconstituted in 2% (v/v) acetonitrile/0.1% (v/v) formic acid in ultrapure water and separated with an Ultimate 3000 RSLCnano System (Thermo Scientific). Subsequently, the sample was loaded on a trap column (C18 PepMap 100, 300 mm x 5 mm, 5 mm particle size, 100 Å pore size; Thermo Scientific) and desalted for 5 min using 0.05% (v/v) TFA/2% (v/v) acetonitrile in ultrapure water with a flow rate of 10 mL*min À1 . Following, peptides were separated on a separation column (Acclaim PepMap100 C18, 75 mm i.D., 2 mm particle size, 100 Å pore size; Thermo Scientific) with a length of 50 cm. General mass spectrometric (MS) parameters are listed in Table 1.
For quantification of glycosyltransferases, PRM (including a target list) was employed on wholecell samples and respective spectra were analyzed with the Skyline software (Pino et al., 2020). For quantification of flagellar proteins, flagella samples were measured in biological quadruplicates in standard, not targeted, data dependent measurements. Following, peptide wise protein abundance ratios (IM/WT) were calculated with ProteomeDiscoverer (normalizing on a set of not membrane standing flagellar proteins) and filtered for proteins identified in at least 11 samples, for proteins having Abundance Ratio Adj. p-value<0.05 for at least one ratio and for proteins appearing in the flagellar proteome ChlamyFPv5 (Pazour et al., 2005).
In order to assign glycopeptides, samples were measured employing In-Source collision induced dissociation (IS-CID) as described previously followed by analysis of data with Ursgal and SugarPy (Kremer et al., 2016;Oltmanns et al., 2019;Schulze et al., 2020).

Microbead measurements
Microbead binding-and transport assays were performed analogous to previous descriptions. (Bloodgood et al., 2019) Monodisperse polystyrene microspheres (0.7 mm diameter) were purchased from Polysciences, Inc Beads were washed with deionized water for three times and resuspended in NFHSM to make a store solution, which was used at 1:10 dilution in adhesion and motility detecting experiment.
To quantify the ability of bead binding, beads were added to 500 mL of cells at a density of 2 Â 10 7 cells*mL À1 . After 5 min, cells were observed with a light microscope (Olympus, U-HGLGPS, 100X oil objective). A flagellum was scored as '+ bead' if beads adhered to it. The percentage of flagellar binding beads was calculated as: Percentage of flagellar binding beads = the number of '+ bead'/ (total number flagella scored) x 100%.
To obtain a kinetic measure of surface motility, cells were mixed with beads as above for 5 min and randomly observed under the light microscope. Each bead adhered to a flagellum was monitored for about 30 s. If beads moved along the flagella, we marked it as 'Moved bead' or it was 'Adhered bead'. The surface motility was calculated as: Percentage of moved beads along with flagella = 'Move bead' x/ ('Moved bead' + 'Adhered bead') 100%.

AFM measurements
C. reinhardtii strains, grown in M1 medium under constant white illumination were grown for 65 hr, were allowed to adhere to a glass slide (immersed in ethanol for overnight, subsequently rinsed with MQ water) in fresh M1 medium for 15 min. Following, cells were incubated in the presence of ciliobrevin D for 1 hr (500 mL M1 supplemented with 200 mM ciliobrevin D). For AFM measurements, only adhered cells in gliding conformation having approximately similar appearance were analyzed. The MLCT-O10 AFM probe (Spring Const.: 0.03 N m À1 , length: 215 mm, width: 20 mm, resonant freq.: 15 kHz, Bruker) was soaked in acetone for 5 min, then subjected to UV illumination (distance to lamp: 3-5 mm) for 15 min. Then, the probe was immersed in 0.01% poly-l-lysine for 1 hr and afterwards rinsed with MQ water. Following, the probe was immersed with 2% glutaraldehyde for 1 hr and rinsed with MQ water before use. The AFM measurement was performed in Force Spectroscopy Mode in liquid at room temperature using a NanoWizard 3 AFM (JPK) equipped with a CellHesion stage (Z range: 100 mm) NanoWizard three head. The spring constant of the cantilever was routinely calibrated using the contact-based thermal noise method. The AFM tip, modified as described, was lowered onto the cell surface at a rate of 10 mm s À1 with a z scale of 25 mm. After contact, the applied force was maintained at 3 nN for 15 s. Then, the cell-attached probe was upraised at a rate of 1 mm s À1 . Force curves were processed with JPK SPM Data Processing (JPK). The forces and energy were determined as described in Figure 3-figure supplement 1 (Liu et al., 2011). Three biological replicates were performed with minimum 5 cells measured per replicate.

Micropipette force measurements
Cell culture growth and micropipette force measurements were performed following established recipes (Kreis et al., 2019;Kreis et al., 2018). In brief, C. reinhardtii strains WT-SAG and CRISP-R XylT1A_1 grew axenically in tris-acetate-phosphate (TAP) medium (Thermo Fisher Scientific) in a Memmert IPP 100Plus incubator on a 12 h day / 12 hr night cycle. The experimental approach is based on the use of a homemade micropipette force sensor, which allows for grasping a living cell by suction (Backholm and Bäumchen, 2019). The micropipette is calibrated by measuring the deflection induced by the weight of an evaporating water droplet at tip of the pipette. The adhesion force is obtained by bringing the flagella into contact with a piece of a silicon wafer (unilateral polished, Si-Mat) cleaned by sonication in ethanol, and by measuring the micropipette deflection during iterative approach and retraction of the substrate moving at 1 mm*s À1 . The substrate approach consists of pushing the cell such that the micropipette is deflected by 10 mm from the cell/substrate contact position, which is then followed by a dwell period of 10 s. The substrate is then retracted by 30 mm from the cell/substrate contact position at the same speed. The overall contact time between the flagella and the substrate is about 30 s. The illumination wavelength for blue and red light was 470 nm and 671 nm respectively, and realized by using narrow band pass interference filters (FWHM: 10 nm) added on top of the condenser of an inverted microscope (Olympus IX-73 and IX-83). During the adhesion force measurements, the cells were illuminated with a constant photon flux of 1019 photons*m À2 * s À1 for both light conditions. For each cell, 10 adhesion force measurements were performed for each wavelength, whereby the order of red and blue light was varied randomly after five consecutive measurements. In order to evaluate the influence of ciliobrevin D on the adhesiveness, a 200 mM stock solution of ciliobrevin D (Merck) was prepared in a 9:1 water: dimethylsulfoxide (DMSO, Purity: 99.9%, Sigma-Aldrich,) mixture. Then, 1.08 mL of this stock solution was added to 30 mL of culture to achieve a final concentration of 7 mM of ciliobrevin D in the cell suspension. The C. reinhardtii suspension containing ciliobrevin D was next incubated for 30 min and then centrifuged at 100 g for ten minutes, followed-up by a minimum of 30 min rest in the incubator. Finally, about 15 mL of the cell suspension was used to fill the liquid chamber. In parallel, a second suspension of C. reinhardtii cells was incubated using the same fraction of DMSO (but without ciliobrevin D), followed by the exact same experimental procedure to serve as a control group.

TIRF imaging
Total internal reflection microscopy (TIRF) was applied to assess IFT and gliding behavior of C. reinhardtii strains expressing YFP-coupled IFT46. Therefore, cell densities were adjusted to 1 Â 10 5 cells*mL À1 . Samples were loaded to a glass bottom microscopy chamber (m-Slide 8 Well Glass Bottom) and refreshed every 20 min while imaging. TIRF microscopy was performed at room temperature with a Nikon Eclipse Ti and a 100x objective. IFT46::YFP was excited at 488 nm and fluorescence was recorded with an iXon Ultra EMCCD camera (Andor). For analysis, images were captured with NIS-Elements software over 30 s at 10 fps and a pixel size of 0.158 mm*pixel À1 . Images were evaluated by use of Fiji via manual evaluation of kymographs. Nett IFT velocities during gliding were calculated by subtracting corresponding gliding velocities. Three biological replicates were performed with 10 cells in gliding configuration analysed per replicate.

Confocal imaging
Cells were incubated with primary antibodies (FMG-1B #8 and #61, available at dshb.com), subsequently incubated with a fluorescently labeled secondary antibody and analyzed by confocal microscopy as described previously (Lv et al., 2017). In brief, cells were plated on 1% poly (ethyleneimine) coated cover glass, decolorized and fixed in methanol at À20˚C for 20 min, permeated cells in PBS buffer for 1 hr, and then blocked in 5% BSA (Biosharp), 10% normal goat serum (Dingguo) and 1% fish gelatin (Sigma) in PBS. Incubated the samples with primary antibodies overnight, washed them, and incubated secondary antibody, washed the samples and mounted them on slides with nail polish. The slides were examined with a Leica confocal microscope (SP8). Images were acquired and processed by LAS X software (Leica) and ImageJ software. The 488 nm laser was used YFP excitation wavelength is 510 nm, the emission wavelength is 525 nm and the exposure time is 200 ms.

Mating and Tetrad analysis
The plus and minus strains was incubated in 2 mL TAP-N medium (2 Â 10 7 cells/mL) under continuous light overnight for gametogenesis. 0.5 mL plus and minus gametes were mixed together and incubated for 2 hr for mating. Then 0.15 mL mixture was dispersed onto mature plate (4% agar). Plate was kept in dark for 5 days, then exposed to light for 24 hr. The unmated gametes were removed from the mature plate with razor blade and were killed using chloroform for 30 s. The agar contained about 30 zygotes was cut and transferred to a germination plate (1% agar). The plates were incubated in bright light till the spores were released from the zygote, then 100 mL water was added to the cut agar and was dispersed on whole plate. Single clones appeared within 3-5 days and were picked for further analysis.

Materials and correspondence
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository with the dataset identifier PXD018353 and will be publically available upon acceptance of the manuscript (Perez-Riverol et al., 2019). For further requests, please contact K. Huang (huangky@ihb.ac.cn) or M. Hippler (mhippler@uni-muenster.de).