Age-mimicking hydrogel stiffness recapitulates the mechanical niche of the hippocampus to regulate neural stem cell senescence

TL;DR

This study shows that making brain-like materials with different stiffness can mimic how aging affects brain stem cells, offering new ways to potentially reverse aging effects.

Contribution

The study introduces age-mimicking hydrogels that replicate hippocampal stiffness and reveal conserved mechanotransduction pathways in NSC aging.

Findings

Hippocampal tissue stiffness increases with age and correlates with reduced neurogenesis.

Stiff hydrogels accelerate NSC aging and impair proliferation and differentiation.

Piezo1 disruption rejuvenates old NSCs in stiff environments across species.

Abstract

Neural stem cell (NSC) aging significantly contributes to reduced neurogenesis, driven by both intrinsic mechanisms and environmental cues. However, the response of hippocampal NSCs to developmental and age-related changes in microenvironmental stiffness remains incompletely understood. Our study showed that hippocampal tissue stiffness increases substantially with age, correlating with diminished neurogenesis. To faithfully model this age-dependent mechanical transition, we engineered hyaluronic acid-laminin hydrogels matching physiological hippocampal stiffness across age groups. Culturing NSCs from different-aged donors on these stiffness-tunable hydrogels revealed that age-related hippocampal stiffening accelerates the NSC aging phenotype and impairs their proliferation and neuronal differentiation. This functional decline was associated with upregulated expression of collagen and integrin genes alongside downregulated expression of cell cycle-promoting genes in NSCs. Our study further revealed that aging alters Piezo1 expression, and disrupting Piezo1 rejuvenated the proliferative capacity of old NSCs while restoring the expression patterns of cell cycle and cell adhesion genes in stiff microenvironments. Moreover, we found that the regulation of NSC aging by niche stiffness is largely conserved from rodents to primates. This conserved mechanism establishes a foundation for novel regenerative strategies that target mechanotransduction pathways, potentially enabling neural tissue repair through biomaterial-assisted cell transplantation.

Image 1

•Engineered laminin-modified hydrogels recapitulate age-dependent hippocampal stiffness.•Hydrogel-mimicked niche stiffness directly regulates NSC aging.•Stiffness modulates NSC aging via cell cycle/ECM gene expression reprogramming.•NSCs perceive niche stiffness via Piezo1 to regulate proliferation and gene expression during aging.•Niche stiffness universally regulates NSC aging across rodents and primates.

Engineered laminin-modified hydrogels recapitulate age-dependent hippocampal stiffness.

Hydrogel-mimicked niche stiffness directly regulates NSC aging.

Stiffness modulates NSC aging via cell cycle/ECM gene expression reprogramming.

NSCs perceive niche stiffness via Piezo1 to regulate proliferation and gene expression during aging.

Niche stiffness universally regulates NSC aging across rodents and primates.

Figures

Full text

Introduction

1

The maintenance of neural stem cell (NSC) self-renewal and differentiation potential is not only crucial for early brain development but also plays a key role in the regulation of postnatal brain functions in both rodents and primates [[1], [2], [3]]. The subgranular zone (SGZ) of the hippocampal dentate gyrus (DG) is recognized as one of the primary neurogenic niches where NSCs reside, and neurogenesis occurs during the postnatal stage [3,4]. Hippocampal neurogenesis is linked to learning and memory [1,4]. Its decline with age is attributed to reduced NSC self-renewal and neural progenitor proliferation, which result from NSC senescence [[5], [6], [7]]. Many intrinsic and environmental factors have been reported to regulate hippocampal NSC proliferation and differentiation during aging, such as transcription factors, epigenetic mechanisms, protein homeostasis, metabolism changes, and niche signaling [5,[8], [9], [10], [11], [12], [13], [14], [15]]. However, it remains unclear whether hippocampal NSCs in the postnatal brain can sense niche stiffness and modulate their neurogenesis potential during aging. Given that tissue maturation and aging alter extracellular stiffness [[16], [17], [18]], we seek to determine whether aging-altered niche stiffness affects stem cell fate. While recent reports have investigated how tissue stiffness regulates the behavior of hematopoietic stem cells, oligodendrocyte progenitors, and epidermal stem cells during aging [[19], [20], [21]], the effects on hippocampal neural stem cells have not been explored. Although several studies suggest that NSC activity and neurogenesis may be associated with the mechanical stiffness of the stem cell niche [[22], [23], [24]], direct evidence and the underlying molecular mechanisms involved in age-related stiffness regulating NSCs remain unclear. Due to the challenge in altering the stiffness of in vivo tissues, we planned to develop an approach that could simulate age-related changes in the stiffness of the NSC in vivo microenvironment, in order to investigate the impact of tissue stiffness on NSC aging and the associated molecular mechanisms.

Hydrogel biomaterials are three-dimensionally structured networks of crosslinked hydrophilic polymer matrices with good biocompatibility, biodegradability, and absorbability. Their adjustable physicochemical properties have made them extensively utilized in tissue engineering and various other biomedical fields [25,26]. Hydrogels offer a favorable microenvironment for cells to acquire nutrients and facilitate growth during tissue repair [[27], [28], [29]]. Furthermore, hydrogels serve as a valuable model for investigating the mechanisms by which the physical properties of the extracellular matrix (ECM) regulate cell fate [[30], [31], [32]]. Early reports indicate that altering hydrogel stiffness might affect the proliferation and differentiation capabilities of NSCs grown on them [33,34]. While previous studies have primarily focused on the behavioral changes of NSCs on matrices of varying stiffness [[33], [34], [35], [36], [37]], they neither accurately mimic the physiological stiffness of the NSC niche during aging nor thoroughly investigate the underlying mechanisms of niche mechanical regulation in the context of NSC aging. Here, we aimed to explore how age-related changes in hippocampal microenvironment stiffness influence NSC aging. To address this, we developed an ECM microenvironment that moves beyond the simplistic classification of “soft” versus “stiff” by precisely replicating the physiological stiffness of the NSC niche at different ages. Specifically, the soft matrix mimics the compliant mechanical properties of young brain tissue in vivo, whereas the stiff matrix reflects the relatively rigid microenvironment of old brain tissue. Using these physiologically relevant matrices, we characterized the NSC aging phenotypes [38], including diminished self-renewal capacity, reduced neuronal differentiation potential, increased astrocytic differentiation propensity, elevated expression of senescence markers (SA-β-gal, p16, p21), and loss of Lamin B1—and further investigated the molecular mechanisms governing NSC fate decisions in response to niche stiffness.

It is well established that neurogenesis declines with age, although the rate of decline varies between primates and rodents [3]. However, whether and how NSCs in primates respond to stiffness cues remains unclear. Due to ethical constraints and the limited availability of human samples, non-human primates, particularly monkeys, share remarkable anatomical and functional similarities with humans, making them valuable models for addressing this question [3]. In this study, we therefore also investigated the conservation of this mechanosensitive mechanism between rodents and primates, aiming to provide new theoretical insights and a foundation for mechanobiology-based neural regeneration therapies.

Materials and methods

2

Animal models and ethical approval statement

2.1

The mice used in this study were housed under standard SPF laboratory conditions with a 12-h light/dark cycle, an ambient temperature of approximately 20 °C, and a relative humidity of 40–60%. All experimental procedures and animal housing protocols were approved by the Animal Ethics Committee of Kunming University of Science and Technology (Approval No. PZWH(Dian)K2023-0024). C57BL/6J mice aged 2, 4, 6, 8, 11, 12, and 13 weeks were used for immunofluorescence analysis of brain tissues and for the culture of NSCs. In vivo NSCs were sorted from 1- and 12-week-old Nestin-GFP mice, originally obtained from the Jackson Laboratory (STOCK Tg(Nes-EGFP)33Enik/J, Strain #: 033927, RRID: IMSR_JAX:033927), and housed at Yunnan University’s Laboratory Animal Center (Registration No. CANS LA0029). Brain tissues used for stiffness measurement were obtained from 1-, 3-, 8-, and 12-week-old C57BL/6J mice purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Certificate No. SCXK(Jing)20160006).

The monkeys used in this study were housed in a controlled environment with a stable temperature of 22 ± 1 °C and relative humidity maintained at 50% ± 5%. A 12-h light/dark cycle was maintained, with lights switched on at 08:00 a.m. Monkeys were fed a commercial diet twice daily, supplemented with fruits and vegetables to ensure balanced nutrition. All experimental procedures involving monkeys were approved by the Animal Ethics Committee of Kunming University of Science and Technology (Approval No. KUST202301005).

Generation of HA-laminin hydrogel

2.2

Hyaluronic acid (HA) powder (5 g; Macklin, H823435) was dissolved in 500 mL of deionized water, and the pH was adjusted to 4.75. Adipic dihydrazide (ADH; Macklin, O815181) was added to the HA solution at three amounts (0.5 g, 0.875 g, and 2.0 g) to obtain the formulations corresponding to Soft, Medium, and Stiff gels, respectively. The mixtures were stirred until fully homogeneous. Subsequently, EDC·HCl (2.0 g; Macklin, N808856) was added, and the reaction mixture was stirred for 20 min while the pH was adjusted to 7.4 to initiate coupling between HA carboxyl groups and ADH hydrazide groups, thereby forming a crosslinked hydrogel network. The formed hydrogels were ultrasonically cleaned five times and then lyophilized. For laminin conjugation, the dried gel polymers were activated with CDI solution (5 mL; Macklin, N805049) under gentle stirring for 15 min. The CDI-activated hydrogels were washed with acetone (Chongqing Chuandong Chemical, No. 1090) at least five times to remove unbound CDI, followed by five washes with 100 mM sodium carbonate buffer (pH 8.5; Macklin, S885295) to completely replace acetone. The activated hydrogels were then immersed in sodium bicarbonate buffer (100 mM, pH 8.5) and incubated with laminin (1 μg/mL; Shanghai Huzhen Biotechnology, 110590-64-2) under stirring for 48 h at room temperature to allow covalent conjugation. After reaction, the HA–laminin hydrogels were thoroughly washed with PBS (pH 7.4; Macklin, P765569) for 4 h, blocked with 100 mM sodium bicarbonate buffer (pH 8.5; Macklin, S885290) for 2 h at room temperature, and lyophilized to obtain HA–laminin hydrogels of different stiffness. The reaction scheme is shown in Fig. 1i.Fig. 1Age-associated hippocampal stiffening and its replication via laminin-modified hydrogels. (a) The strategy of in vivo EdU labeling and marker immunostaining for analyzing NSC proliferation and neurogenesis across various mouse age groups. (b) Co-staining of GFAP, EdU, and DCX in the hippocampus across different ages. Representative images showing a reduction in active radial glia-like stem cells and neuroblasts/newborn neurons with increasing age. GFAP (green), DCX (red), EdU (gray), and DAPI (blue). Scale bar, 100 μm. (c-e) Quantification of active radial glia-like stem cells (GFAP^+^EdU^+^) (c), neuroblasts (DCX^+^EdU^+^) (d), and newborn neurons (DCX^+^) (e) in the SGZ area as in (b). n = 3 or 4 mice per group. (f) Schematic showing the measurement of hippocampal tissue stiffness using the Pavone nanoindenter and the design of hyaluronic acid (HA)–laminin hydrogels with tunable stiffness to mimic hippocampal mechanical properties at different postnatal ages. Soft, medium, and stiff hydrogels correspond to the mechanical characteristics of hippocampal tissues from 1-, 4-, and 12-week-old mice, respectively. (g) Representative images of the dentate gyrus in mouse brain slices across age groups, captured under Pavone nanoindentation microscopy. The SGZ regions measured by the nanoindentation probe are demarcated by paired colorful dashed lines. (h) Quantification of Young’s modulus in the hippocampal SGZ region of mice at different ages using Pavone nanoindentation. Brain slices were obtained from four mice per age group. Measurements were taken from n = 227 spots (1 week), n = 149 spots (4 weeks), n = 241 spots (8 weeks), n = 157 spots (12 weeks). (i) Schematic illustration of the synthesis of HA@HA and HA@HA–Laminin hydrogels. Hyaluronic acid (HA) was first crosslinked with adipic dihydrazide (ADH) using EDC/HCl activation under acidic conditions (pH 3–4) to form HA@HA. Subsequently, laminin was conjugated to the HA network via CDI-mediated coupling to generate HA@HA–Laminin hydrogels. (j) Quantification of Young’s modulus of Soft, Medium, and Stiff HA-laminin hydrogels using the same Pavone nanoindentation used for tissue (Soft hydrogel, n = 44 spots; Medium hydrogel, n = 37 spots; Stiff hydrogel, n = 29 spots). (k) Immunostaining of YAP1 protein shows the subcellular localization of YAP1 in NSCs cultured on HA-laminin hydrogels of varying stiffness. Dashed lines indicate representative cells with YAP1 predominantly in the cytoplasm (indicated by arrowheads), while solid lines represent cells with YAP1 primarily in the nucleus (indicated by arrows). Scale bar, 20 μm. (l) Quantification of YAP1 distribution in NSCs as in (k) (n = 4 wells). For all quantification data, statistical significance was determined using one-way ANOVA with Tukey’s multiple comparison tests. Data are presented as mean ± SD (∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001).Fig. 1

Hyaluronic acid hydrogel degradation performance test

2.3

To assess the degradation behavior of hyaluronic acid (HA) hydrogels with varying stiffness, lyophilized HA hydrogels of equal dry mass were prepared. Each hydrogel sample was first fully swollen in phosphate-buffered saline (PBS; Solarbio, P1020), and the excess surface liquid was gently removed using sterile absorbent paper. The initial swollen weight was recorded as M. The hydrogels were then transferred into 50 mL centrifuge tubes containing 40 U/mL hyaluronidase solution (Solarbio, H8030) prepared in PBS. Samples were incubated at 37 °C in a constant temperature shaker. Every 24 h, hydrogels were removed, the surface liquid was blotted with sterile absorbent paper, and the weight was recorded as m. This procedure was repeated until the hydrogel samples were completely degraded. The degradation ratio at each time point was calculated using the following equation:

$$[eqn]$$

Hyaluronic acid hydrogel swelling performance test

2.4

To evaluate swelling behavior, lyophilized HA hydrogels with different stiffnesses but equal dry weight were weighed and recorded as the initial weight W1. Samples were immersed in excess PBS at room temperature to allow swelling. At 10-min intervals, each hydrogel was removed, gently blotted with lint-free paper to remove surface fluid, and weighed to obtain the swollen weight W2. This process was repeated every 10 min for a total of seven measurements or until the swelling reached equilibrium. The swelling ratio was calculated as:

$$[eqn]$$

Rheological characterization of HA–laminin hydrogels

2.5

Soft, Medium, and Stiff hydrogels were placed between the upper and lower parallel plates of the rheometer, and the measurement geometry was carefully adjusted to ensure full contact between the samples and the plates while avoiding excessive compression of the hydrogels. To characterize the viscoelastic properties of the hydrogels, oscillatory rheological measurements were performed using a rotational rheometer (TA Instruments HR10) equipped with a parallel-plate geometry at 37 °C, including frequency sweep and strain (amplitude) sweep tests.

Frequency sweep: The storage modulus (G′) and loss modulus (G″) were measured over an angular frequency range of 0.1–10 rad/s (logarithmic spacing) at a constant strain amplitude of 0.5%, which lies within the linear viscoelastic region of the hydrogels.

Strain sweep: G′ and G″ were recorded over a strain amplitude range of 0.1–1000% (logarithmic spacing) at a constant angular frequency of 1 rad/s to determine the linear viscoelastic region and to evaluate the yielding behavior of the hydrogels.

Crystal violet diffusion assay

2.6

Ten milliliters of hyaluronic acid (HA) hydrogels with different stiffness were prepared in 20 mL glass vials (NINGKE, ZH10006-02). After complete crosslinking, gelation was confirmed by vial inversion. Subsequently, 500 μL of crystal violet solution (0.5 mg/mL in deionized water; Solarbio, Cat# G1059) was gently added dropwise onto the surface of each hydrogel. The vials containing hydrogels of different stiffness were then placed in a 37 °C incubator, and the diffusion of crystal violet within the hydrogels was recorded at the indicated time points (0, 0.5, 1, 2, 4, 6, 8, 10, 12, and 24 h).

Scanning electron microscopy (SEM)

2.7

HA–laminin hydrogels from the Soft, Medium, and Stiff groups were rapidly frozen in liquid nitrogen and subsequently subjected to freeze-drying to preserve the internal porous structure of the hydrogels. These samples were then fractured to expose the internal cross-sections. The fractured specimens were mounted on SEM stubs and sputter-coated with a thin conductive layer prior to imaging. The microstructure of the hydrogels was examined using a scanning electron microscope (SEM, Nova NanoSEM 450, FEI). SEM imaging was performed in secondary electron (SE) mode with an accelerating voltage of 10.0 kV.

Culture of neural stem cells on hyaluronic acid hydrogels

2.8

HA-laminin hydrogels with varying stiffnesses were equilibrated in NSC maintenance medium (DMEM/F12 [Gibco, 10565018] supplemented with 2% B27 [Gibco, 17504-044], 1% penicillin-streptomycin [Beyotime, C0222], 20 ng/mL EGF [Peprotech, AF-100-15], and 20 ng/mL bFGF [Peprotech, AF-100-18B]) for 24 h prior to cell encapsulation. The swollen hydrogels were evenly distributed in culture wells at a thickness of 2–3 mm. NSCs were then plated at a density of 2 × 10^4^ cells/cm^2^ to prevent multicellular aggregation and to allow for proliferation and growth assessment. The culture medium was refreshed every 48 h by replacing half of the volume.

Cell culture on CytoSoft® elastic silicone gel plates

2.9

CytoSoft® Discovery Kit six-well plates (Advanced BioMatrix, Cat. No. 5190-7 EA) with elastic moduli of 0.2, 0.5, and 2 kPa were used for cell culture. The silicone gel plates were first coated with poly-L-ornithine (PLO; Sigma-Aldrich, P3655) diluted in PBS (Solarbio, P1020) to a final concentration of 100 μg/mL and incubated at 37 °C for 4 h. The silicone gel plates were then washed four times with PBS (5 min each). Subsequently, laminin (Sigma-Aldrich, L2020) was applied at a concentration of 25 μg/mL and incubated at 37 °C for 2 h. After coating, the excess solution was removed prior to cell seeding. NSCs isolated from 8-week-old adult mice were seeded onto the prepared silicone gel plates at a density of 2 × 10^5^ cells per well and maintained at 37 °C in a humidified atmosphere containing 5% CO_2_. Cell density was recorded at 24, 48, and 72 h after seeding. At 72 h, 15 μM EdU (Beyotime, C0075L) was added to the culture medium and incubated for 4 h. EdU signals were visualized using a Nikon ECLIPSE Ts2 microscope. Cell density and EdU-positive cells were counted and analyzed using ImageJ software (version 1.53t, Win64).

Measurement of brain tissue and hydrogel hardness using a nanoindenter

2.10

Male C57BL/6J mice, aged 1, 4, 8, and 12 weeks, were placed in a pre-prepared euthanasia chamber with a flat, clean, and well-ventilated base. Isoflurane gas (RWD, R510-22-10) was gradually introduced to induce anesthesia until the mice reached a deep anesthetic state. Euthanasia was then performed following standard protocols [39], and the brains were rapidly extracted and immediately placed in pre-chilled, oxygenated artificial cerebrospinal fluid (ACSF, Solarbio, H7701) for preservation. The brains were subsequently transferred to a pre-chilled (4 °C) mouse brain slicing mold (RWD, 68713), and coronal brain slices (500 μm thick) were prepared. Brain slices containing the hippocampal region were then placed in polyethylene glycol-coated petri dishes and immersed in pre-chilled, oxygenated ACSF to maintain tissue viability. The SGZ of the mouse hippocampus was identified using the 20 × objective of the Pavone nanoindenter. A 3 × 3 indentation array (10 μm spacing) was performed, and the entire procedure was completed within 30 min. Hydrogels with different stiffnesses were fully swollen in PBS and similarly placed in polyethylene glycol-coated petri dishes for subsequent measurements. The Young’s modulus values of both the mouse hippocampus and hydrogels were calculated based on the Hertz contact model. The specific formula of the Hertz model is as follows:

$$[eqn]$$

Image 1where E is the Young’s modulus of the sample, ν is the Poisson’s ratio, R is the radius of the spherical indenter, F is the applied load, and δ is the indentation depth. Young’s modulus was measured using a Pavone nanoindenter equipped with a probe (Serial Number: PV231048) featuring a geometrical factor in air of 2.67, stiffness of 0.44 N/m, and a tip radius of 25 μm. The Poisson’s ratio was set to 0.5, and the maximum indentation depth was 10 μm. Data were analyzed using DataViewer v2.7 software.

Primary culture of mouse and monkey NSCs

2.11

The hippocampal dentate gyrus was dissected from mice aged 1, 8, and 12 weeks, as well as from macaques aged 4 days and 16 years. Single-cell suspensions were prepared by digesting the tissue with Papain (Worthington, LS003119) following passage through a 40 μm cell strainer (Biosharp, BS-40-CS). The cells were then seeded into culture plates. The digestion and cell extraction procedures were performed according to protocols described in Refs. [40,41]. The digestion protocol was optimized, with the digestion solution containing DNase I (Invitrogen, 18047019, 100 U/mL), Papain (Worthington, LS003119, 1.22 mg/mL), L-Cystine (Sigma, C7352-10 MG, 2 mM), and EDTA (Sigma, E6758-500g). The digestion was terminated using a solution composed of OVO (Sigma, T9253-1g, 1.125 ng/mL), BSA (Sigma, A1933-5G, 0.525 mg/mL), DNase I (Invitrogen, 18047019), and DMEM/F12 (Gibco, 10565018). Mouse NSCs were cultured in DMEM/F12 (Gibco, 10565018) supplemented with B27 (Gibco, 17504-044, 2%), PS (Beyotime, C0222, 1%), EGF (Peprotech, AF-100-15, 20 ng/mL), and bFGF (Peprotech, AF-100-18B, 20 ng/mL). For macaque NSCs, the culture medium was adapted from reference [42], comprising Neurobasal® (Gibco, 21103049), L-Glutamine (Gibco, 25030081), NEAA (Gibco, 11140050), B27 (Gibco, 17504044, 2%), N2 (Gibco, 17502048), PS (Beyotime, C0222, 1%), and bFGF (Peprotech, AF-100-18B, 10 ng/mL).

Flow cytometry sorting of NSCs in hippocampal tissue

2.12

The hippocampal dentate gyrus of Nestin-GFP mice at 1 and 12 weeks of age was dissected and digested into a single-cell suspension. Cells were stained with Cy5-7-AAD (Beyotime, C1053S) to distinguish dead cells. The suspension was filtered through a 40 μm cell strainer (Biosharp, BS-40-CS) and resuspended in DMEMF12 medium (without phenol red, 21041025) containing 1% BSA (Sigma, A1933-5G) for flow cytometry sorting. A 100 μm nozzle was used for sorting. FITC^+^ Cy5^-^ cell populations (live Nestin^+^GFP^+^ NSCs) were collected. 200-500 cells were lysed in Smart-seq lysis buffer (10 × Lysis Buffer + RNase inhibitor, Vazyme, N712) on ice for 15 min. Then, the lysate was snap-frozen in dry ice and stored at −80 °C.

Brain tissue immunostaining and quantification

2.13

After perfusion of mice with 4% PFA (Biosharp, BL539A), the brains were fixed in 4% PFA for 12 h and then dehydrated in 30% sucrose (Sangon Biotech, A502792-0005) for 48 h. The brains were then embedded in OCT (SAKURA, 4583). Coronal brain sections (30 μm thickness) were cut using a Leica CM1950 UV cryostat. Slicing started at the anterior hippocampus and sequentially distributed across 12 collection wells. Each well contained eight consecutive sections spanning the dentate gyrus region, with adjacent sections spaced 330 μm apart [43]. All eight sections from one collection well were used for the immunofluorescence staining experiments. The sections were incubated in blocking solution containing 10% NDS and 0.4% Triton X-100 (prepared in PBS) at room temperature for 1 h. Following this, primary antibodies were diluted to the required concentration in an antibody dilution buffer (5% NDS, 0.4% Triton X-100, PBS). The sections were incubated overnight at 4 °C with primary antibodies: GFAP (Sigma, AB5541, 1:1000), GFAP (Invitrogen, 13-0300, 1:1000) Doublecortin (Abcam, AB18723, 1:500) and Yap1 (Proteintech, 13584-1-AP, 1:800). After washing with PBS three times, the sections were incubated with secondary antibodies at room temperature for 2 h, protected from light. The secondary antibodies used were donkey anti-chicken Alexa 488 (Jackson, 703-545-155, 1:2000), donkey anti-rabbit Cy3 (Jackson, 711-165-152, 1:1000), donkey anti-rat Cy3 (Jackson, 712-165-153, 1:1000) and DAPI (Abcam, ab228549, 1:1000). After removing the secondary antibodies and washing the sections three times with PBS, the sections were transferred onto glass slides and mounted with anti-fade mounting reagent (Beyotime, P0126). These sections were then imaged using a Nikon AX laser confocal microscope. The anatomically matched coronal sections spanning the anterior-posterior axis of the dentate gyrus (DG) were analyzed per biological replicate to account for regional heterogeneity. The granule cell layer (GCL) was manually delineated in ImageJ v1.53t (Fiji distribution) based on DAPI counterstaining, and the traced GCL area was used to calculate cellular density. All positive cell counts were normalized to unit area (cells in GCL/mm^2^), as indicated in Fig. 1c–e.

Cell culture immunostaining

2.14

Neural stem cells, neurons, and astrocytes were fixed at room temperature for 10 min with 4% PFA (Life-iLab, AC28L113), followed by three washes with PBS. The cells were then blocked and permeabilized at room temperature for 1 h with 10% NDS and 0.2% Triton X-100. Primary antibodies were diluted to specific concentrations in an antibody dilution buffer (5% NDS, 0.2% Triton X-100): Map2 (CST, 4542s, 1:500), GFAP (Sigma, AB5541, 1:1000), GFAP (Invitrogen, 13-0300, 1:1000), LaminB1 (Abcam, ab229025, 1:1000), Yap1 (Proteintech,13584-1-AP, 1:500), Nestin(BD Pharmingen, 611658, 1:1000), Tuj1 (Abcam, AB78078, 1:1000) and Piezo1 (Proteintech, 15939-1-AP, 1:500). Cells were incubated with primary antibodies at 4 °C for 12 h. After primary antibody incubation, cells were washed three times with PBS. Secondary antibodies, including donkey anti-chicken Alexa 488 (Jackson, 703-545-155, 1:2000), donkey anti-rabbit Cy3 (Jackson, 711-165-152, 1:2000), donkey anti-mouse Alexa 488 (donkey anti-mouse Alexa 488, 715-545-151, 1:2000), and donkey anti-rat Cy3 (Jackson, 712-165-153, 1:2000), were diluted in 5% NDS and 0.2% Triton X-100 and applied to cells for 2 h at room temperature, protected from light. After washing three times with PBS, cells were incubated with DAPI for 10 min. Finally, cells were washed three times with PBS, and imaging was performed using a Nikon AX laser confocal microscope.

EdU incorporation assay and quantification

2.15

Hippocampal NSCs isolated from mice and macaques of different ages were cultured on hydrogels with varying stiffness. EdU (Beyotime, C0075S) was incorporated into the cultures at different time points and durations. Proliferating cells were analyzed using a cell proliferation assay kit (Beyotime, C0075S) to detect the EdU signal in labeled cells. EdU signals were visualized using a Nikon AX laser confocal microscope. The proportion of EdU^+^ cells was quantified using ImageJ (version win64-1.53t).

Calcein/PI assay for cell viability detection

2.16

NSCs cultured from the hippocampus of 8-week-old mice were seeded at a density of 2 × 10^4^ cells/cm^2^ onto HA hydrogels of varying stiffness (soft, medium, Stiff). After 48 h, the cells were stained using the Calcein/PI cell viability and cytotoxicity assay kit (Beyotime, C2015M) to assess the viability of NSCs on the different stiffness hydrogels. Cell immunofluorescence signals were captured using a Nikon AX laser confocal microscope, while live and dead cell counts based on Calcein/PI staining were quantified with ImageJ (version win64-1.53t).

SA-β-gal staining of neurospheres

2.17

Young (1-week-old) and old (12-week-old) mouse hippocampal NSCs were seeded evenly at a density of 2 × 10^4^ cells/cm^2^ onto soft and stiff HA hydrogels. Each condition was set in triplicate wells, and the culture medium was changed every 48 h. When the diameter of the neurospheres reached approximately 100 μm, SA-β-galactosidase staining was performed following the manufacturer’s instructions (Beyotime, C0602). Cells were washed three times with PBS and fixed with 4% paraformaldehyde (Life-iLab, AC28L113) for 15 min. After fixation and another three PBS washes, neurospheres were incubated overnight at 37 °C (in a non-CO_2_ incubator) with the β-galactosidase staining solution containing X-Gal. Following staining, neurospheres were washed three more times with PBS, imaged using an inverted microscope (Nikon, Eclipse Ts2), and at least 10 random fields per well were captured. Quantification of SA-β-gal signal mean density was performed using ImageJ software (version win64-1.53t).

Neuronal differentiation of mouse and monkey NSCs

2.18

Mouse NSCs at a density of 3 × 10^4^ cells/cm^2^ were seeded onto HA hydrogels of varying stiffness. After 24 h of culture, the medium was replaced with differentiation medium for mouse NSCs: DMEM/F12 (Gibco, 10565018), B27 (Gibco, 17504-044, 2%), and PS (Beyotime, C0222, 1%). The differentiation medium was changed every 48 h. After 6 days of free differentiation, the cells were fixed and stained.

Macaque NSCs at a density of 3 × 10^4^ cells/cm^2^ were seeded onto HA hydrogels of varying stiffness. After 24 h of culture, the medium was replaced with differentiation medium for monkey NSCs [40], consisting of insulin (Sigma, I2643, 5 μg/mL), Forskolin (Selleck, S2449, 10 μM), cAMP (MCE, HY-B1511, 50 μM), B27 (Gibco, 17504-044, 2%), PS (Beyotime, C0222, 1%), N2 (Gibco, 17502048, 1%), L-Glutamine (Gibco, 25030081, 1%), CNTF (B&D Systems, 257-NT010, 20 ng/mL), GDNF (STEMCELL, 78058, 20 ng/mL), and BDNF (STEMCELL, 78005, 20 ng/mL). The medium was replaced every 48 h. After 21 days of neuronal differentiation, the cells were fixed and stained.

To quantify the proportion of differentiated neurons and astrocytes, we performed immunocytochemical analysis using the Cell Counter plugin in ImageJ. First, the DAPI channel image was loaded to manually count all DAPI-positive nuclei, representing the total cell population. Next, in the merged DAPI/GFAP and DAPI/MAP2 images, GFAP^+^ astrocytes and MAP2^+^ neurons were identified based on colocalization with DAPI staining, ensuring careful exclusion of double-counted cells. The percentages of GFAP^+^ astrocytes and MAP2^+^ neurons relative to the total DAPI^+^ cells were then calculated to determine differentiation efficiency.

RNA-seq library construction and data analysis

2.19

Neurospheres approximately 100 μm in diameter and in the proliferative state were carefully aspirated from HA hydrogels with varying stiffnesses using a glass pipette, washed 3-4 times with PBS, and immediately lysed on ice for 15 min using Smart-seq lysis buffer (10 × Lysis Buffer + RNase inhibitor, Vazyme, N712).

After flow cytometry sorting, Nestin-GFP^+^/7-AAD^-^ NSCs were counted using a hemocytometer, and 350 cells were collected. The cells were then lysed in Smart-seq lysis buffer on ice for 15 min.

The first strand cDNA and full-length cDNA of neurospheres and FACS-sorted NSCs were synthesized using the Single Cell Full Length mRNA Amplification Kit (Vazyme, N712). cDNA amplification products were then purified using VAHTS DNA Clean Beads (BECKMAN, B23318), which were equilibrated at room temperature for 30 min. The concentration of the purified cDNA was measured using the Invitrogen Qubit 4 Fluorometer. Following quantification, 1 ng of cDNA was used for library preparation using the TruePrep® DNA Library Prep Kit V2 for Illumina (Vazyme, TD503), and sequencing was performed using the Novaseq-6000 platform.

For sequencing data analysis, adapters were trimmed, and the resulting data were mapped to the GRCm39 genome using Hisat2. Feature Counts was used to summarize read counts. Differentially expressed genes (DEGs) were calculated using DESeq2. Principal Component Analysis (PCA) was used to visualize the similarity of gene expression patterns among different samples. Venn diagrams identified genes that change specifically and commonly during aging and matrix stiffening, including both upregulated (Log2 FC > 1; p-value <0.05) and downregulated genes (Log2 FC <-0.5; p-value <0.05). The overlapping genes identified by the Venn diagrams were analyzed for Gene Ontology Biological Process (GOBP) enrichment using the DAVID v2023q4 website [44]. Heatmaps were used to display changes in cell cycle and adhesion genes. The gene interaction network was created using STRING version 12.0 [45].

qPCR

2.20

NSCs from in vivo (sorted by FACS) and in vitro (cultured on hydrogels of varying stiffness) were collected, and RNA was extracted using the RNA Easy Fast Tissue/Cell Kit (TIANGEN, DP451). A total of 500 ng of RNA was reverse-transcribed using Go Script™ Reverse Transcriptase (Promega, A2801). Real-time PCR was performed on a Bio-Rad CFX96 system using ChamQ Universal SYBR qPCR Master Mix (Vazyme, Q711-02). The fold change in gene expression was calculated using the 2-ΔΔCt method, with normalization to the transcript level of the housekeeping gene Gapdh. The primer sequences for qPCR are listed in Table S4.

Yoda1 treatment

2.21

Mouse NSCs at a density of 3 × 10^4^ cells/cm^2^ were seeded onto HA hydrogels of varying stiffness (soft, stiff). After 24 h of culture, the medium was replaced with medium containing 2 μM Yoda1 (MCE, HY-18723) to treat the cells for 48 h. Subsequently, 15 μM EdU was added to the medium for 4 h to label proliferating cells. Cell proliferation was then assessed using the EdU-555 Cell Proliferation Detection Kit (Beyotime, C0075S).

pSicoR-Piezo1 shRNA plasmid construction

2.22

To knock down the target gene Piezo1, we used the pSicoR plasmid vector (AXYBIO, GS-1447) as the delivery system. The target sequences for Piezo1 knockdown were designed as follows: #1, CACCGGCATCTACGTCAAATA; #4, CGGAATCCTGCTGCTGCTATA [46].

Lentiviral packaging and infection

2.23

HEK293T cells were transfected with pSicoR-Piezo1 shRNA, pSicoR-Scramble, VSVG, and psPAX2 plasmids using PEI 40K transfection reagent (MKBio, MX2203-250 MG) to package lentivirus. After 48 and 72 h post-transfection, the culture supernatant containing the virus was collected and filtered through a 0.45 μm filter (Merck-Millipore, SLHU033RB). The viral supernatant was then centrifuged at 25,000 rpm at 4 °C for 2.5 h using a Beckman Coulter XPN-100 ultracentrifuge to collect the lentiviral particles. The purified lentiviral particles were used to infect mouse NSCs, and after 24 h, the medium was replaced with fresh culture medium. Cells were harvested for subsequent experiments once they reached approximately 80% confluency.

Data presentation and statistical analysis

2.24

Data presentation was performed using GraphPad Prism software version 8.0.2 (263). The corresponding statistical methods are described in the figure legends. Differences between the two groups were assessed using a two-tailed unpaired t-test. Data are presented as mean values ± SD or SEM. One-way ANOVA was used to compare differences among more than two groups. Two-way ANOVA was used to compare differences between groups with two independent variables and to assess both main effects and their interaction. Statistical significance was indicated as follows: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Results

3

Age-related hippocampal stiffening correlates with reduced neurogenesis and is faithfully recapitulated by laminin-modified hydrogels

3.1

It is known that neurogenesis declines with aging due to reduced NSC proliferation and neuronal differentiation, which indicates NSC aging [7,47,48]. Niche signals, including matrix stiffness, have been recognized as important regulators of stem cell functions [20,21,49]. Our study aims to elucidate the role of tissue stiffness in regulating NSC senescence and its underlying mechanisms. We first validated the correlation between niche stiffness and age-related declines in neurogenic capacity. Specifically, we quantified neurogenesis levels and measured the SGZ stiffness within the hippocampal dentate gyrus across multiple age groups in mice. To examine neurogenesis levels in the hippocampal dentate gyrus across different age groups in mice, we performed EdU incorporation assays together with GFAP and DCX immunostaining to identify proliferating cells and NSC lineage (Fig. 1a and b), and quantified their numbers in the dentate gyrus of mice at different ages (Fig. 1c–e). Consistent with prior studies [47], we noted a gradual decline in the proliferation of NSCs with advancing age, as indicated by the diminished number of active radial glia-like stem cells (EdU^+^GFAP^+^ RGLs) (Fig. 1c). Furthermore, there was a marked reduction in neurogenesis with age, indicated by fewer neuroblasts (EdU^+^DCX^+^) and newborn neurons (DCX^+^) (Fig. 1d and e). In parallel, we measured the stiffness of hippocampal tissues at different postnatal ages using a Pavone nanoindenter (Fig. 1f). Coronal brain slices from mice aged 1, 4, 8, and 12 weeks were subjected to nanoindentation tests to evaluate tissue stiffness. Representative indentation maps of the dentate gyrus (DG) at each stage, with dashed outlines indicating the areas within the SGZ, were measured (Fig. 1g). Quantitative analysis revealed a progressive increase in Young’s modulus from 1 to 12 weeks (129.6 Pa, 494.8 Pa, 716.1 Pa, and 959.2 Pa, respectively; Fig. 1h), indicating that hippocampal tissue stiffness markedly increases during postnatal development. Collectively, these data suggest that the age-associated decline in NSC proliferation and neurogenesis is closely correlated with the increased stiffness of the hippocampal microenvironment.

To circumvent the challenges of directly probing the effects of hippocampal tissue stiffness on NSCs in vivo without compromising tissue integrity, we used an HA hydrogel cross-linked with laminin as an ECM scaffold. This material, owing to its excellent biocompatibility with neural tissue and strong cell adhesion properties [[50], [51], [52], [53]], effectively mimicked the NSC microenvironment, enabling us to explore the role of stiffness in NSC aging. Based on in vivo tissue stiffness measurements, we synthesized a series of HA hydrogels with tunable stiffness by systematically varying the adipic dihydrazide (ADH) to HA ratio, and covalently crosslinked laminin to the HA hydrogels (Fig. 1i). These HA-laminin hydrogels were then characterized using the same Pavone Nanoindenter to evaluate their mechanical stiffness (Fig. 1f). As measured with the Pavone Nanoindenter, we selected soft (112 Pa), medium (345.2 Pa), and stiff (1048 Pa) hydrogels to simulate the newborn (1 week), juvenile (4 weeks), and adult (12 weeks) hippocampal SGZ niche, respectively (Fig. 1j).

To systematically characterize the physical properties of hydrogels with varying stiffness, we evaluated their degradation kinetics, equilibrium swelling, rheological behavior, diffusion property and microstructural morphology. Hydrogels with higher stiffness exhibited slower degradation rates and lower equilibrium swelling ratios compared to their softer counterparts (Fig. S1a and b). Rheological measurements indicated a predominantly elastic response for all hydrogels, with the storage modulus (G′) surpassing the loss modulus (G″) across the entire frequency range tested. The Stiff hydrogels exhibited the highest G′, the broadest linear viscoelastic region, and the greatest resistance to yielding, whereas soft hydrogels showed the lowest elastic response and the highest sensitivity (Fig. S1c and d). In Crystal Violet diffusion assays, small-molecule transport was not substantially impeded over 24 h, although diffusion rates were modestly faster in Soft hydrogels (Fig. S1e). To preserve the native microstructure of the HA–laminin hydrogels, samples were rapidly frozen in liquid nitrogen and freeze-dried before scanning electron microscopy (SEM) analysis. The resulting micrographs confirmed that all hydrogels maintained porous structures. While the average pore size slightly decreased with increasing stiffness, the differences were minimal. Specifically, average pore diameters were 46.15 μm (Soft), 40.04 μm (Medium), and 39.67 μm (Stiff) (Fig. S1f and g).

Despite these differences in hydrogel physical properties, neural stem cells (NSCs) cultured on substrates of varying stiffness maintained their stemness and undifferentiated state (Nestin^+^/Map2^-^) when grown in growth factor-supplemented medium (Fig. S2a). Under appropriate differentiation conditions, they exhibit the capacity to differentiate into neurons and astrocytes (Fig. S2b). The survival ratio of NSCs grown on all these selected HA-laminin hydrogels exceeded 95%, revealed by cell viability assay based on Calcein-AM and PI staining (Fig. S2c and e). In addition, NSCs exhibited comparable adhesion across HA-laminin hydrogels of varying stiffness, with no observable influence of adhesion differences on cellular growth (Fig. S2d and f). These results demonstrate that HA-laminin hydrogels with tunable stiffness support normal survival and growth of hippocampal NSCs, providing a suitable platform to investigate how microenvironmental stiffness regulates NSC aging. It is also known that Yes-associated protein (YAP1) undergoes cytoplasm to nucleus translocation in response to mechanical stress [35,54,55]. To determine whether hippocampal NSCs respond to matrix stiffness variations, we examined YAP1 subcellular localization in NSCs cultured on soft, medium, and stiff HA-laminin hydrogels. We observed stiffness-dependent nuclear translocation of YAP1, with increasing substrate stiffness promoting YAP1 accumulation in the nucleus rather than the cytoplasm (Fig. 1k and l). Together, these findings demonstrate that hippocampal NSCs are mechanosensitive to substrate stiffness and that our engineered HA-laminin hydrogels faithfully recapitulate the mechanical properties of native hippocampal tissue. This system provides an ideal platform for investigating how niche stiffness governs NSC fate decisions.

Hydrogels mimicking age-dependent hippocampal stiffness remodel neural stem cell proliferation and differentiation potential

3.2

To examine how niche stiffness affects the fate of hippocampal NSCs, we first cultured the adult hippocampal NSCs derived from 8-week-old mice. We then compared their proliferation and differentiation capacities on hydrogels of varying stiffness—soft (112 Pa), medium (345.2 Pa), and stiff (1048 Pa). The proliferation of NSCs on these hydrogels was evaluated at different time points (Day 1, 3, 5, and 7) through the EdU incorporation assay (Fig. S3a). The EdU incorporation ratios in the NSCs cultured on HA-laminin hydrogels with differential stiffness revealed that the NSCs on soft hydrogels exhibited the most robust proliferation at all observed time points. Conversely, the NSCs on medium and stiff hydrogels demonstrated progressively diminished proliferation rates (Fig. S3b and c). Consistent with these findings, the NSCs cultured on soft hydrogels generated larger-sized neurospheres compared to those on stiff hydrogels (Fig. S3d). Next, the differentiation potential of NSCs was analyzed by removing growth factors to induce spontaneous differentiation (Fig. S3e). Immunostaining and quantification of GFAP^+^ astrocytes and MAP2^+^ neurons showed that, on soft HA-laminin hydrogels, approximately 50% of NSCs differentiated into neurons and 27% into astrocytes. Conversely, on medium stiffness hydrogels, differentiation resulted in 43% neurons and 37% astrocytes. On stiff hydrogels, 25% neurons and 53% astrocytes were observed (Fig. S3f and g). Collectively, these data indicated that softer substrates promote robust proliferation and neuronal differentiation, whereas stiffer substrates favor astrocytic differentiation.

Since there are subtle differences in pore size and architecture among the three hydrogels, and pore size could regulate cell behavior [56], we further employed commercial CytoSoft® silicone gel plates with comparable stiffness to rigorously validate that the observed changes in cell phenotypes were mainly due to stiffness. These plates feature uniform nanoscale pores with no significant differences across stiffness conditions, thereby minimizing the confounding influence of surface pore morphology. NSCs were seeded onto soft (matching the stiffness of soft hydrogels, 0.2 kPa), medium (matching the stiffness of medium hydrogels, 0.5 kPa), and stiff (matching the stiffness of stiff hydrogels, 2 kPa) CytoSoft® six-well plates, and cell growth was monitored at 24, 48, and 72 h (Fig. S4a). Phase-contrast microscopy (Fig. S4b) revealed higher cell densities on soft substrates at all time points. Quantitative analysis confirmed this trend (Fig. S4c), with soft substrates exhibiting ∼1600 cells/mm^2^ at 72 h—significantly higher than medium (∼1200 cells/mm^2^) and stiff (∼800 cells/mm^2^) substrates, particularly at later stages. EdU incorporation assays (Fig. S4d and e) further showed the highest percentage of EdU-positive cells on soft substrates, followed by medium and stiff substrates. Together, these data suggest that substrate stiffness is a key determinant of hippocampal NSC proliferation.

Given that hippocampal neurogenesis declines and hippocampal tissue stiffness increases with age, we aimed to investigate whether a soft hydrogel mimicking the young niche could rejuvenate old NSCs, while a stiff hydrogel mimicking the old niche might impair the self-renewal and differentiation of young NSCs. To this end, we isolated young and old NSCs from the hippocampal DG of 1-week-old and 12-week-old mice and cultured them separately on soft and stiff HA-laminin hydrogels. We then measured the ratio of proliferating cells using the EdU incorporation assay (Fig. 2a). For young NSCs (1-week-old), we found that the soft hydrogel, which mimics the in vivo young niche, effectively supports their proliferation, while the stiff hydrogel, which mimics the old niche, significantly restricts their proliferation (Fig. 2b and c). For old NSCs (12-week-old), the stiff hydrogel, which mimics the in vivo old niche, impaired their proliferation, while the soft hydrogel, which mimics the young niche, partially rejuvenated their proliferation (Fig. 2d and e).Fig. 2Hydrogels mimicking age-dependent hippocampal stiffness modulate the proliferation and differentiation potentials of NSCs from different ages. (a) The schematic of the EdU labeling assay used to evaluate the proliferation and differentiation capacity of young (1-week-old) and old (12-week-old) NSCs cultured on soft (112 Pa) and stiff (1048 Pa) HA-laminin hydrogels. (b) Representative images showing EdU incorporation in young NSCs cultured on 112 Pa (mimicking soft niche) and 1048 Pa (mimicking old niche) HA-laminin hydrogels for 3 and 5 days. (c) Quantification of the ratio of EdU^+^ young NSCs as in (b) (n = 3 wells). (d) Representative images showing EdU incorporation in old NSCs cultured on 112 Pa (mimicking soft niche) and 1048 Pa (mimicking old niche) HA-laminin hydrogels for 3 and 5 days. (e) Quantification of the ratio of EdU^+^ old NSCs as in (d) (n = 3 wells). (f) The strategy for assessing young and old NSCs differentiation potential on soft (112 Pa) and stiff (1048 Pa) HA-laminin hydrogels. All NSCs were induced for spontaneous differentiation for 7 days in DMEM/F12 medium without growth factors. (g, i) Immunostaining of neurons and astrocytes differentiated from young (g) and old (i) NSCs cultured on soft (112 Pa) and stiff (1048 Pa) HA-laminin hydrogels, respectively. Map2^+^ (red) and GFAP^+^ (green) cells indicate neurons and astrocytes, respectively. Scale bar, 50 μm. (h, j) Quantification of the proportion of neurons (Map2^+^) and astrocytes (GFAP^+^) differentiated from young (g) and old (i) NSCs. n = 3 or 4. Statistical significance was determined using two-way ANOVA, with Bonferroni’s Multiple Comparisons Test. Data are presented as mean ± SD (c,e). Statistical significance was determined using unpaired Student’s t tests. Data are presented as mean ± SD (h, j) (∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001).Fig. 2

Next, we investigated whether environmental stiffness can also remodel the differentiation potential of young and old NSCs. To test this, NSCs from both age groups were cultured on soft and stiff hydrogels for 7 days under growth factor-free conditions to permit spontaneous differentiation (Fig. 2f). Upon analyzing the differentiation potential of young NSCs using lineage marker immunostaining, we found that the stiff hydrogel, which mimics the old hippocampal niche, strongly impairs the neuronal differentiation of young NSCs (Fig. 2g and h); while the soft hydrogel, which mimics the young hippocampal niche, significantly enhances the neuronal differentiation of old NSCs (Fig. 2i and j). In summary, both young and old NSCs exhibit increased neuronal differentiation potential when cultured on soft hydrogels mimicking the young hippocampal niche, while stiff hydrogels replicating the old niche promote preferential astrocytic differentiation. These findings indicate that age-related neurogenic decline is partially mediated by niche stiffening and suggest that mechanical softening of the old niche could partially restore proliferation capacity in old NSCs while promoting neuronal differentiation.

Hydrogels mimicking age-dependent hippocampal stiffness regulate NSC aging phenotypes

3.3

Age-related changes in NSCs are characterized not only by diminished proliferative capacity and altered differentiation potential, but also by established cellular senescence markers, including LaminB1 loss, increased senescence-associated β-galactosidase (SA-β-gal) activity, and elevated p16 (Cdkn2a)/p21 (Cdkn1a) expression. To investigate how young (soft, 112 Pa) versus old (stiff, 1048 Pa) hippocampal niche-mimicking hydrogels influence NSC senescence phenotypes, we systematically quantified these hallmark senescence markers [38].

Lamin B1 is an intermediate filament protein that forms a key structural component of the nuclear lamina and plays critical roles in regulating cellular senescence and aging-related processes. Notably, its expression demonstrates significant age-dependent decline across multiple cell types [[57], [58], [59]]. Immunofluorescence analysis revealed significantly reduced Lamin B1 intensity in both young (1-week-old) (Fig. 3a and b) and old (12-week-old) (Fig. 3c and d) NSCs cultured on stiff hydrogels (mimicking the old niche) compared to soft hydrogels (simulating the young niche). We next assessed SA-β-gal activity in young and old NSCs cultured on soft versus stiff hydrogels (Fig. 3e–h). This lysosomal enzyme, which hydrolyzes β-galactosides into monosaccharides, exhibits elevated activity in lysosomes at low pH and serves as a well-established biomarker of cellular senescence [60]. The experimental results indicate that, regardless of whether the NSCs are derived from young or old mice, those cultured on stiff hydrogels mimicking the old hippocampal niche exhibit stronger SA-β-gal positive staining signals (Fig. 3f–h). Since p16 and p21 are senescence-associated markers closely linked to cell cycle arrest [[61], [62], [63], [64]], we analyzed their expression both in vivo (in FACS-sorted 1-week-old and 12-week-old NSCs) and in vitro (in 12-week-old NSCs cultured on soft versus stiff hydrogels (Fig. 3i,j, Fig. S3). The results demonstrated significantly upregulated p16 and p21 expression levels in both old NSCs and those cultured on stiff hydrogels, indicating that age-dependent niche stiffening modulates senescence-associated gene expression.Fig. 3Senescence marker analysis of young and old NSCs cultured on soft versus stiff hydrogels reveals age-related remodeling. (a, c) Lamin B1 immunofluorescence (green) of young (1 week) and old (12 weeks) NSCs cultured on soft (112 Pa) versus stiff (1048 Pa) HA-laminin hydrogels. White dashed boxes indicate magnified regions. Scale bars: 50 μm (main), 25 μm (insets). (b, d) Quantification of Lamin B1 intensity of young NSCs and old NSCs on soft and stiff HA hydrogels (young NSCs on soft, n = 40 fields; young NSCs on stiff, n = 49 fields; old NSCs on soft, n = 63 fields; old NSCs on stiff, n = 56 fields). Data pooled from 3 independent experiments. (e,g) SA-β-gal (blue) staining of young (1 week) and old (12 weeks) NSCs cultured on soft and stiff HA hydrogels. Scale bars, 50 μm. (f,h) SA-β-gal signal density quantification of young NSCs and old NSCs on soft and stiff HA hydrogels (young NSCs on soft, n = 85 fields; young NSCs on stiff, n = 73 fields; old NSCs on soft, n = 46 fields; old NSCs on stiff, n = 66 fields). Old NSC neurospheres on soft aged NSCs on soft (n = 46 fields), stiff (n = 66 fields). (i,j) qPCR analysis of p16 and p21 mRNA levels in FACS-isolated young/old NSCs and hydrogel-cultured groups (n = 4 biological replicates). For all quantification data, statistical significance was determined using unpaired Student’s t tests. Data are presented as mean ± SD (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001).Fig. 3

Niche stiffness regulates the cell cycle and adhesion gene expression of NSCs during aging

3.4

To investigate the molecular mechanisms by which niche stiffness regulates NSC aging, we performed RNA sequencing on NSCs sorted from the hippocampal tissues of 1-week-old and 12-week-old Nestin-GFP mice (Fig. S5), as well as on 12-week-old NSCs cultured on soft (112 Pa) and stiff (1048 Pa) hydrogels, mimicking the stiffness of young and old NSC niches (Fig. S6a). We compared the transcriptomic changes of NSCs from different ages and those cultured on hydrogels of varying stiffness (Fig. S6a, Supplementary Table 1). Spearman sample correlation analysis revealed that the gene expression profile of NSCs cultured on soft hydrogels more closely resembled that of NSCs sorted from 1-week-old mice than those cultured on stiff hydrogels (Fig. S6b). This finding indicates that substrate stiffness significantly influences NSC gene expression: old NSCs cultured on soft hydrogels display greater transcriptional similarity to young NSCs than those cultured on stiff hydrogels.

We next compared gene expression between old NSCs (12-week-old) on soft and stiff hydrogels, as well as between young NSCs sorted from 1-week-old mice and old NSCs sorted from 12-week-old mice. Using Venn diagrams, we identified the common upregulated and downregulated genes during NSC aging and niche stiffening, and presented the number and proportion of these genes (Fig. 4a and b; Supplementary Tables 2 and 3). These common upregulated and downregulated genes during NSC aging and niche stiffening were further analyzed by Gene Ontology (GO) categories. The top 10 enriched Gene Ontology Biological Process (GOBP) terms for upregulated and downregulated genes were shown separately (Fig. 4c and d**).** The top 10 upregulated GOBPs during aging and niche stiffening include signal transduction, cell differentiation, and cell adhesion (Fig. 4c). Conversely, the top 10 downregulated GOBPs are related to the cell cycle and cell division (Fig. 4d). We examined the expression of cell adhesion and cell cycle genes and visualized their expression in sorted young and old NSCs, as well as in old NSCs cultured on soft and stiff hydrogels, using a heatmap. We observed that the expression patterns of cell adhesion and cell cycle genes in old NSCs cultured on soft hydrogels are more similar to those in young NSCs, with cell adhesion genes downregulated and cell cycle genes upregulated. In contrast, the gene expression of NSCs cultured on stiff hydrogels resembles the old NSC expression pattern (Fig. 4e and f**). The mRNA expression alterations of cell cycle (such as Cdk4, Haus1, and Tipin) and cell adhesion genes (such as Col6a4, Col6a6 and Itgad) during aging and niche stiffening have also been confirmed by RT-qPCR (Fig. 4g–l). Together, these findings establish niche stiffness as a regulator of aging-associated transcriptional programs in NSCs, with predominant effects on cell cycle and adhesion pathways. However, the mechanosensing mechanisms by which NSCs perceive stiffness changes and transduce these signals into gene expression alterations remain unknown.Fig. 4Comprehensive transcriptional analysis of NSCs in aging hippocampus and on niche stiffness–mimicking hydrogels (a,b**) The Venn diagram illustrates the proportion and number of genes that change respectively and commonly during aging and niche stiffening, including both upregulated (Log_2_FC > 1; p-value <0.05) and downregulated genes (Log_2_FC <-0.5; p-value <0.05). (c,d) GOBP analysis of 1563 commonly upregulated genes and 251 commonly downregulated genes during aging and niche stiffening was conducted using DAVID. The top 10 enriched GO terms are shown. Signal transduction and cell adhesion genes were enriched in commonly upregulated GOBP terms, while cell cycle and cell division genes were enriched in commonly downregulated GOBP terms. (e,f) The heatmap displays the clustering results of expression patterns for GOBP-enriched cell cycle genes and cell adhesion genes in young and old NSCs, as well as old NSCs cultured on soft and stiff hydrogels. NSCs cultured on soft hydrogels exhibit similar gene expression patterns to young NSCs in vivo, while those on stiff hydrogels resemble old NSCs in vivo. Genes validated by qPCR in the extended data are highlighted in red. (g-i) Gene expression validation of Cdk4, Haus1, Tipin in the cell cycle pathway (n = 3). (j–l) Gene expression validation of Col6a4, Col6a6 and Itgad in the cell adhesion pathway (n = 3). Statistical significance was determined using unpaired Student’s t tests. Data are presented as mean ± SD (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001).Fig. 4

Niche stiffness regulates hippocampal NSC proliferation during aging through Piezo1 activity

3.5

Since Piezo1 (Piezo-type mechanosensitive ion channel component 1), which has been reported as an upstream sensor that detects matrix stiffness and initiates mechano-transduction [20,24,65], we first performed STRING network analysis integrating Piezo1 with differentially expressed genes (DEGs) enriched in cell cycle and cell adhesion pathways. The resulting interaction network suggests potential functional connections between these DEGs and Piezo1-mediated mechanotransduction (Fig. S6c and d). Notably, both qPCR and RNA-seq results revealed significantly elevated Piezo1 expression in both old NSCs and those cultured on stiff hydrogels (Fig. 5a, Fig. S7a, Supplementary Tables 1 and 2), strongly supporting this mechanistic link.Fig. 5Intervention of Piezo1 activity altered NSC proliferation on hydrogels mimicking different niche stiffness. (a) The mRNA expression level of Piezo1 in FACS-Young, FACS-Old, Hydrogel-Soft, and Hydrogel-Stiff NSCs groups (n = 3). (b) The strategy for treating 8-week-old adult NSCs cultured on soft and stiff hydrogels with the Piezo1 agonist Yoda1 and performing EdU labeling assays. (c) EdU detection in adult NSCs on soft and stiff hydrogels treated with 2 μm Yoda1 for 48 h, compared to DMSO mock. Scale bar, 50 μm. (d) Quantification of EdU^+^ proliferating NSCs as in (c) (n = 3). (e) Piezo1 protein immunostaining in the scramble control and Piezo1 knockdown NSCs. Scale bars, 100 μm and 50 μm as indicated. (f) qPCR analysis of Piezo1 mRNA expression levels in the scramble control and Piezo1 knockdown NSCs (n = 3). (g) EdU detection in scramble control and Piezo1 knockdown NSCs cultured on stiff hydrogels. Scale bar, 100 μm. (h) Quantification of EdU^+^ NSCs as in (g) (n = 3). (i-l) qPCR analysis of Cdk4, Haus1, Col6a4, and Col6a6 genes in the DMSO and 2 μM Yoda1 treatment groups, n = 3. (m-p) qPCR analysis of Cdk4, Haus1, Col6a4, and Col6a6 genes in the Scramble and Piezo1 knockdown groups, n = 3. Statistical significance was determined using unpaired Student’s t tests (a, f, h-p). Statistical significance was determined using one-way ANOVA with Tukey’s multiple comparison tests (d). Data are presented as mean ± SEM (a) and mean ± SD(d, f, h, i-p) (∗p < 0.05, ∗∗p < 0.01).Fig. 5

To further examine whether Piezo1-mediated mechanotransduction regulates NSC proliferation during aging, we treated adult NSCs (8-week-old) cultured on soft and stiff hydrogels with Yoda1 (a Piezo1 agonist) and labeled them with EdU to assess its effect on proliferation (Fig. 5b). Yoda1 directly activates Piezo1 ion channels by binding to specific sites on the Piezo1 protein, inducing conformational changes that promote the influx of calcium and other cations, thereby mimicking mechanical stimulation [66]. We found that NSCs grown on soft hydrogels showed no significant difference in proliferation with or without Yoda1 treatment. However, NSCs grown on stiff hydrogels became more sensitive to stiffness with Yoda1 treatment, resulting in a more pronounced inhibition of proliferation (Fig. 5c and d). Furthermore, we knocked down Piezo1 using shRNA in 8-week-old NSCs and observed reduced mRNA and protein expression levels of Piezo1 (Fig. 5e and f). We found that cell proliferation increased in Piezo1 knockdown NSCs cultured on stiff hydrogels and plastic plates (∼1 GPa, mimicking an extremely stiff surface) compared to the scramble control (Fig. 5g,h and Fig. S7b and c). Consistently, NSCs cultured on hydrogels and treated with Yoda1 generated more small-sized neurospheres compared to those without treatment (Fig. S7d), while Piezo1 knockdown NSCs generated more large-sized neurospheres compared to the scramble control (Fig. S7e). Next, we examined whether knockdown or specific activation of the upstream mechanosensor PIEZO1 in NSCs cultured on stiff hydrogels and plastic plates (∼1 GPa) influences the expression of downstream cell cycle– and cell adhesion genes. qPCR analysis revealed that specific activation of Piezo1 with 2 μM Yoda1 in NSCs on these stiff substrates downregulated the cell cycle–associated genes Cdk4and Haus1, while upregulating the ECM-related genes Col6a4 and Col6a6 (Fig. 5i–l, Fig. S7f–i). Conversely, Piezo1 knockdown increased Cdk4 and Haus1 expression but decreased Col6a4 and Col6a6 levels in NSCs on both stiff hydrogels and plastic plates (Fig. 5m–p, Fig. S7j–m). Together, these results identify Piezo1 as a key mechanosensory regulator that links extracellular stiffness cues during aging to the modulation of NSC proliferation and coordinated transcriptional changes in downstream targets.

Conserved regulation of NSC aging by niche stiffness in monkeys

3.6

Building on our rodent studies demonstrating that age-dependent hippocampal stiffening contributes to NSC aging, we investigated whether this mechanism is conserved in primates. Using rhesus monkeys (Macaca mulatta), we investigated age-related changes in hippocampal NSCs, with a focus on their responses to niche stiffness, specifically alterations in proliferative capacity, differentiation potential, and gene expression profiles (Fig. 6a). We cultured young versus old monkey NSCs on soft and stiff hydrogels modeling their respective niches. Subsequently, we conducted proliferation and differentiation assays to analyze their proliferative and differentiation capacities. In the experiment using EdU incorporation to detect proliferation, adult monkey NSCs (old) exhibited much slower proliferation compared to neonatal NSCs (young). Therefore, we labeled the neonatal NSCs with EdU for a short duration (4 h) and the adult NSCs with EdU for a longer duration (24 h) (Fig. 6b and c). When young monkey NSCs were cultured on stiff hydrogel (1048 Pa), their proliferation was evidently restricted (Fig. 6d and e). Conversely, when old monkey NSCs were cultured on soft hydrogel (112 Pa), which mimics a young niche, their proliferation capacity was partially rejuvenated (Fig. 6f and g). For the differentiation assay (Fig. 6h), young monkey NSCs cultured on soft hydrogel, closely resembling their in vivo niche, differentiated well into neurons, as indicated by strong Map2 and Tuj1 immunostaining (Fig. 6i). In contrast, when young monkey NSCs were cultured on stiff hydrogel, which mimics the old niche, they differentiated more into astrocytes and could not spread well (Fig. 6i). Regarding old monkey NSCs, the soft niche could improve their differentiation status slightly, but not to the extent observed in mice (Fig. 2i and j; Fig. 6j). The ability of old monkey NSCs to differentiate into neurons still faces significant barriers and does not show substantial improvement simply due to the softening of the niche. To test the underlying molecular mechanism, we examined the gene expression of young and old monkey NSCs from primary neurospheres, as well as old monkey NSCs cultured on soft and stiff hydrogels. We found that PIEZO1 was upregulated in both old monkey NSCs and NSCs cultured on stiff hydrogels (Fig. 6k). Meanwhile, cell cycle gene HAUS1 was downregulated (Fig. 6l), and cell adhesion genes, such as ITGAD and COL6A6, were upregulated (Fig. 6m and n). These results were similar to those observed in mice. Thus, niche stiffness regulating hippocampal NSC aging is largely conserved between mice and monkeys. The soft niche, resembling a young environment, can partially rejuvenate the activity of old monkey NSCs by rescuing cell cycle and cell adhesion gene expression patterns. However, the rejuvenation effect is more limited for monkey NSCs compared to mouse NSCs.Fig. 6Hydrogels mimicking niche stiffness modulate hippocampal NSC potential and gene expression in monkeys of different ages. a, The schematic illustrates the experimental workflow, including the isolation of NSCs from the hippocampi of young (4-day-old) and old (16-year-old) monkeys, followed by their culture on soft or stiff HA-laminin hydrogels to assess proliferation and differentiation. b,c, The EdU labeling strategy to evaluate the proliferation capacity of young (b) and old (c) monkey hippocampal NSCs cultured on soft and stiff HA-laminin hydrogels. d, Representative images showing 4-h EdU incorporation in young monkey hippocampal NSCs cultured on soft and stiff HA-laminin hydrogels at Day 9. Scale bar, 50 μm. e, Quantification of EdU^+^ young monkey NSCs as in (d) (n = 4). f, Representative images showing 24-h EdU incorporation in old monkey hippocampal NSCs cultured on soft and stiff HA-laminin hydrogels at Day 11. Scale bar, 50 μm g, Quantification of EdU^+^ old monkey NSCs as in (f) (n = 4 or 5). h, The strategy for assessing the differentiation potential of young and old monkey NSCs on soft and stiff HA-laminin hydrogels. Monkey NSCs were induced for neuronal differentiation for 21 days in serum-free DMEM/F12 medium (growth factors withdrawn and supplemented with BDNF/GDNF/CNTF). i,j Immunostaining of neurons and astrocytes differentiated from young (i) and old (j) NSCs on soft and stiff HA-laminin hydrogels. Neurons are Map2 (gray) and Tuj1 (red) positive cells; astrocytes are GFAP (green) positive cells. Scale bars, 100 μm and 50 μm as indicated. k-n, Gene expression validation of PIEZO1, HAUS1, ITGAD, and COL6A6 in monkey NSCs derived from primary cultures of young and old monkeys, as well as old monkey NSCs cultured on soft and stiff hydrogels (n = 3). Statistical significance was determined using unpaired Student’s t tests. Data are presented as mean ± SD (∗p < 0.05, ∗∗p < 0.01).Fig. 6

Discussion

4

In this study, we initially assessed the stiffness of the hippocampal neurogenic niche across neonatal, juvenile, and adult stages. Guided by these measurements, we engineered HA-laminin hydrogels with varied stiffness levels to replicate the diverse hippocampal niche stiffness observed during aging, for the cultivation of NSCs derived from mice and monkeys across different age groups. To investigate how altered niche stiffness influences the aging process of NSCs and the underlying regulatory mechanisms, we examined the senescence makers and stem cell potential of these NSCs on hydrogels with different age-matched stiffness and compared their transcriptional profiles. We discovered that an aging-related niche stiffening impairs the proliferation and neuronal differentiation of hippocampal NSCs and induces their aging phenotype, accompanying with the upregulation of Piezo1 and ECM genes, and the downregulation of cell cycle genes. A soft HA-laminin hydrogel, designed to replicate the young NSC niche, was found to partially restore cell cycle and cell adhesion gene expression and the vitality of old NSCs derived from the hippocampi of mice and monkeys. Conversely, a stiff HA-laminin hydrogel, designed to replicate the old NSC niche, could accelerate the aging process of young NSCs, including increased senescence markers and decreased NSC proliferation and neuronal generation. Furthermore, we also provide evidence for species-conserved yet nuanced differences in how matrix stiffness influences NSC aging between rodents and primates.

Beyond the characterization of hippocampal tissue stiffness in modulating NSC aging phenotypes, our study reveals that PIEZO1 serves as a critical mechanosensor in hippocampal NSC aging, showing upregulation in response to age-related niche stiffening. While previous work established Piezo1’s importance in NSC biology during embryonic development [24,46,67], our findings extend this understanding to the aging context by demonstrating dynamic changes in Piezo1 expression and its downstream pathway genes. The progressive accumulation of collagen and integrins in the aging extracellular matrix drives tissue stiffening [[68], [69], [70], [71]], which in turn upregulates Piezo1 expression. Importantly, knockdown of Piezo1 partially restored the proliferative capacity of old NSCs in stiff microenvironments, accompanied by corresponding changes in cell cycle and cell adhesion gene expression. These results collectively demonstrate that hippocampal NSCs utilize the PIEZO1 to perceive and transduce mechanical signals from their aging niche, ultimately regulating transcriptional programs that govern NSC aging processes. This mechanotransduction pathway represents a previously underappreciated regulatory mechanism underlying age-related NSC functional decline.

Although our study has established a mechanistic link between Piezo1-mediated matrix stiffness sensing and the regulation of NSC aging, demonstrating its capacity to either accelerate or delay the NSC aging phenotypes, direct experimental evidence through hippocampal tissue stiffness modulation remains unavailable. This limitation primarily arises from the technical difficulty of precisely adjusting tissue stiffness in vivo without disrupting structural integrity. Consequently, targeting mechanosensory pathways involved in stiffness perception may represent a more feasible strategy for in vivo regulation. For in vitro applications, our study suggests that optimizing cell or organoid culture systems by altering the stiffness of the culture matrix could be a key to obtaining cells or organoids to meet different requirements. Additionally, using hydrogels with appropriate stiffness for cell or organoid-laden transplantation into damaged tissues holds promising potential for tissue repair and regeneration.

Funding

This work was supported by Brain Science and Brain-like Intelligence Technology-National Science and Technology Major Project (2022ZD0207700), the 10.13039/501100001809National Natural Science Foundation of China (32470846, 32070695, and 52202350), and the Yunnan Province Research Project (202301BF070001-012 and 202503AP140015).

CRediT authorship contribution statement

Luyao Guo: Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Longjiao Ge: Investigation, Methodology, Validation, Writing – review & editing. Yong Li: Formal analysis, Investigation, Methodology. Shouye Wang: Formal analysis, Investigation, Methodology. Huitong Li: Data curation, Formal analysis, Software. Xiaoyu Wang: Formal analysis, Investigation. Weiliang Qian: Formal analysis, Investigation. Yu Zhang: Methodology. Liuhanhui Guo: Formal analysis. Luxuan Guo: Methodology. Ruihong Cheng: Methodology. Weizhi Ji: Funding acquisition, Project administration, Resources. Wenxiang Fu: Funding acquisition, Investigation, Methodology, Resources, Supervision, Writing – review & editing. Lei Zhang: Funding acquisition, Methodology, Resources, Writing – review & editing. Runrui Zhang: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Runrui Zhang reports financial support was provided by National Natural Science Foundation of China. Lei Zhang reports financial support was provided by National Natural Science Foundation of China. Wenxiang Fu reports financial support was provided by National Natural Science Foundation of China. Wenxiang Fu reports financial support was provided by Natural Science Foundation of Yunnan. No additional relationships or activities to declare. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.