Three separable therapeutic categories derived from mesenchymal stem cells (MSCs) have been investigated for intervertebral disc degeneration (IVDD): MSC transplantation (live cell therapy), MSC-conditioned medium (MSC-CM, crude cell-free secretome), and MSC-derived extracellular vesicles/exosomes (MSC-Exos/EVs, purified nanoscale secretory cargo carriers). All three exert regenerative effects primarily via paracrine signaling to suppress inflammation, oxidative damage, nucleus pulposus (NPC) apoptosis and senescence, while boosting extracellular matrix (ECM) anabolism; however, they differ fundamentally in formulation, safety profiles, manufacturing complexity, therapeutic persistence, and clinical translational maturity, as summarized in Table 1.

Table 1. Comparison of key features, mechanisms, advantages, limitations, and clinical translational stages of the three main MSC-based therapeutic modalities for IVDD.

IVDD is a multifactorial process involving complex interactions between genetic predisposition, mechanical loading, aging, and environmental factors [20]. The hallmark of IVDD is a reduction in NPC numbers and function, leading to impaired ECM homeostasis. Apoptosis, autophagy, pyroptosis, and ferroptosis have all been implicated in NPC death [21].

Cellular senescence plays a pivotal role in IVDD, with senescent NPCs secreting pro-inflammatory cytokines and matrix-degrading enzymes that create a vicious cycle of tissue degradation [22]. p16INK4a and p53/p21 pathways are key regulators of senescence, with their activation contributing to age-related disc degeneration [22].

Oxidative stress is a critical driver, with excessive reactive oxygen species (ROS) promoting NPC apoptosis, senescence, and ECM degradation [23]. Emerging evidence implicates ferroptosis, an iron-dependent form of regulated cell death characterized by lipid peroxidation and GPX4 inactivation, in IVDD pathogenesis [21][24].

During embryonic development, the NP is populated by notochordal cells (NCs), which are essential for disc formation [25]. In humans, NCs disappear during early childhood and are replaced by chondrocyte-like NPCs. This transition correlates with the initiation of degenerative changes, suggesting NCs play a protective role [26].

Notochordal cells have been shown to secrete factors that stimulate NPC proliferation, proteoglycan synthesis, and ECM production [27][28]. Notochordal cell conditioned medium (NCCM) promotes MSC differentiation toward an NP-like phenotype, with enhanced expression of NP markers including aggrecan, collagen II, and SOX9 [27]. The loss of NCs during aging is thought to deprive the NP of these trophic factors, contributing to IVDD progression.

MSCs can be isolated from various tissues including bone marrow (BM-MSCs), adipose tissue (ADSCs), umbilical cord (UC-MSCs), and the NP itself [29]. BM-MSCs are the most extensively studied and have demonstrated safety and efficacy in preclinical models [30]. ADSCs offer advantages of abundant availability and less invasive harvest [31]. UC-MSCs represent an attractive allogeneic source with high proliferative capacity and low immunogenicity [32].

MSCs exert therapeutic effects primarily through paracrine signaling, secreting trophic factors that promote endogenous repair [14]. MSC-derived conditioned medium enhances NPC viability, proliferation, and ECM synthesis while inhibiting apoptosis and inflammation [15]. The MSC secretome contains growth factors (TGF-β, IGF-1, FGF-2), cytokines (IL-10, IL-6), chemokines, and extracellular vesicles that collectively modulate the disc microenvironment.

Several clinical trials have evaluated MSC transplantation for IVDD. A pilot study by Orozco et al. [12] demonstrated that autologous BM-MSC injection improved pain and disability scores at 12-month follow-up. Subsequent studies have reported similar improvements with allogeneic MSCs [33].

Despite promising results, challenges remain. MSC survival after intradiscal injection is poor due to the harsh disc microenvironment [9]. Cell leakage can lead to osteophyte formation [13]. These limitations have driven the development of cell-free approaches utilizing MSC-derived exosomes and conditioned media.

Exosomes are nano-sized (30 - 150 nm) extracellular vesicles of endosomal origin, formed through invagination of the endosomal membrane to produce intraluminal vesicles within multivesicular bodies [34]. Biogenesis is regulated by the ESCRT machinery and ESCRT-independent pathways [35]. MSC-Exos are characterized by exosomal markers including CD9, CD63, CD81, TSG101, and Alix [34].

MSC-Exos contain diverse bioactive molecules including proteins, lipids, and nucleic acids [36]. Proteomic analyses have identified proteins involved in cell adhesion, migration, proliferation, and immunomodulation [37]. Nucleic acid content includes mRNA, miRNA, lncRNA, circRNA, and DNA [11]. MicroRNAs are particularly important as exosomal effectors mediating therapeutic effects in IVDD [16][38][39].

Exosomes are taken up through multiple mechanisms including ligand-receptor interaction, endocytosis, and direct membrane fusion [40]. Upon internalization, exosomal cargo modulates cellular functions through post-transcriptional regulation and signaling pathway modulation [41]. MSC-Exos are efficiently internalized by NPCs, where they exert anti-apoptotic, anti-inflammatory, and pro-anabolic effects [16][42].

MSC-Exos potently inhibit NPC apoptosis through multiple mechanisms. Cheng et al. [16] demonstrated that BM-MSC-Exos deliver miR-21 to NPCs, inhibiting apoptosis by targeting PTEN and activating the PI3K/Akt pathway. MSC-Exos containing miR-532-5p inhibit TNF-α-induced apoptosis by targeting RASSF5 [38], while miR-142-3p attenuates IL-1β-induced apoptosis by targeting MLK3 [39].

Ferroptosis, an iron-dependent form of regulated cell death characterized by lipid peroxidation and GPX4 inactivation, has emerged as a critical mechanism in IVDD [21]. MSC-Exos counteract ferroptosis through two complementary mechanisms:

First, Chen et al. [17] demonstrated that MSC-Exos alleviate oxidative stress-induced ferroptosis by regulating the p62/KEAP1/NRF2 pathway. Mechanistically, MSC-Exos promote H3K27ac modification of the FTO promoter via increased p300/CBP histone acetyltransferase activity, enhancing chromatin accessibility and FTO expression. This reduces Nrf2 mRNA m6A methylation, enhancing NRF2 stability and transcriptional activity, which in turn upregulates antioxidant genes and restores GPX4 expression [17].

Second, Wu et al. [18] showed that microenvironment-educated exosomes (D-EVs) are enriched with GPX4 protein, which is delivered directly to senescent NPCs to inhibit lipid peroxidation and prevent ferroptosis. This direct protein delivery mechanism represents a paradigm shift in exosome therapy, complementing the epigenetic regulation of endogenous antioxidant pathways.

MSC-Exos demonstrate potent anti-senescence effects through multiple pathways. Sun et al. [43] showed that induced pluripotent stem cell-derived MSC exosomes deliver miR-105-5p to senescent NPCs, targeting PDE4D and activating the Sirt6 pathway. Guo et al. [44] demonstrated that urine-derived stem cell exosomes containing MATN3 exert anti-senescence effects through TGF-β pathway activation. Additionally, Wu et al. [18] demonstrated that D-EVs exhibit enhanced targeting toward senescent NPCs through the CXCL10-CXCR3 chemokine axis, enabling selective delivery of anti-senescence cargo without activating NF-κB-mediated pro-inflammatory cascades.

MSC-Exos demonstrate potent anti-inflammatory effects through modulation of multiple pathways. Xia et al. [45] showed that BM-MSC-Exos reduce ROS, NLRP3 inflammasome activation, and inflammatory cytokine production. Exosomal miR-410 targets NLRP3, reducing NPC pyroptosis [46], while miR-223 downregulates IRAK1 to inhibit inflammatory responses [47].

MSC-Exos promote ECM synthesis through multiple mechanisms. Lu et al. [42] showed that BM-MSC-Exos upregulate anabolic genes (aggrecan, collagen II, SOX9) and downregulate catabolic genes (MMP-1, MMP-3). Exosomal miR-199a promotes ECM preservation by targeting GREM1 [48], while miR-532-5p inhibits ECM degradation by targeting RASSF5 [38]. The AMPK/SIRT1 pathway, activated by MOTS-c peptide in functional hydrogels, further promotes anti-apoptotic and anti-senescence effects through regulation of mitochondrial function and oxidative stress [49].

The therapeutic effects of MSC-Exos are mediated through an integrated network of signaling pathways. The PI3K/Akt pathway (via miR-21/PTEN), MAPK pathway (via miR-142-3p/MLK3), and NF-κB pathway are modulated to inhibit apoptosis and inflammation. The Nrf2 antioxidant pathway (via epigenetic regulation of FTO) and GPX4 (via direct protein delivery) synergize to inhibit ferroptosis. The CXCL10-CXCR3 axis enables targeted delivery to senescent cells, while TGF-β and Sirt6 pathways regulate senescence and ECM synthesis. This multi-targeted approach distinguishes exosome therapy from single-agent interventions and underlies its therapeutic potency.

Wu et al. [18] developed “domesticated exosomes” (D-EVs) by educating MSCs with senescent NPC conditioned medium. This approach yielded exosomes with enhanced targeting capabilities toward senescent NPCs, mediated by the CXCL10-CXCR3 chemokine axis. D-EVs were enriched with GPX4 protein, which directly inhibited ferroptosis and alleviated NPC senescence.

Genetic modification of MSCs has been explored to enhance therapeutic potential. Khalid et al. [50] demonstrated that overexpression of SOX9 and TGF-β1 promoted chondrogenic differentiation and regenerated IVDD in vivo. Kim et al. [51] developed MSCs with a tetracycline-off system expressing TGF-β1, IGF-1, and BMP-7, demonstrating superior therapeutic effects.

Lin et al. [49] engineered a MOTS-c-modified functional self-assembling peptide hydrogel that enhanced NP-MSC activity by activating the AMPK/SIRT1 pathway. This system demonstrated sustained MOTS-c release and promoted ECM synthesis, with in vivo validation showing preserved disc height and reduced degeneration.

Therapeutic efficacy depends on cell source, cargo, and delivery system. BM-MSCs are most studied, but urine-derived and iPSC-MSCs may offer superior anti-senescence effects. miRNAs targeting PI3K/Akt, NF-κB, and Nrf2 are consistently linked to benefit, yet GPX4 protein cargo directly inhibits ferroptosis. Smart hydrogels outperform bolus injection by enabling sustained release and targeting. Collectively, combinatorial engineering of cargo and delivery platform, rather than cell source alone, emerges as the priority.

Exosome-based therapies for IVDD are currently in preclinical stages, with no approved clinical trials specifically targeting IVDD. However, MSC-Exos have been evaluated in clinical trials for other conditions, including osteoarthritis, myocardial infarction, and wound healing [52][53]. These studies have demonstrated safety and preliminary efficacy, supporting the potential for IVDD applications.

Several clinical trials have evaluated MSC transplantation for IVDD, providing proof-of-concept for cell-based approaches. Noriega et al. [33] conducted a randomized controlled trial of allogeneic BM-MSCs for IVDD, demonstrating safety and potential efficacy at 12 months. Pettine et al. [54] reported 3-year follow-up data for autologous BM-MSC injection, showing sustained pain relief and functional improvement.

Standardization of isolation, characterization, and quality control is essential [19]. Cell source, culture conditions, and isolation methods affect exosome yield, purity, and bioactivity. Regulatory pathways are still evolving, with classification as biological products or advanced therapy medicinal products [55].

A major translational barrier is the striking lack of reproducibility across studies. Isolation methods (ultracentrifugation, size-exclusion chromatography, precipitation) yield exosome populations with divergent purity and cargo [19], while culture conditions and harvest timing further compound variability [34]. Potency assays lack consensus, with disparate readouts (apoptosis, ECM synthesis, antioxidant effects) obscuring cross-study comparison [16]-[18]. Dose reporting is inconsistent (mass, particle count, arbitrary units), and outcome measures in animal models are non-uniform [56]. Harmonized reference materials, standardized potency assays, and consensus reporting guidelines are urgently needed to bridge mechanistic insights and clinical translation [19].

Studies have used exosome doses ranging from 10 - 100 μg/mL [16]. Intradiscal injection is the primary route, but retention is poor, necessitating sustained-release systems [18]. Hydrogel delivery systems show promise in improving retention and controlled release.

MSC-Exos are generally considered safe due to low immunogenicity and lack of tumorigenic potential [10]. However, careful evaluation of biodistribution, clearance, and off-target effects is essential. Standardized donor screening and quality control ensure product safety and consistency.

Based on patient-specific disease characteristics, exosomes could be engineered to deliver optimal cargo combinations. Yuan et al. [57] demonstrated that exosomes derived from human placental MSCs carrying antagomiR-4450 alleviate IVDD through upregulation of ZNF₁₂₁.

Combining MSC-Exos with biomaterials, growth factors, or small molecules may enhance therapeutic efficacy. Zhang et al. [58] demonstrated that combined hydrogel and MSC therapy improves outcomes in moderate-severity disc degeneration.

Preclinical validation in large animal models is essential before clinical translation. The porcine model is particularly relevant due to similarities in disc anatomy and biomechanics [56]. Clinical trials are needed to establish safety and efficacy in humans.

Synthetic miRNAs or miRNA inhibitors could be delivered directly or through engineered exosomes to modulate specific pathways. Zhu et al. [38][39] demonstrated that exosomal miR-532-5p and miR-142-3p target key pathways in IVDD.

Comprehensive multi-omics analysis of MSC-Exos and their effects on NPCs will accelerate understanding of mechanisms and identification of therapeutic targets. Wu et al. [18] used transcriptomic and proteomic analysis to identify GPX4 as a key effector molecule in D-EVs.