CD146-positive mesenchymal stromal cells: A promising subtype for enhanced acute respiratory distress syndrome therapy

Abstract

Acute respiratory distress syndrome (ARDS) remains without effective targeted reparative therapies and continues to carry a high mortality rate. Here, we comment on a recent study by Zhang et al, who identify a CD146+ subpopulation of mesenchymal stromal cells (MSCs) with superior reparative function in lipopolysaccharide-induced ARDS in mice. Compared with CD146- MSCs, CD146+ MSCs secreted higher levels of reparative paracrine mediators, including hepatocyte growth factor, vascular endothelial growth factor, prostaglandin E2 (PGE2), and angiopoietin 1, better preserved endothelial junctional proteins (VE-cadherin, zonula occludens-1), and more effectively modulated T cell and macrophage phenotypes. Mechanistic studies link these effects to a nuclear factor kappa B (NF-κB)/cyclooxygenase-2/PGE2 signaling axis, as pharmacologic blockade of NF-κB (caffeic acid phenethyl ester) abrogated the benefits. We place these results in the context of MSC heterogeneity research, highlight strengths (mechanistic depth, in vivo validation) and limitations (single animal model, reliance on cell lines rather than primary human cells), and propose next steps: Testing efficacy across diverse ARDS etiologies (viral, aspiration), validating effects in primary human alveolar and endothelial cells, delineating the CD146/NF-κB cascade, developing potency biomarkers (e.g., PGE2), and performing rigorous safety profiling. Strategies to enrich or prime MSCs for CD146-associated NF-κB/cyclooxygenase-2/PGE2 program activity may provide a practical route to higher potency.

Key Words: CD146; Mesenchymal stromal cell; Acute respiratory; Distress syndrome; Nuclear factor kappa B; Cyclooxygenase-2; Prostaglandin E2

Core Tip: This article presents evidence that CD146+ mesenchymal stromal cells define a high-potency subpopulation for acute respiratory distress syndrome therapy by engaging a nuclear factor kappa B → cyclooxygenase-2 paracrine signaling program. Enriching or priming mesenchymal stromal cells for CD146 expression and nuclear factor kappa B/cyclooxygenase-2 activity may yield practical potency biomarkers and enable more consistent, mechanism-driven cell therapies that more effectively restore alveolar-capillary barrier integrity.

TO THE EDITOR

This commentary specifically addresses the study by Zhang et al[1]. The strengths and limitations discussed below refer to the experimental design and reporting in the study. Acute respiratory distress syndrome (ARDS) remains a major clinical challenge with high mortality and limited targeted therapies[2,3]. A recent study by Zhang et al[1] reported that a CD146+ subpopulation of mesenchymal stromal cells (MSCs) substantially outperforms CD146- MSCs in a lipopolysaccharide (LPS)-induced ARDS mouse model. CD146+ MSCs show enhanced nuclear factor kappa B (NF-κB)/cyclooxygenase-2 (COX-2) signaling, which increases secretion of reparative paracrine mediators, including hepatocyte growth factor (HGF), vascular endothelial growth factor (VEGF), prostaglandin E2 (PGE2), and angiopoietin 1, and better preserves endothelial junction proteins, like VE-cadherin and zonula occludens-1 (ZO-1). Pharmacologic blockade of NF-κB with caffeic acid phenethyl ester (CAPE) abrogates these benefits, implicating a CD146-associated NF-κB/COX-2 program axis in MSC-mediated lung repair.

These observations align with prior literature on MSC heterogeneity and potency. Multiple reports indicate superior immunomodulatory and regenerative properties of CD146+ MSCs[4,5]. For example, Wu et al[4] found that CD146+ adipose-derived MSCs more effectively suppressed T helper type 17-driven inflammation in an arthritis model, and Zhang et al[5] reported that CD146+ umbilical-cord MSCs accelerated sepsis resolution by expanding regulatory T cells and reparative macrophages. Consistent with these studies, the new ARDS data show that CD146+ cells secrete higher VEGF and PGE2 and better restore endothelial barrier function. Mechanistically, inflammatory stimuli trigger MSC NF-κB signaling to upregulate VEGF and HGF[6], and COX-2-derived PGE2 is a well-established mediator of MSC immunomodulation[7]. Overall, CD146 enrichment or priming of MSCs for NF-κB/COX-2/PGE2 activity may provide a practical potency biomarker and a route toward more consistent, mechanism-driven cell therapies for ARDS.

Notably, phase I clinical trials indicate that MSC therapy is safe and produces transient improvements in lung injury scores and inflammatory cytokine profiles[7]; selecting CD146+ MSCs may further potentiate these effects. The study’s strengths include in vivo validation and mechanistic insight into MSC heterogeneity. Given the greater complexity of human ARDS, the reliance on murine models and in vitro assays is a significant limitation. Future work should evaluate sorted CD146+ MSCs across additional ARDS models, including viral and aspiration pneumonia, and delineate downstream NF-κB-COX-2 signaling. Clinically, enriching MSC preparations for CD146 expression or priming MSCs, for example, with PGE2, could optimize therapeutic potency. Below, we summarize key distinguishing features of CD146+ vs CD146- MSCs and outline priorities for preclinical-to-clinical translation.

KEY FINDINGS AND INTERPRETATION

In the study by Zhang et al[1], human umbilical cord MSCs were cultured in alternative media formulations; cells expanded in YF medium exhibited substantially higher CD146 expression and superior therapeutic efficacy in an LPS-induced ARDS mouse model. CD146+ MSCs were enriched by magnetic sorting and compared with the CD146- fraction. The CD146+ subpopulation secreted significantly greater amounts of pro-angiogenic and anti-inflammatory mediators HGF, VEGF, PGE2, and angiopoietin 1. In functional assays, CD146+ MSCs more effectively preserved endothelial barrier integrity, reducing LPS-induced permeability and maintaining VE-cadherin and ZO-1 junctional localization in human lung endothelial cells. They also exerted stronger immunomodulatory effects, increasing regulatory T cells and reparative (M2-like) macrophages while suppressing proinflammatory cytokine production.

Mechanistic profiling by transcriptomic and proteomic analyses showed enrichment of NF-κB signaling and upregulation of its downstream effector COX-2 in CD146+ MSCs. COX-2 catalyzes PGE2 synthesis, consistent with the higher PGE2 release measured from CD146+ cells. Pharmacologic inhibition of NF-κB with CAPE markedly suppressed COX-2 expression and reduced secretion of PGE2, VEGF, and HGF, and it abolished the CD146+ MSC-mediated preservation of endothelial barrier integrity. Together, these data implicate a CD146-NF-κB-COX-2 signaling axis as central to the enhanced reparative, paracrine phenotype of the CD146+ subpopulation. Accordingly, CD146 may mark an MSC subset that is preactivated or primed for inflammation-triggered, paracrine-mediated tissue repair.

Zhang et al[1] cultured human umbilical-cord MSCs in alternative media and reported that cells expanded in a YF culture system (YF medium supplied by Shandong Qilu Cell Therapy Engineering Technology Co., Ltd) exhibited substantially higher CD146 expression and superior efficacy in an intratracheal LPS (7 mg/kg)-induced ARDS mouse model. CD146+ MSCs were enriched using the EasySep™ PE positive selection kit (PE-anti-CD146, BioLegend, clone/catalog 361006). Pre-sort proportions ranged from approximately 12% to 32%, and post-sort purities were reported in the accompanying flow cytometry data. In vitro assays employed immortalized human lung microvascular endothelial cells (HULEC-5a) and SV40-transformed human type II alveolar epithelial cells (HPAEpic-II-SV40) to evaluate permeability, junctional VE-cadherin/ZO-1 localization, apoptosis, and cytokine responses. Mechanistically, CAPE (5 μM) inhibition of NF-κB and shRNA p65 knockdown reduced COX-2 expression and secretion of HGF, PGE2 and VEGF in CD146+ MSCs, and CAPE partially reversed the endothelial/epithelial protective effects in vitro and in vivo. These concrete methodological details strengthen reproducibility but also highlight specific translational caveats, as described below.

COMPARISON WITH EXISTING LITERATURE

Multiple groups have identified CD146 as a marker of functionally potent MSCs. In rheumatoid arthritis models, Wu et al[4] reported that CD146+ adipose-derived MSCs exerted greater immunosuppressive activity than CD146- cells, more effectively suppressing T helper type 17 responses and reducing joint inflammation. Zhang et al[7] found that CD146+ umbilical-cord MSCs accelerated recovery in LPS-induced sepsis by increasing regulatory T cells and phagocytic M2 macrophages and by producing higher levels of VEGF, thereby abbreviating the inflammatory phase. Bikorimana et al[8] showed that CD146^high MSCs displayed enhanced T-cell suppression in vitro and improved survival in graft-vs-host disease models in vivo. Across these studies, CD146+ MSCs consistently secrete elevated amounts of immunomodulatory and trophic factors (e.g., interleukin-6, VEGF) and demonstrate superior therapeutic efficacy.

It is important to emphasize that the primary functional data discussed here derive from umbilical-cord MSCs. Although CD146 is detectable on MSCs derived from multiple tissues, including bone marrow and adipose tissue[9,10], and has been associated with enhanced migration, angiogenesis and immunomodulatory capacity in several reports[5,11,12], the extent to which it reliably identifies the same high-potency functional subset across distinct tissue origins remains unresolved. Tissue-specific microenvironments, donor variability, and culture conditions can influence CD146 expression and associated programs, such that CD146 positivity in one tissue source may not predict identical transcriptional or functional states in another. Therefore, claims that CD146 is a universal functional marker for MSC potency should be made cautiously and require direct head-to-head validation across bone marrow, adipose, umbilical cord, and other MSC sources.

The current report aligns with this literature: CD146+ MSCs outperform CD146- counterparts in inflammatory injury models, but it extends prior work in three important ways. First, previous studies primarily addressed immune-mediated diseases (arthritis, graft-vs-host disease, sepsis); the present study is among the first to show a clear functional advantage of CD146+ MSCs in a primary lung-injury/ARDS model. Second, it uniquely links CD146 to an NF-κB-driven COX-2/PGE2 paracrine program that underpins the reparative phenotype; earlier work did not identify NF-κB/COX-2 signaling as the principal mechanism. Additionally, this study highlights direct preservation of alveolar-capillary junction integrity (VE-cadherin/ZO-1), whereas prior reports emphasized immunosuppression and engraftment. Table 1 summarizes the consensus: CD146+ MSCs secrete more pro-regenerative factors and exert stronger anti-inflammatory effects. The current study confirms that these enhancements translate into improved preservation of the lung endothelial barrier, a central therapeutic target in ARDS.

Table 1 Comparison of CD146+ and CD146- mesenchymal stromal cell properties relevant to acute respiratory distress syndrome therapy.

Feature/outcome	CD146+ MSCs	CD146- MSCs	Ref.
Paracrine factor secretion	↑ HGF, VEGF, PGE2, Ang-1 (pro-repair)	Lower levels	Zhang et al[5], 2023; Hezam et al[13], 2023
Immunomodulation (T cells)	↑ Treg induction; suppress Th17 proliferation	Less immunosuppressive	Zhang et al[5], 2023; Wu et al[4], 2016
Macrophage effects (M1/M2)	Accelerated M2 polarization, ↑ phagocytosis	Slower M2 shift	Zhang et al[5], 2023
NF-κB/COX-2 pathway activity	Upregulated (↑COX-2, p-NF-κB)	Baseline levels	Current study
Endothelial barrier support	Enhanced VE-cadherin/ZO-1 stability, reduced leak	Weaker barrier protection	Current study

MSC: Mesenchymal stromal cell; HGF: Hepatocyte growth factor; VEGF: Vascular endothelial growth factor; PGE2: Prostaglandin E2; Ang-1: Angiopoietin 1; Treg: Regulatory T cell; Th17: T helper type 17; NF-κB: Nuclear factor kappa B; COX-2: Cyclooxygenase-2; ZO-1: Zonula occludens-1.

MECHANISTIC INSIGHTS: NF-ΚB/COX-2/PGE2 AXIS

Zhang et al[1] present transcriptomic and protein data linking the CD146+ phenotype to enrichment of NF-κB signaling and upregulation of COX-2, with consequent increases in PGE2, VEGF and HGF secretion. Pharmacologic inhibition of NF-κB (CAPE, 5 μM) and p65 knockdown reduced COX-2 expression and attenuated the paracrine and barrier-protective effects, supporting a role for NF-κB → COX-2 → PGE2 in the observed phenotype. Notably, however, Zhang et al[1] did not directly manipulate CD146 expression, for example via knockdown or overexpression, to test whether CD146 is a causal upstream mediator vs a marker of a pre-activated MSC subpopulation. Therefore, these data are evidence for a CD146-associated NF-κB/COX-2/PGE2 program rather than proof that CD146 itself is the proximal upstream receptor driving NF-κB activation. Targeted perturbation experiments (CD146 siRNA/shRNA or overexpression) would clarify causality.

COX-2 induction is mechanistically important because COX-2 catalyzes synthesis of PGE2, a pleiotropic mediator that promotes macrophage M2 polarization and regulatory-T-cell responses[7]. Hezam et al[13] showed that PGE2-primed MSCs yield superior outcomes in LPS-induced lung injury, largely via enhanced M2 polarization, and Zhang et al[7] demonstrated that MSC-derived PGE2 activates the cholinergic anti-inflammatory pathway to ameliorate ARDS. By producing higher PGE2 via NF-κB/COX-2, CD146+ MSCs may therefore deliver a dual benefit of immunomodulation and vascular protection. Indeed, PGE2 and other eicosanoids can strengthen endothelial tight junctions through cAMP-mediated Rac activation[13], which could explain the observed upregulation of VE-cadherin and ZO-1 and the reduction in permeability. In summary, the proposed CD146/NF-κB/COX-2/PGE2 signaling axis is biologically plausible and concordant with existing MSC literature, supporting the view that CD146 marks an MSC subset primed for inflammation-triggered, paracrine repair.

STRENGTHS, LIMITATIONS, AND FUTURE DIRECTIONS

The results presented here originate from Zhang et al’s work[1]; before translating these findings into clinical practice, however, it is essential to take note of the study’s limitations. First, the in vivo efficacy was assessed in an intratracheal LPS-induced ARDS mouse model (7 mg/kg), which reflects endotoxin-driven acute injury but fails to encompass other clinically important ARDS etiologies, such as viral pneumonia, aspiration- or trauma-related injury or subsequent fibrosis. Second, key in vitro assays relied on immortalized cell lines (HULEC-5a and HPAEpic-II-SV40), which may diverge from primary human pulmonary microvascular endothelial and alveolar epithelial cells in barrier dynamics and responsiveness to inflammatory stimuli. Third, Zhang et al[1] reported enhanced CD146 expression in MSCs cultured in a YF system; however, the absence of detailed disclosure of the medium’s molecular composition limits reproducibility and hinders progress towards GMP translation. Fourth, while CAPE treatment (5 μM) and p65 knockdown implicate NF-κB/COX-2/PGE2 signaling in the reparative properties attributed to the CD146+ phenotype, whether CD146 directly drives NF-κB activation remains unresolved, as CD146 itself was not experimentally manipulated. Finally, certain methodological factors raise concerns about potential bias. In particular, the authors note that animal experiments were conducted without blinding, which may have influenced outcome assessment. Together, these limitations indicate that the results are promising but should be regarded as preliminary. To accelerate translation from preclinical studies to clinical application, we recommend several steps, outlined in detail below.

Expanded models

Future studies should evaluate CD146+ and CD146- MSCs in additional ARDS paradigms (e.g., viral pneumonia and acid-aspiration models) and in larger mammalian species to establish generalizability beyond the LPS system.

Validation in primary human cells

Permeability, junctional VE-cadherin/ZO-1 localization, apoptosis and cytokine assays must be replicated in primary human pulmonary microvascular endothelial cells and primary alveolar epithelial cells, ideally from ARDS patients, to determine whether these results translate to primary human cells and hold true in a human disease context.

Cross-source validation

Comparative studies of CD146+ and CD146- subpopulations from diverse MSC sources (bone marrow, adipose tissue, umbilical cord and placenta), conducted under standardized sorting and culture conditions, will be critical to evaluate reproducibility across tissue origins. Key readouts should include: (1) Transcriptomic and proteomic profiling to determine whether a shared NF-κB/COX-2/PGE2 signature is present across sources; (2) Parallel functional assays of paracrine secretion, endothelial/epithelial barrier protection and immunomodulation; and (3) Stability testing of CD146 expression during expansion and after cryopreservation. Together, such experiments will determine whether CD146 denotes a conserved high-potency program or instead reflects a tissue-specific subset whose functional meaning depends on cellular origin and manufacturing conditions.

Establishing causality

Causality should be addressed by manipulating CD146 in MSCs (siRNA/shRNA, CRISPRi or overexpression) and quantifying downstream effects on NF-κB activation, COX-2 induction, PGE2 secretion and barrier function.

Mapping downstream signaling in target cells

EP receptor (EP1-EP4) pharmacology together with phosphoproteomic and time-course studies should be employed to identify endothelial/epithelial effectors (e.g., Rac1/RhoA, cAMP/PKA, etc.).

Standardizing manufacturing inputs

Future work should disclose the YF medium formulation or perform reductionist testing of candidate components (growth factors, serum replacements, etc.) to identify drivers of CD146 expression and ensure batch-to-batch reproducibility under GMP-compatible conditions.

Reporting enrichment metrics and comparing strategies

Comprehensive reporting of CD146 enrichment parameters (antibody clone, labeling conditions, magnetic activated cell sorting vs fluorescence activated cell sorting method, post-sort purity and viability, and expansion conditions), together with comparisons between sorting-based enrichment and culture/priming approaches, will be required to identify the most scalable approach.

Prerequisites for clinical testing

Validated potency assays (e.g., standardized PGE2 enzyme-linked immunosorbent assay), GMP-compatible enrichment/priming workflows and comprehensive safety profiling, including angiogenesis and tumorigenicity assessments, should be completed before early-phase clinical trials.

CONCLUSION

This commentary underscores Zhang et al’s finding[1] that CD146 marks an MSC subpopulation with superior reparative and immunomodulatory activity in ARDS. The delineation of a CD146-associated NF-κB/COX-2/PGE2 program provides both a candidate potency biomarker and a mechanistic target for cell-product optimization. These findings support the concept that CD146 identifies an MSC subset with enhanced paracrine and immunomodulatory properties and that a CD146-associated NF-κB/COX-2/PGE2 program contributes to this phenotype. However, given the limitations described above, further mechanistic dissection, replication in diverse models and primary human cells, and standardization of manufacturing inputs are required before clinical translation can be responsibly pursued.