Holdfast adhesion in the kelp Alaria marginata: Cell wall polysaccharides and phenolics Authors: Maisie R. Roy‐Musor, Christa Agoawike, A. Lacey Samuels, Patrick T. Martone Published: 2026 ABSTRACT AbstractAdhesion is a critical process in the survival of seaweeds, especially in wave‐swept environments. Kelp (Laminariales) sporophytes secrete adhesives from specialized rhizoid cells to anchor themselves to the substrate and withstand immense forces applied by waves. Although the composition of macroalgal adhesive has been described primarily in the Fucales, the current adhesive model is incomplete, and similarities between kelp and fucoid adhesives are unknown. Moreover, the cellular organization of the attachment interface of kelps is poorly understood. In this study, we examined the composition of rhizoid cells and adhesive from developing holdfasts of the kelp Alaria marginata using brightfield, fluorescence, and scanning electron microscopy. Rhizoid cells form a thick, intertwined mat cemented with adhesive. We provide the highest resolution immunofluorescent analysis of kelp adhesives to date using confocal microscopy: Kelp adhesives are composed of fucose‐containing sulfated polysaccharides (FCSPs) and alginate, dominated by BAM3 and BAM4 epitopes. Holdfast adhesives are biochemical extensions of the cell wall with distinct composition from the wall proper. This work advances our basic understanding of kelp adhesion and highlights gaps in our current understanding of adhesive composition, raising questions, for example, about the roles of phenolics and proteins. Biochemical and histological methods presented here set the stage for future examination of the developmental, ecological, and evolutionary dynamics of adhesion. Abbreviations Attenuated Total Reflectance Fourier Transform Infrared gold/palladium brown algal monoclonal confocal laser scanning microscope Centre National de la Recherche Scientifique differential interference contrast electron microscopy sciences fucose‐containing sulfated polysaccharide 1,4‐α‐D‐guluronic acid monoclonal antibody mixed block of 1,4‐β‐D‐mannuronic acid and 1,4‐α‐D‐guluronic acid 1,4‐β‐D‐mannuronic acid numerical aperture non‐fat dry milk Natural Sciences and Engineering Research Council scanning electron microscopy tris‐buffered saline with tween INTRODUCTION Kelp are habitat‐forming, keystone species that play myriad ecological roles in temperate and arctic coastal ecosystems (e.g., Teagle et al., 2017). In order for kelps to provide critical habitat, they must first successfully recruit as spores, then as germinating gametophytes, and finally as developing sporophytes, forming a fast and secure attachment by secreting adhesive onto the substratum (Fletcher & Callow, 1992). Dislodgement in wave‐swept habitats is often a cause of early mortality in seaweeds, selecting for stronger and quicker adhesion (Taylor & Schiel, 2003; Vadas et al., 1990) and suggesting that adhesive composition and attachment traits may help inform ecological patterns in recruitment. Additionally, there is some evidence that adhesion in some seaweeds may be delayed or weakened by ocean acidification (Guenther et al., 2018), suggesting that clarifying adhesive composition may improve our ability to predict dislodgement in future oceans. An improved understanding of seaweed adhesion can also inform various industry applications, such as the design of anti‐biofouling coatings (Petrone, 2013), novel bioinspired marine adhesives (Li & Zeng, 2016), and substrate selection in commercial aquaculture to minimize seaweed dislodgement (Kerrison et al., 2019). Despite the broad ecological and industrial importance of adhesion in seaweed, little is known about the composition or deposition of kelp adhesives. Most brown algal adhesion research has been carried out on species in the Fucales due to their utility as a model system for developmental biology (Hatchett et al., 2022; Tarakhovskaya, 2014). Zygote adhesion in fucoids is described in two steps: “Primary adhesion,” when non‐covalently bound gelling polysaccharides are secreted and form a weak association with the substratum, and “secondary adhesion,” when there is irreversible covalent cross‐linking of the gel matrix by another component that immobilizes the gel and forms strong associations with the substratum (Vreeland et al., 1998). It can be argued that these are not separate phases but a maturation of a calcium‐alginate gel by the addition of polyphenols (Tarakhovskaya, 2014). In brown algae, the cell wall contains cellulose as well as the matrix polysaccharides alginate and fucose‐containing sulfated polysaccharides (FCSPs; reviewed by Beuder & Braybrook, 2023). Alginate, the dominant matrix polysaccharide in brown algal cell walls, is composed of 1,4‐β‐D‐mannuronic acid (MM) and 1,4‐α‐D‐guluronic acid (GG). Alginate is organized as a linear block copolymer, consisting of homogenous MM and GG blocks and mixed MG blocks of both residues along any given polysaccharide chain (Haug et al., 1967). Alginate ionically cross‐links and gels through chelation of Ca2+ and Mg2+ by carboxylate groups in adjacent guluronic acid residues (Draget et al., 1994; Smidsrød, 1974), conferring cohesive strength. The FCSPs are a compositionally diverse group of gelling matrix and wall polysaccharides in brown algae consisting of polysaccharides containing α‐L‐fucose residues that are often highly sulfated (Mabeau et al., 1990). The FCSPs are associated with several structural and physiological roles in brown algae, such as cross‐linking cellulose fibers and regulating osmotic stress (Deniaud‐Bouët et al., 2017), but direct evidence for FCSP in adhesion is limited (Beuder & Braybrook, 2023; Forbes & Hallam, 1979; Tarakhovskaya, 2014). Cross‐linking of fucoid adhesive is thought to occur by deposition of phlorotannins, a highly branched phenolic polymer of phloroglucinol (1,3,5‐hydroxyphenol; McInnes et al., 1985; Lemesheva et al., 2020). In cell walls, phlorotannins may tune mechanical properties (Imbs & Zvyagintseva, 2018) by cross‐linking to alginate through radical oxidation by vanadium‐dependent haloperoxidases (Bitton et al., 2006; Lemesheva et al., 2020). Alternatively, or in combination with the phenolics, proteins have also been localized in fucoid adhesives (Dimartino et al., 2016; Wagner et al., 1992), and other marine adhesive systems such as diatoms and mussels utilize various proteins as complex structural components of their adhesives (Poulsen et al., 2014; Waite, 2017; Zackova Suchanova et al., 2023). Despite several uncertainties, our understanding of fucoid adhesion provides a starting place to consider kelp. There is less information available about the chemical composition of Laminarian kelp adhesives. Transmission electron microscopy of spores of the kelps Saccharina latissima and Nereocystis luetkeana showed membrane‐bound vesicles containing electron‐dense adhesive plaques composed of anionic polysaccharides, with granules of proteins around the periphery of the plaque (Oliveira et al., 1980). Attenuated Total Reflectance Fourier Transform Infrared (ATR‐FTIR) spectroscopy showed that proteins and sulfated moieties were secreted in the adhesive of the kelp Undaria pinnatifida (Petrone et al., 2010). Given our understanding of adhesives in fucoids, it is likely that polysaccharides in the gel matrix are either composed of alginate, FCSPs, or both. If phenolics, such as phlorotannins, do indeed play important roles in fucoid adhesion and we assume the fucoid model is relevant to kelp, then perhaps phenolics are also present in kelp adhesives. The presence or identities of proteins are unclear, and refined proteomic characterizations are required before hypotheses can be made. Finally, studies of adhesion in kelp, fucoids, and other systems have been confined to single‐celled dispersal stages, such as spores and zygotes. Due to boundary‐layer effects and vast differences in scale, mature macroscopic kelp and their spores likely experience quite different environmental constraints. It is unclear how single‐celled spores and large, complex holdfasts compare in terms of adhesion. The histological organization of the rhizoid mat and attachment interface has not been examined previously due to the difficulty of achieving longitudinal sections through the hard substrates to which kelps attach (e.g., rocks and glass slides). We use a variety of novel tools and approaches to investigate adhesion in the kelp Alaria marginata, a common kelp found in the intertidal zone of wave‐exposed habitats along the west coast of North America (McConnico & Foster, 2005) and of interest for commercial farm cultivation (Raymond & Stekoll, 2021; Umanzor & Stephens, 2023). Through scanning electron microscopy, histology, and immunofluorescence microscopy, we describe the cellular organization of the attachment interface in developing holdfasts, clarify matrix carbohydrate identities with monoclonal antibodies (mAbs), investigate the presence of phenolics and proteins in the adhesive, and clarify the compositional differences between cell walls and kelp adhesives. MATERIALS AND METHODS Culturing Alaria marginata To establish a gametophyte culture, mature sporophylls of Alaria marginata were collected intertidally at low tide in April 2022 from Whiffin Spit Park in Sooke, British Columbia (48°21′22″ N 123°43′32″ W). Sporophylls were transported to the University of British Columbia in plastic bags on ice. Sporophylls were wiped clean with paper towels to remove large epiphytes and debris from the surface. Whole sporophylls were mixed in a 3% betadine/seawater solution at 10°C for 1 min to kill epiphytes, then rinsed thoroughly in seawater. Sporophylls were then stored in the dark on dry paper towels at 10°C overnight to allow dehydration for proper release of spores. Spores were released from sori by rehydrating them in sterile seawater. The spore solution was successively filtered through 100‐μm and 18‐μm mesh to remove particulate organic matter, plankton, and epiphytes larger than spores before quantifying spore density by hemocytometer. Spore solutions were diluted in seawater to bring concentrations to approximately 1.0 × 106 spores · mL−1. Cultures were established through dilution series of spores into 50‐mL Falcon tubes and selecting clean gametophyte cultures that developed. Mixed‐sex gametophyte cultures were stored vegetatively under ~20 μmol photons · m−2 · s−1 red light at 10°C in half‐strength f/2 media with monthly changes to minimize maintenance of gametophytes and growth of contamination. To bulk up the biomass of gametophytes, cultures were aerated in 250‐mL flasks under the same conditions. Gametophytes were fragmented using a mortar and pestle, diluted, and grown under ~50 μmol photons · m−2 · s−1 of white light to induce sexual reproduction and sporophyte production. Diluted gametophyte fragments were pipetted onto and cultured on various surfaces, including coverslips for light microscopy and scanning electron microscopy (SEM) or pre‐polymerized discs of the acrylic resin LR White for subsequent infiltration and embedding in resin for histology and immunolabeling. Pellets of LR White were made by polymerizing a block of the plastic and cutting a pellet from the block with a jewelers saw, smoothing the surface with a razor blade, and soaking in milliQ water overnight to remove particles and catalyst from the surface. This provided a substrate that could be sectioned, once sporophytes were cultured on these pellets (Figure 1). Juvenile Alaria marginata sporophytes grown on LR White resin substrates. (a) Sporophytes adhered to a pellet of the acrylic resin, (b) a close‐up of the developing holdfast adhered to the pellet, (c) holdfast and pellet embedded in additional LR White for histology. Gametophyte fragments were grown and monitored daily for sporophyte appearance for 1–2 weeks in quarter‐strength f/2 media at 10°C, ~50 μmol photons · m−2 · s−1, and no water motion to allow fragments to firmly reattach to the substrate. Clean samples with sporophytes were selected and transferred to small, sterile flasks to be cultured under the same conditions but with vigorous bubbling from air stones. Sporophytes for SEM were cultured to 0.1–1 mm long before imaging, after rhizoids appeared but before holdfasts had developed. Sporophytes for light microscopy and embedding were cultured until blades were ~1–2 cm in length with a small, not‐well‐defined holdfast, called “young holdfasts” hereafter. Some sporophytes were gently dislodged and cultured for one more week for live cell observations and embedding of what will be referred to as “older holdfasts,” with holdfasts ~1–2 mm wide. Scanning electron microscopy Gametophytes were fragmented on coverslips as previously described and cultured to produce sporophytes. Contamination was minimal on gametophyte cultures, and initial observations of gametophyte adhesives helped orient images taken of more heavily contaminated sporophyte rhizoids. Coverslips with adhered gametophytes or sporophytes were fixed overnight at 4°C in 5% Formalin in seawater adjusted to pH 7.4. Samples were brought to room temperature and washed three times with Tris‐Buffered Saline with Tween (TBST) buffer for 5 min, briefly rinsed with MilliQ water, and then dehydrated in an ethanol series (1%, 2.5%, 5%, 7.5%, 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, and 95%) at 30 min per solution, then 100% overnight and fresh 100% again for 30 min before critical point drying using a Tousimis Autosamdri 815B critical point dryer. Samples were sputter coated with a 5‐nm coating of gold/palladium (Au/Pd) using a Leica EM ACE600 coater for high‐resolution SEM images. Images were collected on a Zeiss Crossbeam 350 SEM. Rhizoid live cell observations Dislodged sporophytes were examined after 1 week of tumble culture to observe rhizoid cells. Sporophytes were rinsed with fresh seawater and mounted on glass slides in seawater for light microscopy. Sporophytes were observed and imaged on an Olympus BX53W1 light microscope with differential interference contrast (DIC) at various magnifications. Rhizoids were imaged by Olympus FV1000 confocal laser scanning microscope (CLSM) to examine phenolic and plastid distributions using intrinsic fluorescence. Phenolics and plastids were imaged with 405‐nm and 635‐nm solid‐state lasers for excitation, respectively, with emissions collected at 425–460 nm and 650–750 nm, respectively. Image z‐stacks were collected to image whole rhizoids, with step sizes optimized by FluoView software Nyquist sampling calculations. Composite images were generated in FIJI/ImageJ. Histology Attached sporophytes on pellets were fixed overnight at 4°C in 5% Formalin in seawater adjusted to pH 7.4. Samples were brought to room temperature and washed three times with TBST buffer for 5 min, rinsed with MilliQ water, and then dehydrated in an ethanol series as previously described for SEM. Samples were infiltrated with gradually higher concentrations of LR White in ethanol (1:2, 1:1, 2:1) for 2 h per solution, 100% LR White overnight, and 100% LR White for another 30 min. Pellets with attached sporophytes were carefully placed sideways into a gelatin capsule (Electron Microscopy Sciences, EMS), and the capsule was marked to maintain the orientation of the holdfast. Capsules were then filled with LR White and incubated at 65°C overnight to embed the holdfast in the plastic. Samples were trimmed, faced, and sectioned to 2 μm thickness with a glass knife longitudinally through the holdfast‐pellet interface to allow visualization of the attachment interface. Sections were collected by a paintbrush hair, mounted on a drop of water on a slide, and then fixed to the slide using a slide warmer. Sections were stained with 0.1% (w/v) Toluidine blue at pH 4.5 for 5 min for polysaccharides and phenolics, 0.05% (w/v) Ruthenium Red for 30 min for anionic polysaccharides, and 0.25% Coomassie blue in methanol:acetic acid:water (5:1:4) for 30 min to stain proteins. Brightfield images were acquired of rhizoid cells and adhesives by an Olympus BX51WI microscope. Images were processed through FIJI/ImageJ to adjust brightness and contrast. Carbohydrate immunolabeling Sections of holdfasts (n = 3) on Teflon‐coated slides with 5‐mm diameter wells were adhered by warming the slide on a hot plate at low heat. Sections were blocked with a 5% non‐fat dry milk (NFDM) solution in TBST buffer for 30 min before washing slides with TBST. Brown Algal Monoclonal (BAM) series antibodies (SeaProbes, Centre National de la Recherche Scientifique, CNRS) were used for primary antibody with specificity for various FCSPs and alginate, with BAM1, BAM2, BAM3, and BAM4 corresponding to unidentified but distinct fractions of FCSPs, and BAM6, BAM7, and BAM10 corresponding to alginate MM, MG, and GG blocks, respectively (Torode et al., 2015, 2016). Primary antibodies were each diluted 1:10 for all experiments. Twenty microliters of each 1:10 primary mAb solution were added to serial sections and incubated in a humidity chamber for 1 h at room temperature. Slides were briefly rinsed with TBST and then washed with 5% NFDM in TBST for 30 min. Slides were rinsed in TBST, gently dried, and then incubated with a 1:250 dilution of Alexa488‐conjugated anti‐rat secondary mAb (Invitrogen) for 1 h. Slides were again washed with TBST to remove excess secondary mAb and placed in the NFDM solutions again for 30 min. Negative control samples were treated identically but were treated with NFDM solution instead of mAb at each step, and another set was treated with no primary mAb in the first step and treated with secondary mAb in the second. Sections were briefly checked by epifluorescence before being mounted in 50% glycerol in water, and the coverslip was sealed with clear nail polish. Quantitative images of the attachment interfaces were collected by CLSM on an Olympus FV1000 Confocal Microscope at 20x and 60x. Transmitted light and excitation with a blue solid‐state laser (473 nm excitation / 488 nm emission) were used to image the tissue structure and mAb fluorescence, respectively, maintaining the same imaging conditions across all samples. All images were analyzed in FIJI/ImageJ to quantify immunofluorescence. Histogram bounds were set to a scale of 0 to 4000, and the color of images was changed to “Green Blue Fire” pre‐set heat map in the default lookup table. Toluidine blue‐stained serial sections aided the identification of rhizoids and adhesives. To quantify fluorescence, 10 randomly selected, 2‐pixel wide, short lines were drawn across adhesives or rhizoid walls and the mean gray value was measured. All values were recorded in a spreadsheet and analyzed in R (4.1.2) using R Studio (2021.09.01.372) using GGPlot2 and lme4 packages to visualize the data and fit linear mixed‐effects models with holdfast as a random factor to compare rhizoid wall and adhesive composition across holdfasts. Phenolic autofluorescence analysis UV autofluorescence was analyzed to test Alaria sections for the presence of phenolics. Untreated holdfast sections were mounted and the coverslips sealed as previously described. Phenolics are autofluorescent under ultraviolet light (Buschmann & Lichtenthaler, 1998), so samples were imaged on a FV1000 Olympus confocal laser scanning microscope (CLSM) with a 40× silicone immersion objective (1.25 numerical aperture, NA) using a 405 nm solid‐state laser. Lambda scans were then taken using 405 nm laser excitation and scanning emissions in 5 nm increments across wavelengths of 430 nm to 690 nm to test whether autofluorescence in adhesives and rhizoid cells is consistent with phenolics. Epidermal autofluorescence in stipe meristoderm cells was used as a positive control for phlorotannins, as the identity of phenolics in the outer wall is well established (Davies et al., 1973; Schoenwaelder, 2002; Shibata et al., 2004). Spectra were quantified in the adhesive, the rhizoid walls, and the epidermal walls using selected regions of interest. Spectral features were compared qualitatively to validate the presence or absence of phenolics. RESULTS Surface features of adhesion sites Filaments of gametophytes growing on a coverslip (Figure 2) were periodically dipped toward the substratum and placed down adhesive pads before curving back up into the water column (Figure 2b–d). The amount of adhesive produced by adhesive pads in gametophytes varied widely, with some filaments having a barely visible residue at the attachment point and others producing extensive gel networks. At every contact point with another filament or the substratum, kelp gametophytes placed down adhesive pads. Adhesive pads appeared fibrillar in nature, extending perpendicular from the cell wall toward the substratum and forming an interlocking mesh of connected fibers (Figure 2c,d). Scanning Electron Microscopy images of (a) Alaria marginata gametophyte colonies, (b, c) gametophyte branches adhered to the coverslip by fibrillar adhesive, (d) high‐magnification image of gametophyte adhesive, (e) Alaria marginata sporophyte blade with multiple rhizoid cells and a contaminant biofilm, (f) high‐magnification image of adhesive secreted by rhizoid cells. In sporophyte rhizoids, samples were more heavily coated in a complex microbial biofilm than gametophytes (Figure 2e), despite attempts to minimize contamination, making visualization of adhesives more challenging. Diatoms were a common contaminant, as well as bacteria closely associated with the walls and adhesives, and a variety of microbial eukaryotes. However, adhesives generally resembled those observed in gametophytes. Rhizoid cells were firmly pressed to the substratum. Adhesive extended from the cell wall toward the substratum in a fibrillar mesh (Figure 2f). Physode‐rich rhizoids and adhesive form the attachment interface By growing sporophytes on blocks of acrylic resin then embedding the sporophyte holdfast in the same resin, it was possible to section through the rhizoid mat and ultimately image the adhesion zone throughout development (Figure 3). In early development, holdfasts were mostly amorphous with a broad base of cortical cells from which small stipes would emerge (Figure 3a–c). Cortical cells in holdfasts were a mix of longitudinally elongated cells with thin anticlinal and thick periclinal walls, and short, round, thick‐walled cells. Basal layers of each holdfast terminated in a layer of small, round, physode‐dense, undifferentiated cells, with physodes staining strongly with toluidine blue (Figure 3b,f,g). Light micrographs of Alaria marginata holdfasts stained with (a–g) Toluidine Blue, (h) Ruthenium Red, or (i) Coomassie Blue. (a) Longitudinal section through stipe, holdfast, and basal mat of rhizoid cells, (b) close up of transition from cortical cells (Co), physode‐dense undifferentiated cells (U), rhizoid cells (R) at different planes of section, and a putative female gametophyte (G) entombed in the rhizoid mat, (c) higher magnification of rhizoid cells within the rhizoid mat resolving the rhizoid cell wall (W) and the adhesive matrix between them (A), (d) older holdfast showing a clearly formed meristoderm layer (M) and rhizoids differentiating from meristoderm cells by elongation, (e) longitudinally oriented rhizoids attached directly to cortical cells with no meristoderm present, (f) higher magnification of rhizoids in (e) showing an accumulation of physodes (Ph) and other similarly sized vesicle‐like bodies, (g) high‐magnification image of physode‐containing rhizoids at the attachment interface forming a tight seal to the pellet surface and showing adhesive filling the space between rhizoids, (h) Ruthenium red stain section of the attachment interface showing brightly stained rhizoid cell walls and faintly stained adhesive between the cells, (i) Coomassie Blue stain localizing intense protein accumulation in the adhesive compared to the cytoplasm. Beneath the holdfast cortex was a brightly staining mat of rhizoid cells. Rhizoid cells in cross‐section were thick‐walled and circular (Figure 3c). Cells in profile were elongated, unbranched tubes, and oval‐shaped cells were seen when rhizoid cells were at angles to the cutting plane. Inside rhizoid mats, the space occupied by adjacent rhizoids stained a faint, hazy purple with toluidine blue, matching previous observations of adhesive infilling of rhizoid mats. In older holdfasts, rhizoids appeared to originate by differentiation from meristoderm cells (Figure 3d–f). Rhizoid cells descended toward the substratum, making contact and running along the topology of the surface, flattening and taking the shape of the substratum (Figure 3g). In semi‐thin sections of the rhizoid mat, the spaces between rhizoids and between rhizoid walls and the substratum were filled with lightly staining adhesive (Figure 3g,h). Staining the rhizoid/substrate interface with ruthenium red revealed the presence of anionic polysaccharides (Figure 3h). The staining pattern was similar to that of toluidine blue (Figure 3g) with brightly stained cell walls and dimly stained adhesive matrix between adjacent rhizoid walls. Conversely, Coomassie blue intensely stained proteins within adhesives compared to walls and cellular spaces (Figure 3i). These histological data indicate that the adhesive is distinct from the rhizoid cell wall, with different polysaccharide profiles and higher protein content in the adhesive. Live cell observations of rhizoids descending from slightly more developed holdfasts (Figure 4a) revealed a number of interesting insights about these cells. At high magnifications with DIC (Figure 4b), numerous particles and vesicle‐like structures were observed streaming distally and proximally along the length of rhizoids. The directional movements inside the cell were consistent with vigorous trafficking and streaming of cellular materials. Rhizoid cells (Figure 4c) contained small red fluorescent bodies that were distributed evenly across the length of rhizoids (Figure 4d,f), localized by far red light (673 nm excitation), where chlorophyll fluorescence is observed, likely indicating numerous small plastids. In addition, numerous physodes, appearing as small, round, highly fluorescent blue spheres, were observed (Figure 4e,f). Physodes were randomly distributed along rhizoid cells, and autofluorescence in the rhizoid cell wall indicated incorporation of phenolics into the wall. Greater autofluorescence was localized at the growing tip of rhizoid cells (Figure 4e). The intrinsic blue fluorescence of the streaming physodes is consistent with the toluidine blue staining of physodes in the histological sections, indicating that physodes are a prominent feature in rhizoid cells of Alaria marginata holdfasts. Brightfield (a, b) and confocal laser scanning microscopy (c, f) micrographs of intrinsic fluorescence of living rhizoids showing (a) entangled mat of rhizoids (R) hanging down from the holdfast (H) and stipe (S) in Alaria marginata, (b) rhizoid cell with debris (D) on the surface and intracellular vesicles (V) that showed intense bidirectional streaming along the rhizoid, (c) transmitted light control image of rhizoids, (d) chlorophyll‐autofluorescence at 673 nm excitation showing regularly spaced plastids along the rhizoids, (e) UV autofluorescence at 405 nm excitation with blue emission in the rhizoid cell wall and physodes (Ph) characteristic of phenolics, and (f) a composite image of c–e. Immunofluorescence of FCSP and alginates in rhizoid cell walls and adhesive As both calcium‐cross‐linked alginates (Vreeland et al., 1993) and FCSP (Dimartino et al., 2016; Forbes & Hallam, 1979) have been proposed as important components of fucoid adhesive, the BAM monoclonal antibodies were applied to sections of the Alaria marginata holdfast adhesion zone. For structural context for immunofluorescence, adjacent serial sections of toluidine blue‐stained holdfasts were used to delineate rhizoid walls and adhesives. Control sections without primary antibody and exposed only to the secondary antibody tagged with Alexa488 demonstrated a lack of signal and low cytoplasmic autofluorescence (Figure 5a). In contrast, when the specific primary antibody and secondary antibody were applied, a green fluorescent signal was detected (Figure 5a). The FCSP epitopes localized by BAM1, BAM2, BAM3, and BAM4 gave surprisingly diverse labeling patterns, with BAM3 and BAM4 signals that were stronger in adhesive than in rhizoid cell walls (Figure 5a). Alginate epitopes detected by BAM6, BAM7, and BAM10 had the opposite pattern, with rhizoid cell walls having a more intense signal than the adhesive (Figure 5a). (a) Immunofluorescence demonstrates FCSP and alginate distribution in rhizoid cell walls (W) and Adhesives (A) in semi‐thin sections Alaria marginata holdfasts. Controls were treated with no primary antibody, FCSPs are localized by BAM1, BAM2, BAM3, and BAM4, and Alginate MM, MG, and GG blocks are localized by BAM6, BAM7, and BAM10, respectively. (b) FCSP dominates adhesive in Alaria marginata holdfasts. Quantification of immunofluorescent label intensity comparing relative polysaccharide composition of adhesives to rhizoid cell walls. Averages were taken of 10 randomly selected points for both rhizoids and adhesives (n = 3 sectioned holdfasts). The FCSPs represented larger fractions than alginate in the adhesive, whereas alginate and FCSPs were somewhat similar in rhizoid walls (Figure 5b). The quantification reflects the qualitative patterns (Figure 5a), with FCSP epitopes BAM3 and BAM4 localized highly in the adhesive matrix between rhizoid cells and BAM1 and BAM2 as minor components, while all FCSP epitopes labeled at similar intensities in rhizoid walls (purple shades, Figure 5b). The strongest alginate label in adhesive was by BAM10, which binds GG blocks that can form highly gelled regions of alginate, although BAM6 (MM) and BAM7 (MG) epitopes were also found (green shades, Figure 5b). Fluorescence intensities for mAb localizations in adhesives and rhizoid walls varied spatially, often over very short distances (Figure 5a). However, fluorescence intensity distributions of every antibody revealed significant differences between adhesives and rhizoid walls with consistent trends across holdfasts (Figure S1). Mean fluorescence intensities of BAM1, BAM2, BAM6, BAM7, and BAM10 were significantly higher in rhizoid walls than adhesives (Linear mixed‐effects models; Wald χ2 = 79.29, 58.29, 52.18, 127.12, and 34.74, respectively; df = 1 and p < 0.001 for all epitopes), whereas the labeling intensities of BAM3 and BAM4 were significantly higher in adhesives than in rhizoid walls (Linear mixed‐effects models; Wald χ2 = 296.73 and 121.04, respectively; df = 1 and p < 0.001 for both epitopes; see Figure S1). Phenolic autofluorescence analysis Although the signal was low compared to immunofluorescence, autofluorescence was observed consistently in unstained sections of holdfast adhesives and rhizoid cell walls (Figure 6), suggesting that phenolic compounds likely accumulate in adhesives and rhizoid walls. The stipe meristoderm layer, which is known to contain phenolics, was used as a positive control (Figure 6a). Within rhizoid cells, there was often saturated autofluorescence occupying much of the cross‐sectional space of the cell (Figure 6b). Given previous light microscopic observations of clusters of blue‐stained objects in sections of rhizoids and live cell fluorescence of rhizoids (see Figures 3f,g, 4d), saturated cellular autofluorescence was interpreted as being due to clusters of dense phenolic‐containing physodes. Phenolic autofluorescence analysis of Alaria marginata. (a) Confocal image (405 nm excitation) showing phenolic signal in meristoderm layer (M) and meristoderm wall (MW) in longitudinal section of the stipe. (b) Confocal image (405 nm excitation) of rhizoid walls (W) and adhesive (A) between the walls of holdfasts. (c) Fluorescence emission spectra collected by Olympus FV1000 Lambda scan comparing signal in adhesive and rhizoid to signal generated by meristoderm phenolics. Lambda scans of adhesives and rhizoid walls produced a spectrum with maximum peaks around 460 nm and 505 nm, minimum peak at 475 nm, and a steady decline across the range of 510–675 nm (Figure 6c). Spectral features of adhesives and rhizoid walls were similar to meristoderms, where phlorotannins are known to localize. DISCUSSION Rhizoid cells and the attachment interface Visualizing the attachment interface has been a major challenge in studying algal adhesives, since they generally exist as thin layers between biochemically similar cell walls and hard, unsectionable substrates. By allowing holdfasts to develop on LR White pellets, embedding holdfasts in more resin, and longitudinally sectioning through the holdfast‐pellet attachment interface, we have captured histological images of rhizoids and adhesives attached to hard substratum, enabling histological and immunofluorescence analyses. Despite this, the adhesive between the rhizoid and pellet formed such a thin layer that it was still difficult to clearly resolve. Adjacent rhizoids at the attachment interface and within the rhizoid mat had faint bluish purple staining with toluidine blue in the corners between round cells, which is consistent with SEM observations and previous holdfast studies as adhesive infilling between cells (Tovey & Moss, 1978). Given these SEM and staining observations of extracellular spaces within the rhizoid mat and at the attachment interface, it is likely that the whole rhizoid mat is cemented with adhesive. The observation of rhizoids flattening themselves against the micro‐topology of the resin pellet is also consistent with past observations (Tovey & Moss, 1978), wherein rhizoids were thought to take the shape of the substratum to increase the surface area of rhizoid‐surface interactions. The dynamic weaving and flattening of these thick‐walled cells along the attachment interface likely allow them to form a perfect seal to the substratum as rhizoid mats become densely packed. Streaming of cellular contents down toward the rhizoid tip and back up toward the holdfast suggests elevated trafficking activity. Resources need to be trafficked down to the rhizoid tip, where adhesive secretion and cell‐wall remodeling for rhizoid tip growth can occur. Given that rhizoid cells specialize in the synthesis and secretion of adhesives, one could envision an extensive Golgi network, as seen in settling spores (Oliveira et al., 1980), for assembling and secreting adhesive and cell‐wall contents. More work should be carried out to clarify the cytological organization, trafficking, and secretion dynamics of kelp rhizoids to better understand how they specialize in secretion. Physodes were localized along the length of rhizoids, and autofluorescence along rhizoid walls and cell apices was consistent with phenolic incorporation into cell walls. Physodes could be involved in the secretion of phloroglucinols that get cross‐linked into phlorotannins by vanadium‐dependent haloperoxidases, as demonstrated in Fucus embryo cell walls and perhaps in adhesives (Bitton et al., 2006; Lemesheva et al., 2020). Adhesive composition Although previous studies on fucoid seaweeds suggested that the primary gel component of brown algal adhesives is alginate (Berglin et al., 2004; Tarakhovskaya, 2014; Vreeland et al., 1993), our results demonstrated that FCSPs compose the bulk of the gel matrix in kelp adhesives, with alginate composing a smaller fraction. This is consistent with ATR‐FTIR data from adhesive secreted by germlings of the fucoid Durvillaea antarctica, which showed the accumulation of sulfated moieties during polysaccharide secretion (Dimartino et al., 2016). Our findings shift the paradigm from kelp adhesives being a simple phenolic cross‐linked alginate to being a complex matrix of biochemically distinct FCSPs, alginate, phenolics, and proteins. Because BAM mAbs were derived in fucoids—not kelps—probes derived directly from kelp polysaccharides in the future may yield additional insights and better resolution on kelp adhesive composition. Furthermore, given that BAM mAbs have not yet been applied to adhesives in fucoids, it is likely there is still more to learn about polysaccharide composition in fucoid adhesives. Our discovery of phenolics in kelp adhesives raises questions about adhesive evolution in macroalgae more broadly. If phenolics function in cross‐linking adhesives in kelp, as assumed by the current fucoid model, then phenolics may represent a core component in a deeply conserved mechanism of brown algal adhesion. More work is needed to investigate whether representatives of other brown algal orders utilize phlorotannins in adhesion. Additionally, phenolic autofluorescence would allow for easy semi‐quantitative characterization of cross‐linker abundance during adhesive deposition, holdfast development, or in response to environmental factors (e.g., flow, acidification). Limitations of current adhesive model That kelp adhesives and cell walls have broadly similar composition raises questions about the proposed mechanism of adhesion in brown algae. Phenolic cross‐linking occurs in many brown algal cell walls (Imbs & Zvyagintseva, 2018), but cell walls do not generally behave like adhesives. The mechanism of phenolic cross‐linking acting to solidify a simple alginate matrix has been performed in vitro but has not been observed to occur in natural adhesives (Berglin et al., 2004; Bitton et al., 2006). Therefore, phenolics may simply act in adhesives as they do in cell walls and regulate the rigidity of the gel matrix, in which case adhesive properties may be derived from another molecular interaction. Additionally, if phenolics facilitate adhesion and catechol‐like residues mediate surface interactions, then the regulation of catechol chemistry and the redox environment of kelp adhesives would need to be carefully examined, since catechols rapidly oxidize to quinones in seawater, diminishing adhesive properties (Waite, 2017). Future work clarifying specific matrix interactions with phenolics is sorely needed. The role of proteins in brown algal adhesion remains a major gap in our understanding. Abundant proteins have been observed, in the present study and others, through staining or digestions of both fucoid and kelp adhesives (Dimartino et al., 2016; Oliveira et al., 1980; Petrone et al., 2010) but have largely been assumed to be vanadium‐dependent haloperoxidases involved in putative phenolic cross‐linking (Bitton et al., 2006; Vreeland et al., 1998). Given that complex protein assemblages are spatially and temporally heterogeneous and mediate adhesive assembly in diatoms, green algae, and invertebrates (Poulsen et al., 2022; Stanley et al., 1999; Waite, 2017), proteins may also play important roles in brown algal adhesives. The characterization of adhesive proteins has been limited by the difficulty in isolating adhesives without cellular contamination for proteomic identification (Lachnit et al., 2019), and few molecular tools currently exist to validate their localization and associations. As tools become increasingly available for brown algae, research efforts should focus on characterizing proteinaceous components of brown algal adhesives. CONCLUSIONS This study represents significant progress in our understanding of adhesion in kelp. By allowing kelp to attach to plastic pellets, we provide the first histological images of the attachment interface of kelp holdfasts to the substratum. Our novel approach has facilitated the investigation of holdfast structure, organization of rhizoid cells, and biochemistry of kelp adhesives, extending our understanding of attachment. Additionally, this study has produced the highest resolution chemical description of kelp adhesives to date, demonstrating that adhesives are composed of FCSPs, alginate, and phenolics, with a gel matrix dominated by BAM3 and BAM4 epitopes. Adhesives are composed of the same basic materials as rhizoid cell walls but with distinct stoichiometries. Adhesive complexity has likely been underestimated and basic assumptions about kelp adhesive chemistry need to be revisited, including the uncertain roles of phenolics and proteins in adhesive assembly. These findings reshape our understanding of kelp adhesives and provide new methods to facilitate future studies of kelp attachment. AUTHOR CONTRIBUTIONS Maisie R. Roy‐Musor: Conceptualization (lead); formal analysis (lead); investigation (lead); methodology (lead); visualization (lead); writing – original draft (lead); writing – review and editing (equal). Christa Agoawike: Formal analysis (supporting); investigation (supporting); writing – review and editing (equal). A. Lacey Samuels: Conceptualization (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); supervision (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal). Patrick T. Martone: Conceptualization (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); supervision (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal). Supporting information Figure S1: Fluorescence intensities of adhesives (closed circles) and rhizoid cell walls (open circles) for FCSPs and alginate monoclonal antibodies in three replicate holdfasts to highlight intra‐ and inter‐holdfast variation. Data from each antibody were analyzed separately with a linear mixed‐effects model, including holdfast as a random factor. Significant differences between adhesives and rhizoid cell walls are indicated (*p < 0.001).