WDR2 regulates the orphan kinesin KIN-G to promote hook complex and Golgi biogenesis in Trypanosoma brucei

1 The flagellum in Trypanosoma brucei plays crucial roles in cell locomotion, cell morphogenesis, and cell 2 division, and its inheritance depends on the faithful duplication of multiple flagellum-associated structures. One 3 of such cytoskeletal structures is a hairpin -like structure termed the hook complex composed of a fishhook -like 4 structure and a centrin arm structure, whose cellular functions remain poorly understood. We recently identified 5 KIN-G, an orphan kinesin that promotes hook complex and Golgi biogenesis. Here we report a WD40 repeats-6 containing protein named WDR2, which interacts with and regulates KIN-G. WDR2 co-localizes with KIN-G at 7 the centrin arm, and knockdown of WDR2 disrupts hook complex integrity and morphology, inhibits flagellum 8 attachment zone elongation and flagellum positioning, and eventually arrests cytokinesis. Knockdown of WDR2 9 also disrupts the maturation of the centrin arm-associated Golgi , thereby impairing Golgi biogenesis . WDR2 10 interacts with KIN-G via its N-terminal unknown motifs, the middle domain containing a coiled coil and a PB1 11 motif, and the C-terminal WD40 domain, and targets KIN-G to its subcellular location. These results uncover a 12 regulatory role for WDR2 in recruiting KIN-G to regulate hook complex and Golgi biogenesis, thereby impacting 13 flagellum inheritance and cell division plane positioning. 14 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 3

1 The flagellum in the parasitic protozoan Trypanosoma brucei, the causative agent of sleeping sickness in 2 humans and nagana in livestock in sub-Saharan Africa, plays essential roles in cell motility, cell division, and cell-3 cell communication. It is nucleated from a centriole-like structure termed the basal body, and it exits the cell 4 through the flagellar pocket and attaches to the cell membrane for most of its length via a specialized cytoskeletal 5 structure termed the flagellum attachment zone (FAZ) (Sunter and Gull, 2016) , which mai ntains flagellum 6 attachment and defines the cell division plane (Vaughan et al., 2008; Zhou et al., 2011) . In the flagellar pocket 7 region, several specialized cytoskeletal structures, including the flagellar pocket collar (FPC) (Lacomble et al., 8 2009) and the hook complex (Esson et al., 2012), associate with the flagellum (Fig. 1A). The FPC is a horseshoe-9 shaped structure that wraps around the flagellum and likely plays roles in restricting protein transport in and out 10 of the cell through the flagellar pocket (Lacomble et al., 2009). Sitting at the top of the FPC (Esson et al., 2012), 11 the hook complex is a hairpin-like structure composed of a fishhook-like structure marked by TbMORN1 and 12 TbLRRP1 (Morriswood et al., 2009; Zhou et al., 2010) and a bar-shaped centrin arm structure marked by two 13 centrin proteins TbCentrin2 and TbCentrin4 (He et al., 2005; Shi et al., 2008). The centrin arm sits alongside the 14 shank part of the fishhook-like structure, with the centrin arm and the shank embedding the proximal end of the 15 intracellular FAZ filament (Esson et al., 2012; Morriswood, 2015) (Fig. 1A). The structural integrity of the hook 16 complex is critical for promoting FAZ elongation and flagellum positioning, thereby impacting the placement of 17 the cell division plane and cytokinesis (Pham et al., 2020; Zhou et al., 2010). 18 Alongside the intracellular FAZ filament runs a specialized set of four microtubules termed the microtubule 19 quartet (MtQ) (Fig. 1A), which originates between the mature and the pro-basal bodies, wraps around the flagellar 20 pocket, and then pass through the FPC and the hook complex to extend to the distal cell tip (Esson et al., 2012; 21 Vaughan and Gull, 2016) . The precise molecular function of the MtQ remains elusive, but thr ee MtQ proximal 22 end-localized proteins, TbSpef1, NHL1, and SNAP1, appear to promote basal body rotation and segregation, 23 thereby facilitating flagellum positioning (Gheiratmand et al., 2013; Pham et al., 2022; Souza Onofre et al., 2023). 24 In close proximity to the centrin arm, there sits the Golgi apparatus (Fig. 1A), which is duplicated through a de 25 novo pathway, by which the new Golgi is assembled next to the old Golgi and the ER exit site (He et al., 2004). 26 Duplication of the single Golgi apparatus in the procyclic (insect) form of T. brucei appears to depend on the 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 4 assembly of an intact centrin arm, as knockdown of TbCentrin2 (He et al., 2005) and the centrin arm-localized 1 TbPLK (T. brucei Polo-like kinase) (de Graffenried et al., 2008) disrupted Golgi biogenesis. Although knockdown 2 of TbCentrin2 and TbPLK exert s distinct effects on Golgi biogenesis, with TbCentrin2 depletion completely 3 inhibiting the formation of new Golgi (He et al., 2005) and TbPLK depletion producing numerous small Golgi 4 (de Graffenried et al., 2008) , the results suggest that the centrin arm determine s the Golgi assembly site. We 5 recently identified and characterized an orphan kinesin named KIN-G, which localizes to the centrin arm and 6 regulates hook complex and Golgi biogenesis in procyclic trypanosomes (Zhou et al., 2024). We also identified 7 CAAP1 as a Golgi peripheral protein, which promotes the association between Golgi and the centrin arm to 8 facilitate Golgi biogenesis (Zhou et al., 2024). KIN-G is a microtubule plus end-directed motor protein; hence, it 9 might transport certain protein and/or non-protein cargos for hook complex and Golgi biogenesis. The cargos that 10 KIN-G transports to the centrin arm and/or the Golgi remain unidentified. 11 In the current work, we attempted to identify the interacting proteins and the potential cargo proteins of KIN-12 G by proximity-dependent biotin identification (BioID). We identified a KIN-G-interacting protein named WDR2, 13 a kinetoplastid-specific protein containing a WD40-repeats domain at the C-terminus, two domains of unknown 14 function (DUF) at the N-terminus, and a small coiled-coil motif and a PB1 (Phox and Bem1) motif in the middle 15 region, and showed that WDR2 plays an essential role in promoting hook complex and Golgi biogenesis in the 16 procyclic form of T. brucei . We further demonstrated that WDR2 executes its function by recruiting and 17 maintaining KIN-G at the centrin arm. These results discovered a new protein implicated in hook complex and 18 Golgi biogenesis in T. brucei and uncovered its mechanistic role in the regulation of KIN-G localization and 19 stability. 20 21

22 Identification of WDR2 as a KIN-G-interacting protein localized at the centrin arm 23 We recently reported the essential role of the orphan kinesin KIN -G in regulating hook complex assembly 24 and Golgi biogenesis in procyclic trypanosomes (Zhou et al., 2024). To identify the interacting partner(s) of KIN-25 G and the potential cargo proteins that KIN-G transports, we carried out BioID using KIN-G as bait. The KIN-G-26 proximal proteins thus identified were screened based on their subcellular locations that were determined 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 5 previously by the TrypTag project (Billington et al., 2023). Since KIN-G localizes to the centrin arm (Zhou et al., 1 2024), only th ose proteins that localize to the close proximity of the centrin arm were considered as KIN-G-2 proximal proteins, which included the proteins localized to the centrin arm, the hook complex, and the MtQ (Fig. 3 1B). Among the KIN-G-proximal proteins, a WD40 repeats-containing protein (accession number: Tb927.7.4700) 4 was the top hit and, thus, was chosen for further characterization . We named this protein WDR2 for WD40 5 Repeats-containing protein 2, following the previously characterized WDR1 (Hu et al., 2017) . Further, we 6 performed BioID with WDR2 as bait, and the WDR2-proximal proteins were similarly screened based on their 7 previously determined subcellular localizations, and were plotted, based on the mass spectrometry Mascot score, 8 together with the KIN-G-proximal proteins to identify common proximal proteins (Fig. 1B). The list of proximal 9 proteins of KIN-G and WDR2 included three centrin arm- and basal body-localized proteins, eight hook complex-10 localized proteins, five MtQ proximal end-localized proteins, one hook shank-localized protein (TbSmee1), and 11 one Golgi-localized protein (CAAP1) (Fig. 1C). Other potential proximal proteins, including FAZ proteins, FPC 12 proteins, and cytoskeletal proteins, were also identified (Supplemental Table S1). However, these proteins were 13 not considered as potential interacting partners of KIN-G and/or WDR2 due to their farther distance to the centrin 14 arm. Nine proximal proteins were identified by both KIN-G BioID and WDR2 BioID (Fig. 1B, C, protein names 15 in blue), including TbCentrin2, CAAP1, TbSmee1, four hook complex-localized proteins, and two MtQ proximal 16 end-localized proteins (Fig. 1B, C). 17 To test whether WDR2 interacts with KIN-G in vivo in trypanosomes, we epitope-tagged KIN-G and WDR2 18 at their respective endogenous locus in the same cell line and then carried out co -immunoprecipitation. 19 Immunoprecipitation of WDR2-PTP was able to pull down KIN-G-3HA from the cell lysate (Fig. 1D), confirming 20 that the two proteins form a complex in trypanosomes. Further, co-immunostaining showed that WDR2 and KIN-21 G co-localized, and co-immunostaining with the centrin arm marker TbCentrin4 and the hook complex marker 22 TbMORN1 confirmed the localization of WDR2 to the centrin arm, alongside the shank part of the hook complex 23 (Fig. 1E). Therefore, WDR2 and KIN-G form a complex and may function together at the centrin arm. 24 25 Knockdown of WDR2 causes defective cytokinesis and FAZ elongation 26 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 6 To investigate the function of WDR2 in T. brucei, we performed RNAi in the procyclic form. Western blotting 1 was first carried out to monitor the protein level of WDR2, which was endogenously tagged with a triple HA 2 epitope, before and after WDR2 RNAi induction. WDR2 protein level was significantly reduced after one day 3 and was undetectable at day 3 of RNAi induction (Fig. 2A). The depletion of WDR2 by RNAi caused severe 4 growth defects and eventual cell death (Fig. 2B), suggesting the essentiality of WDR2 for cell viability in procyclic 5 trypanosomes. Knockdown of WDR2 caused a substantial increase of xNyK (x>2, y≥1) cells and the emergence 6 of abnormal cell types, such as 2N1K cells and 0N1K cells (Fig. 2C), which could be derived from the aberrant 7 division of 2N2K cells. These results demonstrated that WDR2 knockdown impaired cytokinesis, although this 8 could be the secondary effects attributed to the defects in FAZ elongation and flagellum positioning (see Fig. 2G-9 I below). Since the earliest time point for WDR2 RNAi to generate growth defects and cytokinesis defects was 48 10 h, all the subsequent phenotypic analyses were carried out with cells induced for WDR2 RNAi for 48 h. 11 After WDR2 RNAi, the bi-nucleated (2N2K and 2N1K) cells with a visible cleavage furrow were significantly 12 increased ( Fig. 2D). Moreover, in the WDR2 RNAi -induced cell population, almost all of the dividing bi-13 nucleated cells appeared to produce a smaller-sized daughter cell containing a long unattached flagellum and a 14 normal-sized daughter cell containing a normally attached flagellum, in striking contrast to that in control cells 15 where the two daughter cells had normally attached flagella and similar cell size (Fig. 2E, F ). In wild -type 16 trypanosomes, the two daughters of a dividing cell are not identical in size (Abeywickrema et al., 2019), but the 17 new-flagellum daughter cell of the dividing WDR2 RNAi cell was much smaller than that of the dividing control 18 cell (Fig. 2E, F). Thus, knockdown of WDR2 caused asymmetrical cytokinesis, likely due to cell division plane 19 misplacement. 20 The production of smaller -sized new flagellum daughter cells after WDR2 RNAi prompted us to examine 21 the effect of WDR2 RNAi on FAZ assembly and flagellum attachment by immunofluorescence microscopy using 22 the anti-CC2D antibody, which labels the intracellular FAZ filament (Zhou et al., 2011) , and the pan -centrin 23 antibody 20H5, which labels the flagellum in addition to the basal body and the centrin arm (He et al., 2005). The 24

showed t hat WDR2 knockdown produced bi -nucleated cells with a shorter new FAZ and a longer 25 unattached new flagellum (Fig. 2G-I). Measurement of the length of the new FAZ and the length of the unattached 26 (free) new flagellum showed that the average length of the ne w FAZ in 2N2K cells and 2N1K cells was 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 7 significantly reduced after WDR2 RNAi induction (Fig. 2H). Consequently, the average length of the unattached 1 new flagellum or free new flagellum in 2N2K cells and 2N1K cells was significantly increased after WDR2 RNAi 2 (Fig. 2I). These results suggest that WDR2 knockdown disrupted the elongation of the new FAZ. Because the 3 length of the new FAZ determines the position of the cell division plane in procyclic trypanosomes (Zhou et al., 4 2011) and WDR2 RNAi disrupted new FAZ elongation (Fig. 2G, H), the aberrant cytokinesis caused by WDR2 5 RNAi (Fig. 2E, F) was attributed to the disruption in FAZ elongation, which impaired the placement of the cell 6 division plane. 7 8 Knockdown of WDR2 disrupts flagellum positioning 9 Since the elongation of the new FAZ is coordinated with the migration of the new flagellum toward cell 10 posterior (Sunter and Gull, 2016), the defective FAZ elongation caused by WDR2 RNAi may inhibit flagellum 11 positioning. Indeed, immunostaining of WDR2 RNAi cells with the 20H5 antibody showed that the base of the 12 new flagellum of the dividing bi -nucleated cells was positioned in close proximity to that of the old flagellum 13 (Fig. 2F). This observation was confirmed by scanning electron microscopy, which showed that the new flagellum 14 was positioned either closer to or in the proximity of the old flagellum in the dividing WDR2 RNAi cells (Fig. 15 3A). These results suggest that WDR2 depletion caused defective segregation and positioning of the new flagellum. 16 Further, we investigated the effect of WDR2 RNAi on the segregation of the basal body, which nucleates the 17 flagellum (Hu et al., 2015a), by immunofluorescence microscopy with the anti-TbSAS-6 antibody that detects the 18 basal body cartwheel protein SAS-6 (Hu et al., 2015a). After WDR2 RNAi, the average inter-basal body distance 19 in bi-nucleated cells was significantly reduced in 2N2K and 2N1K cells (Fig. 3B, C). Finally, we investigated the 20 effect of WDR2 RNAi on the segregation of flagellum-associated structures, such as the centrin arm and the FPC 21 by immunofluorescence microscopy with the 20H5 antibody and the anti-TbBILBO1 antibody, respectively (Fig. 22 3D-G). The results showed that the average inter-centrin arm distance and the average inter-FPC distance were 23 both significantly reduced (Fig. 3D-G). These results provided further evidence that WDR2 depletion disrupted 24 flagellum positioning. 25 26 Knockdown of WDR2 disrupts hook complex morphology and integrity 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 8 Since WDR2 localizes to the centrin arm and forms a complex with KIN-G, which is required for hook 1 complex assembly (Zhou et al., 2024) , we asked whether WDR2 plays similar roles. We first performed co-2 immunofluorescence microscopy using the pan -centrin antibody 20H5 and the anti -TbCentrin4 antibody and 3 examined the integrity of the centrin arm in bi-nucleated cells before and after WDR2 RNAi induction. While the 4 average length of the old centrin arm was not significantly affected by WDR2 RNAi , the average length of the 5 new centrin a rm was reduced by ~43% from ~1.3µm to ~0.7µm after RNAi induction (Fig. 4A, B). Those bi-6 nucleated cells with the new centrin arm longer than 1.0µm were significantly reduced, whereas those bi-nucleated 7 cells with the new centrin arm shorter than 1.0µm were significantly increased after RNAi induction (Fig. 4C). 8 These results suggest that WDR2 is required to maintain centrin arm integrity. We next asked whether knockdown 9 of WDR2 disrupted the integrity of the fishhook-like structure of the hook complex, and by immunofluorescence 10 microscopy using the anti -TbMORN1 antibody, we showed that t he old fishhook-like structure remained its 11 typical morphology, but the new fishhook-like structure lost the shank part of the structure in ~65% of the 2N2K 12 cells and ~86% of the 2N1K cells after RNAi (Fig. 4D, E), indicating the disruption of the morphology of the 13 fishhook-like structure. Together, these results demonstrated that WDR2 is required for hook complex assembly 14 by maintaining its morphology and integrity. 15 16 Knockdown of WDR2 impairs Golgi biogenesis 17 To investigate whether WDR2 RNAi also disrupts Golgi biogenesis, like its interacting partner KIN-G (Zhou 18 et al., 2024), we performed co-immunofluorescence microscopy using the anti-TbGRASP antibody to detect the 19 Golgi matrix protein TbGRASP and the anti-HA antibody to detect the 3HA-tagged CAAP1, a Golgi peripheral 20 protein (Zhou et al., 2024), and then counted the number of Golgi in control and WDR2 RNAi cells (Fig. 5A, B). 21 Knockdown of WDR2 resulted in a ~50% reduction of the bi-nucleated cells with two, three, or four Golgi and a 22 corresponding increase of bi-nucleated cells with more than four Golgi from ~4% to ~52% (Fig. 5B). Knockdown 23 of WDR2 caused similar effects on the Sec13-marked ERES (ER exit site), a specialized ER zone involved in the 24 ER-to-Golgi cargo transport (Kurokawa and Nakano, 2019), producing many bi-nucleated cells with more than 25 four ERES foci ( Fig. 5C). Many of the Golgi detected in the WDR2 RNAi cells were smaller in size, having 26 weaker TbGRASP and CAAP1 signal ( Fig. 5A), so w ere some of the ERESs, which had weaker Sec13 signal 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 9 (Fig. 5C). Such smaller Golgi were also detected in control cells ( Fig. 5A, arrows ), as reported previously (de 1 Graffenried et al., 2008; He et al., 2004), which often disappear upon cytokinesis, likely being integrated into the 2 existing Golgi or being disassembled (He et al., 2004). Because the centrin arm determines the assembly site for 3 new Golgi or regulates the size of the new Golgi (de Graffenried et al., 2008; He et al., 2005) and WDR2 disrupted 4 centrin arm integrity (Fig. 4A -C), we wondered whether the shorter new centrin arm in WDR2 RNAi cells 5 associates with a smaller Golgi. Thus, we performed co-immunofluorescence microscopy with the anti-TbGRASP 6 antibody to label the Golgi and 20H5 to label the centrin arm (Fig. 5D). In the non-induced control cells, the old 7 centrin arm-associated and the new centrin arm-associated Golgi had similar TbGRASP signal intensity , but in 8 the WDR2 RNAi cells, the new centrin arm-associated Golgi had weaker TbGRASP signal than the old centrin 9 arm-associated Golgi ( Fig. 5D, E). Together, these results demonstrated that WDR2 depleti on disrupted Golgi 10 biogenesis, likely by impairing the maturation of the new centrin arm-associated Golgi. This phenotype is similar 11 to the knockdown of KIN -G (Zhou et al., 2024) and the knockdown of TbPLK (de Graffenried et al., 2008) , 12 suggesting that WDR2, KIN-G, and TbPLK may function in the same pathway to regulate Golgi biogenesis. 13 14 Knockdown of WDR2 disrupts KIN-G localization and destabilizes KIN-G 15 Because WDR2 forms a complex with KIN-G (Fig. 1) and the phenotype of WDR2 knockdown mimicked 16 that of KIN-G knockdown ( Figs. 2 -5), we asked whether WDR2 regulates KIN -G or vice versa. 17 Immunofluorescence microscopy showed that knockdown of WDR2 disrupted KIN-G localization at the centrin 18 arm (Fig. 6A). After WDR2 RNAi for 24 h, KIN -G signal at the new centrin arm either became weaker or was 19 undetectable in ~44% of the 1N2K cells and ~45% of the 2N2K cells (Fig. 6A, B ). Moreover, KIN-G was 20 undetectable at both the new and the old centrin arms in ~ 11% each of the 1N2K and 2N2K cells (Fig. 6A, B). 21 After WDR2 RNAi for 48 h, KIN-G was undetectable at both the new and the old centrin arms in ~83% of the 22 1N2K cells and ~97% of the 2N2K cells ( Fig. 6A, B). These results suggest that WDR2 is required for the 23 recruitment of KIN-G to the new centrin arm and for the mainte nance of KIN-G at the old centrin arm. Western 24 blotting showed that the level of KIN-G protein was gradually decreased after WDR2 RNAi induction , and 25 treatment of WDR2 RNAi-induced cells (day 2) with the proteasome inhibitor MG132 for 8 h restored KIN-G 26 protein level (Fig. 6C). Immunofluorescence microscopy showed that KIN-G remained undetectable at both the 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 10 new and the old centrin arms in the WDR2 RNAi cells (day 2) treated with MG132, albeit the overall cytosolic 1 KIN-G signal was substantially increased (Fig. 6A), suggesting that KIN-G was degraded in the cytosol of WDR2 2 RNAi cells. Conversely, immunofluorescence microscopy showed that WDR2 was still detected at the centrin 3 arm in KIN -G RNAi cells (Fig. 6D), and western blotting showed that WDR2 protein le vel was unaffected by 4 KIN-G RNAi (Fig. 6E). Altogether, these results suggest that WDR2 is not a KIN-G cargo, but instead it functions 5 to recruit and stabilize KIN-G. 6 7 Determination of WDR2 structural motifs required for interaction with KIN-G 8 The structural motifs in WDR2 were predicted by homology modeling and AlphaFold2 prediction (Fig. 7A-9 C). WDR2 contains in its C-terminal domain (CTD) a WD40 domain composed of seven WD40 repeats, which 10 adopts a circularized, seven-bladed -propeller structure, like a circular solenoid (Fig. 7A-C). WD40 domain is 11 known to serve as a rigid scaffold for protein -protein interaction, and WD40 domain-containing proteins often 12 function in coordinating multi-protein complex assembly (Smith et al., 1999). Thus, the WD40 domain in WDR2 13 may play roles in mediating protein complex assembly in trypanosomes. AlphaFold2 also predicted two domains 14 of unknown function (DUF) in the N-terminal domain (NTD) of WDR2, each of which comprises two -sheets 15 and three -helices (Fig. 7A-C). Homology modeling by SWISS-MODEL (Arnold et al., 2006; Biasini et al., 16 2014) detected a PB1 (Phox and Bem1) domain in the middle domain (MD) of WDR2, which comprises about 80 17 amino acids and exhibits a ubiquitin-like -grasp fold with two -helices and five -sheets (Fig. 7A-C). The PB1 18 domain is evolutionarily conserved among animals, fungi, plants, and protozoa, and functions as a protein-protein 19 interaction module through PB1-mediated heterodimerization/homodimerization or interaction with other protein 20 domains (Sumimoto et al., 2007). Within the middle domain of WDR2, a short coiled-coil (CC) motif was detected 21 between DUF2 and PB1 (Fig. 7A, B). 22 As WDR2 forms a complex with KIN-G, which contains a kinesin motor domain (KMD) in the N-terminus 23 and three CC motifs (CC1-CC3) in the C-terminus (Fig. 7D), we predicted the structure of the WDR2-KIN-G 24 complex using AlphaFold3 (Abramson et al., 2024), which showed that the two proteins may interact via multiple 25 domains (Fig. 7E). The WD40 domain of WDR2 may interact with the loop #1 (L1) of the KMD of KIN -G, the 26 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 11 CC motif and the DUF2 domain of WDR2 may interact with the CC3 motif of KIN-G, and the PB1 domain of 1 WDR2 may interact with the -helix #4 (4) of the KMD of KIN-G (Fig. 7E). To test the potential contribution 2 of WDR2 structural domains to the interaction with KIN-G, we ectopically expressed wild-type WDR2 and three 3 domain-deletion mutants, WDR2-NTD, WDR2-MD, and WDR2-CTD, each of which was tagged with a C-4 terminal triple HA epitope, in the WDR2-3’UTR RNAi cell line (see below for details), and then performed co-5 immunoprecipitation followed by mass spectrometry. Wild-type WDR2, WDR2-NTD, and WDR2-CTD each 6 pulled down a protein of ~70 kDa (Fig. 7F), which was identified as KIN-G by mass spectrometry (Supplemental 7 Fig. S 1 and Supplemental Table S2). However, WDR2 -MD pulled down a protein with a slightly smaller 8 molecular mass than that pulled down by wild-type WDR2 and the other two WDR2 mutants ( Fig. 7F), and this 9 protein was also identified as KIN-G by mass spectrometry (Supplemental Fig. S1 and Supplemental Table S3), 10 suggestive of a potential non -phosphorylated form of KIN -G pulled down by WDR2 -MD only. To test this 11 possibility, trypanosome cell lysate was treated with Lambda protein phosphatase ( PPase) prior to co -12 immunoprecipitation, and wild-type WDR2 and all the three WDR2 mutants pulled down the KIN -G protein of 13 the same molecular mass (Fig. 7F). These results suggest two possibilities. First, KIN-G is de-phosphorylated in 14 the WDR2-MD mutant cells, but the dephosphorylated KIN-G still interacts with WDR2-MD. Second, WDR2-15 MD only interacts with de-phosphorylated KIN-G, whereas wild-type WDR2 and the other two WDR2 mutants 16 interact with phosphorylated KIN-G. Nonetheless, these results suggest that none of the three domains in WDR2 17 are exclusively required for the interaction with KIN-G and all of the three domains (NTD, MD, and CTD) mediate 18 WDR2-KIN-G interaction, in agreement with the interaction interfaces predicted by AlphaFold3 (Fig. 7E). 19 20 Structural motifs required for WDR2 cellular function and KIN-G localization 21 To investigate the potential function of the DUF domains in the NTD, the CC and the PB1 in the MD, and the 22 WD40 domain in the CTD, we deleted the NTD, the MD, or the CTD of WDR2 (Fig. 8A) for complementation 23 of WDR2 deficiency. To this end, we first generated a WDR2 -3’UTR RNAi cell line by targeting against the 24 3’UTR of WDR2, and western blotting demonstrated the knockdown of endogenous , PTP-tagged WDR2 (Fig. 25 8B). This knockdown caused growth defects (Fig. 8C), similar to the RNAi against the coding region of WDR2 26 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 12 (Fig. 2B). Subsequently, we ectopically expressed wild-type and the domain-deletion mutants of WDR2, each of 1 which was tagged with a C-terminal triple HA epitope, in the WDR2-3’UTR RNAi cell line (Fig. 8D). Expression 2 of wild-type WDR2 was able to restore the growth defects of WDR2 -3’UTR RNAi cells, whereas expression of 3 WDR2-NTD partially restored the growth defects , and expression of WDR2-MD or WDR2-CTD failed to 4 restore the growth defects of the WDR2 -3’UTR RNAi cells (Fig. 8E). These results suggest that the NTD is 5 important for WDR2 function, whereas the MD and the CTD are essential for WDR2 function. 6 We next examined the subcellular localization of wild-type and the three domain-deletion mutants of WDR2 7 by immunofluorescence microscopy. Ectopically expressed, 3HA-tagged WDR2 localized to the centrin arm but 8 extended to the region between the basal body and the centrin arm , reminiscent of the proximal end of the MtQ 9 (Fig. 8F), in contrast to the endogenously 3HA-tagged WDR2, which was detected only at the centrin arm (Fig. 10 1E). Similarly, e ctopically expressed, 3HA-tagged KIN-G was detected at both the centrin arm and the MtQ 11 proximal end (Zhou et al., 2024), in contrast to the endogenously 3HA-tagged KIN-G, which was detected only 12 at the centrin arm (Figs. 1E and 6A). The discrepancy could be due to the ectopic expression that produced more 13 proteins to be spread to the MtQ proximal end . Nonetheless, the WDR2-NTD mutant was only detected at the 14 centrin arm, the WDR2-CTD mutant was only detected at the MtQ proximal end, and the WDR2-MD mutant 15 was mis-localized to the cytosol (Fig. 8F). These results suggest that the CTD is required for localization of the 16 ectopically overexpressed WDR2 to the centrin arm , the NTD is required for localization to the MtQ proximal 17 end, and the MD is required for localization to both the centrin arm and the MtQ proximal end. 18 Because WDR2 knockdown disrupted KIN-G localization (Fig. 6), we wondered whether WDR2 can actively 19 target KIN-G to its subcellular location. To test this possibility, we examined the localization of endogenous, PTP-20 tagged KIN-G in WDR2-3’UTR RNAi cells expressing ectopic WDR2 and its domain-deletion mutants, which 21 localized to both the centrin arm and the MtQ proximal end, the centrin arm, the cytosol, or the MtQ proximal 22 end, respectively (Fig. 8F). In non -induced cells, endogenously PTP-tagged KIN-G was detectable only at the 23 centrin arm (Fig. 8G), as expected. However, in cells induced to ectopically express WDR2, KIN-G co-localized 24 with WDR2 at both the centrin arm and the MtQ proximal end ( Fig. 8G). In cells induced to ectopically express 25 WDR2-NTD, KIN -G co -localized with WDR 2-NTD at the centrin arm ( Fig. 8 G). In cells induced to 26 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 13 ectopically express WDR2 -MD, KIN -G was mis-localized to cytosol (Fig. 8 G), and in cells induced to 1 ectopically express WDR2-CTD, KIN-G co-localized with WDR2-CTD at the MtQ proximal end ( Fig. 8G). 2 Thus, the endogenously PTP -tagged KIN-G was targeted to the same locations where ectopically expressed 3 WDR2 and its domain-deletion mutants were localized, suggesting that WDR2 can actively recruit KIN-G. 4 5

6 The centrin arm structure in T. brucei is defined by two centrin proteins, TbCentrin2 and TbCentrin4, and was 7 originally termed the bilobe structure involved in Golgi biogenesis, with one lobe associating with the existing 8 Golgi and the other associating with the growing Golgi (He et al., 2005). Based on the findings from knockdown 9 of TbCentrin2, the centrin arm is hypothesized to determine the site where the new Golgi is assembled (He et al., 10 2005). However, knockdown of TbCentrin4 had no effect on centrin arm assembly and Golgi biogenesis, but it 11 disrupted the coordination of karyokinesis and cytokinesis (Shi et al., 2008). Three other proteins, TbPLK, WDR1, 12 and KIN-G, also localize to the centrin arm. TbPLK regulates centrin arm assembly and Golgi biogenesis, but its 13 regulation of Golgi biogenesis is opposite to that of TbCentrin2, despite that TbCentrin2 is a substrate of TbPLK 14 (de Graffenried et al., 2008). WDR1 is required for centrin arm assembly and basal body segregation, and it plays 15 these roles by controlling the abundance of TbPLK at the centrin arm and the basal body (Hu et al., 2017). Whether 16 WDR1 regulates Golgi biogenesis was not investigated previously, but given its role in regulating TbPLK, it is 17 likely that WDR1 also regulates Golgi biogenesis through an indirect means via TbPLK . KIN-G is an orphan 18 kinesin, and it promotes the biogenesis of the hook complex and the Golgi (Zhou et al., 2024), similar to TbPLK 19 (de Graffenried et al., 2008) . KIN-G plays a role in Golgi biogenesis by recruiting the Golgi peripheral protein 20 CAAP1, which promotes the association of Golgi with the centrin arm (Zhou et al., 2024). As a microtubule plus 21 end-directed motor protein, KIN-G may transport cargos along the MtQ to their destination at the centrin arm and 22 the Golgi to facilitate hook complex and Golgi biogenesis. All in all, it appears that the above-mentioned centrin 23 arm-localized proteins play diverse functions, although some of them may act in the same pathway. 24 We have attempted to identify the potential cargo proteins of KIN-G by proximity biotinylation and have 25 identified a subset of KIN-G-proximal proteins , including three centrin arm -localized proteins (TbCentrin2, 26 TbCentrin4, and WDR2) , four hook complex -localized proteins (Hypo.1-Hypo.4), the hook shank -localized 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 14 protein TbSmee1, and two MtQ proximal end-localized proteins (Hypo.8 and Hypo.10) (Fig. 1B, C). Although 1 the hook complex-localized proteins and MtQ proximal end-localized proteins are in close proximity to KIN-G, 2 they are unlikely to be cargo proteins of KIN -G due to the enrichment of KIN-G in the distal part of the centrin 3 arm (Zhou et al., 2024). We postulate that the most likely cargo proteins of KIN-G are those proteins that localize 4 to the centrin arm-Golgi region, including the hook shank-localized TbSmee1, Golgi peripheral protein CAAP1, 5 and centrin arm-localized proteins. The centrin arm-localized WDR2 was not affected by KIN-G RNAi (Fig. 6), 6 ruling out its candidacy as a KIN -G cargo protein. TbSmee1 was only slightly reduced in some (~20%) of the 7 KIN-G RNAi cell s (Zhou et al., 2024) ; thus, it is unlikely to be a cargo for KIN -G. In contrast, CAAP1 and 8 TbCentrin4 were significantly reduced from the new Golgi and the new centrin arm, respectively , in KIN -G 9 knockdown cells (Zhou et al., 2024) , suggesting that they are potential cargo proteins of KIN-G, although their 10 candidacy as KIN -G cargos remains to be verified by additional experimental approaches . Whether KIN -G 11 transports any other types of cargos, such as vesicles, lipids, and /or organelles, remains to be explored , and is 12 beyond the scope of this study. 13 We have provided evidence to support a regulatory role for WDR2 in the recruitment and maintenance of 14 KIN-G at the centrin arm (Fig. 6). Knockdown of WDR2 disrupted the localization of KIN-G to both the new and 15 the old centrin arms, resulting in the eventual degradation of KIN-G in the cytosol (Fig. 6). Since the new centrin 16 arm undergoes de novo biogenesis during early cell cycle, the disrupted localization of KIN-G to the new centrin 17 arm by WDR2 knockdown (Fig. 6A, B) suggests that WDR2 is required for recruiting KIN-G to the new centrin 18 arm. This notion was further supported by the finding that native KIN-G interacted with and co-localized with the 19 ectopically overexpressed wild -type or the domain -deletion mutants of WDR2 ( Figs. 7F and 8G), which 20 demonstrated an active action of WDR2 in recruiting KIN-G. Further, since KIN-G had already been recruited to 21 the old centrin arm prior to WDR2 RNAi induction, the disrupted localization of KIN -G at the old centrin arm 22 (Fig. 6A, B) suggests that WDR2 is also required for maintaining KIN-G at the old centrin arm. It should be noted 23 that the observed loss of KIN-G signal at the old centrin arm after WDR2 RNAi can also be due to defect s in 24 recruitment. In this scenario, the 2N2K cells that failed to recruit KIN-G to the new centrin arm divided to produce 25 a 1N1K new-flagellum daughter cell that further progressed through the cell cycle to become a 1N2K cell and 26 then a 2N2K cell, in which KIN-G was lost at both the new and the old centrin arms. Because this 1N1K new-27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 15 flagellum daughter cell has a long unattached flagellum, its derived 1N2K cell and 2N2K ce ll will have long 1 unattached new and old flagella. However, we examined the 1N2K and 2N2K cells that had a normal old flagellum 2 and found that they lost KIN -G signal at the old centrin arm ( Fig. 6A). This result suggests that WDR2 plays a 3 role in maintaining KIN-G at the old centrin arm. Together, these results uncovered a mechanistic role for WDR2 4 in recruiting and maintaining KIN-G at the centrin arm. Because KIN-G was degraded in the cytosol of the WDR2 5 RNAi cells ( Fig. 6A, C ), whereas the cytosolically mis-localized KIN -G in the WDR2-3’UTR RNAi cells 6 expressing WDR2-MD was stable (Figs. 7F and 8G), it suggests that interaction with WDR2 stabilizes KIN-G. 7 Knockdown of WDR2 phenocopies all of the defects caused by knockdown of KIN -G in procycli c 8 trypanosomes. First, depletion of WDR2 arrested cytokinesis progression, producing a normal-sized daughter cell 9 and a small -sized daughter cell with a long unattached flagellum, due to the inhibition of new FAZ elongation 10 (Fig. 2). Secondly, knockdown of WDR2 disru pted the positioning of the newly assembled flagellum and the 11 segregation of flagellum-associated cytoskeletal structures ( Fig. 3). Thirdly, deficiency in WDR2 disrupted the 12 typical morphology of the hook complex by reducing the length of the centrin arm and eliminating the shank part 13 of the fishhook-like structure (Fig. 4). Finally, knockdown of WDR2 impaired Golgi biogenesis by disrupting the 14 maturation of the new centrin arm-associated Golgi (Fig. 5). Since WDR2 localizes to the centrin arm, the primary 15 defect caused by WDR2 knockdown is the disruption of hook complex integrity, which led to secondary defects 16 on the elongation of the new FAZ and the positioning of the new flagellum. Consequently, the defects in FAZ 17 elongation and flagellum positioning caused the mis-positioning of the cell division plane, leading to asymmetrical 18 cytokinesis (Fig. 2). All of these cellular defects caused by WDR2 RNAi could be attributed to the disrupted 19 localization and destabilization of KIN-G (Fig. 6). Thus, the primary function for WDR2 is to recruit KIN -G to 20 the centrin arm for the latter to play its role in regulating hook complex and Golgi biogenesis. 21 The endogenous, epitope-tagged WDR2 and KIN-G were detected only at the centrin arm (Fig. 1E). However, 22 the ectopically expressed, epitope-tagged WDR2 and KIN-G were detected at both the centrin arm and the MtQ 23 proximal end (Fig. 8 and (Zhou et al., 2024)), similar to the endogenous, mNeoGreen-tagged WDR2 and KIN-G 24 reported in the TrypTag project (Billington et al., 2023). It is likely that native WDR2 and KIN-G proteins localize 25 to both the centrin arm and the MtQ proximal end, but for unknown reasons, the anti-HA and the anti-Protein A 26 antibodies used in our work were not able to detect the two proteins at the MtQ proximal end . Nonetheless, 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 16 localization of KIN-G to the centrin arm depends on its motor activity (Zhou et al., 2024), whereas localization of 1 WDR2 to the centrin arm requires its WD40 domain ( Fig. 8F). Intriguingly, localization of WDR2 to the MtQ 2 proximal end requires the N-terminal DUF domains, and localization of WDR2 to bo th the centrin arm and the 3 MtQ proximal end requires the CC and the PB1 motif in the middle domain (Fig. 8F). It is possible that WDR2 is 4 first recruited to the MtQ proximal end and the centrin arm, where WDR2 recruits KIN-G. Alternatively, WDR2 5 and KIN-G may form a complex in the cytosol, and the complex is then recruited to the two structures in a WDR2-6 dependent manner. 7 The finding that wild-type WDR2 and the WDR2-NTD and WDR2 -CTD mutants, but not the WDR2-8 MD mutant, pulled down phosphorylated KIN -G from trypanosome cell lysate (Fig. 7F) suggests that either 9 KIN-G is phosphorylated at the centrin arm or phosphorylation of KIN -G targets it to the centrin arm . In the 10 cytosol, KIN-G is non-phosphorylated, but it is still capable of forming a complex with the WDR2-MD mutant 11 (Fig. 7F ), suggesting that complex formation is independent of KIN -G phosphorylation. Because KIN -G 12 localization depends on WDR2 ( Figs. 6 and 8 ), it suggests that phosphorylation of KIN-G is unlikely to play a 13 role in targeting itself to the centrin arm. The best candidate protein kinase responsible for KIN-G phosphorylation 14 is TbPLK, which localizes to the centrin arm and is required for hook complex and Golgi biogenesis (de 15 Graffenried et al., 2008) , similar to the role of KIN -G. Whether TbPLK phosphorylates KIN -G and how the 16 phosphorylation of KIN -G may impact the biochemical function of KIN -G, i.e., the motor activity and/or 17 microtubule-binding activity, and the physiological function of KIN -G in regulating the biogenesis of the hook 18 complex and the Golgi are currently under investigation. 19 In closing, we identified WDR2 as a KIN-G-interacting partner protein, which co-localizes with KIN-G at the 20 centrin arm in the procyclic for m of T. brucei and regulates KIN -G by recruiting the latter to the centrin arm, 21 thereby promoting the biogenesis of the hook complex and the Golgi apparatus. The disruption of hook complex 22 biogenesis further impaired FAZ elongation and flagellum positioning, leading to mis-placement of the cell 23 division plane and, consequently, asymmetrical cytokinesis. 24 25

26 Structural modeling and AlphaFold prediction of protein structure 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 17 Structural modeling of WDR2 structural motifs was carried out by SWISS-MODEL 1 (https://swissmodel.expasy.org/) (Arnold et al., 2006; Biasini et al., 2014) . The WDR2 protein structure was 2 obtained directly from AlphaFold protein structure database (https://alphafold.ebi.ac.uk/), and the two N-terminal 3 domains of unknown function in WDR2 were predicted by AlphaFold2 (Jumper et al., 2021; Varadi et al., 2022). 4 The structure of the WDR2 -KIN-G complex was predicted by AlphaFold3 (Abramson et al., 2024) using the 5 following webserver: https://alphafoldserver.com/. 6 7 Trypanosome cell culture and RNAi 8 The procyclic T. brucei Lister427 strain was grown in SDM-79 medium containing 10% heat-inactivated fetal 9 bovine serum (Millipore-Sigma) at 27oC, and the procyclic T. brucei strain 29-13 (Wirtz et al., 1999) was cultured 10 in SDM-79 medium supplemented with 10% heat-inactivated fetal bovine serum, 15 µg/ml G418, and 50 µg/ml 11 hygromycin at 27oC. 12 To generate the WDR2 RNAi cell line, a 453-bp DNA fragment (nucleotides 1407-1859) of the WDR2 gene 13 was amplified from trypanosome genomic DNA by PCR and cloned into the pZJM vector (Wang et al., 2000) . 14 The resulting plasmid, pZJM-WDR2, was used to transfect the 29-13 strain by electroporation. Transfectants were 15 selected with 2.5 µg/ml phleomycin, and successful transfectants were cloned by limiting dilution in a 96 -well 16 plate containing the SDM-79 medium supplemented with 20% fetal bovine serum and appropriate antibiotics. The 17 KIN-G RNAi cell line was generated previously (Zhou et al., 2024). 18 RNAi was induced by incubating the RNAi cell line with 1.0 µg/ml tetracycline. Three clonal RNAi cell lines 19 were chosen for phenotypic analysis, and because the three clonal cell lines showed almost identical phenotypes, 20 only one clonal RNAi cell line was used for characterization throughout the work. 21 22 In situ epitope tagging of proteins 23 Epitope tagging of WDR2, CAAP1, KIN-G, and Sec13 from their respective endogenous locus was carried 24 out by the PCR-based epitope-tagging method reported previously (Shen et al., 2001). For WDR2 and KIN-G co-25 localization experiments, WDR2 was endogenously tagged with a C -terminal PTP epitope , and KIN-G was 26 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 18 endogenously tagged with a C-terminal triple HA epitope in the same cell line. Transfectants were selected with 1 appropriate antibiotics, and clonal cell lines were obtained by limiting dilution as described above. 2 3 Generation of WDR2 RNAi complementation cell lines 4 To generate WDR2 RNAi complementation cell lines, we first created a WDR2-3’UTR RNAi cell line for 5 ectopic expression of wild -type and the structure domain-deletion mutants of WDR2. To this end, a 564-bp 6 fragment of the 3’UTR of WDR2 was cloned into the pZ JM-PAC vector, and t he resulting plasmid was 7 electroporated into the procyclic 29-13 strain. Transfectants were selected with 1.0 µg/ml puromycin in addition 8 to 15 µg/ml G418 and 50 µg/ml hygromycin, and clonal cell lines were generated by limiting dilution in a 96-well 9 plate as described above . Subsequently, the full -length WDR2 gene and the mutant WDR2 gene lacking the 10 sequence encoding the NTD, the MD, or the CTD were each cloned into the pLew100-3HA-BLE vector (Wei et 11 al., 2014). The resulting plasmids, pLew100-WDR2-3HA-BLE, pLew100-WDR2-NTD-3HA-BLE, pLew100-12 WDR2-MD-3HA-BLE, and pLew100 -WDR2-CTD-3HA-BLE, were used to transfect the WDR2-3’UTR 13 RNAi cell line. Transfectants were selected with 2.5 µg/ml phleomycin in additional to1.0 µg/ml puromycin, 15 14 µg/ml G418, and 50 µg/ml hygromycin, and then cloned by limiting dilution in a 96-well plate as described above. 15 To induce WDR2-3’UTR RNAi and ectopic expression of WDR2-3HA and its domain-deletion mutants, cells 16 were incubated with 1.0 μg/ml tetracycline. 17 18 Co-immunoprecipitation, PPase treatment, silver staining, and mass spectrometry 19 Trypanosomes cells expressing KIN -G-3HA, WDR2 -PTP, or both KIN -G-3HA and WDR2 -PTP at their 20 respective endogenous locus were lysed in 0.5 ml immunoprecipitation (IP) buffer (25 mM Tris-HCl, pH 7.4, 100 21 mM NaCl, 1 mM DTT, 0.1% NP-40, 5% glycerol, and protease inhibitor cocktail) on ice for 30 min. Cell lysate 22 was cleared by centrifugation at the highest speed in a table -top microcentrifuge. The cleared cell lysate was 23 incubated with 25 μl settled IgG Sepharose beads (GE Healthcare) at 4°C for 60 min. After centrifugation at 4°C 24 for 10 min in a microcentrifuge, the IgG Sepharose b eads were washed five times with the IP buffer, and bound 25 proteins were eluted by boiling the beads in 1x SDS sampling buffer for 5 min. Eluted proteins were separated by 26 SDS-PAGE, transferred onto a PVDF membrane, and immunoblotted with anti-HA monoclonal antibody (clone 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 19 HA-7, H9658, Sigma -Aldrich, 1:5,000 dilution) to detect 3HA -tagged KIN-G, and anti -Protein A polyclonal 1 antibody (anti-ProtA; P3775, Sigma-Aldrich, 1:5,000 dilution) to detect PTP-tagged WDR2. 2 Trypanosome cells expressing ectopic WDR2 -3HA or its domain -deletion mutants in the WDR2 -3’UTR 3 RNAi cell line were lysed in the IP buffer on ice for 30 min. For Lambda protein phosphatase (PPase) treatment, 4 100 units of PPase (New England Biolabs, Cat#: P0753), 1× NEBuffer for Protein MetalloPhosphatases, and 1 5 mM MnCl2 were added into the cell lysate and incubated at 30°C for 30 min. Cell lysate was cleared by 6 centrifugation, the cleared supernatant was incubated with 15 μl settled anti-HA agarose beads (Millipore-Sigma) 7 at 4°C for 60 min. Beads were then washed five times with the IP buffer, and bound proteins were eluted for SDS-8 PAGE followed by silver staining using the Pierce™ Silver Stain Kit (Cat# 24612, ThermoFisher Scientific) 9 according to manufacturer’s instructions. 10 Co-immunoprecipitated protein b ands were excised from the SDS -PAGE gel and analyzed by mass 11 spectrometry using the USCF in-gel digestion protocol described in our previous publication (An et al., 2018). 12 After overnight trypsin digestion, peptides were extracted from the gel with 50% acetonitrile and 5% formic acid, 13 dried using speed-vac, and then reconstituted in 2% acetonitrile with 0.1% formic acid. Peptides were injected on 14 to Thermo LTQ Orbitrap XL (Thermo-Fisher Scientific, Bremen, Germany) interfaced with an Eksigent nano-LC 15 2D plus ChipLC system (Eksigent Technologies, Dublin, CA). Raw data files were processed and searched using 16 the Mascot search engine against the T. brucei proteome database. Mass spectrometry and data analysis were 17 carried out at the Clinical and Translational Proteomics Service Center of the University of Texas Health Science 18 Center at Houston. 19 20 Proximity-dependent biotin identification and mass spectrometry 21 To identify proximal proteins of KIN-G and WDR2 by BioID, the full-length sequence of KIN-G and WDR2 22 genes was each cloned into the pLew100 -BirA*-HA vector (Hu et al., 2015b) , and the resulting plasmid was 23 linearized by NotI restriction digestion and used to transfect the procyclic 29-13 strain. Transfectants were selected 24 with 2.5 µg/ml phleomycin and cloned by limiting dilution as described above. To express KIN-G-BirA*-HA or 25 WDR2-BirA*-3HA, cells were incubated with 0.5 µg/ml tetracycline for 24 h and subsequently 50 µM biotin 26 were added and further incubated for 24 h. Cells were harvested and treated with PEME buffer (100 mM PIPES 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 20 pH 6.9, 2 mM EGTA, 1 mM MgSO4, 0.1 mM EDTA) containing 0.5% Nonidet P-40, and the cytosolic (soluble) 1 and cytoskeletal (pellet) fractions were prepared by centrifugation. To solubilize cytoskeletal proteins, t he 2 cytoskeletal fraction was extracted with lysis buffer (0.4% SDS, 500 mM NaCl, 5 mM EDTA, 1 mM DTT , 50 3 mM Tris-HCl, pH 7.4) , and the cytoskeletal extract w as incubated with 500 µl pre -washed streptavidin-coated 4 Dynabeads (Invitrogen) at 4oC for 4 h. The Dynabeads were washed five times with PBS and then five time with 5 50 mM ammonium bicarbonate before re-suspending in 100 mM ammonium bicarbonate. Finally, 10% DTT, 50% 6 iodoacetamide, and 5% DTT were sequentially added to the Dynabeads re-suspension. The Dynabeads 7 resuspension was treated with trypsin at 37 oC for 16 hours , and trypsin-digested peptides were desalted and 8 analyzed on an LTQ Orbitrap XL mass spectrometer (Thermo-Fisher Scientific) interfaced with an Eksgent nano-9 LC 2D plus chipLC system (Eksigent Technologies). Raw mass spectrometry data was searched against the T. 10 brucei proteome database using the Mascot search engine. The search conditions used peptide tolerance of 10 11 ppm and MS/MS tolerance of 0.8 Da with the enzyme trypsin and two missed cleavages. 12 13 Immunofluorescence microscopy 14 Trypanosome cells were settled onto glass coverslips, fixed with methanol at -20oC for 30 min, and then 15 rehydrated with PBS for 10 min at room temperature. Cells on the coverslips were blocked with 3% BSA in PBS 16 at room temperature for 1 h, and incubated at room temperature for 1 h with the following primary antibodies: 17 anti-CC2D polyclonal antibody (1:1,000 dilution) (Zhou et al., 2011) , L8C4 (anti -PFR2) monoclonal antibody 18 (1:50 dilution) (Kohl et al., 1999) , 20H5 monoclonal antibody (1:400 dilution) (He et al., 2005) , anti -19 TbCentrin4/LdCentrin1 polyclonal antibody (1:1000 dilution), anti -TbMORN1 antibody (1: 5000 dilution) 20 (Morriswood et al., 2009) , anti-TbBILBO1 antibody (1: 400 dilution) (Bonhivers et al., 2008) , anti-TbGRASP 21 polyclonal antibody (1:400 dilution) (He et al., 2004), anti-TbSAS-6 antibody (1:1,000 dilution) (Hu et al., 2015a), 22 FITC-conjugated anti-HA monoclonal antibody (1:400 dilution, Sigma -Aldrich), and anti-Protein A polyclonal 23 antibody (1:400 dilution, Sigma-Aldrich). Cells were washed with PBS and then incubated at room temperature 24 for 1 h with the following secondary antibodies: FITC-conjugated anti-mouse IgG, Cy3-conjugated anti-rabbit 25 IgG, Cy3-conjugated anti-mouse IgG, and FITC -conjugated anti-rabbit IgG, all of which were purchased from 26 Sigma-Aldrich. Cells were washed with PBS for three times, mounted with DAPI -containing Vec taShield 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 21 mounting medium (Vector Labs), and imaged with the Olympus IX71 fluorescence microscope. Images were 1 acquired and processed using the Slidebook software. 2 3 Scanning electron microscopy 4 Scanning electron microscopy was performed as described previously (Souza Onofre et al., 2023). Cells were 5 fixed with 2.5% (v/v) glutaraldehyde, and fixed cells were harvested by cen trifugation at 750 × g for 10 min, 6 washed with PBS for three times, and then settled onto glass coverslips. Subsequently, cells on the coverslip were 7 dehydrated in alcohol solutions (30%, 50%, 70%, 90%, and 100%) each for 5 min, and dried by critical point 8 drying. Finally, cell samples were coated with a 5-nm metal film (Pt:Pd 80:20, Ted Pella Inc.) using a sputter -9 coater (Cressington Sputter Coater 208 HR, Ted Pella Inc.) , and cells were imaged using Nova NanoSEM 230 10 (FEI) with the accelerating high voltage set at 8 kV and the scanning work distance set at 5 mm. 11 12 Data analysis and statistical analysis 13 ImageJ (National Institutes of Health, Bethesda, MD; http://imagej.nih.gov/ij/) was used to measure the FAZ 14 length, the flagellum length, the centrin arm length, the inter-centrin arm distance, the inter-FPC distance, and the 15 inter-basal body di stance, and to measure the fluorescence intensity of TbGRASP. D ata were exported to 16 Microsoft Excel or the GraphPad Prism9 for analysis. Statistical analysis was conducted using the two-tailed 17 student’s t-test or one-way ANOVA (analysis of variance). Error bars represent standard deviation (SD) from the 18 mean of three independent experiments. 19 20

We thank Dr. Cynthia Y. He of National University of Singapore for providing anti-CC2D, 21 anti-TbCentrin4, and anti-TbGRASP antibodies, Dr. Brooke Morriswood of University of Wurzburg for providing 22 anti-TbMORN1 antibody, and Dr. Derrick Robinson of University of Bordeaux for providing anti-TbBILBO1 23 antibody. We are also grateful to Dr. Jianhua Gu of Houston Methodist Research Institute for assistance with 24 scanning electron microscopy and to Dr. Li Li of the Clinical and Translational Service Center at University of 25 Texas Health Science Center at Houston for assistance with mass spectrometry. This work was supported by the 26 NIH R01 grants AI118736 and AI101437 to Z. L and in part by the Clinical and Translational Proteomics Service 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 22 Center at the University of Texas Health Science Center at Houston. The content is solely the responsibility of the 1 authors and does not necessarily represent the official views of the National Institutes of Health. 2 3 Competing interests: The authors declare that they have no conflicts of interest with the contents of this article. 4 5 Data availability: All data are contained within the manuscript. 6 7 Author contributions: Qing Zhou: Conceptualization, Methodology, Visualization, Investigation, Formal 8 analysis, Writing – Reviewing and Editing. Huiqing Hu: Methodology, Visualization, Investigation, Writing – 9 Reviewing and Editing. Ziyin Li: Conceptualization, Supervision, Project administration, Funding acquisition, 10 Writing – Original Draft, Writing – Reviewing and Editing. 11 12

13 Abeywickrema, M., H. Vachova, H. Farr, T. Mohr, R.J. Wheeler, D.H. Lai, S. Vaughan, K. Gull, J.D. Sunter, and 14 V. Varga. 2019. Non-equivalence in old- and new-flagellum daughter cells of a proliferative division in 15 Trypanosoma brucei. Mol Microbiol. 112:1024-1040. 16 Abramson, J., J. Adler, J. Dunger, R. Evans, T. Green, A. Pritzel, O. Ronneberger, L. Willmore, A.J. Ballard, J. 17 Bambrick, S.W. Bodenstein, D.A. Evans, C.C. Hung, M. O'Neill, D. Reiman, K. Tunyasuvunakool, Z. 18 Wu, A. Zemgulyte, E. Arvaniti, C. Beattie, O. Bertolli, A. Bridgland, A. Cherepanov, M. Congreve, A.I. 19 Cowen-Rivers, A. Cowie, M. Figurnov, F.B. Fuchs, H. Gladman, R. Jain, Y.A. Khan, C.M.R. Low, K. 20 Perlin, A. Potapenko, P. Savy, S. Singh, A. Stecula, A. Thillaisundaram, C. Tong, S. Yakneen, E.D. Zhong, 21 M. Zielinski, A. Zidek, V. Bapst, P. Kohli, M. Jaderberg, D. Hassabis, and J.M. Jumper. 2024. Accurate 22 structure prediction of biomolecular interactions with AlphaFold 3. Nature. 630:493-500. 23 An, T., Y. Liu , S. Gourguechon, C.C. Wang, and Z. Li. 2018. CDK Phosphorylation of Translation Initiation 24 Factors Couples Protein Translation with Cell-Cycle Transition. Cell Rep. 25:3204-3214 e3205. 25 Arnold, K., L. Bordoli, J. Kopp, and T. Schwede. 2006. The SWISS -MODEL workspace: a web -based 26 environment for protein structure homology modelling. Bioinformatics. 22:195-201. 27 Biasini, M., S. Bienert, A. Waterhouse, K. Arnold, G. Studer, T. Schmidt, F. Kiefer, T. Gallo Cassarino, M. 28 Bertoni, L. Bordoli, and T. Schwede. 2014. SWISS-MODEL: modelling protein tertiary and quaternary 29 structure using evolutionary information. Nucleic Acids Res. 42:W252-258. 30 Billington, K., C. Halliday, R. Madden, P. Dyer, A.R. Barker, F.F. Moreira-Leite, M. Carrington, S. Vaughan, C. 31 Hertz-Fowler, S. Dean, J.D. Sunter, R.J. Wheeler, and K. Gull. 2023. Genome -wide subcellular protein 32 map for the flagellate parasite Trypanosoma brucei. Nat Microbiol. 8:533-547. 33 Bonhivers, M., S. Nowacki, N. Landrein, and D.R. Robinson. 2008. Biogenesis of the trypanos ome endo -34 exocytotic organelle is cytoskeleton mediated. PLoS Biol. 6:e105. 35 de Graffenried, C.L., H.H. Ho, and G. Warren. 2008. Polo-like kinase is required for Golgi and bilobe biogenesis 36 in Trypanosoma brucei. J Cell Biol. 181:431-438. 37 Esson, H.J., B. Mor riswood, S. Yavuz, K. Vidilaseris, G. Dong, and G. Warren. 2012. Morphology of the 38 trypanosome bilobe, a novel cytoskeletal structure. Eukaryot Cell. 11:761-772. 39 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 23 Gheiratmand, L., A. Brasseur, Q. Zhou, and C.Y. He. 2013. Biochemical characterization of the bi-lobe reveals a 1 continuous structural network linking the bi-lobe to other single-copied organelles in Trypanosoma brucei. 2 J Biol Chem. 288:3489-3499. 3 He, C.Y., H.H. Ho, J. Malsam, C. Chalouni, C.M. West, E. Ullu, D. Toomre, and G. Warren. 2004. Golgi 4 duplication in Trypanosoma brucei. J Cell Biol. 165:313-321. 5 He, C.Y., M. Pypaert, and G. Warren. 2005. Golgi duplication in Trypanosoma brucei requires Centrin2. Science. 6 310:1196-1198. 7 Hu, H., Y. Liu, Q. Zhou, S. Siegel, and Z. Li. 2015a. The Centriole Car twheel Protein SAS-6 in Trypanosoma 8 brucei Is Required for Probasal Body Biogenesis and Flagellum Assembly. Eukaryot Cell. 14:898-907. 9 Hu, H., Q. Zhou, X. Han, and Z. Li. 2017. CRL4 WDR1 Controls Polo-like Kinase Protein Abundance to Promote 10 Bilobe Duplication, Basal Body Segregation and Flagellum Attachment in Trypanosoma brucei. PLoS 11 Pathog. 13:e1006146. 12 Hu, H., Q. Zhou, and Z. Li. 2015b. SAS -4 Protein in Trypanosoma brucei Controls Life Cycle Transitions by 13 Modulating the Length of the Flagellum Attachment Zone Filament. J Biol Chem. 290:30453-30463. 14 Jumper, J., R. Evans, A. Pritzel, T. Green, M. Figurnov, O. Ronneberger, K. Tunyasuvunakool, R. Bates, A. Zidek, 15 A. Potapenko, A. Bridgland, C. Meyer, S.A.A. Kohl, A.J. Ballard, A. Cowie, B. Romera -Paredes, S. 16 Nikolov, R. Jain, J. Adler, T. Back, S. Petersen, D. Reiman, E. Clancy, M. Zielinski, M. Steinegger, M. 17 Pacholska, T. Berghammer, S. Bodenstein, D. Silver, O. Vinyals, A.W. Senior, K. Kavukcuoglu, P. Kohli, 18 and D. Hassabis. 2021. Highly accurate protein s tructure prediction with AlphaFold. Nature. 596:583-19 589. 20 Kohl, L., T. Sherwin, and K. Gull. 1999. Assembly of the paraflagellar rod and the flagellum attachment zone 21 complex during the Trypanosoma brucei cell cycle. J Eukaryot Microbiol. 46:105-109. 22 Kurokawa, K., and A. Nakano. 2019. The ER exit sites are specialized ER zones for the transport of cargo proteins 23 from the ER to the Golgi apparatus. J Biochem. 165:109-114. 24 Lacomble, S., S. Vaughan, C. Gadelha, M.K. Morphew, M.K. Shaw, J.R. McIntosh, and K. Gul l. 2009. Three-25 dimensional cellular architecture of the flagellar pocket and associated cytoskeleton in trypanosomes 26 revealed by electron microscope tomography. J Cell Sci. 122:1081-1090. 27 Morriswood, B. 2015. Form, Fabric, and Function of a Flagellum-Associated Cytoskeletal Structure. Cells. 4:726-28 747. 29 Morriswood, B., C.Y. He, M. Sealey-Cardona, J. Yelinek, M. Pypaert, and G. Warren. 2009. The bilobe structure 30 of Trypanosoma brucei contains a MORN-repeat protein. Mol Biochem Parasitol. 167:95-103. 31 Pham, K.T .M., H. Hu, and Z. Li. 2020. Maintenance of hook complex integrity and centrin arm assembly 32 facilitates flagellum inheritance in Trypanosoma brucei. J Biol Chem. 295:12962-12974. 33 Pham, K.T.M., Q. Zhou, K.J. Lee, and Z. Li. 2022. A Spef1 -interacting microtu bule quartet protein in 34 Trypanosoma brucei promotes flagellar inheritance by regulating basal body segregation. J Biol Chem. 35 298:102125. 36 Shen, S., G.K. Arhin, E. Ullu, and C. Tschudi. 2001. In vivo epitope tagging of Trypanosoma brucei genes using 37 a one step PCR-based strategy. Mol Biochem Parasitol. 113:171-173. 38 Shi, J., J.B. Franklin, J.T. Yelinek, I. Ebersberger, G. Warren, and C.Y. He. 2008. Centrin4 coordinates cell and 39 nuclear division in T. brucei. J Cell Sci. 121:3062-3070. 40 Smith, T.F., C. Gaitatzes , K. Saxena, and E.J. Neer. 1999. The WD repeat: a common architecture for diverse 41 functions. Trends Biochem Sci. 24:181-185. 42 Souza Onofre, T., K.T.M. Pham, Q. Zhou, and Z. Li. 2023. The microtubule quartet protein SNAP1 in 43 Trypanosoma brucei facilitates flagellum and cell division plane positioning by promoting basal body 44 segregation. J Biol Chem. 299:105340. 45 Sumimoto, H., S. Ka makura, and T. Ito. 2007. Structure and function of the PB1 domain, a protein interaction 46 module conserved in animals, fungi, amoebas, and plants. Sci STKE. 2007:re6. 47 Sunter, J.D., and K. Gull. 2016. The Flagellum Attachment Zone: 'The Cellular Ruler' of T rypanosome 48 Morphology. Trends Parasitol. 32:309-324. 49 Varadi, M., S. Anyango, M. Deshpande, S. Nair, C. Natassia, G. Yordanova, D. Yuan, O. Stroe, G. Wood, A. 50 Laydon, A. Zidek, T. Green, K. Tunyasuvunakool, S. Petersen, J. Jumper, E. Clancy, R. Green, A. Vora, 51 M. Lutfi, M. Figurnov, A. Cowie, N. Hobbs, P. Kohli, G. Kleywegt, E. Birney, D. Hassabis, and S. 52 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 24 Velankar. 2022. AlphaFold Protein Structure Database: massively expanding the structural coverage of 1 protein-sequence space with high-accuracy models. Nucleic Acids Res. 50:D439-D444. 2 Vaughan, S., and K. Gull. 2016. Basal body structure and cell cycle-dependent biogenesis in Trypanosoma brucei. 3 Cilia. 5:5. 4 Vaughan, S., L. Kohl, I. Ngai, R.J. Wheeler, and K. Gull. 2008. A repetitive protein essential for the flagellum 5 attachment zone filament structure and function in Trypanosoma brucei. Protist. 159:127-136. 6 Wang, Z., J.C. Morris, M.E. Drew, and P.T. Englund. 2000. Inhibition of Trypanosoma brucei gene expression 7 by RNA interference using an integratable vector with opposing T7 promoters. J Biol Chem. 275:40174-8 40179. 9 Wei, Y., H. Hu, Z.R. Lun, and Z. Li. 2014. Centrin3 in trypanosomes maintains the stability of a flagellar inner -10 arm dynein for cell motility. Nat Commun. 5:4060. 11 Wirtz, E., S. Leal, C. Ochatt, and G.A. Cross. 1999. A tightly regulated inducible expression system for conditional 12 gene knock-outs and dominant-negative genetics in Trypanosoma brucei. Mol Biochem Parasitol. 99:89-13 101. 14 Zhou, Q., L. Gheiratmand, Y. Chen, T.K. Lim, J. Zhang, S. Li, N. Xi a, B. Liu, Q. Lin, and C.Y. He. 2010. A 15 comparative proteomic analysis reveals a new bi -lobe protein required for bi -lobe duplication and cell 16 division in Trypanosoma brucei. PLoS ONE. 5:e9660. 17 Zhou, Q., Y. Kurasawa, H. Hu, T. Souza Onofre, and Z. Li. 2024 . An orphan kinesin in Trypanosoma brucei 18 regulates hook complex assembly and Golgi biogenesis. mBio:e0263424. 19 Zhou, Q., B. Liu, Y. Sun, and C.Y. He. 2011. A coiled -coil- and C2-domain-containing protein is required for 20 FAZ assembly and cell morphology in Trypanosoma brucei. J Cell Sci. 124:3848-3858. 21 22 Figure Legends: 23 Figure 1. Identification of WDR2 as a n interacting protein of KIN -G. (A). Schematic drawing of a 24 trypanosome cell showing the flagellum and its associated cytoskeletal structures. mBB: mature basal body; pBB: 25 pro-basal body; FPC: flagellar pocket collar; MtQ: microtubule quartet; FAZ: flagellum attachment zone; HC: 26 hook complex; CA: centrin arm. (B). KIN-G- and WDR2-proximal proteins identified by BioID. Bait proteins are 27 highlighted in red or green, and common proximal proteins are highlighted in blue. (C). List of KIN -G- and 28 WDR2-proximal proteins and their subcellular local izations. Bait proteins are in red or green, and KIN-G- and 29 WDR2-shared proximal proteins are in blue. CA: centrin arm; BB: basal body; HC: hook complex; MtQ: 30 microtubule quartet. (D). Co-immunoprecipitation to test the in vivo interaction between KIN -G and WDR2 . 31 Endogenous KIN-G-3HA and WDR2-PTP were detected by anti-HA and anti-Protein A antibodies, respectively. 32 (E). Subcellular localization of WDR2 , endogenously tagged with PTP or 3HA, and its co -localization with 33 endogenously 3HA-tagged KIN-G, TbCentrin4, and TbMORN1 . Insets show the zoom -in view of the selected 34 region. Scale bar: 5 μm. 35 36 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 25 Figure 2. Knockdown of WDR2 causes asymmetrical cytokinesis and inhibits FAZ elongation . (A). 1 Knockdown of WDR2 by RNAi. Endogenous WDR2-3HA was detected by the anti-HA antibody. TbPSA6 served 2 as a loading control. (B). WDR2 knockdown caused growth defects in procyclic trypanosomes. (C). Effect of 3 WDR2 knockdown on cell cycle progression. Shown is the quantitation of cells with different numbers of nuclei 4 (N) and kinetoplasts (K) before and after tetracycline (Tet) -induced WDR2 RNAi induction. Error bars indicate 5 S.D. from three independent experiments. (D). Effect of WDR2 knockdown on cleavage furrow ingression. Shown 6 is the percentage of bi -nucleated cells with or without a visible cleavage furrow. Error bars indicate S.D. from 7 three independent experiments. ***, p<0.001. (E). Control and WDR2 knockdown cells undergoing cytokinesis. 8 Scale bar: 5 μm. (F). Scanning electron microscopic analysis of control and WDR2 RNAi cells undergoing 9 cytokinesis. Scale bar: 5 μm. (G). WDR2 knockdown disrupted FAZ elongation and flagellum positioning. Cells 10 were co-immunostained with 20H5, which labels the flagellum, and anti-CC2D antibody, which labels the FAZ 11 filament. NF: new flagellum; OF: old flagellum; nFAZ: new FAZ; oFAZ: old FAZ. Scale bar: 5 μm. ( H). 12 Measurement of the length of the new FAZ in bi-nucleated cells before and after WDR2 RNAi. ***, p<0.001. (I). 13 Measurement of the length of the free new flagellum in bi -nucleated cells before and after WDR2 RNAi. ***, 14 p<0.001. 15 16 Figure 3. Knockdown of WDR2 disrupts flagellum positioning. (A). Scanning electron microscopic analysis 17 of the dividing control and WDR2 RNAi cells, showing the effect of WDR2 knockdown on flagellum positioning. 18 Solid arrowheads and open arrowheads indicate the proximal bases of the new and the old flagella, respectively. 19 Scale bar: 5 μm. (B). Immunofluorescence microscopy to detect the basal body in control and WDR2 RNAi cells 20 with the anti-TbSAS-6 antibody. Solid arrowheads and open arrowheads indicate the new and the old basal bodies, 21 respectively. Scale bar: 5 μm. (C). Measurement of the inter-basal body distance in bi-nucleated cells before and 22 after WDR2 RNAi. ***, p<0.001. (D). Immunofluorescence microscopy to detect the centrin arm in control and 23 WDR2 RNAi cells with the pan-centrin antibody 20H5. Solid arrowheads and open arrowheads indicate the new 24 and the old centrin arms, respectively. Scale bar: 5 μm. (E). Measurement of the inter-centrin arm distance in bi-25 nucleated cells before and after WDR2 RNAi. ***, p<0.001. (F). Immunofluorescence microscopy to detect the 26 PFC in control and WDR2 RNAi cells with the anti-TbBILBO1 antibody. Solid arrowheads and open arrowheads 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 26 indicate the new and the old FPCs, respectively. Scale bar: 5 μm. (G). Measurement of the inter-PFC distance in 1 bi-nucleated cells before and after WDR2 RNAi (48 h). ***, p<0.001. 2 3 Figure 4. Knockdown of WDR2 impairs hook complex integrity. (A). Immunofluorescence microscopy to 4 examine the centrin arm with the anti-TbCentrin4 antibody and the pan-centrin antibody 20H5. Solid arrowheads 5 indicate the new centrin arm (nCA), and open arrowheads indicate the old centrin arm (oCA). Scale bar: 5 μm. 6 (B). Measurement of the length of the new centrin arm and the old centrin arm in control and WDR2 RNAi cells. 7 (C). Quantitation of bi-nucleated cells with normal-length centrin arm or short-length centrin arm in control and 8 WDR2 RNAi cells . Error bars indicate S.D. from three independent experiments. ***, p<0.001. (D). 9 Immunostaining of the hook complex with anti-TbMORN1 antibody. Solid arrowheads indicate the new hook 10 complex (nHC), and open arrowheads indicate the old hook complex (oHC). Depicted on the right is the 11 morphology of the new and the old hook complexes marked by anti -TbMORN1 antibody. Scale bar: 5 μm. (E). 12 Quantitation of bi-nucleated cells with the hook complex displaying a typical fishhook-like morphology or a non-13 fishhook-like morphology in control and WDR2 RNAi cells. Error bars indicate S.D. from three independent 14 experiments. ***, p<0.001. 15 16 Figure 5. WDR2 is required for Golgi biogenesis . (A). Immunostaining of TbGRASP and CAAP1 in control 17 and WDR2 RNAi cells. TbGRASP was detected by the anti-TbGRASP antibody, and endogenously 3HA-tagged 18 CAAP1 was detected by the anti-HA antibody. Arrows indicate small Golgi. Scale bar: 5 μm. (B). Quantitation 19 of the numbers of Golgi in control and WDR2 RNAi -induced bi-nucleated cells. Error bars indicate S.D. from 20 three independent experiments. ***, p<0.001. (C). Fluorescence microscopic analysis of Sec13 -mCherry in 21 control and WDR2 RNAi cells. TbGRASP served as a marker for the Golgi. Scale bar: 5 μm. (D). Effect of WDR2 22 RNAi on Golgi-centrin arm association. Centrin arm was immunostained by the 20H5 antibody, and Golgi was 23 detected by the anti-TbGRASP antibody. Solid arrowheads indicate the new centrin arm-associated Golgi, and 24 open arrowheads indicate the old centrin arm -associated Golgi. The white line indicates the transect for 25 quantitative measurement of TbGRASP signal shown in panel E. Insets in the merged panel show the zoom-in 26 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 27 view of the selected region. Scale bar: 5 μm. (E). Quantitation of TbGRASP signal intensity in the centrin arm-1 associated new and old Golgi in control and WDR2 RNAi cells presented in panel D. 2 3 Figure 6. WDR2 is required for recruitment and maintenance of KIN-G at the centrin arm . (A). 4 Immunofluorescence microscopic analysis of KIN-G-3HA in control, WDR2 RNAi cells, and WDR2 RNAi cells 5 treated with MG132. nCA: new centrin arm; oCA: old centrin arm. Scale bar: 5 μm. (B). Quantitation of 1N2K 6 and 2N2K cells with different localization patterns of KIN -G at the new and the old centrin arms in control and 7 WDR2 RNAi cells. Error bars indicate S.D. from three independent experiments. nCA+ /oCA+: KIN-G-positive 8 at both the new and the old centrin arms; nCA-/oCA+: KIN-G-negative at the new centrin arm but positive at the 9 old centrin arm; nCA -/oCA-: KIN -G-negative at both the new and the old centrin arms. ***, p<0.001. (C). 10 Western blotting to detect the level of endogenously 3HA-tagged KIN-G before and after WDR2 RNAi induction 11 and in WDR2 RNAi -induced cells (2 d) treated with MG132 (8 h). TbPSA6 served as a loading control. (D). 12 Immunofluorescence microscopic analysis of WDR2 -3HA in control and KIN -G RNAi cells. Solid arrowhead 13 and open arrowhead indicate the WDR2 fluorescence signal at the new and the old centrin arms, respectively. 14 Scale bar: 5 μm. (E). Western blotting to detect the level of endogenously 3HA-tagged WDR2 before and after 15 KIN-G RNAi induction. TbPSA6 served as a loading control. 16 17 Figure 7. Structural domains of WDR2 required for interaction with KIN-G. (A). Schematic drawing of the 18 structural domains in WDR2. DUF: domain of unknown function; CC: coiled coil; PB1: Phox and Bem1; WD: 19 Tryptophan (W)-Aspartate D); NTD: N-terminal domain; MD: middle domain; CTD: C -terminal domain. (B). 20 Structure of WDR2 predicted by AlphaFold 2. The structura l domains were indicated. (C). Structural domains 21 predicted by SWISS-MODEL or AlphaFold2. The templates used for structural modeling of the PB1 domain and 22 the WD40 domain in WDR2 are 2PPH and 8OKF, respectively. (D). Schematic drawing of the structural domains 23 in KIN-G. KMD: kinesin motor domain; CC: coiled coil. (E). Structure of the WDR2-KIN-G complex predicted 24 by AlphaFold3. The structural domains in each protein and the Loop #1 (L1) and the -helices #4 ( 4) of the 25 KMD domain of KIN -G were indicated. (F). Co-immunoprecipitation to detect the in vivo interaction between 26 native KIN-G and wild -type and the domain-deletion mutants of WDR2. Protein bands were stained by silver 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 28 staining and identified by mass spectrometry. Solid arrows indicate the ectopically expressed, 3HA-tagged WDR2 1 and its domain-deletion mutants. p-KIN-G: phosphorylated KIN-G; PPase: Lambda protein phosphatase. 2 3 Figure 8. Structural domains of WDR2 required for WDR2 cellular function. (A). Schematic drawing of 4 WDR2 and its domain -deletion mutants used for genetic complementation . (B). Western blotting to detect 5 endogenously PTP-tagged WDR2 in WDR2 -3’UTR RNAi cell line induced for 3 days. TbPSA6 served as a 6 loading control. (C). Growth curve of the WDR2-3’UTR RNAi cell line. (D). Western blotting to detect 7 ectopically expressed, 3HA-tagged wild-type and mutant WDR2 in the WDR2-3’UTR RNAi cell line induced for 8 3 days. TbPSA6 served as a loading control. ( E). Growth curves of WDR2-3’UTR RNAi complementation cell 9 lines expressing wild-type or the domain-deletion mutants of WDR2. (F). Subcellular localization of ectopically 10 expressed, 3HA-tagged WDR2, WDR2-NTD, WDR2-MD, and WDR2-CTD. Cells were co-immunostained 11 with the anti-HA antibody and the anti-TbCentrin4 antibody. BB: basal body; CA: centrin arm. Solid arrowheads 12 indicate the WDR2 signal at the MtQ proximal end, and open arrowheads indicate the WDR2 signal at the centrin 13 arm. Insets show the zoom -in view of the selected region. Scale ba r: 5 μm. (G). Co -immunostaining of 14 endogenously expressed, PTP-tagged KIN-G and ectopically expressed, 3HA-tagged wild-type and the domain-15 deletion mutants of WDR2. Solid arrowheads indicate the WDR2 signal at the MtQ proximal end, and open 16 arrowheads indicate the WDR2 signal at the centrin arm. Insets show the zoom -in view of the co -localization. 17 Scale bar: 5 μm. 18 19 Supplemental Table S1. List of proximal proteins of KIN-G and WDR2 identified by BioID. 20 21 Supplemental Table S2. Mass spectrometry analysis of the phosphorylated KIN -G band pulled down by 22 wild-type WDR2. 23 24 Supplemental Table S3. Mass spectrometry analysis of the non-phosphorylated KIN-G band pulled down 25 by the WDR2-MD mutant. 26 27 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 29 Supplemental Figure S1. Mass spectrometry identification of native KIN-G protein co-immunoprecipitated 1 by 3HA-tagged wild-type WDR2 and the WDR2-MD mutant. 2 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint Figure 1 B C E KIN-G-3HA WDR2-PTP KIN-G-3HA +WDR2-PTP KIN-G-3HA WDR2-PTP KIN-G-3HA +WDR2-PTP Input IP: IgG beads KIN-G-3HA WDR2-PTP IB: anti-HA IB: anti-ProtA D WDR2-3HA TbCentrin4DIC DAPI WDR2-3HA TbMORN1DIC DAPI Merge Merge WDR2 BioID (Log2 Mascot score) KIN-G BioID (Log2 Mascot score) WDR2-PTP KIN-G-3HADIC DAPI Merge Name Acc # Localization KIN-G WDR2 WDR1 TbCentrin2 TbCentrin4 CAAP1 TbSmee1 TbLRRP1 Hypo.1 Hypo.2 Hypo.3 Hypo.4 Hypo.5 Hypo.6 Hypo.7 Tb927.6.1770 Tb927.7.4700 Tb927.10.15080 Tb927.8.1080 Tb927.7.3410 Tb927.10.1450 Tb927.10.8820 Tb927.11.8950 Tb927.10.3010 Tb927.8.7060 Tb927.10.850 Tb927.8.7800 Tb927.7.5610 Tb927.5.2460 Tb927.7.7000 Tb927.4.4180 Hypo.8 Hypo.9 Hypo.10 Hypo.11 NHL1 Tb927.11.14650 Tb927.8.8160 Tb927.10.250 Tb927.9.6160 CA CA CA, BB CA, BB CA, BB Golgi Hook shank HC HC HC HC HC HC HC HC MtQ MtQ MtQ MtQ MtQ KIN-G WDR2 Hypo.1 CAAP1 Hypo.2 Hypo.3 TbCentrin2 TbLRRP1 TbCentrin4 TbSmee1 Hypo.4 Hypo.5 WDR1 Hypo.6 Hypo.7 Hypo.8 Hypo.9 Hypo.10 Hypo.11NHL1 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16 A mBB pBB MtQ FPC CAFAZ Golgi Flagellum HC .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint 0 1 2 3 (days) WDR2 RNAi 75- 30- TbPSA6 kDa WDR2-3HA A 5 6 7 8 0 1 2 3 4 5 B Log (cells/ml) Time (days) WDR2 RNAi -Tet +Tet C Control WDR2 RNAi (48 h) Symmetrical cytokinesis Asymmetrical cytokinesis F Normal daughter cell Normal daughter cell Small daughter cell Normal daughter cell % Cells 0 20 40 60 80 100 No furrow Visible furrow Control (2N2K) WDR2 RNAi (2N2K) WDR2 RNAi (2N1K) % Cells ****** D WDR2 RNAi 0 20 40 60 80 100 0 h 24 h 48 h 72 h G H ControlWDR2 RNAi (48 h) DIC 20H5DAPI CC2D 2N2K 2N2K 2N1K 2N2K 2N2K 2N1K Control WDR2 RNAi Length of new FAZ (µm)0 5 10 15 Length of free new flagellum (µm)0 5 10 15 20 25 2N2K 2N2K 2N1K Control WDR2 RNAi I Merge OF OF OF NF NF NF oFAZ nFAZ oFAZ nFAZ oFAZ nFAZ ****** ****** E Control WDR2 RNAi (48 h) 2N2K furrowDIC/DAPI 2N2K furrow Figure 2 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint G WDR2 RNAi (48 h) Control DIC DAPIF 2N2K 2N2K 2N1K Inter-FPC distance (µm) 0 2 4 6 10 8 2N2K 2N2K 2N1K Control WDR2 RNAi TbBILBO1 Figure 3 Inter-centrin arm distance (µm)0 2 4 6 8 2N2K 2N2K 2N1K Control WDR2 RNAi E 10 WDR2 RNAi (48 h) Control DIC 20H5DAPID 2N2K 2N2K 2N1K C WDR2 RNAi (48 h) Control TbSAS-6DIC DAPIB 2N2K 2N2K 2N1K Inter-basal body distance (µm)0 2 4 6 10 8 2N2K 2N2K 2N1K Control WDR2 RNAi ****** ****** ****** WDR2 RNAi (48 h)ControlA .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint DIC DAPI 20H5 TbCentrin4 Merge ControlWDR2 RNAi (48 h) A 0 0.4 0.8 1.2 1.6 2 0 0.4 0.8 1.2 1.6 2 Control WDR2 RNAi BNew CA length (µm) Old CA length (µm) Figure 4 % Cells (2N2K) C 0 20 40 60 80 100 Normal new CA Short new CA Control WDR2 RNAi D E nCA oCA 0 20 40 60 80 100 Fishhook Non-fishhook Control (2N2K) WDR2 RNAi (2N2K) WDR2 RNAi (2N1K) % Cells New HC morphology ControlWDR2 RNAi (48 h) DIC DAPI TbMORN1 nHC oHC nHC oHC oHC nHCnHC oHCnCA oCA *** *** *** .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint DIC DAPI CAAP1-3HATbGRASP Merge ControlWDR2 RNAi (48 h) Figure 5 % Cells Golgi number A B 0 10 20 30 40 50 60 2 3 4 >4 Control (2N2K) WDR2 RNAi (2N2K) WDR2 RNAi (2N1K) DIC DAPI Sec13-mCherryTbGRASP Merge ControlWDR2 RNAi(48 h) C 2N2K 2N2K 2N2K 2N1K 2N2K 2N2K 2N1K D DIC DAPI 20H5TbGRASP Merge ControlWDR2 RNAi (48 h) E 0 20 40 60 80 100 2N2K 2N2K 0 20 40 60 80 100 New Golgi Old Golgi Distance Distance Control WDR2 RNAi TbGRASP intensity (a.u.) *** *** ****** ****** ****** .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint A DIC KIN-G-3HADAPI E KIN-G RNAi 75- 30- TbPSA6 kDa WDR2-3HA 0 1 2 3 4 5 (days) WDR2 RNAi (24 h) Control TbCentrin4 WDR2 RNAi (48 h) +MG132 -MG132 -MG132 strong strong strong strong weak strong weak strong 1N2K 2N2K 1N2K 2N2K 1N2K 2N2K 2N2K C KIN-G-3HA TbPSA6 WDR2 RNAi kDa 75- 30- d0 d1 d2 d3 d4 d5 d2+MG132 (8 h) PTP-WDR2100- 0 20 40 60 80 100 nCA+/oCA+ nCA-/oCA+ nCA-/oCA- 0 h 24 h 48 h 0 20 40 60 80 100 nCA+/oCA+ nCA-/oCA+ nCA-/oCA- 0 h 24 h 48 h B WDR2 RNAi % Cells (1N2K)% Cells (2N2K) KIN-G localization at the centrin arm ****** D DIC WDR2-3HADAPI Control KIN-G RNAi 2N2K 2N2K Figure 6 KIN-G localization at the centrin arm nCA oCA ****** .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint A DUF1 DUF2 PB1 WD40 B C Model confidence Very high (pLDDT>90) High (90>pLDDT>70) Low (70>pLDDT>50) Very low (pLDDT<50) 0 100 200 300 0 200 400 400 600 Scored residue Aligned residue 0 10 3020 Expected position error (Å) Predicted aligned error 500 600 DUF1 DUF2 PB1 WD40 CC D E KIN-G 12 366 628 a.a. CC1 CC2 CC3 KMD F p-KIN-G Figure 7 DUF2 DUF1 WD40 PB1 KMD CC CC3 CC2 CC1KIN-G WDR2 L1 4 75- 50- kDa WDR2 DUF1 DUF2 PB1 WD1 WD2 WD3 WD4 WD5 WD6 WD7CC 4 85 100 181 280 360 402 697 a.a.WD40 domain C-terminal domain (CTD)Middle domain (MD)N-terminal domain (NTD) 215 231 - PPase WDR2-3HA WDR2-NTD-3HA WDR2-MD-3HA WDR2-CTD-3HA KIN-G + PPase 75- 50- KIN-G IP: -HA .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint A 4 85 100 181 280 360 401 a.a. WDR2-CTD 182 280 360 402 697 a.a. WDR2-NTD 4 85 100 181 402 697 a.a. WDR2-MD WD40 domain WDR2 4 85 100 181 280 360 402 697 a.a.WD40 domain C-terminal domain (CTD)Middle domain (MD)N-terminal domain (NTD) FC WDR2-CTDWDR2-NTDWDR2 DICDAPI-HATbCen4Merge 5 6 7 8 9 0 1 2 3 4 5 RNAi: WDR2-3’UTR Log (cells/ml) Time (days) 5 6 7 8 9 0 1 2 3 4 5 Time (days) RNAi: WDR2-3’UTR OE: WDR2 RNAi: WDR2-3’UTR OE: WDR2-NTD RNAi: WDR2-3’UTR OE: WDR2-MD 5 6 7 8 9 0 1 2 3 4 5 Log (cells/ml) Time (days) 5 6 7 8 9 0 1 2 3 4 5 Time (days) -Tet +Tet -Tet +Tet -Tet +Tet -Tet +Tet ectopically expressed, 3HA-tagged WDR2-MD G RNAi: WDR2-3’UTR OE: WDR2-CTD Log (cells/ml) PTP- WDR2 TbPSA630- 100- WDR2-NTD-3HADIC DAPI KIN-G-PTP DIC DAPI KIN-G-PTP -Tet WDR2-MD-3HADIC DAPI KIN-G-PTP WDR2-CTD-3HADIC DAPI KIN-G-PTP 20H5 +Tet+Tet+Tet Time (days) WDR2-3HADIC DAPI KIN-G-PTP +Tet BB CA CA BB CA BB CA BB Figure 8 BB CA 5 6 7 8 9 0 1 2 3 4 5 -Tet +Tet D WDR2-CTD-3HA WDR2-NTD-3HA WDR2-3HA WDR2-MD-3HA 0 1 2 3 (days) +Tet 75- 50- 50- 50- kDa 0 1 2 3 (days) +Tet kDa B E 30- TbPSA6 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 29, 2025. ; https://doi.org/10.1101/2025.01.29.635568doi: bioRxiv preprint