Abstract

   Dry eye disease (DED) is characterized by chronic inflammation and an
   unstable tear film. Stem cells have shown potential for DED treatment,
   but the main challenge lies in improving cell delivery effectiveness.
   Here, we developed eye drops for autoimmune DED treatment using porous
   arginine–glycine–aspartic acid (RGD)–modified alginate microcarriers
   with mesenchymal stem/stromal cells (MSCs) (RGD-Alg@MSCs). These
   microcarriers provided a favorable microenvironment for large-scale
   cell expansion while maintaining stemness with ideal mechanical
   properties for ocular application. In vitro, RGD-Alg@MSCs demonstrated
   significantly enhanced therapeutic effects compared to conventional
   MSCs, including improved cell viability, reduced apoptosis and reactive
   oxygen species, and enhanced release of immunomodulatory factors.
   Transcriptomic analysis revealed distinct molecular mechanisms
   underlying these enhanced therapeutic effects. In the mouse model,
   RGD-Alg@MSCs exhibited prolonged ocular retention and enhanced tear
   production, promoted corneal healing, and suppressed inflammation by
   inhibiting dendritic cell activation and T[H]17 differentiation. Our
   microcarrier system substantially improves stem cell delivery
   efficiency for treating autoimmune DED.
     __________________________________________________________________

   MSC-loaded porous microcarriers improve retention and immune regulation
   for treating autoimmune dry eye disease.

INTRODUCTION

   Dry eye disease (DED) is a prevalent ocular disease characterized by an
   unstable tear film and persistent inflammation, which cause discomfort,
   visual impairment, and damage to the surface of the eye. DED not only
   involves a lack of tears but also an immune-mediated inflammatory
   response that affects both the innate and adaptive immune systems.
   Autoimmune-related DED, such as that associated with Sjögren’s syndrome
   (SS), represents a particularly challenging form of the disease with
   substantial inflammatory components. The global burden of
   autoimmune-related DED is considerable, with epidemiological studies
   revealing remarkably high prevalence rates across various autoimmune
   conditions. Notably, up to 95% of patients with SS experience dry eye
   symptoms, while notable rates are also observed in rheumatoid arthritis
   (38 to 47%), systemic lupus erythematosus (10 to 21%), and systemic
   sclerosis (37 to 79%) patients ([34]1, [35]2). This high prevalence
   translates into considerable impact on the life quality of patients,
   with many experiencing daily symptoms that markedly interfere with
   routine activities such as reading, driving, and digital screen use.
   Beyond the physical discomfort, autoimmune-related DED imposes a
   profound psychological burden, with anxiety and depression rates
   ranging from 26.5% to 83.8% among affected patients ([36]3),
   representing a 37% higher risk compared to conventional dry eye
   patients ([37]4). Moreover, the economic burden is equally concerning,
   encompassing both direct medical costs for treatments and substantial
   indirect costs from reduced workplace productivity and absenteeism. For
   example, a study on patients with SS in East China found that their
   annual total expenditure on treatment was approximately 5.5 times
   higher than that of patients with non-Sjögren’s dry eye ([38]4). These
   epidemiological insights underscore the urgent clinical need for more
   effective therapeutic interventions for autoimmune-related DED.
   Existing therapies, including artificial tears, anti-inflammatory
   medications, and surgical procedures, frequently offer only short-term
   alleviation of symptom and cannot effectively target the fundamental
   immunological malfunction. Notably, stem cells are a potential
   treatment for autoimmune DED because they can control immune responses,
   decrease inflammation, and facilitate tissue repair and regeneration
   ([39]5). Mesenchymal stem/stromal cells (MSCs) have the ability to heal
   the tissue, reduce proinflammatory cytokines, halt the activation of
   dendritic cells (DCs), and decrease T cell proliferation. As a result,
   an anti-inflammatory environment that encourages healing is created
   ([40]6). Cell-based therapies represent a promising frontier in
   ophthalmology, offering potential for treating conditions resistant to
   conventional therapies. However, effective implementation faces several
   challenges, including limited cell expansion capability, poor survival
   in harsh environments, and inefficient delivery methods. Current
   approaches for MSC administration typically rely on intravenous
   injection, presenting considerable limitations for ocular therapy. When
   delivered systemically, only a small fraction of stem cells reach the
   eye, while direct ocular injection can be invasive and may compromise
   cell survival in the harsh ocular microenvironment. Additionally, the
   dynamic conditions of the ocular surface, including blinking and tear
   flow, rapidly clear conventional cell-based eye drops, limiting
   therapeutic effectiveness. Therefore, innovative delivery methods are
   urgently needed to enhance MSC retention, survival, and therapeutic
   efficacy for treating ocular conditions like autoimmune DED.

   Herein, we propose using porous hydrogel microcarriers to deliver MSCs
   as eye drops for treating autoimmune DED. Hydrogels are a promising
   method for stem cell delivery due to their compatibility with living
   tissue, customizable nature, and ability to mimic the natural
   environment of extracellular matrix ([41]7, [42]8). Hydrogels can be
   modified and engineered to improve cell distribution ([43]9).
   Nevertheless, conventional hydrogel systems often lack porous
   structures, which restricts nutrient diffusion, impedes therapeutic
   factor release, and limits interactions between encapsulated stem cells
   and their environment ([44]10–[45]12). The method of microfluidic
   electrospray makes it possible to produce consistent hydrogel
   microcarriers with exact control over their size and composition,
   overcoming the limitations of conventional systems and enhancing the
   effectiveness of stem cell treatments ([46]13). These microcarriers
   create an ideal milieu for encapsulating MSCs, guaranteeing consistency
   and scalability in manufacturing ([47]14). However, this method remains
   largely unexplored for DED and offers a groundbreaking approach with
   considerable potential to enhance the delivery of MSCs, providing an
   innovative solution for DED treatment.

   Here, we fabricated porous microparticles encapsulating MSCs based on
   microfluidic electrospray for DED treatment ([48]Fig. 1). The
   microcarriers were created by electrospraying the solution of
   arginine–glycine–aspartic acid (RGD)–modified sodium alginate with
   polyethylene oxide (PEO) and MSCs into the calcium chloride (CaCl[2])
   solution. As PEO dissolved in water, it created channels within the
   microcarriers that facilitate material exchange, nutrient diffusion,
   and release of bioactive compounds from MSCs, resulting in persistent
   therapeutic advantages. These microcarriers incorporate RGD peptide
   modification to enhance cell attachment and stability, significantly
   improving the therapeutic potential of the encapsulated MSCs. Their
   mechanical properties balance structural integrity with softness, while
   their shear-thinning behavior resembles conventional eye drop
   solutions, ensuring ocular comfort and adaptability to the dynamic eye
   environment. Our stability studies demonstrated that RGD-Alg@MSCs
   maintained structural integrity and proper cell viability for at least
   24 hours when stored at 4°C, showing superior stability compared to
   MSCs in direct suspension. The porous microcarriers were evaluated in
   vitro using a hyperosmotic corneal epithelial cell culture that
   replicates dry eye circumstances. The results demonstrated notable
   enhancements in cell viability, reductions in apoptosis and reactive
   oxygen species (ROS), and diminished expression of proinflammatory
   cytokines. RGD-Alg@MSCs showed significantly superior protective
   effects for apoptosis and enhanced secretory factors compared to
   conventional MSCs. Additionally, transcriptomic analysis through RNA
   sequencing (RNA-seq) revealed distinct molecular mechanisms underlying
   the enhanced therapeutic effects of microcarrier-encapsulated MSCs. In
   the in vivo study, comprehensive biocompatibility assessments confirmed
   the safety of RGD-Alg@MSCs, with terminal deoxynucleotidyl transferase
   (TdT)–mediated deoxyuridine triphosphate (dUTP) nick end labeling
   (TUNEL) staining revealing no significant apoptosis in ocular tissues
   following treatment. In vivo imaging system demonstrated that
   RGD-Alg@MSCs exhibited substantially enhanced retention at the ocular
   surface. An autoimmune-related DED model was established using male
   nonobese diabetic (NOD) mice. The results demonstrated that MSC-loaded
   microcarriers significantly improved tear production, corneal healing,
   restoration of conjunctival goblet cells, and reduction of
   inflammation. Furthermore, the therapy led to a noteworthy reduction in
   DC activation and T helper 17 (T[H]17) cell differentiation in vivo and
   in vitro, both of which are critical elements in the development of
   autoimmune DED. Thus, our MSC-encapsulated porous microcarrier system
   showed great potential for improving stem cell delivery over
   traditional methods and providing a more effective, long-term
   therapeutic option for autoimmune DED.

Fig. 1. Schematic of RGD-Alg microcarrier fabrication and application.

   [49]Fig. 1.
   [50]Open in a new tab

   Schematic of the fabrication of the porous RGD-Alg microcarriers with
   MSC encapsulation and their application in promoting autoimmune DED in
   NOD mice.

RESULTS

Synthesis and characterization of RGD-alginate

   RGD-alginate (RGD-Alg) was synthesized by grafting RGD peptides onto
   the alginate with 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC)
   and N-hydroxysuccinimide (NHS) coupling chemistry, following
   established protocols in the literature ([51]15). This process ensures
   the effective grafting of RGD peptides, known for enhancing cell
   adhesion, to the alginate structure. The composite RGD-Alg was
   characterized using proton nuclear magnetic resonance spectroscopy (^1H
   NMR) spectroscopy. The RGD-modified alginate showed proton peaks
   corresponding to arginine residues at 1.4 to 1.8 parts per million
   (ppm), indicating that the RGD peptide was successfully grafted onto
   the alginate ([52]Fig. 2A and fig. S1). Further structural analysis was
   done using Fourier transform infrared (FTIR) ([53]Fig. 2B) to ensure
   the successful alteration of RGD. The distinctive infrared absorption
   peaks of the alginate appeared at around 1400 and 1600 cm^−1. New
   characteristic peaks were observed in the RGD-Alg samples at
   approximately 1670 and 1560 cm^−1, corresponding to the amide I and
   amide II bands, respectively, which confirmed the successful formation
   of amide bonds between RGD peptides and alginate chains. The influence
   of RGD-Alg hydrogels on cell adhesion was analyzed and contrasted with
   unmodified alginate hydrogels (fig. S2). As expected, RGD-Alg hydrogel
   significantly promoted cell adhesion, demonstrating the effectiveness
   of RGD modification in enhancing the cell-friendly properties of the
   hydrogel (fig. S3). We adjusted the pore size by varying the PEO ratio.
   The final hydrogel precursor solution contained 2% RGD-Alg with PEO
   concentrations of 0.01%, 0.05%, 0.1%, 0.5%, 1%, 1.5%, and 2%. After
   crosslinking in 2% CaCl[2] solution, the gel was transferred to
   phosphate-buffered saline (PBS) for 5 min to dissolve the PEO and form
   pores. Scanning electron microscopy (SEM) images showed the
   microstructure and quantified pore sizes. As PEO increased, pore size
   grew from 1 to 15 μm, forming interconnected, honeycomb-like pores
   (fig. S4). Considering MSC size and efficient water and nutrient
   exchange, 0.1% PEO was selected as the final concentration for
   preparing porous hydrogel microcarriers.

Fig. 2. Characterization and analysis of RGD-Alg hydrogel and MSC-laden
microcarriers.

   [54]Fig. 2.
   [55]Open in a new tab

   (A) NMR and (B) FTIR spectra of RGD-Alg hydrogel. (C) SEM picture of
   the porous RGD-Alg hydrogel. Scale bar, 10 μm. (D) Image of
   monodisperse RGD-Alg microparticles in the bright field. Scale bar, 100
   μm. (E) Distribution of RGD-Alg microcarrier particle sizes. (F)
   Quantification of cell proliferation of the laden MSCs (n = 4). (G)
   Fluorescence images of MSCs loaded on days 1, 3, and 7 within the
   microcarriers. Calcein AM and PI were used for staining. Scale bar, 50
   μm.

Preparation of porous RGD-Alg microcarriers

   During the experiment, an RGD-Alg solution containing 0.1% PEO was
   precisely injected through a microfluidic device during the
   electrospray procedure. The combined pregel solution experienced
   surface tension and transformed into microdroplets due to the
   high-voltage electric field generated by a direct current apparatus.
   The droplets then polymerized following the cross-linking process upon
   immersion in the CaCl[2] solution and were then soaked in PBS to
   eliminate the PEO. Using optical microscopy, we confirmed the
   morphological characteristics of the microcarriers. The optimal
   parameter was assessed by measuring the diameters of several
   microcarriers under diverse circumstances. The findings demonstrate
   that the particle size of microcarriers may be accurately regulated by
   modifying three critical parameters: the electric field voltage between
   the microfluidic outlet and the collecting buffer surface, the
   concentration of the RGD-Alg solution, and the flow velocity (fig. S5).
   By fine-tuning these variables, we identified optimal conditions for
   producing microcarriers that meet our design specifications. More
   specifically, porous RGD-Alg microcarriers with diameters in about 130
   μm were produced using an electric field voltage of 8 kV, a 2% RGD-Alg
   solution, and a flow rate of 100 μl/min ([56]Fig. 2, D and E). The
   microcarriers were successfully synthesized, and the mass, stable, and
   uniform synthesis of these microcarriers may be achieved using our
   microfluidic electrospray fabrication method.

Mechanical and rheological properties of the porous RGD-Alg microcarriers

   To characterize the mechanical properties of the porous RGD-Alg
   microcarriers, Young’s modulus was measured using a micrometer-scale
   mechanical testing system. The microcarriers exhibited a Young’s
   modulus of approximately 50 kPa at 25% deformation (fig. S6),
   indicating a soft and elastic nature suitable for ocular applications.
   Rheological properties of the porous RGD-Alg microcarriers were
   assessed and compared with common eye drop solutions including
   polyvinyl alcohol (PVA), sodium hyaluronate, and sodium carboxymethyl
   cellulose (CMC) (fig. S7). Steady shear tests were performed over a
   shear rate range of 350 to 2350 s^−1, approaching the blink shear
   frequency of approximately 4250 s^−1 or higher ([57]16), within the
   measurement capabilities of the test machine. The porous RGD-Alg
   microcarriers exhibited a viscosity (η) decrease from 250 to 50 mPa·s,
   demonstrating a shear-thinning behavior. This aligns with the ideal
   viscosity range of 30 to 300 mPa·s for effective ophthalmic drug
   delivery, balancing drug release, comfort, and corneal residence time
   ([58]17). At higher shear rates, the viscosity may further decrease,
   potentially nearing that of common eye drops. This shear-thinning
   property, combined with the soft elastic nature of the microcarriers,
   suggests reduced friction during blinking and reduced potential foreign
   body sensation.

Biocompatibility of the porous RGD-Alg microcarriers

   To assess the biocompatibility of the porous RGD-Alg microspheres,
   human corneal epithelial cells (HCECs) and MSCs were cultured normally
   as the control group, while cultured in a medium exposed to the extract
   from the porous microspheres as the experimental group. Over a 3-day
   culture period, both control and experimental groups exhibited high
   confluence and robust proliferation (fig. S8), with cell viability
   consistently above 95% (fig. S9), as determined by calcein AM/propidium
   iodide (PI) staining. Besides, the viability of encapsulated cells
   depends on the ability of RGD-Alg hydrogel to retain water, as
   demonstrated by their ability to achieve a swelling ratio of 1300 wt %
   after incubation in deionized water (fig. S10). MSCs were loaded into
   porous RGD-Alg microcarriers (RGD-Alg@MSCs) by mixing them into the
   pregel solution. Further testing was conducted using calcein AM/PI
   staining to evaluate the cell viability and cell proliferation of the
   MSCs encapsulated in the microcarriers over a 7-day culture period. The
   RGD-Alg group exhibited a similar cell viability as did the porous
   alginate microcarrier control group (fig. S11) and a significantly
   higher cell proliferation than did the Alg group ([59]Fig. 2G),
   indicating a superior biocompatible environment provided by the porous
   RGD-Alg microcarriers. The quantitative fluorescence data ([60]Fig. 2F)
   corroborated this, highlighting the pro-proliferative effect of the RGD
   modification and the porous structure of the microcarriers. Beyond cell
   viability, maintaining cell stemness is crucial for therapeutic
   efficacy. Cell stemness is essential for efficient tissue regeneration
   and repair. To evaluate the stemness of MSCs cultured in the
   microcarriers, flow cytometry was performed after a 7-day coculture
   period. The expression levels of surface marker proteins in MSCs
   released from the microcarriers were statistically comparable to those
   of MSCs cultured on traditional tissue culture plates (Figs. S12 and
   S13). This result confirmed that the RGD-Alg microcarriers can
   effectively maintain the stemness of MSCs, preserving their therapeutic
   potential ([61]18).

Stability of RGD-Alg@MSCs

   To evaluate the potential for clinical application, we assessed the
   stability of RGD-Alg@MSC eye drops during storage at 4°C in PBS. The
   microcarriers maintained their spherical morphology over 24 hours with
   limited degradation (fig. S14). Flow cytometry analysis revealed that
   MSCs encapsulated in the porous RGD-Alg microcarriers retained higher
   viability (83.6%) after 24 hours of cold storage compared to
   conventional suspended MSCs in PBS (69.3%), demonstrating the
   protective effect of the microcarrier system (fig. S15).

   For clinical application, RGD-Alg@MSC eye drops are packaged in sterile
   dropper bottles under aseptic conditions and stored at 4°C to preserve
   microcarrier integrity and MSC bioactivity (fig. S16). Patients should
   gently shake the bottle before use to ensure uniform microcarrier
   distribution and then apply several drops as directed. This protocol
   maintains therapeutic efficacy while ensuring convenient
   administration.

In vitro effect of RGD-Alg@MSCs

   The in vitro hypertonic model simulates the critical process of tear
   film instability ([62]19), which is the central feature of DED. To
   assess how RGD-Alg@MSCs affected HCECs under hyperosmotic stress, we
   cultured HCECs in a hypertonic medium and cocultured them with RGD-Alg,
   MSCs, and RGD-Alg@MSCs. Cell viability assays demonstrated that both
   MSCs and RGD-Alg@MSCs significantly rescued the viability of HCECs in
   the hypertonic environment ([63]Fig. 3A). Moreover, the rate of HCEC
   apoptosis was evaluated using flow cytometry. When compared to MSCs,
   the percentage of late apoptotic cells was significantly lower with
   RGD-Alg@MSCs ([64]Fig. 3, B and D). Furthermore, the mRNA expression of
   tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), and IL-1β in
   HCECs is inhibited by RGD-Alg@MSCs ([65]Fig. 3C), which indicated a
   strong anti-inflammatory effect, crucial for mitigating dry eye
   symptoms ([66]20, [67]21). The ROS level was assessed by fluorescence
   microscopy and the quantity of the fluorescence intensity ([68]Fig. 3,
   E and F). RGD-Alg@MSCs demonstrated a significant reduction in ROS
   levels. ELISA assays further quantified the levels of typical cytokines
   in the coculture medium. The results showed that compared to MSCs,
   RGD-Alg@MSCs stably produced and released significantly higher levels
   of TNF-stimulated gene-6 (TSG-6), CC motif chemokine ligand 20
   (CCL-20), indoleamine 2,3-dioxygenase (IDO), and prostaglandin E[2]
   (PGE[2]) ([69]Fig. 3G), which are vital for enhancing anti-inflammation
   and immune regulation and promoting tissue repair ([70]22–[71]25).

Fig. 3. RGD-Alg@MSCs exhibited an in vitro protective effect.

   [72]Fig. 3.
   [73]Open in a new tab

   (A) A CCK-8 kit was used to evaluate HCEC viability under hyperosmotic
   stress (n = 6). (B and D) Flow cytometry was used to determine the late
   apoptosis rate (n = 3). (C) mRNA levels of typical inflammatory factors
   in the HCECs (n = 4). (E) Fluorescence microscopy images of ROS in
   HCECs. (F) Quantification of ROS fluorescence intensity (n = 4). (G)
   Levels of factors released measured by ELISA (n = 6). HO,
   hyperosmolarity.

Transcriptomic analysis of MSCs and HCECs in coculture

   To understand the molecular mechanisms mediating therapeutic actions of
   MSCs encapsulated in microcarriers, we performed RNA-seq analysis on
   both MSCs and HCECs after coculture under hyperosmotic conditions. RNA
   was extracted from HCECs cocultured with either MSCs encapsulated in
   porous RGD-Alg microcarriers (RGD-Alg@MSCs-Treated) or MSCs
   (MSCs-Treated), as well as from the HCECs under hyperosmolarity without
   MSC coculture (Untreated). Simultaneously, RNA was extracted from MSCs
   in both RGD-Alg@MSC and MSC groups after coculture with HCECs under
   hyperosmolarity, with normal MSCs (not cocultured with HCECs) serving
   as the Control group. Differential expression analysis revealed
   distinct transcriptomic profiles between the treatment groups. In
   HCECs, the RGD-Alg@MSCs-Treated versus Untreated comparison identified
   204 up-regulated and 1955 down-regulated genes, while the MSCs-Treated
   versus Untreated comparison yielded 611 up-regulated and 1433
   down-regulated genes ([74]Fig. 4A). Comprehensive functional enrichment
   analysis of the RGD-Alg@MSCs-Treated group and the MSCs-Treated group
   revealing respective molecular mechanisms is presented in fig. S17.

Fig. 4. Transcriptomic analysis of HCECs and MSCs in coculture systems.

   [75]Fig. 4.
   [76]Open in a new tab

   (A) Volcano plots comparing gene expression in HCECs under
   hyperosmolarity across different treatment conditions, including
   RGD-Alg@MSCs-Treated versus Untreated and MSCs-Treated versus
   Untreated. Untreated refers to HCECs exposed to hyperosmolar conditions
   without any treatment. (B) Volcano plots comparing gene expression in
   MSCs across different groups, including RGD-Alg@MSCs versus Control and
   conventional MSCs versus Control. Control represents non-cocultured
   MSCs grown in 2D conditions. (C) Venn diagrams showing
   up-regulated/down-regulated DEGs in HCECs under hyperosmolarity after
   coculture with RGD-Alg@MSCs or MSCs, compared to the untreated group.
   (D) Venn diagrams showing up-regulated/down-regulated DEGs in MSCs of
   different groups compared to the control. (E) GO enrichment
   scatterplots and (G) KEGG pathway enrichment scatterplots of specific
   up-regulated and down-regulated DEGs identified in hyperosmotic-induced
   HCECs cocultured with RGD-Alg@MSCs. (F) GO enrichment scatterplots and
   (H) KEGG pathway enrichment scatterplots of specific up-regulated and
   down-regulated DEGs identified in RGD-Alg@MSCs. Bubble size represents
   gene count, and color indicates P value.

   To investigate the unique molecular mechanisms associated with
   microcarrier-encapsulated MSCs, we performed Venn diagram analysis
   ([77]26) of the up-regulated and down-regulated differentially
   expressed genes (DEGs) ([78]Fig. 4C). This analysis revealed that the
   RGD-Alg@MSCs-Treated group exhibited 79 uniquely up-regulated and 1087
   uniquely down-regulated DEGs that were not present in the MSCs-Treated
   group, suggesting specific molecular mechanisms that may contribute to
   the enhanced therapeutic efficacy of microcarrier-encapsulated MSCs.

   To elucidate the unique biological significance of these DEGs, we
   performed Gene Ontology (GO) ([79]Fig. 4E) and Kyoto Encyclopedia of
   Genes and Genomes (KEGG) pathway enrichment analyses ([80]Fig. 4G),
   which revealed distinctive molecular signatures that highlight the
   enhanced therapeutic potential of RGD-Alg@MSCs over conventional MSCs.
   Up-regulated DEGs in RGD-Alg@MSCs were significantly enriched in
   biological processes related to cell-cell signaling and positive
   regulation of cell migration, alongside DNA binding transcription
   factor activity. These up-regulated functions represent a key advantage
   of the RGD-Alg microcarrier system, as they enhance the ability of MSCs
   to coordinate cellular responses during tissue repair and regeneration.
   The KEGG pathway analysis further identified unique up-regulation in
   p53 and FoxO signaling pathways, which confer superior cell survival,
   stress resistance, and cellular homeostasis compared to conventional
   MSCs. The most notable advantage of RGD-Alg@MSCs was demonstrated in
   their inflammatory profile. These microencapsulated MSCs exhibited
   substantial down-regulation of multiple inflammatory response pathways,
   including positive regulation of proinflammatory cytokines (IL-1β,
   IL-6, IL-8, IL-12), positive regulation of NLRP3 inflammasome complex
   assembly, and reduced neutrophil chemotaxis and TNF production. This
   represents a marked improvement over conventional MSCs in controlling
   excessive inflammation. Key inflammatory signaling pathways were also
   uniquely down-regulated, including Janus kinase (JAK)–signal transducer
   and activator of transcription (STAT), mitogen-activated protein kinase
   (MAPK), and nuclear factor κB (NF-κB) signaling pathways, collectively
   indicating superior immunomodulatory capabilities. Additionally,
   pathways involved in mediator regulation of transient receptor
   potential (TRP) channels, arachidonic acid metabolism, and neutrophil
   extracellular trap formation were more effectively suppressed in
   RGD-Alg@MSCs. Another distinctive advantage was the down-regulation of
   cellular response to hypoxia, suggesting that RGD-Alg@MSCs have
   improved adaptation to low-oxygen environments often present in damaged
   tissues, representing a critical advancement over conventional MSCs for
   therapeutic applications. These findings collectively indicate that
   RGD-Alg microcarrier system confers superior immunomodulatory
   properties to MSCs through the more effective suppression of multiple
   proinflammatory pathways and mediators while simultaneously enhancing
   cell migration and intercellular communication necessary for tissue
   repair. This enhanced immunoregulatory profile provides the molecular
   basis for understanding the superior therapeutic efficacy of
   RGD-Alg@MSCs over conventional MSCs in promoting tissue regeneration
   and dampening excessive inflammatory responses.

   At the same time, we analyzed the transcriptomic changes in MSCs of
   different groups after coculture with HCECs. Compared to control MSCs,
   RGD-Alg@MSCs showed 2631 up-regulated and 1902 down-regulated genes,
   while MSCs exhibited 1661 up-regulated and 1228 down-regulated genes
   ([81]Fig. 4B). The overall functional enrichment analysis of
   RGD-Alg@MSCs and MSCs is presented in fig. S18. Venn diagram analysis
   revealed that RGD-Alg@MSCs had 2020 uniquely up-regulated and 1492
   uniquely down-regulated genes that were not shared with the MSC group
   ([82]Fig. 4D).

   The GO analysis ([83]Fig. 4F) of uniquely up-regulated DEGs in
   RGD-Alg@MSCs revealed significant enrichment in T cell–related immune
   functions, including T cell differentiation involved in immune response
   and CD4-positive, α-β T cell activation. Additionally, up-regulation in
   cytokine binding further enhances the immunomodulatory capabilities of
   these cells through improved paracrine signaling. Structurally,
   RGD-Alg@MSCs exhibited up-regulation in cell adhesion and cytoskeletal
   organization processes including filopodium membrane, regulation of
   actin filament–based process, homotypic cell-cell adhesion, focal
   adhesion, stress fiber, cell-substrate junction, and positive
   regulation of supramolecular fiber organization. These structural
   adaptations reflect the advantages of three-dimensional (3D)
   microcarrier culture, which promotes enhanced cytoskeletal organization
   and cell-matrix interactions, likely contributing to improved cell
   survival and functionality. The KEGG pathway analysis ([84]Fig. 4H)
   further demonstrated enrichment in multiple signaling cascades
   essential for cell communication and survival, including Rap1, Ras, and
   Hippo signaling pathways. Notably, the up-regulation of C-type lectin
   receptor signaling pathway is particularly important for immune
   regulation, as these receptors play crucial roles in immune cell
   recognition and response. The down-regulation of TNF signaling pathway
   in RGD-Alg@MSCs indicates enhanced anti-inflammatory properties,
   critical for reducing excessive inflammation in damaged tissues. The
   balance between up-regulated Wnt signaling pathway regulation and
   down-regulated negative regulation of Wnt signaling pathway suggests
   optimized control of this key developmental and regenerative pathway.
   These transcriptomic changes collectively explain the superior
   therapeutic efficacy of RGD-Alg@MSCs in tissue repair and immune
   modulation. The 3D microenvironment provided by the RGD-modified
   alginate microencapsulation appears to optimize the structural
   organization of MSCs, intercellular communication, and signaling
   networks, ultimately enhancing their immunomodulatory function and
   tissue regenerative potential when compared to conventional MSC
   cultures.

Therapeutic effects of RGD-Alg@MSC eye drops in the DED mouse model

   The in vivo biocompatibility was conducted before the therapeutic
   effect test. No toxicity was observed in the treated mice of all groups
   (fig. S19). To further evaluate whether our microcarriers might cause
   ocular discomfort or potential damage due to the size, TUNEL staining
   was performed on the eyeballs. No significant increase in
   TUNEL-positive cells was observed in the RGD-Alg@MSC–treated group
   compared to the control group, confirming the ocular safety of the
   RGD-Alg@MSC eye drops (fig. S20).

   To evaluate ocular retention capabilities, we used red fluorescent
   protein–labeled MSCs (RFP-MSCs) for track. Conventional suspended
   RFP-MSCs or RGD-Alg@RFP-MSCs were topically administered to the ocular
   surface of mice and monitored using in vivo fluorescence imaging (fig.
   S21). Conventional RFP-MSCs showed rapid clearance from the corneal
   surface, with substantial signal reduction by 40 min and near-complete
   disappearance by 80 min post-application. In contrast, RGD-Alg@
   RFP-MSCs demonstrated substantially enhanced retention at the ocular
   surface, with fluorescence signals maintained for up to 80 to 120 min
   after application. This substantially improved retention time suggests
   that RGD-Alg encapsulation provides a critical advantage by effectively
   prolonging the presence of therapeutic MSCs at the ocular surface,
   thereby extending their therapeutic window and enhancing treatment
   efficacy.

   To assess the efficacy of RGD-Alg@MSC eye drops on the autoimmune
   disease–related DED, we used 12-week male NOD mice ([85]Fig. 5A). Male
   ICR mice of the same age with no treatment served as the control. As
   shown in our results, the untreated NOD mice exhibited significant
   symptoms of DED compared to the ICR control mice, including reduced
   tear production, increased corneal fluorescein staining, decreased
   conjunctival goblet cell density, and pronounced inflammatory
   infiltration in the lacrimal and harderian glands. These observations
   confirmed the presence of dry eye condition in our NOD mouse model,
   consistent with previous reports ([86]27) of autoimmune-related DED.
   The administration of RGD-Alg@MSC eye drops successfully enhanced tear
   production ([87]Fig. 5B) and decreased the score of sodium fluorescein
   staining ([88]Fig. 5, C and D) in NOD mice. The treatment with
   RGD-Alg@MSCs increased the density of conjunctival goblet cells
   compared to no treatment ([89]Fig. 5, E and F). We evaluated
   inflammatory cytokine mRNA expression in the ocular surface to learn
   more about whether RGD-Alg@MSCs can reduce inflammation in DED. The
   TNF-α, IL6, and IL-1β mRNA level of the cornea and conjunctiva tested
   by real-time quantitative polymerase chain reaction (RT-qPCR) analysis
   was shown to be lower in the RGD-Alg@MSC–treated mice as compared to
   the untreated group ([90]Fig. 5H). Hematoxylin and eosin (H&E) staining
   was used on the lacrimal glands of NOD mice to assess the extent of
   inflammation infiltration ([91]Fig. 5G), which is a crucial aspect of
   SS-like disease in the model used as well as a characteristic of
   autoimmune dacryoadenitis ([92]28). Quantitative analysis of the
   H&E-stained sections ([93]Fig. 5J) revealed significant differences in
   the degree of lymphocytic infiltration among the groups. The
   nontreatment and RGD-Alg groups exhibited extensive lymphocytic
   infiltration, characterized by densely packed lymphocyte clusters
   within the lacrimal gland tissue. Conversely, the MSC and RGD-Alg@MSC
   groups showed a marked reduction in lymphocytic infiltration,
   suggesting a substantial alleviation of autoimmune-induced
   inflammation. Likewise, RGD-Alg@MSC eye drops alleviated lymphatic
   infiltration in harderian glands in NOD mice ([94]29) ([95]Fig. 5, I
   and K), which were responsible for secreting tear lipids in mouse
   harderian glands. These findings confirmed the therapeutic efficacy of
   RGD-Alg@MSC eye drops in treating SS-like DED.

Fig. 5. In NOD mice, the effectiveness of RGD-Alg@MSC eye drops was
investigated.

   [96]Fig. 5.
   [97]Open in a new tab

   (A) Schematic description of animal experiment. (B) Tear volume
   measurements (n = 10) using cotton threads impregnated with phenol red.
   (C) Images of the fluorescein staining of the cornea. (D) Scores of the
   corneal fluorescein staining (n = 10). (E) Images of the conjunctival
   tissue stained by PAS. Scale bar, 50 μm. (F) Quantification of the
   conjunctival goblet cell density (n = 5) in the conjunctiva. (H)
   Typical inflammatory cytokine expression levels in the cornea and
   conjunctiva tissue (n = 4). H&E staining and histological analysis of
   the lacrimal glands (G and J) (n = 5) and harderian glands (I and K)
   (n = 5). Scale bar, 100 μm. White arrows point to the inflammatory
   infiltrate area.

Immunomodulatory effects of RGD-Alg@MSCs on DCs and T[H]17 differentiation in
vitro

   To explore the potential of RGD-Alg@MSCs in modulating immune
   responses, we designed in vitro experiments to assess their impact on
   DC maturation and T[H]17 differentiation based on reported studies
   ([98]26, [99]30). DCs were derived from human peripheral blood
   mononuclear cells (PBMCs). Immature DCs (iDCs) were generated by
   culturing PBMCs with interleukin-4 (IL-4) and granulocyte-macrophage
   colony-stimulating factor (GM-CSF), followed by stimulation with
   lipopolysaccharide (LPS) to induce their maturation into mature DCs
   (mDCs). These mDCs were then cocultured with RGD-Alg@MSCs using a
   Transwell chamber. To evaluate the expression of key surface markers,
   flow cytometry analysis was used. The successful induction of DCs was
   confirmed by low CD14 expression (fig. S22, A and D) and high CD11c
   expression (fig. S22, B and E) across all groups. Results showed that
   coculture with RGD-Alg@MSCs significantly reduced the expression of
   maturation markers HLA-DR ([100]Fig. 6, A and D), CD86 ([101]Fig. 6, B
   and E), and CD83 (fig. S22, C and F) compared to the untreated mDC
   group, indicating effective inhibition of DC maturation. Furthermore,
   RGD-Alg@MSCs decreased CD40 expression ([102]Fig. 6, C and F), a
   critical molecule involved in DC–T cell interaction, significantly
   outperforming the simple MSC group. To assess the impact on T[H]17
   differentiation, CD4^+ naïve T cells were introduced after the MSCs
   were removed from the coculture. The RGD-Alg@MSC–treated DCs exhibited
   a significantly reduced ability to induce T[H]17 differentiation
   ([103]Fig. 6, G and H), with a statistically lower proportion of CD4^+
   IL-17A^+ cells compared to both the untreated mDCs and the MSC-only
   groups. These findings underscore the superior capacity of RGD-Alg@MSCs
   to inhibit DC maturation and suppress T[H]17 differentiation,
   disrupting the inflammatory loop associated with DED pathogenesis more
   effectively than conventional MSCs alone.

Fig. 6. Effects of immune regulation resulted from RGD-Alg@MSCs in vitro.

   [104]Fig. 6.
   [105]Open in a new tab

   Flow cytometry showing the expression of representative cell surface
   markers HLA-DR (A), CD86 (B), and CD40 (C) of mDCs in different
   coculture treatment groups. The uninduced iDC group was used as the
   control. Quantification of the percentage of cells positively
   expressing these cell surface markers (D to F) (n = 4). (G) DCs treated
   with different methods were cocultured with naïve CD4^+ T cells
   isolated from PBMCs, and flow cytometry was used to detect the
   proportion of T[H]17 cells (CD11c^− CD4^+ IL-17A^+) in them. (H)
   Percentage of T[H]17 cells (CD11c^− CD4^+ IL-17A^+ cells) (n = 4).
   Isotype, isotype control.

Immunomodulatory effects of RGD-Alg@MSCs in vivo

   Next, we evaluated the effects of RGD-Alg@MSCs on the immune responses
   in the cornea and ocular draining lymph nodes (dLNs) in NOD mice.
   Following topical treatment with RGD-Alg@MSCs, RGD-Alg, MSCs, or no
   treatment at all, the corneas and dLNs were harvested. To determine how
   many DCs were in each cornea, we performed immunofluorescence staining.
   Using flow cytometry, the number and the maturity of DCs in dLNs were
   evaluated. The results of immunofluorescence staining ([106]Fig. 7A)
   demonstrated a significant rise in dendritic maturity and DC density in
   the cornea of NOD mice. Treatment with RGD-Alg@MSC eye drops
   significantly reduced the density and dendritic maturity of DCs
   compared to the NOD mice with no treatment group ([107]Fig. 7, C and
   D). Flow cytometric analysis further elucidated the immunomodulatory
   impact of RGD-Alg@MSCs. In the cornea of NOD mice, the percentage of
   T[H]17 cells was remarkably high, which are essential to the
   pathophysiology of DED ([108]31) ([109]Fig. 7, B and E). However,
   treatment with RGD-Alg@MSCs substantially decreased the percentage of
   T[H]17 cells in the cornea, significantly outperforming the reductions
   observed with MSCs alone. This finding suggested that the RGD-Alg@MSC
   eye drops more effectively mitigate the inflammatory response
   associated with T[H]17 cells in DED. In addition, we checked the
   immunomodulatory impact in the dLNs. As in many previous studies
   ([110]32, [111]33), we used the percentage of CD45^+CD11c^+MHC-II^+
   cells in lymph nodes to represent the quantity of DCs, and the
   percentage of CD45^+CD11c^+CD86^+ cells to represent the maturity of
   DCs. Treatment with RGD-Alg@MSCs significantly down-regulated both the
   number ([112]Fig. 7G and fig. S23A) and maturity ([113]Fig. 7H and fig.
   S23B) of DCs in the dLNs, with a particularly pronounced effect on DC
   maturation compared to conventional MSC treatment. In the dLNs of NOD
   mice, there was a considerable rise in the percentage of CD4^+IL-17A^+
   T[H]17 cells ([114]Fig. 7, F and I). This percentage was significantly
   reduced by RGD-Alg@MSC treatment, demonstrating the strong
   immunoregulatory effect of the treatment modality. Overall, these
   results demonstrate that RGD-Alg@MSC eye drops exert a substantial
   immunomodulatory effect in NOD mice, more potently than conventional
   MSCs in reducing both DC maturation and T[H]17-mediated inflammation,
   which are key contributors to the disease pathology ([115]34, [116]35).

Fig. 7. Effects of immune regulation resulted from RGD-Alg@MSC eye drops in
NOD mice.

   [117]Fig. 7.
   [118]Open in a new tab

   (A) Representative images of corneal DCs in different groups (scale
   bar, 50 μm). (B) Flow cytometry showing the level of T[H]17 cells in
   the cornea. Quantification of DC density (C) (n = 5) and dendritic
   complexity (D) (n = 4) in the cornea. (E) Quantification of T[H]17 cell
   percentage in the cornea (n = 3) via flow cytometry. (F) Flow cytometry
   showing the level of T[H]17 cells in the dLNs. Quantification of cell
   number (G) and maturity (H) of DCs in dLNs (n = 4). (I) Percentage of
   T[H]17 cells in dLNs (n = 4). The number of dendritic points per cell
   serves as a proxy for the dendritic complexity of DCs.

DISCUSSION

   Here, we developed an MSC-encapsulated porous microcarrier eye drop
   system for the treatment of autoimmune DED. Our key findings
   demonstrate that RGD-Alg porous microcarriers encapsulating MSCs
   exhibited significant therapeutic effects both in vitro and in vivo.
   Under hyperosmotic conditions in vitro, RGD-Alg@MSCs enhanced corneal
   epithelial cell viability, reduced apoptosis and ROS levels, and
   decreased proinflammatory cytokine expression. In the NOD mouse model
   of autoimmune-related DED, RGD-Alg@MSCs significantly increased tear
   production, promoted corneal healing, restored conjunctival goblet
   cells, decreased proinflammatory cytokine expression in the cornea and
   conjunctiva, and alleviated inflammatory infiltration in the lacrimal
   and harderian glands.

   Our study demonstrated the advantages of microcarrier technology for
   stem cell therapy in DED. Compared to traditional MSC delivery methods,
   the RGD-Alg@MSC system addresses several critical challenges: (i) The
   porous structure and RGD modification enhance MSC adhesion, survival,
   and function while providing a protective microenvironment for
   sustained therapeutic factor release; (ii) RGD-Alg@MSCs demonstrate
   significantly stronger immunomodulatory capabilities than conventional
   MSCs, more effectively reducing DC maturation and T[H]17 cell
   infiltration both in vitro and in vivo; (iii) transcriptomic analysis
   revealed distinct gene expression patterns in RGD-Alg@MSC–treated
   HCECs, with unique enrichment in cytokine-receptor interaction and
   IL-17 signaling pathways; (iv) the system achieves prolonged retention
   in a noninvasive eye drop form; and (v) the microcarrier platform
   enables efficient large-scale expansion of MSCs while preserving their
   stemness. These findings not only provide a therapeutic strategy for
   autoimmune-related DED but also establish a platform for cell expansion
   and delivery for cell therapy in ocular surface diseases.

   Our findings reveal several key mechanisms by which RGD-Alg@MSCs
   alleviate DED symptoms. The porous microcarrier structure provides a
   protected microenvironment for MSCs while allowing efficient exchange
   of nutrients and therapeutic factors with the surrounding tissue. This
   is evidenced by the sustained release of therapeutic factors including
   TSG-6, CCL-20, IDO, and PGE[2] observed in our in vitro studies, which
   are crucial for immunomodulation, anti-inflammatory effects and tissue
   repair. Our immunological analyses demonstrated that RGD-Alg@MSCs more
   effectively modulate the immune environment in DED compared to
   conventional MSCs. In vitro coculture experiments showed that the
   RGD-Alg@MSC group exhibited a stronger reduction in CD40 expression on
   mDC surfaces and more potently inhibited mDC-induced differentiation of
   CD4^+ naïve T cells into T[H]17 cells. These enhanced effects were
   confirmed in vivo, where RGD-Alg@MSC treatment was significantly more
   effective at reducing both the number and maturity of DCs in the dLNs
   and decreasing corneal T[H]17 cell infiltration compared to
   conventional MSCs. This represents a notable advantage as the
   microcarrier system enhances the immunomodulatory capacity of MSCs,
   leading to more effective disruption of the inflammatory feedback loop
   characteristic of DED.

   Furthermore, transcriptomic analysis revealed distinct gene expression
   patterns in HCECs cocultured with RGD-Alg@MSCs compared to those
   cocultured with conventional MSCs. The RGD-Alg@MSCs-Treated group
   exhibited unique molecular signatures not present in the MSCs-Treated
   group, particularly in immunomodulatory pathways and inflammatory
   responses. Specifically, GO and KEGG pathway analyses demonstrated
   enhanced down-regulation of proinflammatory cytokines (IL-1β, IL-6,
   IL-8, IL-12) and signaling pathways (JAK-STAT, MAPK, and NF-κB) in the
   RGD-Alg@MSC group. Simultaneously, pathways related to cell migration,
   intercellular communication, and stress resistance were uniquely
   up-regulated. The MSCs themselves showed significant transcriptomic
   differences when cultured in the porous RGD-Alg microcarriers, with
   up-regulation in T cell immunomodulation functions, cytokine binding,
   and cell-matrix interactions, while showing down-regulation of TNF
   signaling pathways. These molecular changes collectively provide the
   mechanistic basis for the enhanced therapeutic efficacy of
   RGD-Alg@MSCs, highlighting how the 3D microenvironment optimizes MSC
   function for both immunomodulation and tissue regeneration.

   The careful formulation of our MSC-encapsulated porous microcarrier eye
   drops was optimized specifically for DED treatment. We selected
   alginate as the primary material due to its biocompatibility, minimal
   immunogenicity, and mucoadhesive properties that extend ocular surface
   residence time. RGD peptide modification enhanced cell attachment and
   survival through integrin receptor interactions. CaCl[2] was chosen as
   the crosslinking agent for its rapid ionic bonding with alginate,
   creating a gentle gelation process that preserves MSC viability while
   providing mechanical stability without introducing toxic chemical
   crosslinkers. PEO, which shares chemical structure with U.S. Food and
   Drug Administration–approved polyethylene glycol (PEG) but with higher
   molecular weight, served as a temporary water-soluble porogen mixed in
   the precursor solution. After gelation, it was removed by dissolution
   when transferred to water, creating essential porous structures while
   avoiding residual toxicity and harsh processing conditions typically
   associated with conventional porogens. The microcarrier was suspended
   in isotonic PBS to ensure stability and minimize irritation.

   The physical properties of our microcarriers contribute substantially
   to their ocular applicability. With a diameter of 130 μm, they fall
   within the standard size range for cell microcarriers (100 to 300 μm)
   used in biomedical applications. Their measured Young’s modulus of
   approximately 50 kPa indicates a soft, elastic nature comparable to
   soft biological tissues, potentially minimizing foreign body sensation
   while maintaining structural integrity. The rheological properties
   further support their suitability for ocular use. Despite instrument
   limitations preventing assessment at the full range of physiological
   blinking shear rates (which can reach up to 100,000 s^−1), our measured
   shear-thinning behavior suggests that viscosity would likely decrease
   further at higher shear rates during actual blinking. With viscosities
   within the optimal 30 to 300 mPa·s range for ophthalmic formulations
   ([119]17), our microcarriers achieve a rheological balance that
   conventional eye drops between retention time and patient comfort.

   For the eye drop formulation, we prepared a suspension of microcarriers
   in PBS at a concentration of 6 × 10^7 MSC-laden microcarriers per
   milliliter, administering 5 μl per eye twice daily. In our protocol,
   freshly prepared microcarriers were used immediately for maximum
   therapeutic efficacy, with storage at 4°C maintained throughout the
   transfer process. Additionally, we conducted stability studies, which
   demonstrated that RGD-Alg@MSCs maintained structural integrity and
   proper cell viability for at least 24 hours when stored at 4°C, showing
   superior stability compared to MSCs in direct suspension, suggesting
   potential for practical clinical application with storage advantages.
   TUNEL staining confirmed the ocular safety of RGD-Alg@MSC eye drops, as
   no significant apoptosis was observed in the eyeballs of treated
   healthy ICR mice, further supporting their biocompatibility for ocular
   applications.

   Compared to existing DED therapies, our approach offers several
   advantages. Unlike artificial tears that provide only temporary
   symptomatic relief, RGD-Alg@MSCs address the underlying inflammatory
   mechanisms driving DED pathogenesis. Conventional MSC therapies face
   challenges in cell survival and retention on the ocular surface,
   whereas our microcarrier system demonstrated improved MSC viability and
   prolonged therapeutic effects. MSC-conditioned medium faces substantial
   limitations including the presence of nonspecific metabolites and waste
   products, lacks the dynamic responsiveness to the disease
   microenvironment, and encounters standardization challenges. In
   contrast, our approach overcomes these limitations through sustained
   delivery of viable cells that can actively respond to the pathological
   conditions. The noninvasive nature of our eye drop formulation also
   represents an advantage over surgical interventions or direct
   intraocular injections, potentially improving patient compliance and
   reducing procedural risks. Our system enhances the therapeutic
   potential of MSCs by optimizing their microenvironment, particularly
   suitable for the long-term management of chronic inflammatory
   conditions like DED.

   Despite the promising results, our study has several methodological
   limitations that must be acknowledged. First, while we demonstrated the
   efficacy of RGD-Alg@MSCs in alleviating DED symptoms in the NOD mouse
   model, future studies should evaluate long-term effects and durability
   of the therapeutic response. The NOD mouse model, while valuable for
   studying autoimmune-related DED due to its spontaneous development of
   lacrimal gland inflammation and ocular surface damage, may not fully
   recapitulate all aspects of human DED, which is often multifactorial.
   Alternative models, such as the scopolamine-induced DED model that
   simulates non-Sjögren’s aqueous-deficient dry eye, or the age-related
   dry eye, could be explored in future studies to evaluate the
   therapeutic potential of our approach across diverse DED etiologies.

   In our mechanism exploration, our in vitro hyperosmotic HCEC model
   cannot fully replicate the complexity of DED or incorporate other
   components of the lacrimal functional unit such as lacrimal glands.
   Additionally, while RNA-seq effectively distinguished molecular
   responses in both MSCs and HCECs, the absence of proteomic analysis
   limits direct characterization of the MSC secretome. However, this
   transcriptomic approach provided valuable insights by clearly
   delineating the distinct responses in both cell populations. In
   comparing RGD-Alg@MSCs with conventional MSCs, we focused on uniquely
   regulated genes rather than analyzing more complex expression patterns
   (such as genes up-regulated in one group but down-regulated in
   another). This analytical approach was intentionally designed to
   prioritize the most statistically significant and consistent pathways,
   maintain analytical clarity, and avoid potential confounding effects
   from more variable expression patterns.

   Additionally, further investigation into the optimal dosing regimen,
   concentration, and frequency, long-term storage stability, and scalable
   manufacturing processes of RGD-Alg@MSCs will be essential for
   successful clinical translation. Future formulation refinements could
   explore alternative vehicles such as artificial tear solutions or
   hyaluronic acid to potentially enhance retention time and therapeutic
   efficacy while maintaining the fundamental microcarrier design. Our
   transcriptomic analysis has identified several potential molecular
   pathways involved in the therapeutic effects of RGD-Alg@MSCs, which
   warrant further investigation to develop more targeted approaches for
   DED treatment.

MATERIALS AND METHODS

Materials

   The chemicals used in this study included sodium alginate
   (Sigma-Aldrich, USA), RGD peptide (Typeptide Biotechnology, China), PEO
   (Sigma-Aldrich), CaCl[2] (Sigma-Aldrich), EDC (Aladdin, China), NHS
   (Aladdin), and sodium citrate (Sigma-Aldrich). The water used in the
   experiment underwent purification through the Millipore Milli-Q
   purification system.

Cells

   HCECs (CRL-11135), human umbilical cord MSCs (Nanjing Drum Tower
   Hospital), and RFP-MSCs (Nanjing Drum Tower Hospital) were both
   cultured in the Dulbecco’s modified Eagle’s medium/F12 medium (HyClone,
   Cytiva) containing 10% fetal bovine serum (FBS; Gibco) and 1%
   penicillin/streptomycin (PS; Pricella). The extraction of PBMCs
   ([120]36) and the induction processes for iDCs and mDCs ([121]37)
   followed a similar protocol to the previous article. Briefly, IL-4
   (PeproTech) and GM-CSF (PeproTech) were used to induce iDCs from PBMCs,
   while LPS (Sigma-Aldrich) was added to generate mDCs. Naïve CD4^+ T
   cells were isolated from PBMCs using magnetic bead sorting (ImunoSep)
   and cultured in the medium containing plate-bound anti-CD3 (Thermo
   Fisher Scientific), soluble anti-CD28 (Thermo Fisher Scientific),
   anti-IL-4 (Abcam), and anti-IFN-γ (Interferon gamma; Abcam) antibodies,
   along with IL-23 (interleukin-23; PeproTech), TGF-β (transforming
   growth factor-β; PeproTech) and IL-6 (PeproTech) for T[H]17
   differentiation followed a similar protocol before ([122]36, [123]38).
   In RPMI 1640 (HyClone), PBMCs, iDCs, mDCs, and naïve CD4^+ T cells were
   cultivated with 10% FBS and 1% PS added. All cells were incubated at
   37°C with 5% CO[2] in a humidified environment.

Synthesis and analysis of RGD-Alg hydrogel

   RGD-Alg hydrogel was fabricated by adapting established protocols to
   meet specific experimental needs ([124]15). Sodium alginate, at a
   concentration of 0.12 g in 10 ml of PBS buffer (pH 6 to 6.5,
   Servicebio), was activated using an equimolar mixture of EDC and NHS.
   RGD peptide was then introduced, and the mixture was stirred
   continuously to promote coupling. Following the reaction, the solution
   was dialyzed using a 14-kDa cutoff membrane (Biosharp) and then
   neutralized to achieve a pH of 7 to 7.5. Subsequent lyophilization
   yielded a dry powder, which was then rehydrated with deionized water to
   obtain a specific concentration of the RGD-Alg precursor solution as
   required. ^1H NMR spectroscopy (Varian Inova 500 MHz) was used to
   identify the characteristic peaks corresponding to the RGD moiety
   integrated within the alginate structure. Using an FTIR system (Bruker,
   USA), the integration of RGD peptides into the alginate hydrogel was
   evaluated via FTIR over a 400 to 4000 cm^−1 spectrum at a 2 cm^−1
   resolution.

Cell adhesion in RGD-modified hydrogel

   To assess cell adhesion on RGD-modified alginate hydrogels, 2% RGD-Alg
   solution was cured in 2% CaCl[2] to form hydrogels. After
   lyophilization, the films were seeded with MSCs ([125]39). Cells were
   stained with calcein AM (C2015, Beyotime) and examined after a 24-hour
   period.

Fabrication of the porous RGD-Alg hydrogel

   A 2% RGD-Alg solution (initial concentration selected to control
   variables while exploring PEO effects) with a final concentration of
   0.01%, 0.05%, 0.1%, 0.5%, 1%, 1.5%, and 2% PEO was crosslinked in a 2%
   CaCl[2] solution. This CaCl[2] concentration was selected based on
   established protocols ([126]40, [127]41). Previous studies confirmed
   that RGD modification did not significantly alter these optimal
   crosslinking conditions ([128]42, [129]43, [130]44). After
   freeze-drying, the surface pore size was observed using SEM.

Swelling characterization of the porous RGD-Alg microcarriers

   The swelling properties were characterized by the swelling of the
   porous RGD-Alg microcarriers. Wd, the appropriate dry weight, was first
   obtained by lyophilizing the microcarriers. Each sample was immersed in
   deionized water, and the sample weight was recorded daily until the
   sample weight became equal to the swelling equilibrium weight, Weq.
   This procedure was repeated three times under the same conditions.
   Swelling ratio (%) = (Weq/Wd) × 100% was the formula used to estimate
   the swelling ratio ([131]45).

Optimization of electrospraying parameters

   The pre-gel solution was electrosprayed into a 2% CaCl[2] solution
   under varying parameters: voltage (4 to 12 kV), flow rate (50 to 400
   μl/min), and RGD-Alg concentration (1.5% to 4%). Their appearance was
   observed under a light microscope, and ImageJ software was used to
   measure and analyze particle size distribution under different
   parameters.

Mechanical testing of microspheres

   The mechanical properties of the microspheres were evaluated using a
   micrometer-scale mechanical testing system (Microsquisher, CellScale)
   ([132]46). A tungsten cantilever beam (modulus = 411 GPa,
   diameter = 203.2 μm) was used to compress the microspheres. All samples
   were tested in PBS at 37°C. Young’s modulus was determined at 25%
   deformation following a previously published method ([133]47). The
   force at 25% deformation was calculated via linear regression of the
   force-displacement curve.

Rheological characterization test

   The rheological properties of 1.5% PVA (Aladdin), 0.3% sodium
   hyaluronate (Aladdin), 0.5% CMC solutions (Aladdin), and RGD-Alg@MSCs
   were characterized using the Rotational Rheometer (Thermo Fisher
   Scientific HAAKE MARS iQ) equipped with a parallel plate geometry
   ([134]48). All measurements were conducted at 37°C in a hydrated
   environment to mimic physiological conditions. Steady shear tests were
   conducted over a shear rate range of 350 to 2350 s^−1 to evaluate the
   viscosity (η) and shear-thinning behavior of the solutions. Each
   solution was tested in triplicate, and data were analyzed using HAAKE
   RheoWin software.

Biocompatibility test

   To assess biocompatibility, MSCs and HCECs were cultured with extracts
   from the porous microcarriers for 48 hours. Calcein AM and PI (C2015,
   Beyotime) were used to stain the cells, which were then photographed.
   The fluorescence was measured, and the cell viability and proliferation
   were assessed ([135]49). The proportion of live cells was determined
   through fluorescence quantification.

Fabrication of MSC-encapsulated porous RGD-Alg hydrogel microspheres

   A 2% RGD-Alg solution containing 0.1% PEO and 6 × 10^7 MSCs/ml was
   prepared. This cell-laden solution was processed using an electrospray
   device with a high-voltage power supply and a syringe pump operating at
   a voltage of 8 kV and a flow rate of 100 μl/min ([136]50, [137]51). The
   solution was electrosprayed and crosslinked in a 2% CaCl[2] solution to
   form microcarriers. After 5 min, microcarriers were harvested and then
   soaked in PBS to remove PEO, resulting in porous RGD-Alg microcarriers
   encapsulated with MSCs (RGD-Alg@MSCs).

Assessment of MSC viability and proliferation in the porous RGD-Alg
microcarriers

   RGD-Alg@MSCs were cultured for 1, 3, and 7 days. Post-incubation, the
   RGD-Alg@MSCs were examined under fluorescence microscopy. Porous Alg
   microcarriers with MSCs served as the control. Calcein AM and PI were
   used to stain the cells. By evaluating the intensity of cell
   fluorescence, the viability and the proliferation rate of the MSCs were
   objectively calculated ([138]39).

Assessment of MSC stemness in the porous RGD-Alg microcarriers

   RGD-Alg@MSCs were dissolved using a mild, nonirritating 55 mM sodium
   citrate solution, followed by centrifugation at 800 rpm for 5 min to
   collect the encapsulated MSCs without compromising cell viability or
   functional state ([139]52, [140]53). Using flow cytometry, cell surface
   marker expression was examined to assess the stemness preservation of
   MSCs grown inside the hydrogel microcarriers. Following 1 week of
   culture, MSCs were extracted from the microcarriers and stained with
   antibodies that were fluorescently tagged against CD105 (BioLegend),
   CD90 (BioLegend), CD73 (BioLegend), CD45 (BioLegend), CD34 (BioLegend),
   CD11b (BioLegend), and CD19 (BioLegend) ([141]54). Flow cytometry was
   performed using the FACSCalibur system (BD, USA). FlowJo (version 10.4)
   was used for data analysis.

Assessment of RGD-Alg@MSC eye drop stability

   The RGD-Alg@MSC microcarriers were suspended in PBS to prepare eye
   drops and stored at 4°C for immediate use. To evaluate the storage
   stability, we assessed the morphological integrity of the eye drops
   during a 24-hour period. Samples were collected at 0, 12, and 24 hours
   and observed under a light microscope (Nikon, Japan) to assess the
   preservation of the spherical morphology over time ([142]55, [143]56).

   The viability of MSCs within the stored microcarriers was assessed
   after 24 hours of storage at 4°C in PBS. RGD-Alg@MSCs were dissolved
   using 55 mM sodium citrate solution, followed by centrifugation to
   collect the encapsulated MSCs. Cell viability was evaluated using flow
   cytometry with calcein AM/PI staining. Three conditions were compared:
   (i) MSCs cultured under standard conditions (37°C, complete medium),
   (ii) MSCs in PBS at 4°C, and (iii) RGD-Alg@MSCs stored in PBS at 4°C.

In vitro DED model

   To mimic an in vitro DED model in HCECs, a medium (450 mOsm/liter) was
   prepared by supplementing NaCl (Shanghai Hushi), following the previous
   method ([144]57).

Evaluation of RGD-Alg@MSCs for DED treatment in vitro

   To simulate DED in vitro, the cells were subjected to hyperosmotic
   stress in a modified medium. Then, RGD-Alg@MSCs were inserted into the
   upper inserts of the Transwell system (Labselect) to coculture with the
   HCECs for 24 hours ([145]58). Following the cocultivation procedure, a
   set of experiments was carried out to examine the potential therapeutic
   benefits of RGD-Alg@MSCs.

Cell viability recovery test

   The vitality of HCECs was tested with an assay of cell counting kit-8
   (CCK-8) (Vazyme) after a 24-hour coculture ([146]20). The optical
   density at 450 nm was measured using a microplate reader (Thermo Fisher
   Scientific), in accordance with protocol from the manufacturer for the
   test.

Evaluation of apoptosis rate

   Using an apoptosis kit (Abbkine), an in vitro evaluation was carried
   out to determine the capacity of RGD-Alg@MSCs to reduce apoptosis
   ([147]59). Flow cytometry using the BD Accuri C6 (BD Biosciences, USA)
   was used. The analysis was done by FlowJo. Figure S24 depicts the flow
   cytometry gating rule.

Real-time quantitative PCR

   TRIzol reagent (Invitrogen, USA) was used to extract the total RNA from
   cells as well as corneal and conjunctival tissues ([148]57). One
   microgram of RNA was reverse-transcribed into complementary DNA (cDNA)
   using reverse transcription kit (Vazyme). The resulting cDNA was then
   amplified through RT-qPCR on the Q5 system using a SYBR Green-based
   master mix (Vazyme). A set of primers was displayed in table S1.

Assessment of cytokine release

   To quantify the release of cytokines specifically from MSCs in the
   porous RGD-Alg microcarriers, both single-culture and coculture
   experiments were conducted. MSCs alone or RGD-Alg@MSCs alone were
   cultured in complete medium as controls, while parallel coculture
   experiments with HCECs were performed as described previously.
   Supernatants from both conditions (single culture and coculture) were
   collected after a 24-hour period and analyzed for the presence of
   therapeutic factors including TSG-6 ([149]60, [150]61, [151]62), CCL20
   ([152]63), IDO ([153]64, [154]65), and PGE[2] ([155]66), and an
   enzyme-linked immunosorbent test (ELISA) was used. Supernatants were
   collected and subjected to specific ELISA kit analysis following a
   24-hour coculture period (table S2).

RNA-seq and bioinformatics analysis

   HCECs and MSCs were cocultured in a Transwell system under hyperosmotic
   conditions to investigate the therapeutic effects of RGD-Alg@MSCs on
   HCECs and explore the potential molecular mechanisms mediating these
   actions. For corneal epithelial cells, the treatment groups included
   cells cocultured with RGD-Alg@MSCs (RGD-Alg@MSCs-Treated) or
   2D-cultured MSCs (MSCs-Treated), while the disease control group
   consisted of HCECs under hyperosmotic conditions without MSC coculture
   (Untreated). For MSCs, the treatment groups included RGD-Alg@MSCs
   (RGD-Alg@MSCs) and 2D-cultured MSCs (MSCs) cocultured with HCECs under
   hyperosmotic conditions, with normal MSCs (not cocultured with HCECs
   under hyperosmotic conditions) serving as the baseline control
   (Control). HCECs from the RGD-Alg@MSCs-Treated, MSCs-Treated, and
   Untreated groups, as well as MSCs from the RGD-Alg@MSCs, MSCs, and
   control MSC groups, were collected for RNA extraction.

   Total RNA was extracted using TRIzol reagent (Invitrogen) following the
   protocol of the manufacturer. RNA quantity and purity were assessed
   with NanoDrop ND-1000 (NanoDrop, USA), and integrity was confirmed
   using Bioanalyzer 2100 [Agilent, USA; RNA integrity number (RIN) > 7.0]
   and denaturing agarose gel electrophoresis. Poly(A) mRNA was isolated
   from 1 μg of total RNA using Dynabeads Oligo (dT)25-61005 (Thermo
   Fisher Scientific, USA) with two purification rounds, fragmented at
   94°C for 5 to 7 min using the Magnesium RNA Fragmentation Module [New
   England Biolabs (NEB); catalog no. E6150, USA], and reverse-transcribed
   into cDNA with SuperScript II Reverse Transcriptase (Invitrogen,
   catalog no. 1896649, USA). Second-strand DNA was synthesized using
   Escherichia coli DNA polymerase I (NEB, catalog no. M0209, USA),
   ribonuclease H (NEB, catalog no. M0297, USA), and dUTP solution (Thermo
   Fisher Scientific, catalog no. R0133, USA). After adding A-bases to the
   blunt ends, indexed adapters were ligated, and size selection was
   performed using AMPureXP beads. The U-labeled second-strand DNA was
   treated with UDG enzyme (NEB, catalog no. M0280, USA), followed by PCR
   amplification (initial denaturation at 95°C for 3 min; eight cycles of
   98°C for 15 s, 60°C for 15 s, 72°C for 30 s; final extension at 72°C
   for 5 min). The resulting cDNA libraries [average insert size: 300 ± 50
   base pairs (bp)] were sequenced on an Illumina NovaSeq 6000 (LC-Bio
   Technology Co. Ltd., Hangzhou, China) using 2 × 150 bp paired-end
   sequencing (PE150) ([156]67).

   Raw sequencing data were quality-filtered using fastp (default
   parameters) to remove adapters, low-quality reads, and undetermined
   bases. Clean reads were aligned to the Homo sapiens GRCh38 reference
   genome using HISAT2 ([157]68). Transcripts were assembled and merged
   across samples using StringTie ([158]69) and gffcompare, respectively.
   Gene expression levels were quantified as FPKM (fragments per kilobase
   of transcript per million mapped reads) using StringTie. Differential
   expression analysis was performed with the R package edgeR ([159]70),
   defining DEGs as those with a fold change of >2 or <0.5 and P-adjust
   value of <0.05 using the Benjamini-Hochberg correction method. The DEGs
   were subjected to GO ([160]71) and KEGG ([161]72) pathway enrichment
   analysis using DAVID software, with a significance threshold of
   P < 0.05.

Coculture of RGD-Alg@MSCs and DCs

   After inducing the maturation of DCs, a Transwell chamber was placed
   into the well plate. The upper layer of the chamber was loaded with
   equal amounts of RGD-Alg empty vector, conventionally cultured MSCs,
   and RGD-Alg@MSCs. The coculture was conducted for 24 hours, with the
   uninduced iDC group serving as the control. To evaluate the expression
   of DC surface markers including CD14, CD11c, CD83, CD86, human
   leukocyte antigen (HLA)–DR, and CD40, flow cytometry was used
   ([162]30).

Coculture of treated DCs with naïve CD4^+ T cells

   After the DC coculture was finished, treated DCs were collected, washed
   with PBS, and cocultured at a 1:10 ratio with naïve CD4^+ T cells
   ([163]30, [164]73). To encourage the development of T[H]17 cells, the T
   cells were stimulated with plate-bound anti-CD3 and soluble anti-CD28
   antibodies, and the medium was supplemented with a cytokine cocktail of
   IL-23, TGF-β and IL-6, along with neutralizing antibodies against |L-4
   and IFN-γ. The coculture lasted for 72 hours. The percentage of T cells
   that developed into T[H]17 cells in each treatment group was assessed
   using flow cytometry, following the gating strategy for T[H]17 cells
   (CD11c^− CD4^+ IL-17A^+), as illustrated in fig. S25.

DED mouse model and treatment

   To assess the efficacy of RGD-Alg@MSC eye drops, we used 12-week-old
   male NOD mice (Huachuang Sino), which naturally develop symptoms
   consistent with autoimmune DED. Male NOD mice were specifically
   selected because they exhibit more severe lacrimal gland dysfunction
   compared to females ([165]74), with pronounced CD4^+ T cell
   infiltration leading to acinar atrophy and ductal obstruction. This
   model is well established for studying SS-like DED, characterized by
   reduced tear secretion, corneal epithelial damage, conjunctival goblet
   cell loss, and inflammatory infiltration of the lacrimal glands. The
   pathological changes in this model closely resemble the dysfunction of
   the lacrimal functional unit observed in human DED patients, making it
   suitable for evaluating therapeutic interventions ([166]27).

   The mice were assigned to four groups of 15 by blinded randomization.
   Eye drops containing 5 μl of either RGD-Alg@MSCs (containing about 3 ×
   10^5 MSCs), porous RGD-Alg microcarriers, or suspended MSCs (containing
   about 3 × 10^5 MSCs) were applied twice daily for 14 days in three
   groups under different treatments. The left group that received no
   treatment was NOD group. Furthermore, a group of male ICR mice of the
   same age (Huachuang Sino) kept in the same setting with no
   modifications served as the control for NOD mice. The study followed
   the guidelines established by the ethics committee of Nanjing Drum
   Tower Hospital (permission number 2022AE01039).

   The effectiveness of RGD-Alg@MSC eye drops was assessed using
   fluorescein staining, where standard sodium fluorescein strips (Tianjin
   Jingming) moistened with 200 μl of sterile water were gently applied to
   the mouse corneal surface and evaluated 2 min post-application,
   yielding a score between 0 and 15 ([167]75). Tear volume was measured
   using phenol red thread test, which was selected for its established
   advantages in small animal models including minimally invasive
   application, precise measurement capability, and suitability for the
   small palpebral fissure size of mice compared to standard Schirmer
   strips ([168]76). Eye photography was also used to document changes.
   After 14 days of treatment, ocular tissues were collected for periodic
   acid–Schiff (PAS) staining and H&E staining to examine the structural
   alterations in corneal tissue.

PAS staining and H&E staining

   The specimens that were gathered were preserved in a solution of 4%
   paraformaldehyde (PFA; Servicebio). After that, these specimens were
   immersed in paraffin and then sectioned into slices. PAS staining was
   performed on the slices of the eyeballs and eyelids using a kit from
   Solarbio. The sections of lacrimal glands and harderian glands
   underwent H&E staining using a kit from Servicebio. Nikon digital light
   microscope was used to take and evaluate images of the stained sections
   ([169]57).

In vitro biocompatibility studies

   To evaluate the in vivo biocompatibility of RGD-Alg@MSC eye drops,
   healthy male ICR mice (12 weeks old, n = 3 per group) were treated with
   RGD-Alg@MSC eye drops (5 μl per eye, twice daily) for 14 days. Other
   groups received the porous RGD-Alg microcarriers, MSCs, or no treatment
   under the same conditions. After treatment, mice were euthanized and
   important organs were removed and stained with an H&E kit. Stained
   sections were imaged using a Nikon digital light microscope to assess
   histological changes and toxicity.

TUNEL staining

   To assess potential ocular tissue damage following RGD-Alg@MSC eye drop
   treatment, TUNEL staining was performed on the collected eyeballs
   ([170]77). Eyeballs were fixed in 4% PFA, embedded in paraffin, and
   sectioned into 5-μm slices. TUNEL staining was conducted using a TUNEL
   assay kit (Servicebio) according to the instructions of the
   manufacturer. Briefly, sections were deparaffinized, rehydrated, and
   treated with proteinase K (20 μg/ml) for 15 min at 37°C to permeabilize
   the tissue. The TUNEL reaction mixture was applied and incubated for 1
   hour at 37°C in a humidified chamber. Positive controls were prepared
   by treating sections with deoxyribonuclease (DNase) I (100 U/ml) for 10
   min to induce DNA strand breaks, while negative controls were incubated
   with the TUNEL reaction mixture lacking TdT. Sections were
   counterstained with 4′,6-diami-dino-2-phenylindole (DAPI) to label
   nuclei and imaged using a scanning microscope (VS200, Olympus).
   Apoptotic cells were identified by colocalization of TUNEL-positive
   signals (green) and DAPI-stained nuclei (blue).

In vivo retention study of RGD-Alg@MSCs

   To evaluate the ocular retention time of MSCs, RFP-MSCs were either
   encapsulated in RGD-Alg@MSCs or left as conventional suspended MSCs.
   For in vivo imaging, mice were briefly anesthetized with isoflurane
   inhalation only during the examination time points, while remaining
   conscious and maintaining normal activity throughout the observation
   period. A volume of 5 μl of eye drops containing 3 × 10^5 RFP-MSCs was
   topically administered to the corneal surface of 12-week-old male ICR
   mice. Fluorescence imaging was performed using a multi-mode in vivo
   animal imaging system (AniView100) at 0, 40, 80, and 120 min
   post-application ([171]78).

Analysis of DCs and T[H]17 cells in the dLNs

   To create a homogeneous single-cell suspension, lymph nodes were
   removed, mechanically cracked open, and passed through 70-μm strainers
   of filter paper ([172]32, [173]79). Using conjugated primary
   antibodies, this suspension was stained to identify DCs, major
   histocompatibility complex class II (MHC-II) [MHC-II–phycoerythrin
   (PE), eBioscience], CD45 (CD45-PE-CYANINE7, eBioscience), CD86
   [CD86-allophycocyanin (APC), eBioscience], and CD11c [CD11c–fluorescein
   isothiocyanate (FITC), BioLegend]. The staining process was then
   prolonged for 30 min at 4°C. Using antibodies of CD45-PE-CYANINE7
   (eBioscience) and CD4-FITC (BioLegend), cell suspensions were treated
   with a cell stimulation cocktail (BD Biosci-ences) for 5 hours at 37°C
   to identify T[H]17 cells. After permeabilization and fixing, cells were
   intracellularly labeled with PE-conjugated anti-IL-17A (IL-17A-PE,
   eBioscience). Isotype controls were both used. With the Accuri C6 flow
   cytometer and FlowJo V10 software, flow cytometric analysis was carried
   out. The gating rule is displayed in fig. S26.

Immunofluorescence staining

   Immunofluorescence staining was conducted according to previously
   established methods ([174]80). In brief, mouse corneas were preserved
   in 4% PFA. These corneas were then treated with 10% Triton-X 100
   (Beyotime) for 1 hour to permeabilize the cell membranes. To minimize
   nonspecific binding, corneas were treated with 5% donkey serum (Jackson
   ImmunoResearch) at 37°C for 1 hour. The corneas were then exposed to
   CD11c antibodies (1:100 dilution, PE conjugate, BioLegend) for an
   entire night at 4°C to measure and examine morphology. Following the
   incubation of the antibodies, the samples were cleaned with PBS and
   coated with an antifade mountant containing DAPI (Abcam). A confocal
   laser scanning microscope (Zeiss, Germany) was used to record the
   staining results, and ImageJ software was used to evaluate them. To
   ascertain dendritic complexity, two blinded evaluators examined
   enhanced magnification confocal maximum projection z-stacks, focusing
   on characteristic dendritic shapes and manually enumerating the
   dendritic tips per cell.

Analytical statistics

   GraphPad Prism (version 10) was used for all statistical analysis, and
   the data are shown as the mean ± SD of a minimum of three biological
   replicates. First, the normalcy was evaluated using the Shapiro-Wilk
   test. If the data were normally distributed, a repeated-measures
   analysis of variance was carried out for comparisons across several
   groups, and Dunnett’s multiple comparison test was then run. Friedman
   test was used to examine group differences for non-normally distributed
   data, with Dunn’s post hoc test used to identify which specific groups
   differ from each other. Adjusted P values were provided to reduce type
   I errors brought on by numerous comparisons and guarantee the
   reliability of the findings. When comparing just two groups, the
   Mann-Whitney U test was used for distributed non-normally data, and
   Welch’s t test was selected for regularly distributed data. Statistical
   significance was established as P < 0.05 in all situations. For the
   purpose of capturing potential differences in both directions, all
   statistical tests were two-tailed. P < 0.05 is denoted by *, P < 0.01
   by **, and P < 0.001 by ***, while ns denotes no significance.

Acknowledgments