Mushroom membranes sliced in different directions steer bone regrowth in rats

Researchers have shown that simply changing the direction in which they slice the stipe of the king oyster mushroom, Pleurotus eryngii, can transform the same fungal tissue into barrier membranes with sharply different functional properties for guided bone regeneration. In a study published in Frontiers in Cell and Developmental Biology, membranes cut lengthwise along the mushroom’s natural hyphal alignment showed stronger directional mechanics, better guidance of bone-forming cells, tighter exclusion of fibroblasts, and more new bone growth in a rat calvarial defect model than membranes cut across the stipe. The work demonstrates that a native biological architecture, preserved rather than artificially engineered, can encode spatially differentiated functions at the scale of a medical membrane.

Guided bone regeneration, or GBR, is a widely used technique in oral implantology and maxillofacial repair. A barrier membrane is placed between a bone defect and the fast-growing surrounding soft tissue, maintaining a protected space where osteogenic cells can migrate and new bone can form. An ideal membrane must do several things at once, and these requirements are not uniform in space: the bone-facing surface should support adhesion and organization of osteogenic cells, while the through-thickness direction should block fibroblasts and other soft-tissue cells from invading the defect. Conventional approaches to building such directionally structured materials, including electrospinning, micro-patterning, directional freezing, and three-dimensional printing, generally demand specialized equipment and multiple fabrication steps.

The Shenzhen-based team, working with collaborators at Anhui Medical University and Sichuan Provincial People’s Hospital, took a different route. The stipe of P. eryngii is built from interconnected tubular hyphae whose cell walls consist mainly of chitin and beta-glucan, and its tissue is naturally organized along the longitudinal axis. The researchers cut roughly 1-millimeter-thick slices either parallel or perpendicular to this axis, then subjected both sets to an identical combined chemical and enzymatic treatment: immersion in sodium carbonate and sodium dodecyl sulfate at 60 degrees Celsius, repeated washing, beta-glucanase digestion, further washing, and a hot-water bath, followed by freeze-drying. The goal was to strip away cellular residues while leaving the tubular hyphal network intact.

The treatment worked as intended. Residual DNA fell from roughly 170 to 220 nanograms per milligram of dry tissue to below 30 nanograms per milligram, a reduction of about 82 to 86 percent, while protein dropped from about 10 micrograms per milligram to below 2.5 micrograms per milligram, a removal of more than 75 percent. Scanning electron microscopy and histological staining confirmed that the tubular hyphal architecture remained recognizable in both cutting orientations. The cross-cut membranes, designated Tr-C, displayed a relatively disordered two-dimensional morphology of transverse and oblique hyphal profiles, whereas the longitudinal-cut membranes, Tr-L, showed tubular structures predominantly aligned in a common direction across the membrane plane.

Despite these structural differences, attenuated total reflectance Fourier-transform infrared spectroscopy showed that the two treated membranes had generally similar chemical profiles, with spectral changes after treatment suggesting an increased relative contribution of chitin and a reduced contribution of beta-glucan. This similarity in bulk chemistry is what makes the subsequent findings striking: the functional differences between the membranes arose almost entirely from how the same native hyphal network was geometrically presented, not from compositional changes. When loaded along the principal hyphal orientation, Tr-L reached a wet-state tensile strength of approximately 1.48 megapascals and an elastic modulus of about 3.0 megapascals, compared with roughly 0.19 and 0.54 megapascals when loaded perpendicular to the hyphae, and about 0.15 and 0.33 megapascals for Tr-C.

Fluid behavior also diverged by orientation. Tr-C, with its exposed transverse and oblique hyphal openings, infiltrated liquid faster and absorbed more water, showing a hydrated-to-dry mass ratio of about 34 grams per gram against roughly 25 grams per gram for Tr-L. Tr-L itself was anisotropic in transport, with liquid migrating faster along the hyphal direction than across it. Over four weeks of incubation in phosphate-buffered saline, in lysozyme-containing PBS, and in hydrogen peroxide-containing PBS, both membranes retained most of their initial mass, with remaining masses ranging from about 71.4 to 85.2 percent; Tr-C generally lost more mass than Tr-L, and the greatest losses occurred under oxidative conditions. The authors note these measurements reflect apparent mass loss and erosion rather than a specific chemical degradation process.

In cell experiments, both membranes supported the metabolic activity of MC3T3-E1 mouse preosteoblasts without differences from controls, but their topographies produced markedly different cell behavior. Cells on Tr-L became elongated and aligned along the membrane’s parallel surface textures, with a median aspect ratio above 3 compared with about 2 on control and Tr-C surfaces, and a cellular orientation order parameter approaching 0.8, versus values below 0.2 elsewhere. Single-cell analysis showed that cells more closely aligned with the substrate’s principal orientation were generally more elongated. After 14 days of osteogenic induction, alkaline phosphatase activity was significantly higher on Tr-L than on controls, and after 21 days, Alizarin Red S quantification of mineralization-associated staining was significantly higher on Tr-L than on both Tr-C and controls. In Transwell assays, both membranes significantly reduced L929 fibroblast migration over five days, with no migrated cells detected beneath Tr-L.

The membranes also elicited distinct macrophage responses under inflammatory stimulation. With RAW264.7 macrophages stimulated by lipopolysaccharide, Tr-C significantly lowered tumor necrosis factor-alpha release at 24 hours, to about 7.26 nanograms per milliliter versus roughly 10.10 in controls, while Tr-L significantly reduced nitrite accumulation at 72 hours, to about 17.0 micromolar versus 39.5 in controls. Neither membrane elevated inflammatory markers without stimulation, and neither reduced macrophage metabolic activity. The authors interpret these differences as time- and marker-dependent modulation tied to the distinct cell-material interfaces created by each hyphal topography, rather than evidence of a generalized anti-inflammatory effect or a specific polarization shift.

In the in vivo test, adult Sprague-Dawley rats received bilateral 5-millimeter full-thickness calvarial defects covered with Tr-C or Tr-L membranes, with untreated defects as controls. Micro-computed tomography at 6 and 12 weeks showed new bone extending centripetally from the defect margins toward the center in both treated groups, most extensively with Tr-L. Bone volume fraction and bone mineral density followed an increasing trend from control to Tr-C to Tr-L at both time points, and trabecular number was highest in Tr-L at 12 weeks, although trabecular thickness did not follow the same pattern. Histology confirmed more extensive new bone formation and greater defect-size reduction with Tr-L, though complete bony bridging was not achieved in any group within 12 weeks, with four defects per group at each time point.

The authors are careful about the limits of their findings. The structure-cell relationships were analyzed mainly at morphological and phenotypic levels; mechanotransduction pathways such as integrin/FAK, RhoA/ROCK, and YAP/TAZ were not directly examined, and FTIR could not resolve the chitin-to-beta-glucan ratio or surface exposure states. Because 5-millimeter rat calvarial defects rarely close spontaneously, the model is considered critical-size, but the unactivated native membrane is better described as an architecture-driven GBR platform than a strongly osteoinductive material. The preserved porous hyphal network may nonetheless offer structural space for future loading of drugs, osteogenic peptides, or mineral components. The central message stands: the cutting plane alone reinterpreted structural information already encoded in the mushroom’s growth, converting one biological template into two functionally distinct membranes, and the longitudinally aligned version best matched the dual demands of in-plane cell guidance and through-thickness soft-tissue exclusion.

Subject of Research: Pleurotus eryngii-derived barrier membranes with cutting-orientation-dependent anisotropic functions for guided bone regeneration

Article Title: Native hyphal alignment enables spatially anisotropic functions in Pleurotus eryngii-derived membranes for guided bone regeneration

Article References: Chen, X., Dong, N., Hou, H., Li, Y., He, H., Huang, J., Li, H., Chen, J., & Li, Q. (2026). Native hyphal alignment enables spatially anisotropic functions in Pleurotus eryngii-derived membranes for guided bone regeneration. Frontiers in Cell and Developmental Biology, 14, Article 1997731. https://doi.org/10.3389/fcell.2026.1997731

Image Credits: AI Generated

DOI: 10.3389/fcell.2026.1997731

Keywords: guided bone regeneration, Pleurotus eryngii, hyphal architecture, anisotropy, barrier membrane, contact guidance, osteogenesis, decellularization, chitin, beta-glucan, rat calvarial defect, biomaterials