Heparan sulfate organizes transmembrane semaphorins at the cell surface to tune plexin signaling

Heparan sulfate organizes transmembrane semaphorins at the cell surface to tune plexin signaling | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Heparan sulfate organizes transmembrane semaphorins at the cell surface to tune plexin signaling Margarita Sobol, Zuzana Cockova, Farahdokht Nourisanami, Michaela Blazikova, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9405124/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Heparan sulfate proteoglycans are cell-surface and extracellular-matrix components that modulate the activity of diverse signaling pathways. A hallmark of heparan sulfate is its domain organization into less sulfated N–acetyl–rich and highly sulfated N–sulfo–rich regions. In the nervous system, semaphorins and their plexin receptors comprise a major axon guidance system, and a growing body of evidence has implicated proteoglycans as key modulators of their function. However, how heparan sulfate domain organization mechanistically tunes semaphorin–plexin signaling at the cell surface remains unclear. Here, we show that heparan sulfate directly organizes transmembrane semaphorin–plexin signaling at the plasma membrane. The Drosophila transmembrane semaphorins Sema1a and Sema1b, together with their receptor PlexA, bind heparan sulfate and preferentially associate with less sulfated N-acetyl–rich domains, in contrast to previously reported secreted semaphorins that favor highly sulfated N-sulfo–rich clusters. The N–acetyl–rich-domain engagement promotes nanoscale clustering of Sema1a and reduces its lateral mobility, consistent with increased local surface density and longer dwell time. These membrane-organizational effects enhance plexin-dependent signaling output, as demonstrated in a collapse assay, in which cell-surface glycosaminoglycans potentiated ligand-induced cell collapse. Together, our findings reveal heparan sulfate as an active organizer of the local plasma membrane microenvironment for transmembrane semaphorins, defining a distinct mode of glycosaminoglycan-dependent regulation that controls protein distribution, encounter probability, and signaling thresholds. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 INTRODUCTION The assembly of neural circuits relies on the precise navigation of growing axons through complex and dynamic environments. This process is controlled by axon guidance cues—secreted or membrane-associated proteins that act through receptors and co-receptors on growth cones to regulate cytoskeletal dynamics and steer navigation. Major guidance cue families include semaphorins, netrins, slits, and ephrins, which together guide axons to their targets (reviewed in 1 ). Proteoglycans have long been recognized as key modulators of axon guidance, yet the molecular mechanisms by which they shape guidance decisions remain poorly understood. Proteoglycans consist of a core protein decorated with one or more covalently attached glycosaminoglycan (GAG) chains—linear polysaccharides composed of repeating disaccharide units. In the nervous system, the predominant GAGs are heparan sulfate and chondroitin sulfate. These polymers are chemically distinct: heparan sulfate is built from repeating uronate–glucosamine units (GlcA/IdoA–GlcNAc/GlcNS), whereas chondroitin sulfate consists of glucuronate–galactosamine repeats (GlcA–GalNAc). They also differ in their modification patterns. Chondroitin sulfate is most often O-sulfated at defined positions in a relatively uniform manner along the chain. In contrast, heparan sulfate is extensively and non-uniformly modified, creating distinct domains along a single chain. As a result, a typical heparan sulfate chain contains regions rich in GlcA and GlcNAc (N-acetyl-rich, NA domains) and regions enriched in IdoA and GlcNS derivatives (N-sulfo-rich, NS domains) 2 , 3 . This molecular heterogeneity dramatically expands the binding repertoire of heparan sulfate, enabling interactions with a wide range of extracellular ligands. Beyond their structural roles in the extracellular matrix, proteoglycans can therefore act as dynamic regulators of intercellular signaling by shaping ligand distribution, stabilizing signaling complexes, and functioning as co-receptors at the cell surface, including in axon guidance pathways (reviewed in 4,5 ). Among guidance cue families, semaphorins exhibit particularly rich and diverse interactions with proteoglycans. Semaphorins comprise secreted, transmembrane, and GPI-anchored proteins and signal primarily through plexin receptors 6 . Multiple studies have demonstrated that proteoglycans modulate semaphorin-dependent guidance outcomes. In vertebrates, the transmembrane Sema5A exemplifies GAG-dependent functional switching, heparan sulfate promotes semaphorin-mediated attraction while chondroitin sulfate induces repulsion during formation of the fasciculus retroflexus 7 . In Drosophila, the heparan sulfate proteoglycan perlecan facilitates motor axon repulsion mediated by the transmembrane Sema1a 8 . Secreted semaphorins are likewise influenced by proteoglycans. Sema3A associates with proteoglycans at the neuronal surface, and exogenous heparin enhances Sema3A-induced growth cone collapse 9 . Subsequent work implicated chondroitin sulfate in Sema3A-dependent migration of cortical interneurons 10 and demonstrated that Sema3A is retained within chondroitin sulfate–rich perineuronal nets in a sulfation-dependent manner 11 . Together, these studies establish proteoglycans as critical modulators of semaphorin signaling but the underlying molecular mechanisms remain poorly understood. We recently provided mechanistic insight into proteoglycan interactions of secreted semaphorins in Drosophila. We showed that the secreted semaphorins Sema2a and Sema2b bind GAG chains in a sulfation-dependent manner via a positively charged C-terminal tail, a motif also conserved in vertebrate class 3 semaphorins. Disrupting this interaction compromises Sema2b function in vivo . These findings led us to propose a “semaphorin bridge” model in which GAG binding concentrates secreted semaphorins at defined anatomical locations, enabling short-range, surface-tethered guidance 12 . While this work established how GAGs spatially organize secreted semaphorins at the cell surface, it did not address whether similar principles apply to membrane-associated semaphorins or to their plexin receptors. In particular, it is unclear whether GAGs directly engage these proteins, whether such interactions are selective for specific heparan sulfate regions, and how they influence nanoscale membrane organization, lateral mobility, and signaling output. These questions are especially relevant for transmembrane semaphorins, which are already membrane tethered and therefore face distinct organizational constraints compared with secreted cues. Here, we address this gap by focusing on the Drosophila class 1 transmembrane semaphorins Sema1a and Sema1b and their shared receptor PlexA. We show that Sema1a, Sema1b, and PlexA bind heparan sulfate chains and preferentially associate with NA domains at the cell surface, in contrast to secreted semaphorins, which favor highly sulfated NS domains. We further demonstrate that heparan sulfate modulates the nanoscale clustering and lateral mobility of these proteins and potentiates plexin-mediated signaling. Together, our findings reveal heparan sulfate as an active organizer of the cis membrane landscape for transmembrane semaphorin–plexin signaling, uncovering a distinct mode of GAG-dependent regulation. RESULTS Sema1a, Sema1b, and their receptor PlexA bind GAG chains To investigate the GAG-binding properties of transmembrane semaphorins, we performed surface plasmon resonance (SPR) binding experiments with purified ectodomains of Drosophila Sema1a and Sema1b, expressed in mammalian cells. Consistent with our previous findings 13 , the Sema1a ectodomain was obtained as a mixture of monomer and disulfide-linked dimer, whereas Sema1b was exclusively monomeric. To ensure monovalent binding and enable direct comparison, we used only the monomeric forms of both proteins in SPR assays. We first compared the GAG-binding properties of Sema1a and Sema1b. For SPR binding experiments, heparin, heparan sulfate, and chondroitin sulfate were immobilized on the sensor chip, and Sema1a and Sema1b were injected at increasing concentrations. Both proteins bound heparin and heparan sulfate with apparent dissociation constants in the micromolar range, but showed no detectable binding to chondroitin sulfate (Fig. 1 a, b), even at the highest concentrations tested (up to 16 µM for Sema1a and 116 µM for Sema1b). Quantitative analysis revealed pronounced differences in GAG-binding affinities between Sema1a and Sema1b: Sema1a bound heparin and heparan sulfate with apparent K D values of 1.5 µM and 1.1 µM, respectively, whereas Sema1b displayed substantially weaker binding, with K D values of 18.4 µM for heparin and 46.5 µM for heparan sulfate. Compared to these transmembrane semaphorins, the previously reported secreted semaphorin Sema2a binds heparin and heparan sulfate with higher affinity (apparent K D = 0.4 µM and 2.3 µM, respectively), while also exhibiting measurable binding to chondroitin sulfate (apparent K D = 19.3 µM) 12 . The stronger GAG binding observed for Sema2a is likely influenced, at least in part, by avidity effects, as Sema2a was analyzed in its dimeric form. We then asked whether PlexA, the shared receptor for Sema1a and Sema1b, directly interacts with GAG chains. The purified PlexA ectodomain bound heparin with an apparent K D of 3.7 µM and heparan sulfate with markedly weaker affinity (apparent K D > 43.7 µM), while no binding to chondroitin sulfate was detected (Fig. 1 c). Thus, PlexA binds GAG chains with affinities intermediate between Sema1a and Sema1b, but is particularly weak for heparan sulfate. Finally, to assess whether Sema1a can bind native heparan sulfate chains of proteoglycans, we immobilized a secreted, GPI-anchor–deleted variant of the Drosophila glypican-family proteoglycan Dally-like protein (Dlp ΔGPI ) on an SPR chip. Monomeric Sema1a bound Dlp ΔGPI with an apparent affinity of 6.5 µM, while no binding was observed to the Dlp core protein lacking heparan sulfate chains (Fig. 1 d). Together, these results indicate that the ectodomains of transmembrane semaphorins and their PlexA receptor selectively recognize heparin and heparan sulfate, but not chondroitin sulfate, and that this interaction extends to native GAG chains of proteoglycans, such as glypicans. Although reliable extraction of individual association (k on ) and dissociation (k off ) rate constants was not feasible under the experimental conditions used, qualitative inspection of the SPR sensorgrams ( Fig. S1 ) nevertheless provides insight into the overall binding behavior of semaphorin–GAG interactions. In all cases, the response increased rapidly upon analyte injection, consistent with relatively fast association. Following the injection phase, dissociation occurred slowly, indicating limited complex decay over the time window of the experiment. For PlexA–GAG interactions, the dissociation phase was particularly shallow with minimal signal loss during dissociation. While these observations do not permit quantitative kinetic analysis, they are consistent with a binding mode characterized by rapid complex formation and slow dissociation, resulting in comparatively long-lived interactions despite micromolar apparent affinities. Sema1a engages GAG chains in both cis and trans SPR experiments indicate that the ectodomains of Sema1a, Sema1b, and PlexA selectively recognize heparin and heparan sulfate, but not chondroitin sulfate. Because heparin is an experimental surrogate and not a native cell-surface proteoglycan, we next focused on heparan sulfate proteoglycans as the physiologically relevant partners. We therefore next asked whether Sema1a, Sema1b, and PlexA engage heparan sulfate chains at the cell surface, and whether such interactions occur in cis —on the same cell—or in trans between neighboring cells. To examine potential cis interactions, we performed colocalization analyses between fluorescently labelled Sema1a, Sema1b, or PlexA and endogenous GAG chains on the surface of CHO cells. For visualization, the cytoplasmic domains of Sema1a, Sema1b, and PlexA were replaced with the fluorescent proteins mClover or mVenus. CHO-K1 cells were chosen because they constitutively express high levels of heparan sulfate proteoglycans, whereas GAG-deficient CHO-PgsA745 cells—harboring a loss-of-function mutation in xylosyltransferase—were used as a negative control. To achieve nanometer-scale resolution, we employed STED super-resolution microscopy using an anti-NS antibody, which specifically detects sulfated-rich NS clusters in heparan sulfate 14 . STED super-resolution microscopy revealed that semaphorins and their PlexA receptor appeared as numerous small and well-defined puncta across the cell surface, consistent with a high density and relatively uniform distribution (Fig. 2 a). In contrast, NS staining produced fewer but noticeably larger puncta. The larger apparent size of NS puncta may reflect the extended and flexible nature of heparan sulfate chains, which present multiple epitopes and can accommodate numerous antibody molecules. Colocalization was quantified by calculating both Pearson’s correlation coefficient (PCC) and Manders’ colocalization coefficient (M) using pixel-by-pixel fluorescence intensity values. The PCC revealed a relatively weak linear correlation between NS clusters and the semaphorins or PlexA (Fig. 2 b-d). However, Manders’ coefficients indicated substantial spatial overlap, suggesting that while the distribution of NS clusters does not strictly predict semaphorin or PlexA localization, the semaphorins are frequently found in regions enriched in NS clusters (Fig. 2 b-d). Moreover, the Manders’ coefficients for Sema1a, Sema1b, and PlexA were consistently higher than those for the NS clusters, indicating that a greater proportion of semaphorin signal overlaps with GAG-rich regions than vice versa. These findings support the conclusion that Sema1a, Sema1b, and PlexA can engage heparan sulfate chains in cis on the same cell surface. To probe the potential for trans interactions, CHO-K1 and CHO-PgsA745 cells were adapted to growth in suspension and examined for cell clumping following expression of Sema1a–mClover driven by a strong promoter. Under these conditions, GAG-deficient CHO-PgsA745 cells expressing Sema1a–mClover showed no obvious tendency to form cell clumps (Fig. 2 e). In contrast, GAG-rich CHO-K1 cells expressing the same construct frequently exhibited multicellular aggregates (Fig. 2 f). These aggregates were classified into three size categories—small (3–10 cells), intermediate (11–50 cells), and large (51–500 cells)—with intermediate and large aggregates being substantially more prevalent in GAG-rich CHO-K1 cells than in GAG-deficient CHO-PgsA745 cells (Fig. 2 g). In addition, these experiments were performed under conditions of elevated Sema1a expression and may therefore not fully reflect physiological expression levels. Together, these results indicate that Sema1a can mediate trans interactions with heparan sulfate chains on neighboring cells when expressed at high levels. Combined with our cis data, this suggests a dual mode of GAG engagement by Sema1a, potentially contributing to its context-dependent signaling roles. Transmembrane semaphorins and PlexA preferentially associate with less sulfated heparan sulfate NA clusters We previously showed that the secreted semaphorins Sema2a and Sema2b associate with endogenous cell-surface heparan sulfate chains and preferentially localize to highly sulfated NS clusters 12 . To test whether transmembrane semaphorins follow the same pattern or instead preferentially associate with less sulfated NA clusters, we performed STED-based colocalization analyses of Sema1a, Sema1b, and their receptor PlexA using an NA-cluster–specific antibody 15 (Fig. 3 a) and compared these results to our NS-cluster analysis (Fig. 2 a-d). For all three proteins, colocalization was consistently stronger with less sulfated NA clusters (Fig. 3 b–d) than with NS clusters. For Sema1a, colocalization with NA clusters yielded a higher median PCC (0.457) compared with NS clusters (0.302) (Fig. 3 e), together with a larger median fraction of Sema1a signal overlapping with NA clusters (M(Sema1a) = 0.779 for NA versus 0.686 for NS) (Fig. 3 f). This distinction was even more pronounced for Sema1b, which exhibited weak spatial association with NS clusters (median PCC = 0.139; median M(Sema1b) = 0.402) but substantially stronger overlap with NA clusters (median PCC = 0.411; median M(Sema1b) = 0.705) (Fig. 3 e-f). PlexA displayed an intermediate behavior, showing moderate colocalization with NS clusters (median PCC = 0.189; median M(PlexA) = 0.527) and a clear increase in both metrics when analyzed against NA clusters (median PCC = 0.351; median M(PlexA) = 0.702) (Fig. 3 e-f). For both cluster types, Manders’ coefficients calculated for the semaphorins or PlexA signal were consistently higher than those calculated for the corresponding NA (Fig. 3 b-d) or NS cluster signal (Fig. 2 b-d), indicating that while a substantial fraction of the protein signal resides within heparan sulfate-enriched regions, only a subset of NA or NS clusters is occupied by these proteins. Together, these data indicate that, in contrast to secreted semaphorins, transmembrane semaphorins and PlexA are more frequently associated with NA clusters than with discrete NS clusters on the same cell surface. GAG chains promote membrane clustering of Sema1a and PlexA To investigate the lateral organization of Sema1a, Sema1b, and PlexA at the plasma membrane, we performed STED microscopy followed by spatial distribution analysis using Ripley’s K function on GAG-rich CHO-K1 and GAG-deficient CHO-PgsA745 cells (Fig. 4 a-d). Ripley’s K function quantifies spatial patterns by comparing the observed number of particles within a given radius to the number expected under complete spatial randomness 16 . The L(r) – r transformation of this function facilitates visual interpretation, where values above zero indicate clustering and values below zero indicate dispersion. The analysis revealed pronounced clustering of Sema1a in CHO-K1 cells, with higher L(r) – r values in CHO-K1 cells compared to CHO-PgsA745 cells across a range of spatial scales, most prominently between 100 and 300 nm, and persisting up to ~ 500 nm (Fig. 4 b). This indicates that GAG chains promote Sema1a clustering at the membrane. In contrast, Sema1b exhibited nearly identical clustering profiles in both cell types, suggesting that its membrane distribution is independent of GAGs (Fig. 4 c). PlexA also showed consistently higher clustering in CHO-K1 cells across the 100–500 nm range (Fig. 4 d), although the difference was moderate—not as strong as for Sema1a—indicating a partial dependence on GAGs for its spatial organization. GAG chains modulate the lateral mobility of semaphorins at the cell surface To assess how GAG chains influence the lateral mobility of Sema1a, Sema1b, and their receptor PlexA at the cell surface, we performed fluorescence recovery after photobleaching (FRAP). GAG-rich CHO-K1 cells were transfected to express either Sema1a, Sema1b, or PlexA, with their cytoplasmic domains replaced by the fluorescent protein mClover. As a GAG-deficient control, we used CHO-PgsA745 cells expressing the same constructs. FRAP measurements were conducted by bleaching a circular region of 1.5 µm in diameter across the plasma membrane and monitoring fluorescence recovery within this area (Fig. 4 e-g, Fig. S2 ). For both Sema1a and Sema1b, effective diffusion coefficients were consistently higher on GAG-deficient CHO-PgsA745 cells compared to CHO-K1 cells (Fig. 4 h-i), indicating that GAG chains slow semaphorin diffusion, likely via interactions between their ectodomains and GAG chains. In the absence of GAG chains, these interactions are reduced or absent, resulting in increased mobility. Notably, Sema1b displayed systematically higher diffusion coefficients than Sema1a across both cell lines (Fig. 4 h-i), consistent with its lower GAG-binding affinity relative to Sema1a (Fig. 1 a-b). In contrast, PlexA exhibited only a minor, albeit statistically significant, difference in diffusion between CHO-K1 and CHO-PgsA745 cells (Fig. 4 j), suggesting that GAG chains exert a substantially weaker influence on PlexA mobility compared to the semaphorins. Together, these data demonstrate that GAG chains selectively modulate the lateral mobility of transmembrane semaphorins at the cell surface, whereas their effect on PlexA is limited. GAG chains on the cell surface enhance Plexin-mediated cell collapse To investigate how GAG chains on the cell surface influence plexin-mediated cell collapse, we compared the extent of collapse between GAG-rich CHO-K1 cells and GAG-deficient CHO-PgsA745 cells. Although our initial aim was to examine Drosophila PlexA and Sema1a, we were unable to identify suitable Drosophila cell lines with defined differences in GAG chain biosynthesis analogous to those of CHO-K1 and CHO-PgsA745 cells. We therefore opted for their mammalian counterparts—mouse PlxnD1 and Sema3E—which are known to robustly induce collapse in neurons 17 and COS7 cells 18 , 19 . To assess the impact of GAG chains on collapse, we transfected CHO-K1 and CHO-PgsA745 cells with a construct encoding PlxnD1 fused to the fluorescent protein mVenus. Live-cell imaging was performed following the addition of purified Sema3E, and individual cells were tracked over time (Fig. 5 a). Cell collapse was quantified by measuring changes in cell surface area. In GAG-rich CHO-K1 cells, Sema3E induced pronounced collapse, with cells shrinking to approximately 65% of their original surface area after 150 minutes (Fig. 5 b). While this collapse was slightly less dramatic than previously reported for COS7 cells 18 —likely due to morphological differences, as COS7 cells exhibit a more spread-out shape—the response was still robust. In contrast, GAG-deficient CHO-PgsA745 cells exhibited a significantly attenuated response, retaining ~ 90% of their original surface area after the same period (Fig. 5 b). Taken together, these results suggest that GAG chains on the cell surface significantly enhance plexin-mediated cell collapse. DISCUSSION Heparan sulfate proteoglycans are broadly implicated as cell-surface modulators of guidance-cue signaling, but the molecular basis by which they tune cell-surface receptor signaling remains incompletely understood. Our data identify heparan sulfate as a direct interaction partner and functional modulator of Drosophila transmembrane semaphorins Sema1a and Sema1b and their shared receptor PlexA. Using SPR with monomeric ectodomains, we showed that Sema1a, Sema1b, and PlexA bind heparin and heparan sulfate with micromolar apparent affinities, while no binding to chondroitin sulfate was detected under the concentrations tested. This selectivity differs from several secreted semaphorins that have been reported to engage both heparan sulfate and chondroitin sulfate, including Drosophila Sema2a and Sema2b 12 and vertebrate Sema3A 9 , 20 . Beyond GAG-class selectivity, our data reveal a difference between secreted and transmembrane semaphorins in their apparent preference for distinct heparan sulfate clusters. In particular, an early study showed that heparan sulfate modification is non-uniform along a single chain and is organized into highly sulfated N-sulfo–rich NS clusters and less sulfated N-acetyl–rich NA clusters 2 , 3 . Another study further indicated that the relative proportion (NA-to-NS ratio) and arrangement of NA and NS clusters can be proteoglycan-specific and vary across tissues 15 , adding an additional layer of molecular patterning and positional information to cell-surface heparan sulfate. Such organization has been implicated in tuning morphogen signaling pathways, including Wnt 15 and BMP 21 . In our previous study, we found that secreted semaphorins (Sema2a and Sema2b) preferentially associate with NS clusters 12 . By contrast, we show here that the transmembrane semaphorins Sema1a and Sema1b, together with PlexA, preferentially associate with NA clusters. This divergence is consistent with the distinct physical and biological constraints faced by secreted versus transmembrane semaphorins. In particular, in line with the semaphorin bridge model 12 , secreted semaphorins are retained near their sites of production to limit dilution and ectopic diffusion. Preferential association with highly sulfated NS clusters, which have a higher negative charge density, would support efficient surface tethering and spatial confinement (Fig. 6 a). By contrast, transmembrane semaphorins and PlexA are already membrane associated and therefore do not require high-affinity capture. Instead, their key challenge is to tune nanoscale organization and dynamics within the plane of the membrane. Preferential association with NA clusters is compatible with broad, reversible cis interactions (Fig. 6 b) distributed across the cell surface—interactions strong enough to influence clustering and lateral mobility, yet sufficiently dynamic to permit rapid reorganization during cell–cell contact. An additional advantage of NA-specific association of transmembrane semaphorins is that it may reduce the risk of constitutive sequestration. Highly sulfated NS clusters could behave as high-affinity “traps” that overly immobilize membrane proteins and thereby dampen responsiveness to trans cues. Preferential engagement of NA clusters by transmembrane semaphorins and PlexA may therefore provide a tunable interaction landscape that supports organizational control while preserving membrane dynamics and signaling flexibility. Building on this NA-cluster-biased cis engagement, three independent lines of evidence indicate that GAG chains actively shape the nanoscale organization, dynamics, and signaling output of membrane-anchored semaphorins and plexins. First, point-pattern analysis (Ripley’s K; L(r)–r) shows that GAG chains promote nanoscale clustering of Sema1a and, to a lesser extent, PlexA, whereas Sema1b displays largely GAG-independent clustering. Second, FRAP measurements demonstrate that GAG chains reduce the lateral mobility of both Sema1a and Sema1b, with only a minor effect on PlexA diffusion. Together, these imaging-based readouts are consistent with multivalent, transient ectodomain–GAG contacts that bias proteins into GAG-enriched membrane territories and reduce effective diffusion, thereby increasing local surface density and encounter probability without requiring strict one-to-one colocalization. This interpretation is further supported by recent structural and biochemical studies showing that heparan sulfate engagement promotes oligomerization of purified Sema5A ectodomains, highlighting a broader principle whereby GAG-mediated multivalency can drive higher-order assembly of transmembrane semaphorins 22 . Third, we link these membrane-organizational effects to functional signaling output. In a quantitative collapse assay, cell-surface GAG chains markedly enhanced plexin-dependent morphological collapse, with GAG-rich cells exhibiting substantially stronger collapse than GAG-deficient cells. This observation is consistent with prior work showing that exogenous heparin can potentiate Sema3A-induced growth cone collapse 9 . More broadly, an essential role for cell-surface heparan sulfate has also been reported for other axon guidance systems; notably, Slit–Robo signaling has been shown to be abrogated in the absence of heparan sulfate chains at the cell surface 23 . Together, these findings support a model in which heparan sulfate acts as a low-affinity, multivalent scaffold that increases the probability and lifetime of productive signaling complexes at the membrane—through local enrichment, altered mobility, and extended effective ligand–receptor engagement—thereby lowering the threshold for plexin activation and amplifying downstream cytoskeletal remodeling. More broadly, our results align with a general role for cell-surface heparan sulfate proteoglycans as organizers that tune signaling by controlling local ligand–receptor engagement and nanoscale membrane organization. Similar principles have been described in other pathways. For example, heparan sulfate can promote ligand retention and signaling efficiency in fibroblast growth factor signaling 24 – 26 , support chemokine function through GAGs-dependent interactions (reviewed in 27 ), and shape morphogen distribution and signaling—such as Wnt—through distinct NA/NS heparan sulfate domains 15 . Heparan sulfate has also been implicated in BMP signaling: recent work supports a cluster-based mechanism in which BMP4 preferentially associates with NS clusters, whereas the secreted antagonist Cerberus binds less sulfated NA-rich regions and broadens BMP4 distribution and signaling range by redistributing ligand across the cell surface 21 . How might these heparan sulfate-dependent organizational effects fit with established semaphorin–plexin signaling mechanisms? Canonically, semaphorin-mediated plexin activation depends on semaphorin-driven plexin dimerization 28 – 30 , which promotes activation of the plexin intracellular GAP domain 31 , leading to downstream cytoskeletal remodeling. Here, we propose that heparan sulfate acts primarily upstream of signaling, serving as a membrane organizer that tunes the probability of productive plexin engagement. We therefore propose the following model (Fig. 6 c). First, in cis , NA clusters within heparan sulfate provide a nanoscale organizational scaffold that enriches transmembrane semaphorins (most prominently Sema1a) and, to a lesser extent, PlexA within heparan sulfate–defined membrane territories, while simultaneously reducing semaphorin mobility. This combination increases local surface density and dwell time. Second, these effects should promote trans signaling at cell–cell contacts by increasing the probability of productive semaphorin-driven plexin dimerization and activation. Third, under conditions of elevated semaphorin abundance, heparan sulfate can also support trans adhesive bridging, as suggested by our aggregation assay. In addition, concentration within heparan sulfate-defined territories may also increase the likelihood of encounters with additional heparan sulfate-binding co-receptors—such as Off-track (Otk), a Sema1a co-receptor 32 —thereby promoting assembly of higher-order signaling complexes. Together, our results support a view in which heparan sulfate does not merely capture guidance cues but tunes the cis membrane landscape—organization, mobility, and encounter kinetics—to set the threshold for plexin activation and downstream signaling. Future in vivo tests will be required to validate this model and to determine which heparan sulfate contexts regulate cis membrane organization versus trans signaling efficiency. METHODS Accession Numbers and Sequence Versions Used The following protein sequence versions were used in this study: Drosophila melanogaster Sema1a (UniParc UPI0000252361; release date: 2004-03-15; formerly UniProt Q24322), Sema1b (UniProt Q7KK54; release date: 2006-10-03), PlexA (UniProt Q9V491; release date: 2000-05-01), mouse PlxnD1 (UniProt Q3UH93; release date: 2005-10-11, version 1), and mouse Sema3E (UniProt P70275; release date: 2011-07-27, version 3). Protein Production Constructs encoding the ectodomains of Sema1a (residues 21N–606E; hereafter Sema1a ecto ) and Sema1b (residues 37D–659S; Sema1b ecto ) were cloned into the pHR-CMV-TetO2-IRES-EmGFP vector, in-frame with a C-terminal 3C-Avi-His₆ tag. Stable cell lines were generated by lentiviral transduction 33 . Sema1a ecto was produced in HEK293T cells (ATCC CRL-3216) cultured in DMEM with 2% FBS at 30°C, while Sema1b ecto was expressed in HEK293S GnTI⁻ cells (ATCC CRL-3022) maintained in FreeStyle 293 Expression Medium at 37°C. Conditioned media were buffer-exchanged and proteins were purified by immobilized metal affinity chromatography (HisTrap FF, Cytiva), followed by size-exclusion chromatography using a HiLoad 16/600 Superdex 200 pg column (Cytiva). The ectodomain of PlexA (PlexA ecto , residues 28–1272) was produced and purified as described previously 13 , 34 . Mouse Sema3E (residues 26A–756E) was produced in fusion with an Fc tag. The Sema3E-Fc construct was cloned into the pHR-CMV-TetO2-IRES-EmGFP vector, with a C-terminal 3C-Avi-His₆ tag. Stable expression was established in Expi293F cells (Thermo Fisher Scientific, Cat. No. A14527) maintained in Expi293 Expression Medium at 37°C. The purification protocol followed the same approach as that used for Sema1a ecto . Surface plasmon resonance (SPR) binding experiments SPR binding experiments were performed as described in our previous study 12 . In brief, glycosaminoglycans (heparin, heparan sulfate, and chondroitin sulfate; Toronto Research Chemicals) were biotinylated and immobilized on SA sensor chips (Cytiva). Biotinylated Dlp constructs were produced in HEK293T cells by co-transfection with pDisplay-BirA-ER and cultured in the presence of 100 µM D-biotin. Three days post-transfection, conditioned media were collected, dialyzed, and used for immobilization on SA sensor chips. SPR measurements were carried out on a Biacore S200 instrument (Cytiva) at 25°C in a running buffer containing 15 mM Tris (pH 7.5), 150 mM NaCl, and 0.05% (v/v) Tween-20. The surface was regenerated between runs using a buffer containing 0.1 M Tris (pH 8.0), 1.0 M NaCl, and 1% CHAPS. Data were analyzed using Biacore Insight Evaluation software (Cytiva), and each experiment was performed in two to three biological replicates. Immunofluorescence Adherent CHO-K1 and CHO-PgsA745 cells were seeded on 18 mm high-precision coverslips (15,000 or 25,000 cells per coverslip) in DMEM supplemented with 2% FBS. After 24 h of growth, cells were transfected with constructs encoding Sema1a, Sema1b, or PlexA using either GeneJuice or Lipofectamine 3000 transfection reagent. The constructs were generated by cloning DNAs encoding Drosophila melanogaster Sema1a (residues 21–633), Sema1b (residues 37–686), or PlexA (residues 28–1311) into the pHLsec vector in-frame with C-terminal fluorescent proteins mVenus or mClover, respectively. Forty-eight hours post-transfection, cells were washed twice with PBS and fixed with 3% paraformaldehyde in PBS for 10 min. After two 5-min PBS washes, cells were permeabilized with 0.1% saponin in PBS for 20 min, followed by three additional 5-min PBS washes. Blocking was performed in 4% BSA in PBS for 1 h. For the clustering assay, cells were incubated with 10 µg/ml anti-GFP antibody (Abcam, ab290; rabbit polyclonal) in PBS containing 0.3% BSA and 0.05% saponin for 1 h. For the colocalization assay, cells were incubated with anti-GFP antibody together with either 1 µg/ml anti-N-sulfated heparan sulfate antibody (Amsbio, 370-255-5; mouse IgM, clone F58-10E4) or anti-N-acetylated heparan sulfate antibody (1:50 dilution; kindly provided by Prof. Masanori Taira and Dr. Yusuke Mii, The University of Tokyo) under the same buffer conditions. After extensive washing, secondary detection was performed for 1 h in PBS with 0.3% BSA and 0.05% saponin using 3.3 µg/ml goat anti-rabbit IgG antibody conjugated with Abberior STAR 635P (clustering assay) or together with 5 µg/ml goat anti-mouse IgM antibody conjugated with Alexa Fluor 594 (colocalization assay). Finally, cells were washed five times for 5 min each in PBS, rinsed thoroughly with ultrapure water, air-dried, and mounted in VectaShield. STED Microscopy Super-resolution images were acquired using an Abberior STED microscope equipped with a 775 nm pulsed STED laser operating at 40 MHz, controlled by Imspector software (version 16.3). A Nikon CFI Plan Apo Lambda oil immersion objective (60×, NA 1.40) was used for image acquisition, focusing on the most well-spread region of the cell membrane. Images were deconvolved using Huygens Professional (version 22.10) with the Classic Maximum Likelihood Estimation (CMLE) algorithm. For the clustering assay, protein distribution within the plasma membrane was analyzed using custom ImageJ 35 macros and R scripts (version 4.3.0); code is available upon request. Briefly, the most extended region of the cell membrane was manually masked, and a 2 × 2 µm² region within this mask was selected for analysis. Automatic thresholding was applied using the Otsu algorithm, and particles of ≥ 2 pixels in size were detected based on an intensity prominence threshold of 3000 using the Find Maxima function in ImageJ. For each detected particle, the maximum intensity was identified, and a binary image of particle maxima was generated. To assess spatial distribution, the L(r) function—a linear transformation of Ripley’s K(r)—was computed the spatstat library (version 3.2.1), comparing the observed distribution within a given radius “r” to the expected distribution under spatial randomness (based on 25 Monte Carlo simulations). Clustering was inferred when L(r) – r > 0, and dispersion when L(r) – r < 0. Statistical significance of protein clustering was assessed using two-way ANOVA in GraphPad Prism (version 10.5.0). For the colocalization assay, colocalization within the most well-spread region of the plasma membrane was quantified using the Coloc2 plugin in ImageJ (version 1.54p). Significance of colocalization was evaluated using the Wilcoxon test in GraphPad Prism (version 10.5.0). Clumping Assay The construct encoding the extracellular and transmembrane domains of Drosophila melanogaster Sema1a (residues 21–633) was cloned into the pHR-CMV-TetO2 vector in-frame with a C-terminal mClover. Stable CHO-K1 and CHO-PgsA745 cell lines expressing Sema1a-mClover were generated by lentiviral transduction 33 . Subsequently, the adherent cultures were adapted to suspension growth and maintained in phenol red-free DMEM supplemented with 10% FBS, under continuous shaking at 130 rpm and 8% CO₂. For the clumping assay, suspension cells were transferred into µ-Slide 8 Well chambers (200,000 cells per well) in phenol red-free DMEM supplemented with 10% FBS. Immediately after cell transfer, a large-field image of each well (5 × 5 fields, 1% overlap; total area 3439.21 × 3439.21 µm²) was acquired and merged using a Nikon CSU-W1 spinning disk confocal microscope equipped with a CFI Plan Apo VC objective (20×, NA 0.75), an environmental chamber (37°C, 5% CO₂), and NIS-Elements AR software (version 5.41.01). Cell clumping was analyzed using a combination of NIS-Elements AR, ImageJ 35 , Cellpose (version 3.0.7) (ref 36 ), and custom Python scripts. Prior to segmentation, images were resized to 50% using averaging. The DIC channel was preprocessed using the Homogenization function in NIS-Elements (resolution degree 10) to reduce illumination inhomogeneity and improve contrast. Cells and clumps were segmented using a custom-trained Cellpose model. The model was trained on 26 image patches (512 × 512 pixels), comprising 2749 manually annotated objects. Training was performed on the DIC channel for 100 epochs with a learning rate of 0.1 and weight decay of 0.0001, using ‘cyto3’ as the initial model. Segmented objects touching image borders were removed using the scikit-image library (version 0.25.2) in Python (version 3.10.19). For each clump, area and mean fluorescence intensity were quantified after background subtraction. Background intensity was estimated as the mean intensity of regions outside segmented objects. Objects with mean fluorescence intensity above a threshold of 80 were retained for analysis. The threshold was determined empirically from the distribution of mean object intensities pooled across all images, corresponding to the transition region between the dominant low-intensity negative population and the positive signal tail. Statistical significance was assessed using an unpaired t -test in GraphPad Prism (version 10.5.0). All experiments were performed in quadruplicate. FRAP (Fluorescence Recovery After Photobleaching) Adherent CHO-K1 and CHO-PgsA745 cells were seeded in µ-Slide 8 Well chambers (40,000 cells per well) in phenol red-free DMEM supplemented with 2% FBS. After 24 h of growth, cells were transfected with constructs encoding Sema1a, Sema1b, or PlexA fused to mClover using either GeneJuice or Lipofectamine 3000 transfection reagents. Construct details are provided in the Immunofluorescence section. Forty-eight hours post-transfection, FRAP experiments were performed using a Nikon CSU-W1 spinning disk confocal microscope equipped with a CF Plan Apo VC water immersion objective (60×, NA 1.20), an environmental chamber (37°C, 5% CO₂), and controlled by NIS-Elements AR software (version 5.41.01). Imaging was focused on the most well-spread region of the plasma membrane. Fluorescence of mClover was excited using a 488 nm laser at 5% power for imaging and 7% for photobleaching. A circular region of interest (ROI) with a 1.5 µm diameter was selected for photobleaching. Pre-bleach acquisition was conducted for 5 s at 500 ms intervals. Photobleaching was performed over 200 ms, followed by post-bleach acquisition for 2 min at 500 ms intervals. FRAP data were analyzed using custom Matlab (R2024a) scripts (available upon request). Briefly, circular ROIs matching the size of the photobleached area were placed in unbleached membrane regions (reference) and in areas without fluorescence signal (background). Fluorescence intensity values were normalized per cell, and recovery curves were fitted globally with shared parameters using nonlinear least-squares solver lsqcurvefit . Effective diffusion coefficients were determined from the recovery curves as described previously 37 . Statistical significance of protein dynamics was assessed using an unpaired t -test in GraphPad Prism (version 10.5.0). Live-Cell Imaging of Sema3E-Induced Collapse of CHO Cells Adherent CHO-K1 and CHO-PgsA745 cells were cultured in µ-Slide 8 Well chambers (10,000 or 20,000 cells per well) in phenol red-free DMEM supplemented with 2% FBS. After 24 hours of growth, cells were transfected with a construct encoding mouse PlxnD1 fused to mClover, using either GeneJuice or Lipofectamine 3000 transfection reagents. This construct was generated by subcloning mouse PlxnD1 (residues 49L–1925A) into the pHLsec vector in-frame with a C-terminal mVenus fluorescent protein. After 48 hours post-transfection, the medium was aspirated and cells were washed twice with DMEM. For the collapse assay, either vehicle (medium) or purified Sema3E was added to the cultures at a final concentration of 0.5 µM. Over the following 150 minutes, large-field images of each well (5 × 5 fields, 1% overlap; total area 3439.21 × 3439.21 µm²) were acquired every 30 minutes and merged. Imaging was performed using a Nikon CSU-W1 spinning disk confocal microscope equipped with a CFI Plan Apo VC objective (20×, NA 0.75), an environmental chamber (37°C, 5% CO₂), and NIS-Elements AR software (version 5.41.01). Cell collapse was analyzed using NIS-Elements AR, ImageJ 35 , Cellpose (version 3.0.7) (ref 36 ), and custom Python scripts. Briefly, the original images were resized to 50% using averaging and aligned across all time points in ImageJ using the BigWarp plugin (version 9.3.1). Time-lapse stacks were then segmented with a custom Cellpose model and analyzed for cell surface area over time with tracking using the TrackMate plugin 38 (version 7.14.0). The model was trained on 19 image patches (512 × 512 pixels), comprising 610 manually annotated objects. Training was performed on the DIC channel for 100 epochs with a learning rate of 0.1 and weight decay of 0.0001, using ‘cyto3’ as the initial model. Object tracking was performed using a Simple LAP tracker with a linking distance of 20 pixels, a gap-closing distance of 5 pixels, and a maximum frame gap of 2 frames. Only tracks containing bright cells with mean fluorescence intensity above a background threshold of 200 in at least 3 frames and starting before frame 2 were retained for collapse analysis. Statistical significance of cell collapse was evaluated using two-way ANOVA in GraphPad Prism (version 10.5.0). All experiments were conducted in triplicate. Declarations Competing interests The authors declare that they have no conflict of interest. Author contributions M.S., and D.R. designed research; M.S., F.N., K.K., J.V., N.K., and D.R. performed research; M.S., Z.C., M.B., and D.R. analyzed data; and M.S., and D.R. wrote the paper. Acknowledgments We thank Prof Masanori Taira and Dr Yusuke Mii from the University of Tokyo for their generous gift of the anti-N-acetylated heparan sulfate antibody. This work was supported by the Czech Science Foundation (Junior Star Grant 21-27204M). Computational resources were provided by the e-INFRA CZ project (ID:90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic. We acknowledge the support and access to the Core Facility of Biomolecular Interactions and Crystallography at Czech Infrastructure for Integrative Structural Biology and Instruct-CZ Centre, supported by the Ministry of Education, Youth and Sports of the Czech Republic (LM2023042) and European Regional Development Fund-Project (No. Z.02.1.01/0.0/0.0/18_046/0015974). We also appreciate the support and assistance of the Imaging Methods Core Facility of Charles University and the Light Microscopy Core Facility of the Institute of Molecular Genetics, both funded by the Ministry of Education, Youth and Sports of the Czech Republic (LM2023050 Czech-BioImaging). Data availability statement All data supporting the findings of this study are available within the paper and its Supplementary Information. All original data used to draw the graphs for SPR assay, colocalization assay, clustering assay, FRAP assay, collapse assay as well as the clustering assay code and FRAP scripts will be available in Figshare. References Dickson BJ (2002) Molecular mechanisms of axon guidance. Science 298:1959–1964. https://doi.org/10.1126/science.1072165 Maccarana M, Sakura Y, Tawada A, Yoshida K, Lindahl U (1996) Domain structure of heparan sulfates from bovine organs. J Biol Chem 271:17804–17810. https://doi.org/10.1074/jbc.271.30.17804 Sarrazin S, Lamanna WC, Esko JD (2011) Heparan sulfate proteoglycans. Cold Spring Harb Perspect Biol 3. https://doi.org/10.1101/cshperspect.a004952 Masu M (2016) Proteoglycans and axon guidance: a new relationship between old partners. J Neurochem 139:58–75. https://doi.org/10.1111/jnc.13508 Fawcett JW, Kwok JCF (2022) Proteoglycan Sulphation in the Function of the Mature Central Nervous System. Front Integr Neurosci 16:895493. https://doi.org/10.3389/fnint.2022.895493 Pasterkamp RJ (2012) Getting neural circuits into shape with semaphorins. Nat Rev Neurosci 13:605–618. https://doi.org/10.1038/nrn3302 Kantor DB et al (2004) Semaphorin 5A is a bifunctional axon guidance cue regulated by heparan and chondroitin sulfate proteoglycans. Neuron 44:961–975. https://doi.org/10.1016/j.neuron.2004.12.002 Cho JY, Chak K, Andreone BJ, Wooley JR, Kolodkin AL (2012) The extracellular matrix proteoglycan perlecan facilitates transmembrane semaphorin-mediated repulsive guidance. Genes Dev 26:2222–2235. https://doi.org/10.1101/gad.193136.112 De Wit J, De Winter F, Klooster J, Verhaagen J (2005) Semaphorin 3A displays a punctate distribution on the surface of neuronal cells and interacts with proteoglycans in the extracellular matrix. Mol Cell Neurosci 29:40–55. https://doi.org/10.1016/j.mcn.2004.12.009 Zimmer G et al (2010) Chondroitin sulfate acts in concert with semaphorin 3A to guide tangential migration of cortical interneurons in the ventral telencephalon. Cereb Cortex 20:2411–2422. https://doi.org/10.1093/cercor/bhp309 Vo T et al (2013) The chemorepulsive axon guidance protein semaphorin3A is a constituent of perineuronal nets in the adult rodent brain. Mol Cell Neurosci 56:186–200. https://doi.org/10.1016/j.mcn.2013.04.009 Nourisanami F et al (2024) Molecular mechanisms of proteoglycan-mediated semaphorin signaling in axon guidance. Proc Natl Acad Sci U S A 121:e2402755121. https://doi.org/10.1073/pnas.2402755121 Rozbesky D et al (2019) Diversity of oligomerization in Drosophila semaphorins suggests a mechanism of functional fine-tuning. Nat Commun 10:3691. https://doi.org/10.1038/s41467-019-11683-y David G, Bai XM, Van der Schueren B, Cassiman JJ, Van den Berghe H (1992) Developmental changes in heparan sulfate expression: in situ detection with mAbs. J Cell Biol 119:961–975. https://doi.org/10.1083/jcb.119.4.961 Mii Y et al (1973) (2017) Roles of two types of heparan sulfate clusters in Wnt distribution and signaling in Xenopus. Nat Commun 8. https://doi.org/10.1038/s41467-017-02076-0 Ripley BD (1979) Tests of 'Randomness' for Spatial Point Patterns. J Royal Stat Soc Ser B: Stat Methodol 41:368–374. https://doi.org/10.1111/j.2517-6161.1979.tb01091.x Burk K et al (2017) Post-endocytic sorting of Plexin-D1 controls signal transduction and development of axonal and vascular circuits. Nat Commun 8:14508. https://doi.org/10.1038/ncomms14508 Gu C et al (2005) Semaphorin 3E and plexin-D1 control vascular pattern independently of neuropilins. Science 307:265–268. https://doi.org/10.1126/science.1105416 Mehta V et al (2020) The guidance receptor plexin D1 is a mechanosensor in endothelial cells. Nature 578:290–295. https://doi.org/10.1038/s41586-020-1979-4 Dick G et al (2013) Semaphorin 3A binds to the perineuronal nets via chondroitin sulfate type E motifs in rodent brains. J Biol Chem 288:27384–27395. https://doi.org/10.1074/jbc.M111.310029 Yamamoto T, Mii Y, Saiga H, Michiue T, Taira M (2025) https://doi.org/10.1101/2025.04.16.649141 Nagy GN et al (2024) Structure and function of Semaphorin-5A glycosaminoglycan interactions. Nat Commun 15:2723. https://doi.org/10.1038/s41467-024-46725-7 Hussain SA et al (2006) A molecular mechanism for the heparan sulfate dependence of slit-robo signaling. J Biol Chem 281:39693–39698. https://doi.org/10.1074/jbc.M609384200 Duchesne L et al (2012) Transport of fibroblast growth factor 2 in the pericellular matrix is controlled by the spatial distribution of its binding sites in heparan sulfate. PLoS Biol 10:e1001361. https://doi.org/10.1371/journal.pbio.1001361 Xue S et al (2022) Phase separation on cell surface facilitates bFGF signal transduction with heparan sulphate. Nat Commun 13:1112. https://doi.org/10.1038/s41467-022-28765-z Venero Galanternik M, Kramer KL, Piotrowski T (2015) Heparan Sulfate Proteoglycans Regulate Fgf Signaling and Cell Polarity during Collective Cell Migration. Cell Rep 10:414–428. https://doi.org/10.1016/j.celrep.2014.12.043 Cambier S, Gouwy M, Proost P (2023) The chemokines CXCL8 and CXCL12: molecular and functional properties, role in disease and efforts towards pharmacological intervention. Cell Mol Immunol 20:217–251. https://doi.org/10.1038/s41423-023-00974-6 Janssen BJ et al (2010) Structural basis of semaphorin-plexin signalling. Nature 467:1118–1122. https://doi.org/10.1038/nature09468 Kong Y et al (2016) Structural Basis for Plexin Activation and Regulation. Neuron 91:548–560. https://doi.org/10.1016/j.neuron.2016.06.018 Lu D, Shang G, He X, Bai XC, Zhang X (2021) Architecture of the Sema3A/PlexinA4/Neuropilin tripartite complex. Nat Commun 12:3172. https://doi.org/10.1038/s41467-021-23541-x Wang Y, Pascoe HG, Brautigam CA, He H, Zhang X (2013) Structural basis for activation and non-canonical catalysis of the Rap GTPase activating protein domain of plexin. Elife 2:e01279. https://doi.org/10.7554/eLife.01279 Rozbesky D et al (2020) Drosophila OTK Is a Glycosaminoglycan-Binding Protein with High Conformational Flexibility. Structure 28, 507–515 e505 https://doi.org/10.1016/j.str.2020.02.008 Elegheert J et al (2018) Lentiviral transduction of mammalian cells for fast, scalable and high-level production of soluble and membrane proteins. Nat Protoc 13:2991–3017. https://doi.org/10.1038/s41596-018-0075-9 Rozbesky D et al (2020) Structural basis of semaphorin-plexin cis interaction. EMBO J 39:e102926. https://doi.org/10.15252/embj.2019102926 Schindelin J et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682. https://doi.org/10.1038/nmeth.2019 Stringer C, Wang T, Michaelos M, Pachitariu M (2020) Cellpose: a generalist algorithm for cellular segmentation. Nat Methods 18:100–106. https://doi.org/10.1038/s41592-020-01018-x Soumpasis DM (1983) Theoretical analysis of fluorescence photobleaching recovery experiments. Biophys J 41:95–97. https://doi.org/10.1016/S0006-3495(83)84410-5 Ershov D et al (2022) TrackMate 7: integrating state-of-the-art segmentation algorithms into tracking pipelines. Nat Methods 19:829–832. https://doi.org/10.1038/s41592-022-01507-1 Supplementary Files SupplementaryFigure.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 26 Apr, 2026 Reviewers invited by journal 23 Apr, 2026 Editor assigned by journal 17 Apr, 2026 First submitted to journal 15 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9405124","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":628538393,"identity":"a9dc65df-0787-4315-b0e3-495910ac248a","order_by":0,"name":"Margarita Sobol","email":"","orcid":"","institution":"Charles University: Univerzita Karlova","correspondingAuthor":false,"prefix":"","firstName":"Margarita","middleName":"","lastName":"Sobol","suffix":""},{"id":628538394,"identity":"d0610354-960f-42e5-9a93-053245340c55","order_by":1,"name":"Zuzana Cockova","email":"","orcid":"","institution":"Charles University: Univerzita Karlova","correspondingAuthor":false,"prefix":"","firstName":"Zuzana","middleName":"","lastName":"Cockova","suffix":""},{"id":628538395,"identity":"0c3d0051-a25a-4b53-a4dc-11c52b23a306","order_by":2,"name":"Farahdokht Nourisanami","email":"","orcid":"","institution":"Charles University: Univerzita Karlova","correspondingAuthor":false,"prefix":"","firstName":"Farahdokht","middleName":"","lastName":"Nourisanami","suffix":""},{"id":628538396,"identity":"299f7316-2cf3-4101-96ae-4ce1823ec733","order_by":3,"name":"Michaela Blazikova","email":"","orcid":"","institution":"Institute of Molecular Genetics Czech Academy of Sciences: Ustav molekularni genetiky Akademie Ved Ceske Republiky","correspondingAuthor":false,"prefix":"","firstName":"Michaela","middleName":"","lastName":"Blazikova","suffix":""},{"id":628538397,"identity":"dc1af060-ec5f-4a2f-9e5c-a35953558ca6","order_by":4,"name":"Karolina Kowalska","email":"","orcid":"","institution":"Charles University: Univerzita Karlova","correspondingAuthor":false,"prefix":"","firstName":"Karolina","middleName":"","lastName":"Kowalska","suffix":""},{"id":628538398,"identity":"2f247c93-cb9f-4dac-8c35-653a50fccf56","order_by":5,"name":"Jonas Vlasak","email":"","orcid":"","institution":"Charles University: Univerzita Karlova","correspondingAuthor":false,"prefix":"","firstName":"Jonas","middleName":"","lastName":"Vlasak","suffix":""},{"id":628538399,"identity":"9a61fc45-f6ce-427a-87d5-1c3ba5bdfa8a","order_by":6,"name":"Nicola Koziskova","email":"","orcid":"","institution":"Charles University: Univerzita Karlova","correspondingAuthor":false,"prefix":"","firstName":"Nicola","middleName":"","lastName":"Koziskova","suffix":""},{"id":628538400,"identity":"2f1083f4-b1f4-4fb0-8997-a5398fa7b63e","order_by":7,"name":"Daniel Rozbesky","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYHACNiBOYOBHCDA2EKdFsgGulFgtBgeIUAoG5u3tzx58+JOWuPl28/PHvHsY5A1uNzcw/PiDW4vMmQPphjPbchK33Tlm2MzzjMFww52DDYw9PLi1SEgkHJPmbahI3HYjAajlAEOCwY3EBmYGCXxaEtuk//ypSNw8I/0jkhYDfFqS2aQZ2HISN0jkINuSgEcLzzF2w962NOMZN3IKZ845IGE4E6jlYM8BPFrYgSH240+ybP+M9A0f3hywkee7kf7wAb4QgwHHBqgRYBKPHQhgT4yiUTAKRsEoGKEAAO01VbuletytAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0001-6546-8219","institution":"Charles University: Univerzita Karlova","correspondingAuthor":true,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Rozbesky","suffix":""}],"badges":[],"createdAt":"2026-04-13 14:04:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9405124/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9405124/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108493713,"identity":"3411a7d3-a7e7-4a3c-bbf2-747b9c34098f","added_by":"auto","created_at":"2026-05-05 10:01:23","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":259917,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTransmembrane semaphorins and their PlexA receptor bind GAG chains.\u003c/strong\u003e\u003cbr\u003e\n(\u003cstrong\u003ea–c\u003c/strong\u003e) Surface plasmon resonance (SPR) binding isotherms showing interactions of the ectodomains of Sema1a (a), Sema1b (b), and PlexA (c) with immobilized heparin (HP, orange), heparan sulfate (HS, blue), and chondroitin sulfate (CS, black). Apparent dissociation constants (K\u003csub\u003eD\u003c/sub\u003e) are indicated above each plot. Error bars represent the standard deviation (SD) of two or three independent biological replicates.\u003cbr\u003e\n(\u003cstrong\u003ed\u003c/strong\u003e) SPR binding isotherms showing interaction of the Sema1a ectodomain with immobilized Dlp core protein lacking heparan sulfate chains (Dlp\u003csup\u003ecore\u003c/sup\u003e; black) and a GPI-anchor–deleted glypican variant retaining heparan sulfate chains (Dlp\u003csup\u003eΔGPI\u003c/sup\u003e; blue). Apparent K\u003csub\u003eD\u003c/sub\u003e values are indicated above the plots. Error bars represent SD of two independent biological replicates.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/8afde591fcf2616b58fa400c.jpeg"},{"id":108494174,"identity":"bd619d12-19a2-4999-90c6-9c12211434f9","added_by":"auto","created_at":"2026-05-05 10:02:47","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1074822,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eColocalization of transmembrane semaphorins and PlexA with heparan sulfate and GAG-dependent cell clumping.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(a)\u003c/strong\u003e Representative STED super-resolution microscopy images of CHO-K1 cells showing the spatial relationship of Sema1a, Sema1b, and PlexA with NS clusters of heparan sulfate, detected using an anti-NS antibody that recognizes highly sulfated NS regions within heparan sulfate chains. Scale bar, 1 µm.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(b–d)\u003c/strong\u003e Quantitative colocalization analysis of Sema1a (\u003cstrong\u003eb\u003c/strong\u003e), Sema1b (\u003cstrong\u003ec\u003c/strong\u003e), and PlexA (\u003cstrong\u003ed\u003c/strong\u003e) with highly sulfated NS clusters. Pearson’s correlation coefficients (PCC) are shown in black, while Manders’ colocalization coefficients are shown in orange for Sema1a, Sema1b, or PlexA and in blue for NS clusters. Each data point represents a single cell. The number of cells analyzed was \u003cem\u003eN\u003c/em\u003e = 75 for Sema1a, \u003cem\u003eN\u003c/em\u003e = 73 for Sema1b, and \u003cem\u003eN\u003c/em\u003e = 75 for PlexA, pooled from two independent experiments. Statistical significance was assessed using the Wilcoxon test. In the violin plots, the solid line indicates the median, and the dashed lines denote the 25th and 75th percentiles.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(e)\u003c/strong\u003e Representative confocal microscopy images of GAG-deficient CHO-PgsA745 cells stably expressing Sema1a, showing no obvious tendency to form cell clusters. Scale bar, 500 µm.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(f)\u003c/strong\u003e Representative confocal microscopy images of GAG-rich CHO-K1 cells stably expressing Sema1a, illustrating the formation of multicellular aggregates.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(g)\u003c/strong\u003e Quantitative analysis of cell clumping. The distribution of cells among four aggregate size categories is shown: no clumps (1–2 cells), small (3–10 cells), intermediate (11–50 cells), and large (51–500 cells). GAG-rich CHO-K1 cells exhibit a markedly higher proportion of intermediate and large aggregates compared with GAG-deficient CHO-PgsA745 cells. The number of cells analyzed was \u003cem\u003eN\u003c/em\u003e = 78521 for CHO-K1 and \u003cem\u003eN\u003c/em\u003e = 86500 for CHO-PgsA745, pooled from four independent experiments.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/a839fa80ea0f95ed99e81397.png"},{"id":108448591,"identity":"ed1cfb6b-33fc-4c36-a02d-11a61c78dbfc","added_by":"auto","created_at":"2026-05-04 18:40:34","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1332296,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePreferential association of transmembrane semaphorins and PlexA with NA clusters of heparan sulfate.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Representative STED super-resolution microscopy images of CHO-K1 cells illustrating the spatial relationship of Sema1a, Sema1b, and PlexA with NA clusters, detected using an anti-NA antibody that recognizes less sulfated regions of heparan sulfate chains. Scale bar, 1 µm.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eb–d\u003c/strong\u003e) Quantitative colocalization analysis of Sema1a (\u003cstrong\u003eb\u003c/strong\u003e), Sema1b (\u003cstrong\u003ec\u003c/strong\u003e), and PlexA (\u003cstrong\u003ed\u003c/strong\u003e) with NA clusters. Pearson’s correlation coefficients (PCC) are shown in black. Manders’ colocalization coefficients are shown in orange for Sema1a, Sema1b, or PlexA and in blue for NA clusters. Each data point represents a single cell. The number of cells analyzed was \u003cem\u003eN\u003c/em\u003e= 60 for Sema1a, \u003cem\u003eN\u003c/em\u003e = 60 for Sema1b, and \u003cem\u003eN\u003c/em\u003e = 60 for PlexA, pooled from two independent experiments. Statistical significance was assessed using the Wilcoxon test. In the violin plots, the solid line denotes the median, and the dashed lines indicate the 25th and 75th percentiles.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ee\u003c/strong\u003e) Comparison of Pearson’s correlation coefficients for colocalization of Sema1a, Sema1b, and PlexA with NA clusters versus NS clusters of heparan sulfate. Statistical significance was assessed using one-way ANOVA test.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ef\u003c/strong\u003e) Comparison of Manders’ colocalization coefficients for Sema1a, Sema1b, and PlexA with NA clusters versus NS clusters of heparan sulfate. Statistical significance was assessed using the one-way ANOVA test.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/88071788244556b63f593602.jpeg"},{"id":108493407,"identity":"29ed9eb5-d53e-4aa0-bc44-bdee0fc2d267","added_by":"auto","created_at":"2026-05-05 10:00:18","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":526177,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGlycosaminoglycan chains promote membrane clustering and modulate lateral mobility of transmembrane semaphorins.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Representative STED super-resolution microscopy images of GAG-rich CHO-K1 and GAG-deficient CHO-PgsA745 cells showing the nanoscale spatial distribution of Sema1a, Sema1b, and PlexA at the cell surface. Scale bar, 500 nm.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eb–d\u003c/strong\u003e) Spatial cluster analysis of Sema1a (\u003cstrong\u003eb\u003c/strong\u003e), Sema1b (\u003cstrong\u003ec\u003c/strong\u003e), and PlexA (\u003cstrong\u003ed\u003c/strong\u003e) on the plasma membrane of CHO-K1 (orange) and CHO-PgsA745 (blue) cells using Ripley’s K function. Data are presented as the L(r) – r transformation, where positive values indicate clustering relative to complete spatial randomness and negative values indicate dispersion. Curves represent the mean clustering behavior across analyzed cells. The number of cells analyzed was \u003cem\u003eN\u003c/em\u003e = 24 (CHO-K1) and \u003cem\u003eN\u003c/em\u003e = 25 (CHO-PgsA745) for Sema1a; \u003cem\u003eN\u003c/em\u003e = 24 (CHO-K1) and \u003cem\u003eN\u003c/em\u003e = 22 (CHO-PgsA745) for Sema1b; and \u003cem\u003eN\u003c/em\u003e= 25 (CHO-K1) and \u003cem\u003eN\u003c/em\u003e = 24 (CHO-PgsA745) for PlexA, pooled from two independent experiments. Statistical significance was assessed using two-way ANOVA test.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ee–g\u003c/strong\u003e) Averaged fluorescence recovery after photobleaching (FRAP) traces for Sema1a (\u003cstrong\u003ee\u003c/strong\u003e), Sema1b (\u003cstrong\u003ef\u003c/strong\u003e), and PlexA (\u003cstrong\u003eg\u003c/strong\u003e) on the plasma membrane of CHO-K1 (orange) and CHO-PgsA745 (blue) cells. FRAP experiments were performed by bleaching a circular region of 1.5 µm diameter, and fluorescence recovery was monitored over time. Data are shown as mean ± 95% confidence interval. The number of cells analyzed was \u003cem\u003eN\u003c/em\u003e = 32 (CHO-K1) and \u003cem\u003eN\u003c/em\u003e = 35 (CHO-PgsA745) for Sema1a, \u003cem\u003eN\u003c/em\u003e = 43 (CHO-K1) and \u003cem\u003eN\u003c/em\u003e = 45 (CHO-PgsA745) for Sema1b, and \u003cem\u003eN\u003c/em\u003e= 32 (CHO-K1) and \u003cem\u003eN\u003c/em\u003e= 48 (CHO-PgsA745) for PlexA, pooled from two independent experiments for Sema1a and four independent experiments for Sema1b and PlexA.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eh–j\u003c/strong\u003e) Effective lateral diffusion coefficients derived from FRAP measurements for Sema1a (\u003cstrong\u003eh\u003c/strong\u003e), Sema1b (\u003cstrong\u003ei\u003c/strong\u003e), and PlexA (\u003cstrong\u003ej\u003c/strong\u003e) on the cell surface of CHO-K1 (orange) and CHO-PgsA745 (blue) cells. Statistical significance was assessed by an unpaired t-test.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/0ff92a9a786009ff9716bb45.jpeg"},{"id":108448595,"identity":"a07705b0-3d9f-4d93-85b9-d28e5979c681","added_by":"auto","created_at":"2026-05-04 18:40:35","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":490728,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCell-surface GAG chains enhance PlxnD1-mediated cell collapse in response to Sema3E.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Representative time-lapse confocal images showing Sema3E-induced collapse of PlxnD1-expressing CHO-K1 (GAG-rich) and CHO-PgsA745 (GAG-deficient) cells. Cells were imaged before stimulation (0 min) and at 30, 60, 90, 120, and 150 min following addition of purified Sema3E (0.5 µM). Vehicle-treated cells are shown as controls. Images illustrate changes in cell morphology and surface area over time.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eb\u003c/strong\u003e) Quantification of cell collapse based on time-lapse imaging shown in (\u003cstrong\u003ea\u003c/strong\u003e). Cell surface area is plotted as a percentage of the initial area (time 0) as a function of time for CHO-K1 and CHO-PgsA745 cells treated with Sema3E or vehicle. Each data point represents a single cell; lines connect mean values from three independent experiments. Statistical significance was assessed using two-way ANOVA.\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/b64e10b9bd0291a1fb64329c.jpeg"},{"id":108448594,"identity":"722b808b-8733-41f6-b791-8be1fec5d652","added_by":"auto","created_at":"2026-05-04 18:40:34","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":615558,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHeparan sulfate domain organization differentially regulates secreted and transmembrane semaphorins.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Secreted semaphorins preferentially associate with highly sulfated N-sulfo–rich (NS) clusters of heparan sulfate at the cell surface, promoting ligand capture and spatial confinement.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eb\u003c/strong\u003e) Transmembrane semaphorins (Sema1a) and their receptor PlexA preferentially associate in \u003cem\u003ecis\u003c/em\u003e with less sulfated N-acetyl–rich (NA) clusters. These interactions enrich both proteins within defined membrane microdomains prior to cell–cell contact.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003ec\u003c/strong\u003e) During \u003cem\u003etrans\u003c/em\u003ecell–cell interaction, Sema1a engages PlexA on the opposing membrane while both proteins remain associated in \u003cem\u003ecis\u003c/em\u003e with NA clusters. NA-domain engagement promotes local enrichment and reduced lateral mobility, increasing the probability of productive Sema–Plex interactions and lowering the activation threshold for plexin signaling.\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/4147f871ed8a5eccf740c479.jpeg"},{"id":108495297,"identity":"4ebbc3a6-73e8-4940-abb8-3c422f5f8623","added_by":"auto","created_at":"2026-05-05 10:09:48","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4634975,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/83130b8f-dd50-4b40-92ec-9c4fadecbebd.pdf"},{"id":108448589,"identity":"b658b1a3-6065-4c6e-bc8c-323f97ba8185","added_by":"auto","created_at":"2026-05-04 18:40:34","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2004482,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure.docx","url":"https://assets-eu.researchsquare.com/files/rs-9405124/v1/0791f51bc4453790bd471929.docx"}],"financialInterests":"","formattedTitle":"Heparan sulfate organizes transmembrane semaphorins at the cell surface to tune plexin signaling","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eThe assembly of neural circuits relies on the precise navigation of growing axons through complex and dynamic environments. This process is controlled by axon guidance cues\u0026mdash;secreted or membrane-associated proteins that act through receptors and co-receptors on growth cones to regulate cytoskeletal dynamics and steer navigation. Major guidance cue families include semaphorins, netrins, slits, and ephrins, which together guide axons to their targets (reviewed in\u003csup\u003e1\u003c/sup\u003e). Proteoglycans have long been recognized as key modulators of axon guidance, yet the molecular mechanisms by which they shape guidance decisions remain poorly understood. Proteoglycans consist of a core protein decorated with one or more covalently attached glycosaminoglycan (GAG) chains\u0026mdash;linear polysaccharides composed of repeating disaccharide units. In the nervous system, the predominant GAGs are heparan sulfate and chondroitin sulfate. These polymers are chemically distinct: heparan sulfate is built from repeating uronate\u0026ndash;glucosamine units (GlcA/IdoA\u0026ndash;GlcNAc/GlcNS), whereas chondroitin sulfate consists of glucuronate\u0026ndash;galactosamine repeats (GlcA\u0026ndash;GalNAc). They also differ in their modification patterns. Chondroitin sulfate is most often O-sulfated at defined positions in a relatively uniform manner along the chain. In contrast, heparan sulfate is extensively and non-uniformly modified, creating distinct domains along a single chain. As a result, a typical heparan sulfate chain contains regions rich in GlcA and GlcNAc (N-acetyl-rich, NA domains) and regions enriched in IdoA and GlcNS derivatives (N-sulfo-rich, NS domains)\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. This molecular heterogeneity dramatically expands the binding repertoire of heparan sulfate, enabling interactions with a wide range of extracellular ligands. Beyond their structural roles in the extracellular matrix, proteoglycans can therefore act as dynamic regulators of intercellular signaling by shaping ligand distribution, stabilizing signaling complexes, and functioning as co-receptors at the cell surface, including in axon guidance pathways (reviewed in\u003csup\u003e4,5\u003c/sup\u003e).\u003c/p\u003e \u003cp\u003eAmong guidance cue families, semaphorins exhibit particularly rich and diverse interactions with proteoglycans. Semaphorins comprise secreted, transmembrane, and GPI-anchored proteins and signal primarily through plexin receptors\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Multiple studies have demonstrated that proteoglycans modulate semaphorin-dependent guidance outcomes. In vertebrates, the transmembrane Sema5A exemplifies GAG-dependent functional switching, heparan sulfate promotes semaphorin-mediated attraction while chondroitin sulfate induces repulsion during formation of the fasciculus retroflexus\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. In Drosophila, the heparan sulfate proteoglycan perlecan facilitates motor axon repulsion mediated by the transmembrane Sema1a\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Secreted semaphorins are likewise influenced by proteoglycans. Sema3A associates with proteoglycans at the neuronal surface, and exogenous heparin enhances Sema3A-induced growth cone collapse\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Subsequent work implicated chondroitin sulfate in Sema3A-dependent migration of cortical interneurons\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e and demonstrated that Sema3A is retained within chondroitin sulfate\u0026ndash;rich perineuronal nets in a sulfation-dependent manner\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Together, these studies establish proteoglycans as critical modulators of semaphorin signaling but the underlying molecular mechanisms remain poorly understood.\u003c/p\u003e \u003cp\u003eWe recently provided mechanistic insight into proteoglycan interactions of secreted semaphorins in Drosophila. We showed that the secreted semaphorins Sema2a and Sema2b bind GAG chains in a sulfation-dependent manner via a positively charged C-terminal tail, a motif also conserved in vertebrate class 3 semaphorins. Disrupting this interaction compromises Sema2b function \u003cem\u003ein vivo\u003c/em\u003e. These findings led us to propose a \u0026ldquo;semaphorin bridge\u0026rdquo; model in which GAG binding concentrates secreted semaphorins at defined anatomical locations, enabling short-range, surface-tethered guidance\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. While this work established how GAGs spatially organize secreted semaphorins at the cell surface, it did not address whether similar principles apply to membrane-associated semaphorins or to their plexin receptors. In particular, it is unclear whether GAGs directly engage these proteins, whether such interactions are selective for specific heparan sulfate regions, and how they influence nanoscale membrane organization, lateral mobility, and signaling output. These questions are especially relevant for transmembrane semaphorins, which are already membrane tethered and therefore face distinct organizational constraints compared with secreted cues.\u003c/p\u003e \u003cp\u003eHere, we address this gap by focusing on the Drosophila class 1 transmembrane semaphorins Sema1a and Sema1b and their shared receptor PlexA. We show that Sema1a, Sema1b, and PlexA bind heparan sulfate chains and preferentially associate with NA domains at the cell surface, in contrast to secreted semaphorins, which favor highly sulfated NS domains. We further demonstrate that heparan sulfate modulates the nanoscale clustering and lateral mobility of these proteins and potentiates plexin-mediated signaling. Together, our findings reveal heparan sulfate as an active organizer of the \u003cem\u003ecis\u003c/em\u003e membrane landscape for transmembrane semaphorin\u0026ndash;plexin signaling, uncovering a distinct mode of GAG-dependent regulation.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSema1a, Sema1b, and their receptor PlexA bind GAG chains\u003c/h2\u003e \u003cp\u003eTo investigate the GAG-binding properties of transmembrane semaphorins, we performed surface plasmon resonance (SPR) binding experiments with purified ectodomains of Drosophila Sema1a and Sema1b, expressed in mammalian cells. Consistent with our previous findings\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, the Sema1a ectodomain was obtained as a mixture of monomer and disulfide-linked dimer, whereas Sema1b was exclusively monomeric. To ensure monovalent binding and enable direct comparison, we used only the monomeric forms of both proteins in SPR assays.\u003c/p\u003e \u003cp\u003eWe first compared the GAG-binding properties of Sema1a and Sema1b. For SPR binding experiments, heparin, heparan sulfate, and chondroitin sulfate were immobilized on the sensor chip, and Sema1a and Sema1b were injected at increasing concentrations. Both proteins bound heparin and heparan sulfate with apparent dissociation constants in the micromolar range, but showed no detectable binding to chondroitin sulfate (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, b), even at the highest concentrations tested (up to 16 \u0026micro;M for Sema1a and 116 \u0026micro;M for Sema1b). Quantitative analysis revealed pronounced differences in GAG-binding affinities between Sema1a and Sema1b: Sema1a bound heparin and heparan sulfate with apparent K\u003csub\u003eD\u003c/sub\u003e values of 1.5 \u0026micro;M and 1.1 \u0026micro;M, respectively, whereas Sema1b displayed substantially weaker binding, with K\u003csub\u003eD\u003c/sub\u003e values of 18.4 \u0026micro;M for heparin and 46.5 \u0026micro;M for heparan sulfate. Compared to these transmembrane semaphorins, the previously reported secreted semaphorin Sema2a binds heparin and heparan sulfate with higher affinity (apparent K\u003csub\u003eD\u003c/sub\u003e = 0.4 \u0026micro;M and 2.3 \u0026micro;M, respectively), while also exhibiting measurable binding to chondroitin sulfate (apparent K\u003csub\u003eD\u003c/sub\u003e = 19.3 \u0026micro;M)\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. The stronger GAG binding observed for Sema2a is likely influenced, at least in part, by avidity effects, as Sema2a was analyzed in its dimeric form.\u003c/p\u003e \u003cp\u003eWe then asked whether PlexA, the shared receptor for Sema1a and Sema1b, directly interacts with GAG chains. The purified PlexA ectodomain bound heparin with an apparent K\u003csub\u003eD\u003c/sub\u003e of 3.7 \u0026micro;M and heparan sulfate with markedly weaker affinity (apparent K\u003csub\u003eD\u003c/sub\u003e \u0026gt; 43.7 \u0026micro;M), while no binding to chondroitin sulfate was detected (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). Thus, PlexA binds GAG chains with affinities intermediate between Sema1a and Sema1b, but is particularly weak for heparan sulfate.\u003c/p\u003e \u003cp\u003eFinally, to assess whether Sema1a can bind native heparan sulfate chains of proteoglycans, we immobilized a secreted, GPI-anchor\u0026ndash;deleted variant of the Drosophila glypican-family proteoglycan Dally-like protein (Dlp\u003csup\u003eΔGPI\u003c/sup\u003e) on an SPR chip. Monomeric Sema1a bound Dlp\u003csup\u003eΔGPI\u003c/sup\u003e with an apparent affinity of 6.5 \u0026micro;M, while no binding was observed to the Dlp core protein lacking heparan sulfate chains (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Together, these results indicate that the ectodomains of transmembrane semaphorins and their PlexA receptor selectively recognize heparin and heparan sulfate, but not chondroitin sulfate, and that this interaction extends to native GAG chains of proteoglycans, such as glypicans.\u003c/p\u003e \u003cp\u003eAlthough reliable extraction of individual association (k\u003csub\u003eon\u003c/sub\u003e) and dissociation (k\u003csub\u003eoff\u003c/sub\u003e) rate constants was not feasible under the experimental conditions used, qualitative inspection of the SPR sensorgrams (\u003cb\u003eFig. S1\u003c/b\u003e) nevertheless provides insight into the overall binding behavior of semaphorin\u0026ndash;GAG interactions. In all cases, the response increased rapidly upon analyte injection, consistent with relatively fast association. Following the injection phase, dissociation occurred slowly, indicating limited complex decay over the time window of the experiment. For PlexA\u0026ndash;GAG interactions, the dissociation phase was particularly shallow with minimal signal loss during dissociation. While these observations do not permit quantitative kinetic analysis, they are consistent with a binding mode characterized by rapid complex formation and slow dissociation, resulting in comparatively long-lived interactions despite micromolar apparent affinities.\u003c/p\u003e \u003cp\u003e \u003cb\u003eSema1a engages GAG chains in both\u003c/b\u003e \u003cb\u003ecis\u003c/b\u003e \u003cb\u003eand\u003c/b\u003e \u003cb\u003etrans\u003c/b\u003e\u003c/p\u003e \u003cp\u003eSPR experiments indicate that the ectodomains of Sema1a, Sema1b, and PlexA selectively recognize heparin and heparan sulfate, but not chondroitin sulfate. Because heparin is an experimental surrogate and not a native cell-surface proteoglycan, we next focused on heparan sulfate proteoglycans as the physiologically relevant partners. We therefore next asked whether Sema1a, Sema1b, and PlexA engage heparan sulfate chains at the cell surface, and whether such interactions occur in \u003cem\u003ecis\u003c/em\u003e\u0026mdash;on the same cell\u0026mdash;or in \u003cem\u003etrans\u003c/em\u003e between neighboring cells. To examine potential \u003cem\u003ecis\u003c/em\u003e interactions, we performed colocalization analyses between fluorescently labelled Sema1a, Sema1b, or PlexA and endogenous GAG chains on the surface of CHO cells. For visualization, the cytoplasmic domains of Sema1a, Sema1b, and PlexA were replaced with the fluorescent proteins mClover or mVenus. CHO-K1 cells were chosen because they constitutively express high levels of heparan sulfate proteoglycans, whereas GAG-deficient CHO-PgsA745 cells\u0026mdash;harboring a loss-of-function mutation in xylosyltransferase\u0026mdash;were used as a negative control. To achieve nanometer-scale resolution, we employed STED super-resolution microscopy using an anti-NS antibody, which specifically detects sulfated-rich NS clusters in heparan sulfate\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. STED super-resolution microscopy revealed that semaphorins and their PlexA receptor appeared as numerous small and well-defined puncta across the cell surface, consistent with a high density and relatively uniform distribution (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). In contrast, NS staining produced fewer but noticeably larger puncta. The larger apparent size of NS puncta may reflect the extended and flexible nature of heparan sulfate chains, which present multiple epitopes and can accommodate numerous antibody molecules. Colocalization was quantified by calculating both Pearson\u0026rsquo;s correlation coefficient (PCC) and Manders\u0026rsquo; colocalization coefficient (M) using pixel-by-pixel fluorescence intensity values. The PCC revealed a relatively weak linear correlation between NS clusters and the semaphorins or PlexA (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-d). However, Manders\u0026rsquo; coefficients indicated substantial spatial overlap, suggesting that while the distribution of NS clusters does not strictly predict semaphorin or PlexA localization, the semaphorins are frequently found in regions enriched in NS clusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-d). Moreover, the Manders\u0026rsquo; coefficients for Sema1a, Sema1b, and PlexA were consistently higher than those for the NS clusters, indicating that a greater proportion of semaphorin signal overlaps with GAG-rich regions than vice versa. These findings support the conclusion that Sema1a, Sema1b, and PlexA can engage heparan sulfate chains in \u003cem\u003ecis\u003c/em\u003e on the same cell surface.\u003c/p\u003e \u003cp\u003eTo probe the potential for \u003cem\u003etrans\u003c/em\u003e interactions, CHO-K1 and CHO-PgsA745 cells were adapted to growth in suspension and examined for cell clumping following expression of Sema1a\u0026ndash;mClover driven by a strong promoter. Under these conditions, GAG-deficient CHO-PgsA745 cells expressing Sema1a\u0026ndash;mClover showed no obvious tendency to form cell clumps (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). In contrast, GAG-rich CHO-K1 cells expressing the same construct frequently exhibited multicellular aggregates (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef). These aggregates were classified into three size categories\u0026mdash;small (3\u0026ndash;10 cells), intermediate (11\u0026ndash;50 cells), and large (51\u0026ndash;500 cells)\u0026mdash;with intermediate and large aggregates being substantially more prevalent in GAG-rich CHO-K1 cells than in GAG-deficient CHO-PgsA745 cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg). In addition, these experiments were performed under conditions of elevated Sema1a expression and may therefore not fully reflect physiological expression levels. Together, these results indicate that Sema1a can mediate \u003cem\u003etrans\u003c/em\u003e interactions with heparan sulfate chains on neighboring cells when expressed at high levels. Combined with our \u003cem\u003ecis\u003c/em\u003e data, this suggests a dual mode of GAG engagement by Sema1a, potentially contributing to its context-dependent signaling roles.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eTransmembrane semaphorins and PlexA preferentially associate with less sulfated heparan sulfate NA clusters\u003c/h3\u003e\n\u003cp\u003eWe previously showed that the secreted semaphorins Sema2a and Sema2b associate with endogenous cell-surface heparan sulfate chains and preferentially localize to highly sulfated NS clusters\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. To test whether transmembrane semaphorins follow the same pattern or instead preferentially associate with less sulfated NA clusters, we performed STED-based colocalization analyses of Sema1a, Sema1b, and their receptor PlexA using an NA-cluster\u0026ndash;specific antibody\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea) and compared these results to our NS-cluster analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea-d). For all three proteins, colocalization was consistently stronger with less sulfated NA clusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb\u0026ndash;d) than with NS clusters. For Sema1a, colocalization with NA clusters yielded a higher median PCC (0.457) compared with NS clusters (0.302) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee), together with a larger median fraction of Sema1a signal overlapping with NA clusters (M(Sema1a)\u0026thinsp;=\u0026thinsp;0.779 for NA versus 0.686 for NS) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef). This distinction was even more pronounced for Sema1b, which exhibited weak spatial association with NS clusters (median PCC\u0026thinsp;=\u0026thinsp;0.139; median M(Sema1b)\u0026thinsp;=\u0026thinsp;0.402) but substantially stronger overlap with NA clusters (median PCC\u0026thinsp;=\u0026thinsp;0.411; median M(Sema1b)\u0026thinsp;=\u0026thinsp;0.705) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee-f). PlexA displayed an intermediate behavior, showing moderate colocalization with NS clusters (median PCC\u0026thinsp;=\u0026thinsp;0.189; median M(PlexA)\u0026thinsp;=\u0026thinsp;0.527) and a clear increase in both metrics when analyzed against NA clusters (median PCC\u0026thinsp;=\u0026thinsp;0.351; median M(PlexA)\u0026thinsp;=\u0026thinsp;0.702) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee-f). For both cluster types, Manders\u0026rsquo; coefficients calculated for the semaphorins or PlexA signal were consistently higher than those calculated for the corresponding NA (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb-d) or NS cluster signal (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-d), indicating that while a substantial fraction of the protein signal resides within heparan sulfate-enriched regions, only a subset of NA or NS clusters is occupied by these proteins. Together, these data indicate that, in contrast to secreted semaphorins, transmembrane semaphorins and PlexA are more frequently associated with NA clusters than with discrete NS clusters on the same cell surface.\u003c/p\u003e\n\u003ch3\u003eGAG chains promote membrane clustering of Sema1a and PlexA\u003c/h3\u003e\n\u003cp\u003eTo investigate the lateral organization of Sema1a, Sema1b, and PlexA at the plasma membrane, we performed STED microscopy followed by spatial distribution analysis using Ripley\u0026rsquo;s K function on GAG-rich CHO-K1 and GAG-deficient CHO-PgsA745 cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea-d). Ripley\u0026rsquo;s K function quantifies spatial patterns by comparing the observed number of particles within a given radius to the number expected under complete spatial randomness\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. The L(r) \u0026ndash; r transformation of this function facilitates visual interpretation, where values above zero indicate clustering and values below zero indicate dispersion. The analysis revealed pronounced clustering of Sema1a in CHO-K1 cells, with higher L(r) \u0026ndash; r values in CHO-K1 cells compared to CHO-PgsA745 cells across a range of spatial scales, most prominently between 100 and 300 nm, and persisting up to ~\u0026thinsp;500 nm (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). This indicates that GAG chains promote Sema1a clustering at the membrane. In contrast, Sema1b exhibited nearly identical clustering profiles in both cell types, suggesting that its membrane distribution is independent of GAGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec). PlexA also showed consistently higher clustering in CHO-K1 cells across the 100\u0026ndash;500 nm range (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed), although the difference was moderate\u0026mdash;not as strong as for Sema1a\u0026mdash;indicating a partial dependence on GAGs for its spatial organization.\u003c/p\u003e\n\u003ch3\u003eGAG chains modulate the lateral mobility of semaphorins at the cell surface\u003c/h3\u003e\n\u003cp\u003eTo assess how GAG chains influence the lateral mobility of Sema1a, Sema1b, and their receptor PlexA at the cell surface, we performed fluorescence recovery after photobleaching (FRAP). GAG-rich CHO-K1 cells were transfected to express either Sema1a, Sema1b, or PlexA, with their cytoplasmic domains replaced by the fluorescent protein mClover. As a GAG-deficient control, we used CHO-PgsA745 cells expressing the same constructs. FRAP measurements were conducted by bleaching a circular region of 1.5 \u0026micro;m in diameter across the plasma membrane and monitoring fluorescence recovery within this area (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee-g, \u003cb\u003eFig. S2\u003c/b\u003e). For both Sema1a and Sema1b, effective diffusion coefficients were consistently higher on GAG-deficient CHO-PgsA745 cells compared to CHO-K1 cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh-i), indicating that GAG chains slow semaphorin diffusion, likely via interactions between their ectodomains and GAG chains. In the absence of GAG chains, these interactions are reduced or absent, resulting in increased mobility. Notably, Sema1b displayed systematically higher diffusion coefficients than Sema1a across both cell lines (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh-i), consistent with its lower GAG-binding affinity relative to Sema1a (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-b). In contrast, PlexA exhibited only a minor, albeit statistically significant, difference in diffusion between CHO-K1 and CHO-PgsA745 cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ej), suggesting that GAG chains exert a substantially weaker influence on PlexA mobility compared to the semaphorins.\u003c/p\u003e \u003cp\u003eTogether, these data demonstrate that GAG chains selectively modulate the lateral mobility of transmembrane semaphorins at the cell surface, whereas their effect on PlexA is limited.\u003c/p\u003e\n\u003ch3\u003eGAG chains on the cell surface enhance Plexin-mediated cell collapse\u003c/h3\u003e\n\u003cp\u003eTo investigate how GAG chains on the cell surface influence plexin-mediated cell collapse, we compared the extent of collapse between GAG-rich CHO-K1 cells and GAG-deficient CHO-PgsA745 cells. Although our initial aim was to examine Drosophila PlexA and Sema1a, we were unable to identify suitable Drosophila cell lines with defined differences in GAG chain biosynthesis analogous to those of CHO-K1 and CHO-PgsA745 cells. We therefore opted for their mammalian counterparts\u0026mdash;mouse PlxnD1 and Sema3E\u0026mdash;which are known to robustly induce collapse in neurons\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e and COS7 cells\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. To assess the impact of GAG chains on collapse, we transfected CHO-K1 and CHO-PgsA745 cells with a construct encoding PlxnD1 fused to the fluorescent protein mVenus. Live-cell imaging was performed following the addition of purified Sema3E, and individual cells were tracked over time (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). Cell collapse was quantified by measuring changes in cell surface area. In GAG-rich CHO-K1 cells, Sema3E induced pronounced collapse, with cells shrinking to approximately 65% of their original surface area after 150 minutes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). While this collapse was slightly less dramatic than previously reported for COS7 cells\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e\u0026mdash;likely due to morphological differences, as COS7 cells exhibit a more spread-out shape\u0026mdash;the response was still robust. In contrast, GAG-deficient CHO-PgsA745 cells exhibited a significantly attenuated response, retaining\u0026thinsp;~\u0026thinsp;90% of their original surface area after the same period (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). Taken together, these results suggest that GAG chains on the cell surface significantly enhance plexin-mediated cell collapse.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eHeparan sulfate proteoglycans are broadly implicated as cell-surface modulators of guidance-cue signaling, but the molecular basis by which they tune cell-surface receptor signaling remains incompletely understood. Our data identify heparan sulfate as a direct interaction partner and functional modulator of \u003cem\u003eDrosophila\u003c/em\u003e transmembrane semaphorins Sema1a and Sema1b and their shared receptor PlexA. Using SPR with monomeric ectodomains, we showed that Sema1a, Sema1b, and PlexA bind heparin and heparan sulfate with micromolar apparent affinities, while no binding to chondroitin sulfate was detected under the concentrations tested. This selectivity differs from several secreted semaphorins that have been reported to engage both heparan sulfate and chondroitin sulfate, including \u003cem\u003eDrosophila\u003c/em\u003e Sema2a and Sema2b\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e and vertebrate Sema3A\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Beyond GAG-class selectivity, our data reveal a difference between secreted and transmembrane semaphorins in their apparent preference for distinct heparan sulfate clusters. In particular, an early study showed that heparan sulfate modification is non-uniform along a single chain and is organized into highly sulfated N-sulfo\u0026ndash;rich NS clusters and less sulfated N-acetyl\u0026ndash;rich NA clusters\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Another study further indicated that the relative proportion (NA-to-NS ratio) and arrangement of NA and NS clusters can be proteoglycan-specific and vary across tissues\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, adding an additional layer of molecular patterning and positional information to cell-surface heparan sulfate. Such organization has been implicated in tuning morphogen signaling pathways, including Wnt\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e and BMP\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn our previous study, we found that secreted semaphorins (Sema2a and Sema2b) preferentially associate with NS clusters\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. By contrast, we show here that the transmembrane semaphorins Sema1a and Sema1b, together with PlexA, preferentially associate with NA clusters. This divergence is consistent with the distinct physical and biological constraints faced by secreted versus transmembrane semaphorins. In particular, in line with the semaphorin bridge model\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, secreted semaphorins are retained near their sites of production to limit dilution and ectopic diffusion. Preferential association with highly sulfated NS clusters, which have a higher negative charge density, would support efficient surface tethering and spatial confinement (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). By contrast, transmembrane semaphorins and PlexA are already membrane associated and therefore do not require high-affinity capture. Instead, their key challenge is to tune nanoscale organization and dynamics within the plane of the membrane. Preferential association with NA clusters is compatible with broad, reversible \u003cem\u003ecis\u003c/em\u003e interactions (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb) distributed across the cell surface\u0026mdash;interactions strong enough to influence clustering and lateral mobility, yet sufficiently dynamic to permit rapid reorganization during cell\u0026ndash;cell contact. An additional advantage of NA-specific association of transmembrane semaphorins is that it may reduce the risk of constitutive sequestration. Highly sulfated NS clusters could behave as high-affinity \u0026ldquo;traps\u0026rdquo; that overly immobilize membrane proteins and thereby dampen responsiveness to \u003cem\u003etrans\u003c/em\u003e cues. Preferential engagement of NA clusters by transmembrane semaphorins and PlexA may therefore provide a tunable interaction landscape that supports organizational control while preserving membrane dynamics and signaling flexibility.\u003c/p\u003e \u003cp\u003eBuilding on this NA-cluster-biased \u003cem\u003ecis\u003c/em\u003e engagement, three independent lines of evidence indicate that GAG chains actively shape the nanoscale organization, dynamics, and signaling output of membrane-anchored semaphorins and plexins. First, point-pattern analysis (Ripley\u0026rsquo;s K; L(r)\u0026ndash;r) shows that GAG chains promote nanoscale clustering of Sema1a and, to a lesser extent, PlexA, whereas Sema1b displays largely GAG-independent clustering. Second, FRAP measurements demonstrate that GAG chains reduce the lateral mobility of both Sema1a and Sema1b, with only a minor effect on PlexA diffusion. Together, these imaging-based readouts are consistent with multivalent, transient ectodomain\u0026ndash;GAG contacts that bias proteins into GAG-enriched membrane territories and reduce effective diffusion, thereby increasing local surface density and encounter probability without requiring strict one-to-one colocalization. This interpretation is further supported by recent structural and biochemical studies showing that heparan sulfate engagement promotes oligomerization of purified Sema5A ectodomains, highlighting a broader principle whereby GAG-mediated multivalency can drive higher-order assembly of transmembrane semaphorins\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThird, we link these membrane-organizational effects to functional signaling output. In a quantitative collapse assay, cell-surface GAG chains markedly enhanced plexin-dependent morphological collapse, with GAG-rich cells exhibiting substantially stronger collapse than GAG-deficient cells. This observation is consistent with prior work showing that exogenous heparin can potentiate Sema3A-induced growth cone collapse\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. More broadly, an essential role for cell-surface heparan sulfate has also been reported for other axon guidance systems; notably, Slit\u0026ndash;Robo signaling has been shown to be abrogated in the absence of heparan sulfate chains at the cell surface\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Together, these findings support a model in which heparan sulfate acts as a low-affinity, multivalent scaffold that increases the probability and lifetime of productive signaling complexes at the membrane\u0026mdash;through local enrichment, altered mobility, and extended effective ligand\u0026ndash;receptor engagement\u0026mdash;thereby lowering the threshold for plexin activation and amplifying downstream cytoskeletal remodeling.\u003c/p\u003e \u003cp\u003eMore broadly, our results align with a general role for cell-surface heparan sulfate proteoglycans as organizers that tune signaling by controlling local ligand\u0026ndash;receptor engagement and nanoscale membrane organization. Similar principles have been described in other pathways. For example, heparan sulfate can promote ligand retention and signaling efficiency in fibroblast growth factor signaling\u003csup\u003e\u003cspan additionalcitationids=\"CR25\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, support chemokine function through GAGs-dependent interactions (reviewed in\u003csup\u003e27\u003c/sup\u003e), and shape morphogen distribution and signaling\u0026mdash;such as Wnt\u0026mdash;through distinct NA/NS heparan sulfate domains\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Heparan sulfate has also been implicated in BMP signaling: recent work supports a cluster-based mechanism in which BMP4 preferentially associates with NS clusters, whereas the secreted antagonist Cerberus binds less sulfated NA-rich regions and broadens BMP4 distribution and signaling range by redistributing ligand across the cell surface\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHow might these heparan sulfate-dependent organizational effects fit with established semaphorin\u0026ndash;plexin signaling mechanisms? Canonically, semaphorin-mediated plexin activation depends on semaphorin-driven plexin dimerization\u003csup\u003e\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, which promotes activation of the plexin intracellular GAP domain\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, leading to downstream cytoskeletal remodeling. Here, we propose that heparan sulfate acts primarily upstream of signaling, serving as a membrane organizer that tunes the probability of productive plexin engagement.\u003c/p\u003e \u003cp\u003eWe therefore propose the following model (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec). First, in \u003cem\u003ecis\u003c/em\u003e, NA clusters within heparan sulfate provide a nanoscale organizational scaffold that enriches transmembrane semaphorins (most prominently Sema1a) and, to a lesser extent, PlexA within heparan sulfate\u0026ndash;defined membrane territories, while simultaneously reducing semaphorin mobility. This combination increases local surface density and dwell time. Second, these effects should promote \u003cem\u003etrans\u003c/em\u003e signaling at cell\u0026ndash;cell contacts by increasing the probability of productive semaphorin-driven plexin dimerization and activation. Third, under conditions of elevated semaphorin abundance, heparan sulfate can also support \u003cem\u003etrans\u003c/em\u003e adhesive bridging, as suggested by our aggregation assay. In addition, concentration within heparan sulfate-defined territories may also increase the likelihood of encounters with additional heparan sulfate-binding co-receptors\u0026mdash;such as Off-track (Otk), a Sema1a co-receptor\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e\u0026mdash;thereby promoting assembly of higher-order signaling complexes.\u003c/p\u003e \u003cp\u003eTogether, our results support a view in which heparan sulfate does not merely capture guidance cues but tunes the \u003cem\u003ecis\u003c/em\u003e membrane landscape\u0026mdash;organization, mobility, and encounter kinetics\u0026mdash;to set the threshold for plexin activation and downstream signaling. Future \u003cem\u003ein vivo\u003c/em\u003e tests will be required to validate this model and to determine which heparan sulfate contexts regulate \u003cem\u003ecis\u003c/em\u003e membrane organization versus \u003cem\u003etrans\u003c/em\u003e signaling efficiency.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eAccession Numbers and Sequence Versions Used\u003c/h2\u003e \u003cp\u003eThe following protein sequence versions were used in this study: \u003cem\u003eDrosophila melanogaster\u003c/em\u003e Sema1a (UniParc UPI0000252361; release date: 2004-03-15; formerly UniProt Q24322), Sema1b (UniProt Q7KK54; release date: 2006-10-03), PlexA (UniProt Q9V491; release date: 2000-05-01), mouse PlxnD1 (UniProt Q3UH93; release date: 2005-10-11, version 1), and mouse Sema3E (UniProt P70275; release date: 2011-07-27, version 3).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eProtein Production\u003c/h2\u003e \u003cp\u003eConstructs encoding the ectodomains of Sema1a (residues 21N\u0026ndash;606E; hereafter Sema1a\u003csub\u003eecto\u003c/sub\u003e) and Sema1b (residues 37D\u0026ndash;659S; Sema1b\u003csub\u003eecto\u003c/sub\u003e) were cloned into the pHR-CMV-TetO2-IRES-EmGFP vector, in-frame with a C-terminal 3C-Avi-His₆ tag. Stable cell lines were generated by lentiviral transduction\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Sema1a\u003csub\u003eecto\u003c/sub\u003e was produced in HEK293T cells (ATCC CRL-3216) cultured in DMEM with 2% FBS at 30\u0026deg;C, while Sema1b\u003csub\u003eecto\u003c/sub\u003e was expressed in HEK293S GnTI⁻ cells (ATCC CRL-3022) maintained in FreeStyle 293 Expression Medium at 37\u0026deg;C. Conditioned media were buffer-exchanged and proteins were purified by immobilized metal affinity chromatography (HisTrap FF, Cytiva), followed by size-exclusion chromatography using a HiLoad 16/600 Superdex 200 pg column (Cytiva). The ectodomain of PlexA (PlexA\u003csub\u003eecto\u003c/sub\u003e, residues 28\u0026ndash;1272) was produced and purified as described previously\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Mouse Sema3E (residues 26A\u0026ndash;756E) was produced in fusion with an Fc tag. The Sema3E-Fc construct was cloned into the pHR-CMV-TetO2-IRES-EmGFP vector, with a C-terminal 3C-Avi-His₆ tag. Stable expression was established in Expi293F cells (Thermo Fisher Scientific, Cat. No. A14527) maintained in Expi293 Expression Medium at 37\u0026deg;C. The purification protocol followed the same approach as that used for Sema1a\u003csub\u003eecto\u003c/sub\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eSurface plasmon resonance (SPR) binding experiments\u003c/h2\u003e \u003cp\u003eSPR binding experiments were performed as described in our previous study\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. In brief, glycosaminoglycans (heparin, heparan sulfate, and chondroitin sulfate; Toronto Research Chemicals) were biotinylated and immobilized on SA sensor chips (Cytiva). Biotinylated Dlp constructs were produced in HEK293T cells by co-transfection with pDisplay-BirA-ER and cultured in the presence of 100 \u0026micro;M D-biotin. Three days post-transfection, conditioned media were collected, dialyzed, and used for immobilization on SA sensor chips. SPR measurements were carried out on a Biacore S200 instrument (Cytiva) at 25\u0026deg;C in a running buffer containing 15 mM Tris (pH 7.5), 150 mM NaCl, and 0.05% (v/v) Tween-20. The surface was regenerated between runs using a buffer containing 0.1 M Tris (pH 8.0), 1.0 M NaCl, and 1% CHAPS. Data were analyzed using Biacore Insight Evaluation software (Cytiva), and each experiment was performed in two to three biological replicates.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eImmunofluorescence\u003c/h2\u003e \u003cp\u003eAdherent CHO-K1 and CHO-PgsA745 cells were seeded on 18 mm high-precision coverslips (15,000 or 25,000 cells per coverslip) in DMEM supplemented with 2% FBS. After 24 h of growth, cells were transfected with constructs encoding Sema1a, Sema1b, or PlexA using either GeneJuice or Lipofectamine 3000 transfection reagent. The constructs were generated by cloning DNAs encoding \u003cem\u003eDrosophila melanogaster\u003c/em\u003e Sema1a (residues 21\u0026ndash;633), Sema1b (residues 37\u0026ndash;686), or PlexA (residues 28\u0026ndash;1311) into the pHLsec vector in-frame with C-terminal fluorescent proteins mVenus or mClover, respectively. Forty-eight hours post-transfection, cells were washed twice with PBS and fixed with 3% paraformaldehyde in PBS for 10 min. After two 5-min PBS washes, cells were permeabilized with 0.1% saponin in PBS for 20 min, followed by three additional 5-min PBS washes. Blocking was performed in 4% BSA in PBS for 1 h. For the clustering assay, cells were incubated with 10 \u0026micro;g/ml anti-GFP antibody (Abcam, ab290; rabbit polyclonal) in PBS containing 0.3% BSA and 0.05% saponin for 1 h. For the colocalization assay, cells were incubated with anti-GFP antibody together with either 1 \u0026micro;g/ml anti-N-sulfated heparan sulfate antibody (Amsbio, 370-255-5; mouse IgM, clone F58-10E4) or anti-N-acetylated heparan sulfate antibody (1:50 dilution; kindly provided by Prof. Masanori Taira and Dr. Yusuke Mii, The University of Tokyo) under the same buffer conditions. After extensive washing, secondary detection was performed for 1 h in PBS with 0.3% BSA and 0.05% saponin using 3.3 \u0026micro;g/ml goat anti-rabbit IgG antibody conjugated with Abberior STAR 635P (clustering assay) or together with 5 \u0026micro;g/ml goat anti-mouse IgM antibody conjugated with Alexa Fluor 594 (colocalization assay). Finally, cells were washed five times for 5 min each in PBS, rinsed thoroughly with ultrapure water, air-dried, and mounted in VectaShield.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eSTED Microscopy\u003c/h2\u003e \u003cp\u003eSuper-resolution images were acquired using an Abberior STED microscope equipped with a 775 nm pulsed STED laser operating at 40 MHz, controlled by Imspector software (version 16.3). A Nikon CFI Plan Apo Lambda oil immersion objective (60\u0026times;, NA 1.40) was used for image acquisition, focusing on the most well-spread region of the cell membrane. Images were deconvolved using Huygens Professional (version 22.10) with the Classic Maximum Likelihood Estimation (CMLE) algorithm. For the clustering assay, protein distribution within the plasma membrane was analyzed using custom ImageJ\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e macros and R scripts (version 4.3.0); code is available upon request. Briefly, the most extended region of the cell membrane was manually masked, and a 2 \u0026times; 2 \u0026micro;m\u0026sup2; region within this mask was selected for analysis. Automatic thresholding was applied using the Otsu algorithm, and particles of \u0026ge;\u0026thinsp;2 pixels in size were detected based on an intensity prominence threshold of 3000 using the \u003cem\u003eFind Maxima function\u003c/em\u003e in ImageJ. For each detected particle, the maximum intensity was identified, and a binary image of particle maxima was generated. To assess spatial distribution, the L(r) function\u0026mdash;a linear transformation of Ripley\u0026rsquo;s K(r)\u0026mdash;was computed the spatstat library (version 3.2.1), comparing the observed distribution within a given radius \u0026ldquo;r\u0026rdquo; to the expected distribution under spatial randomness (based on 25 Monte Carlo simulations). Clustering was inferred when L(r) \u0026ndash; r\u0026thinsp;\u0026gt;\u0026thinsp;0, and dispersion when L(r) \u0026ndash; r\u0026thinsp;\u0026lt;\u0026thinsp;0. Statistical significance of protein clustering was assessed using two-way ANOVA in GraphPad Prism (version 10.5.0). For the colocalization assay, colocalization within the most well-spread region of the plasma membrane was quantified using the Coloc2 plugin in ImageJ (version 1.54p). Significance of colocalization was evaluated using the Wilcoxon test in GraphPad Prism (version 10.5.0).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eClumping Assay\u003c/h2\u003e \u003cp\u003eThe construct encoding the extracellular and transmembrane domains of \u003cem\u003eDrosophila melanogaster\u003c/em\u003e Sema1a (residues 21\u0026ndash;633) was cloned into the pHR-CMV-TetO2 vector in-frame with a C-terminal mClover. Stable CHO-K1 and CHO-PgsA745 cell lines expressing Sema1a-mClover were generated by lentiviral transduction\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Subsequently, the adherent cultures were adapted to suspension growth and maintained in phenol red-free DMEM supplemented with 10% FBS, under continuous shaking at 130 rpm and 8% CO₂. For the clumping assay, suspension cells were transferred into \u0026micro;-Slide 8 Well chambers (200,000 cells per well) in phenol red-free DMEM supplemented with 10% FBS. Immediately after cell transfer, a large-field image of each well (5 \u0026times; 5 fields, 1% overlap; total area 3439.21 \u0026times; 3439.21 \u0026micro;m\u0026sup2;) was acquired and merged using a Nikon CSU-W1 spinning disk confocal microscope equipped with a CFI Plan Apo VC objective (20\u0026times;, NA 0.75), an environmental chamber (37\u0026deg;C, 5% CO₂), and NIS-Elements AR software (version 5.41.01).\u003c/p\u003e \u003cp\u003eCell clumping was analyzed using a combination of NIS-Elements AR, ImageJ\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, Cellpose (version 3.0.7) (ref\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e), and custom Python scripts. Prior to segmentation, images were resized to 50% using averaging. The DIC channel was preprocessed using the Homogenization function in NIS-Elements (resolution degree 10) to reduce illumination inhomogeneity and improve contrast. Cells and clumps were segmented using a custom-trained Cellpose model. The model was trained on 26 image patches (512 \u0026times; 512 pixels), comprising 2749 manually annotated objects. Training was performed on the DIC channel for 100 epochs with a learning rate of 0.1 and weight decay of 0.0001, using \u0026lsquo;cyto3\u0026rsquo; as the initial model. Segmented objects touching image borders were removed using the scikit-image library (version 0.25.2) in Python (version 3.10.19). For each clump, area and mean fluorescence intensity were quantified after background subtraction. Background intensity was estimated as the mean intensity of regions outside segmented objects. Objects with mean fluorescence intensity above a threshold of 80 were retained for analysis. The threshold was determined empirically from the distribution of mean object intensities pooled across all images, corresponding to the transition region between the dominant low-intensity negative population and the positive signal tail. Statistical significance was assessed using an unpaired \u003cem\u003et\u003c/em\u003e-test in GraphPad Prism (version 10.5.0). All experiments were performed in quadruplicate.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eFRAP (Fluorescence Recovery After Photobleaching)\u003c/h2\u003e \u003cp\u003eAdherent CHO-K1 and CHO-PgsA745 cells were seeded in \u0026micro;-Slide 8 Well chambers (40,000 cells per well) in phenol red-free DMEM supplemented with 2% FBS. After 24 h of growth, cells were transfected with constructs encoding Sema1a, Sema1b, or PlexA fused to mClover using either GeneJuice or Lipofectamine 3000 transfection reagents. Construct details are provided in the Immunofluorescence section. Forty-eight hours post-transfection, FRAP experiments were performed using a Nikon CSU-W1 spinning disk confocal microscope equipped with a CF Plan Apo VC water immersion objective (60\u0026times;, NA 1.20), an environmental chamber (37\u0026deg;C, 5% CO₂), and controlled by NIS-Elements AR software (version 5.41.01). Imaging was focused on the most well-spread region of the plasma membrane. Fluorescence of mClover was excited using a 488 nm laser at 5% power for imaging and 7% for photobleaching. A circular region of interest (ROI) with a 1.5 \u0026micro;m diameter was selected for photobleaching. Pre-bleach acquisition was conducted for 5 s at 500 ms intervals. Photobleaching was performed over 200 ms, followed by post-bleach acquisition for 2 min at 500 ms intervals. FRAP data were analyzed using custom Matlab (R2024a) scripts (available upon request). Briefly, circular ROIs matching the size of the photobleached area were placed in unbleached membrane regions (reference) and in areas without fluorescence signal (background). Fluorescence intensity values were normalized per cell, and recovery curves were fitted globally with shared parameters using nonlinear least-squares solver \u003cem\u003elsqcurvefit\u003c/em\u003e. Effective diffusion coefficients were determined from the recovery curves as described previously\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Statistical significance of protein dynamics was assessed using an unpaired \u003cem\u003et\u003c/em\u003e-test in GraphPad Prism (version 10.5.0).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eLive-Cell Imaging of Sema3E-Induced Collapse of CHO Cells\u003c/h2\u003e \u003cp\u003eAdherent CHO-K1 and CHO-PgsA745 cells were cultured in \u0026micro;-Slide 8 Well chambers (10,000 or 20,000 cells per well) in phenol red-free DMEM supplemented with 2% FBS. After 24 hours of growth, cells were transfected with a construct encoding mouse PlxnD1 fused to mClover, using either GeneJuice or Lipofectamine 3000 transfection reagents. This construct was generated by subcloning mouse PlxnD1 (residues 49L\u0026ndash;1925A) into the pHLsec vector in-frame with a C-terminal mVenus fluorescent protein. After 48 hours post-transfection, the medium was aspirated and cells were washed twice with DMEM. For the collapse assay, either vehicle (medium) or purified Sema3E was added to the cultures at a final concentration of 0.5 \u0026micro;M. Over the following 150 minutes, large-field images of each well (5 \u0026times; 5 fields, 1% overlap; total area 3439.21 \u0026times; 3439.21 \u0026micro;m\u0026sup2;) were acquired every 30 minutes and merged. Imaging was performed using a Nikon CSU-W1 spinning disk confocal microscope equipped with a CFI Plan Apo VC objective (20\u0026times;, NA 0.75), an environmental chamber (37\u0026deg;C, 5% CO₂), and NIS-Elements AR software (version 5.41.01). Cell collapse was analyzed using NIS-Elements AR, ImageJ\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, Cellpose (version 3.0.7) (ref\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e), and custom Python scripts. Briefly, the original images were resized to 50% using averaging and aligned across all time points in ImageJ using the BigWarp plugin (version 9.3.1). Time-lapse stacks were then segmented with a custom Cellpose model and analyzed for cell surface area over time with tracking using the TrackMate plugin\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e (version 7.14.0). The model was trained on 19 image patches (512 \u0026times; 512 pixels), comprising 610 manually annotated objects. Training was performed on the DIC channel for 100 epochs with a learning rate of 0.1 and weight decay of 0.0001, using \u0026lsquo;cyto3\u0026rsquo; as the initial model. Object tracking was performed using a Simple LAP tracker with a linking distance of 20 pixels, a gap-closing distance of 5 pixels, and a maximum frame gap of 2 frames. Only tracks containing bright cells with mean fluorescence intensity above a background threshold of 200 in at least 3 frames and starting before frame 2 were retained for collapse analysis. Statistical significance of cell collapse was evaluated using two-way ANOVA in GraphPad Prism (version 10.5.0). All experiments were conducted in triplicate.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare that they have no conflict of interest.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eM.S., and D.R. designed research; M.S., F.N., K.K., J.V., N.K., and D.R. performed research; M.S., Z.C., M.B., and D.R. analyzed data; and M.S., and D.R. wrote the paper.\u003c/p\u003e\u003ch2\u003eAcknowledgments\u003c/h2\u003e \u003cp\u003eWe thank Prof Masanori Taira and Dr Yusuke Mii from the University of Tokyo for their generous gift of the anti-N-acetylated heparan sulfate antibody. This work was supported by the Czech Science Foundation (Junior Star Grant 21-27204M). Computational resources were provided by the e-INFRA CZ project (ID:90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic. We acknowledge the support and access to the Core Facility of Biomolecular Interactions and Crystallography at Czech Infrastructure for Integrative Structural Biology and Instruct-CZ Centre, supported by the Ministry of Education, Youth and Sports of the Czech Republic (LM2023042) and European Regional Development Fund-Project (No. Z.02.1.01/0.0/0.0/18_046/0015974). We also appreciate the support and assistance of the Imaging Methods Core Facility of Charles University and the Light Microscopy Core Facility of the Institute of Molecular Genetics, both funded by the Ministry of Education, Youth and Sports of the Czech Republic (LM2023050 Czech-BioImaging).\u003c/p\u003e\u003ch2\u003eData availability statement\u003c/h2\u003e \u003cp\u003eAll data supporting the findings of this study are available within the paper and its Supplementary Information. All original data used to draw the graphs for SPR assay, colocalization assay, clustering assay, FRAP assay, collapse assay as well as the clustering assay code and FRAP scripts will be available in Figshare.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eDickson BJ (2002) Molecular mechanisms of axon guidance. Science 298:1959\u0026ndash;1964. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.1072165\u003c/span\u003e\u003cspan address=\"10.1126/science.1072165\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaccarana M, Sakura Y, Tawada A, Yoshida K, Lindahl U (1996) Domain structure of heparan sulfates from bovine organs. J Biol Chem 271:17804\u0026ndash;17810. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/jbc.271.30.17804\u003c/span\u003e\u003cspan address=\"10.1074/jbc.271.30.17804\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSarrazin S, Lamanna WC, Esko JD (2011) Heparan sulfate proteoglycans. Cold Spring Harb Perspect Biol 3. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/cshperspect.a004952\u003c/span\u003e\u003cspan address=\"10.1101/cshperspect.a004952\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMasu M (2016) Proteoglycans and axon guidance: a new relationship between old partners. J Neurochem 139:58\u0026ndash;75. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jnc.13508\u003c/span\u003e\u003cspan address=\"10.1111/jnc.13508\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFawcett JW, Kwok JCF (2022) Proteoglycan Sulphation in the Function of the Mature Central Nervous System. Front Integr Neurosci 16:895493. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fnint.2022.895493\u003c/span\u003e\u003cspan address=\"10.3389/fnint.2022.895493\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePasterkamp RJ (2012) Getting neural circuits into shape with semaphorins. Nat Rev Neurosci 13:605\u0026ndash;618. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nrn3302\u003c/span\u003e\u003cspan address=\"10.1038/nrn3302\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKantor DB et al (2004) Semaphorin 5A is a bifunctional axon guidance cue regulated by heparan and chondroitin sulfate proteoglycans. Neuron 44:961\u0026ndash;975. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.neuron.2004.12.002\u003c/span\u003e\u003cspan address=\"10.1016/j.neuron.2004.12.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCho JY, Chak K, Andreone BJ, Wooley JR, Kolodkin AL (2012) The extracellular matrix proteoglycan perlecan facilitates transmembrane semaphorin-mediated repulsive guidance. Genes Dev 26:2222\u0026ndash;2235. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/gad.193136.112\u003c/span\u003e\u003cspan address=\"10.1101/gad.193136.112\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Wit J, De Winter F, Klooster J, Verhaagen J (2005) Semaphorin 3A displays a punctate distribution on the surface of neuronal cells and interacts with proteoglycans in the extracellular matrix. Mol Cell Neurosci 29:40\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.mcn.2004.12.009\u003c/span\u003e\u003cspan address=\"10.1016/j.mcn.2004.12.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZimmer G et al (2010) Chondroitin sulfate acts in concert with semaphorin 3A to guide tangential migration of cortical interneurons in the ventral telencephalon. Cereb Cortex 20:2411\u0026ndash;2422. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/cercor/bhp309\u003c/span\u003e\u003cspan address=\"10.1093/cercor/bhp309\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVo T et al (2013) The chemorepulsive axon guidance protein semaphorin3A is a constituent of perineuronal nets in the adult rodent brain. Mol Cell Neurosci 56:186\u0026ndash;200. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.mcn.2013.04.009\u003c/span\u003e\u003cspan address=\"10.1016/j.mcn.2013.04.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNourisanami F et al (2024) Molecular mechanisms of proteoglycan-mediated semaphorin signaling in axon guidance. Proc Natl Acad Sci U S A 121:e2402755121. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.2402755121\u003c/span\u003e\u003cspan address=\"10.1073/pnas.2402755121\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRozbesky D et al (2019) Diversity of oligomerization in Drosophila semaphorins suggests a mechanism of functional fine-tuning. Nat Commun 10:3691. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-019-11683-y\u003c/span\u003e\u003cspan address=\"10.1038/s41467-019-11683-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavid G, Bai XM, Van der Schueren B, Cassiman JJ, Van den Berghe H (1992) Developmental changes in heparan sulfate expression: in situ detection with mAbs. J Cell Biol 119:961\u0026ndash;975. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1083/jcb.119.4.961\u003c/span\u003e\u003cspan address=\"10.1083/jcb.119.4.961\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMii Y et al (1973) (2017) Roles of two types of heparan sulfate clusters in Wnt distribution and signaling in Xenopus. \u003cem\u003eNat Commun\u003c/em\u003e 8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-017-02076-0\u003c/span\u003e\u003cspan address=\"10.1038/s41467-017-02076-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRipley BD (1979) Tests of 'Randomness' for Spatial Point Patterns. J Royal Stat Soc Ser B: Stat Methodol 41:368\u0026ndash;374. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.2517-6161.1979.tb01091.x\u003c/span\u003e\u003cspan address=\"10.1111/j.2517-6161.1979.tb01091.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurk K et al (2017) Post-endocytic sorting of Plexin-D1 controls signal transduction and development of axonal and vascular circuits. Nat Commun 8:14508. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/ncomms14508\u003c/span\u003e\u003cspan address=\"10.1038/ncomms14508\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu C et al (2005) Semaphorin 3E and plexin-D1 control vascular pattern independently of neuropilins. Science 307:265\u0026ndash;268. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.1105416\u003c/span\u003e\u003cspan address=\"10.1126/science.1105416\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMehta V et al (2020) The guidance receptor plexin D1 is a mechanosensor in endothelial cells. Nature 578:290\u0026ndash;295. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41586-020-1979-4\u003c/span\u003e\u003cspan address=\"10.1038/s41586-020-1979-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDick G et al (2013) Semaphorin 3A binds to the perineuronal nets via chondroitin sulfate type E motifs in rodent brains. J Biol Chem 288:27384\u0026ndash;27395. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/jbc.M111.310029\u003c/span\u003e\u003cspan address=\"10.1074/jbc.M111.310029\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYamamoto T, Mii Y, Saiga H, Michiue T, Taira M (2025) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2025.04.16.649141\u003c/span\u003e\u003cspan address=\"10.1101/2025.04.16.649141\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNagy GN et al (2024) Structure and function of Semaphorin-5A glycosaminoglycan interactions. Nat Commun 15:2723. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-024-46725-7\u003c/span\u003e\u003cspan address=\"10.1038/s41467-024-46725-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHussain SA et al (2006) A molecular mechanism for the heparan sulfate dependence of slit-robo signaling. J Biol Chem 281:39693\u0026ndash;39698. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/jbc.M609384200\u003c/span\u003e\u003cspan address=\"10.1074/jbc.M609384200\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuchesne L et al (2012) Transport of fibroblast growth factor 2 in the pericellular matrix is controlled by the spatial distribution of its binding sites in heparan sulfate. PLoS Biol 10:e1001361. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pbio.1001361\u003c/span\u003e\u003cspan address=\"10.1371/journal.pbio.1001361\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXue S et al (2022) Phase separation on cell surface facilitates bFGF signal transduction with heparan sulphate. Nat Commun 13:1112. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-022-28765-z\u003c/span\u003e\u003cspan address=\"10.1038/s41467-022-28765-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenero Galanternik M, Kramer KL, Piotrowski T (2015) Heparan Sulfate Proteoglycans Regulate Fgf Signaling and Cell Polarity during Collective Cell Migration. Cell Rep 10:414\u0026ndash;428. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.celrep.2014.12.043\u003c/span\u003e\u003cspan address=\"10.1016/j.celrep.2014.12.043\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCambier S, Gouwy M, Proost P (2023) The chemokines CXCL8 and CXCL12: molecular and functional properties, role in disease and efforts towards pharmacological intervention. Cell Mol Immunol 20:217\u0026ndash;251. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41423-023-00974-6\u003c/span\u003e\u003cspan address=\"10.1038/s41423-023-00974-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanssen BJ et al (2010) Structural basis of semaphorin-plexin signalling. Nature 467:1118\u0026ndash;1122. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nature09468\u003c/span\u003e\u003cspan address=\"10.1038/nature09468\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKong Y et al (2016) Structural Basis for Plexin Activation and Regulation. Neuron 91:548\u0026ndash;560. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.neuron.2016.06.018\u003c/span\u003e\u003cspan address=\"10.1016/j.neuron.2016.06.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu D, Shang G, He X, Bai XC, Zhang X (2021) Architecture of the Sema3A/PlexinA4/Neuropilin tripartite complex. Nat Commun 12:3172. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-021-23541-x\u003c/span\u003e\u003cspan address=\"10.1038/s41467-021-23541-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y, Pascoe HG, Brautigam CA, He H, Zhang X (2013) Structural basis for activation and non-canonical catalysis of the Rap GTPase activating protein domain of plexin. Elife 2:e01279. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7554/eLife.01279\u003c/span\u003e\u003cspan address=\"10.7554/eLife.01279\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRozbesky D et al (2020) Drosophila OTK Is a Glycosaminoglycan-Binding Protein with High Conformational Flexibility. \u003cem\u003eStructure\u003c/em\u003e 28, 507\u0026ndash;515 e505 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.str.2020.02.008\u003c/span\u003e\u003cspan address=\"10.1016/j.str.2020.02.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElegheert J et al (2018) Lentiviral transduction of mammalian cells for fast, scalable and high-level production of soluble and membrane proteins. Nat Protoc 13:2991\u0026ndash;3017. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41596-018-0075-9\u003c/span\u003e\u003cspan address=\"10.1038/s41596-018-0075-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRozbesky D et al (2020) Structural basis of semaphorin-plexin cis interaction. EMBO J 39:e102926. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.15252/embj.2019102926\u003c/span\u003e\u003cspan address=\"10.15252/embj.2019102926\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchindelin J et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676\u0026ndash;682. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nmeth.2019\u003c/span\u003e\u003cspan address=\"10.1038/nmeth.2019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStringer C, Wang T, Michaelos M, Pachitariu M (2020) Cellpose: a generalist algorithm for cellular segmentation. Nat Methods 18:100\u0026ndash;106. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41592-020-01018-x\u003c/span\u003e\u003cspan address=\"10.1038/s41592-020-01018-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoumpasis DM (1983) Theoretical analysis of fluorescence photobleaching recovery experiments. Biophys J 41:95\u0026ndash;97. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0006-3495(83)84410-5\u003c/span\u003e\u003cspan address=\"10.1016/S0006-3495(83)84410-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eErshov D et al (2022) TrackMate 7: integrating state-of-the-art segmentation algorithms into tracking pipelines. Nat Methods 19:829\u0026ndash;832. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41592-022-01507-1\u003c/span\u003e\u003cspan address=\"10.1038/s41592-022-01507-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"cellular-and-molecular-life-sciences","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"life","sideBox":"Learn more about [Cellular and Molecular Life Sciences](https://link.springer.com/journal/18)","snPcode":"18","submissionUrl":"https://www.editorialmanager.com/life/default2.aspx","title":"Cellular and Molecular Life Sciences","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Open","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9405124/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9405124/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eHeparan sulfate proteoglycans are cell-surface and extracellular-matrix components that modulate the activity of diverse signaling pathways. A hallmark of heparan sulfate is its domain organization into less sulfated N\u0026ndash;acetyl\u0026ndash;rich and highly sulfated N\u0026ndash;sulfo\u0026ndash;rich regions. In the nervous system, semaphorins and their plexin receptors comprise a major axon guidance system, and a growing body of evidence has implicated proteoglycans as key modulators of their function. However, how heparan sulfate domain organization mechanistically tunes semaphorin\u0026ndash;plexin signaling at the cell surface remains unclear. Here, we show that heparan sulfate directly organizes transmembrane semaphorin\u0026ndash;plexin signaling at the plasma membrane. The Drosophila transmembrane semaphorins Sema1a and Sema1b, together with their receptor PlexA, bind heparan sulfate and preferentially associate with less sulfated N-acetyl\u0026ndash;rich domains, in contrast to previously reported secreted semaphorins that favor highly sulfated N-sulfo\u0026ndash;rich clusters. The N\u0026ndash;acetyl\u0026ndash;rich-domain engagement promotes nanoscale clustering of Sema1a and reduces its lateral mobility, consistent with increased local surface density and longer dwell time. These membrane-organizational effects enhance plexin-dependent signaling output, as demonstrated in a collapse assay, in which cell-surface glycosaminoglycans potentiated ligand-induced cell collapse. Together, our findings reveal heparan sulfate as an active organizer of the local plasma membrane microenvironment for transmembrane semaphorins, defining a distinct mode of glycosaminoglycan-dependent regulation that controls protein distribution, encounter probability, and signaling thresholds.\u003c/p\u003e","manuscriptTitle":"Heparan sulfate organizes transmembrane semaphorins at the cell surface to tune plexin signaling","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-04 18:40:29","doi":"10.21203/rs.3.rs-9405124/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2026-04-26T23:14:47+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-23T18:04:15+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-17T13:31:05+00:00","index":"","fulltext":""},{"type":"submitted","content":"Cellular and Molecular Life Sciences","date":"2026-04-15T10:49:42+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"cellular-and-molecular-life-sciences","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"life","sideBox":"Learn more about [Cellular and Molecular Life Sciences](https://link.springer.com/journal/18)","snPcode":"18","submissionUrl":"https://www.editorialmanager.com/life/default2.aspx","title":"Cellular and Molecular Life Sciences","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Open","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"e6af2bbe-0f72-4098-988a-54c6c4079a0c","owner":[],"postedDate":"May 4th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-04T18:40:29+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-04 18:40:29","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9405124","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9405124","identity":"rs-9405124","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}