Histone lactylation is an emerging epigenetic modification that links metabolic changes, particularly in the tumor microenvironment, to gene expression alterations driving cancer development. This study leverages a multi-omics approach, integrating bulk RNA sequencing, single-cell RNA sequencing (scRNA-seq), and spatial transcriptomics, to dissect the role of lactylation in GBM at unprecedented resolution. By probing datasets from GEO and TCGA, the team identified lactylation-related genes that are markedly upregulated in GBM tissues and are associated with immunosuppressive microenvironments and poor clinical outcomes.

A key finding centers around the discovery of distinct malignant tumor cell subpopulations exhibiting high levels of lactylation, which reside predominantly within hypoxic regions of the tumor core. These hypoxic niches are well-known for fostering aggressive tumor phenotypes that evade immune surveillance. Single-cell analyses revealed that these lactylated clusters undergo profound metabolic reprogramming, tailoring their gene expression to survive and thrive under oxygen-deprived conditions, while concurrently orchestrating mechanisms to suppress the surrounding immune response.

Spatial transcriptomics added another critical dimension to the findings by mapping the precise localization of these lactylated tumor cells within the heterogeneous tumor architecture. In particular, cells expressing high levels of S100A6, a gene intimately linked to lactylation, were found concentrated in aggressive tumor regions notorious for rapid proliferation and invasion. This spatial information underscores the functional heterogeneity within GBM and provides a tangible target for therapeutic interventions.

To translate these molecular insights into clinical practice, the researchers developed a prognostic risk model based on nine lactylation-associated genes. Using LASSO-Cox regression—a powerful statistical method for feature selection—they stratified GBM patients into distinct high- and low-risk groups. Strikingly, this model demonstrated impressive predictive accuracy with area under the curve (AUC) values ranging from 0.77 to 0.87, suggesting its potential utility as a robust biomarker panel for patient prognosis and treatment stratification.

The compelling prognostic value of the lactylation signature is further supported by experimental validation. In vitro functional assays targeting S100A6 demonstrated that silencing this gene significantly impaired GBM cell proliferation, migration, and invasion, highlighting its pivotal role in maintaining tumor aggressiveness. These findings position S100A6 not merely as a biomarker but as a potential therapeutic target for disrupting lactylation-driven malignant programs.

Underpinning these discoveries is the innovative application of SCENIC transcriptional network inference and CellChat intercellular communication modeling. These computational tools enabled the authors to uncover regulatory networks and cell-cell interactions modulated by lactylation, providing mechanistic insights into how tumor cells rewire signaling pathways to foster immune suppression and metabolic adaptation in GBM.

Pseudotime trajectory analyses further delineated the dynamic states of tumor cell populations, tracing the evolutionary paths from less aggressive to more malignant lactylated states. This temporal framework enriches our understanding of tumor progression and highlights critical junctures where therapeutic interventions might be most effective.

The study also sheds light on the tumor immune microenvironment, revealing that lactylation-associated clusters contribute to the establishment of immunosuppressive niches. This finding dovetails with accumulating evidence that metabolic reprogramming in tumors orchestrates immune evasion, a major challenge for immunotherapies in GBM.

Moreover, the emergence of lactylation as a key metabolic-epigenetic axis opens avenues for novel therapeutic strategies. Targeting enzymes responsible for lactylation or the downstream effectors, such as S100A6, could potentially disrupt malignant metabolic circuits, sensitize tumors to immune attack, or enhance the efficacy of existing treatments.

Beyond its immediate clinical relevance, this research marks a significant advance in cancer biology by employing integrated single-cell and spatial transcriptomics to parse tumor complexity. This multidimensional profiling affords a holistic view of cellular heterogeneity, spatial organization, and functional states within tumors—an approach likely to become foundational in precision oncology.

Despite these promising findings, challenges remain. Validation of the prognostic model and therapeutic targets in larger, independent patient cohorts and in vivo models will be essential to confirm their utility. Additionally, translating knowledge of lactylation into safe and effective clinical interventions will require comprehensive understanding of the broader systemic effects of modulating this epigenetic mark.

Nevertheless, this study underscores the transformative potential of marrying metabolic insights with high-resolution transcriptomic technologies to redefine our understanding of glioblastoma. By pinpointing lactylation as a central player in tumor cell clustering, metabolic adaptation, and immune modulation, it opens a new chapter in the fight against one of the deadliest brain cancers.

In summary, this pioneering work reveals that lactylation is more than a metabolic footnote in glioblastoma biology; it is a defining feature of tumor heterogeneity and aggressiveness. The identification of lactylation-enriched tumor cell clusters, spatially anchored in hypoxic niches and regulated by signatures including S100A6, provides a powerful prognostic tool and therapeutic target. This research paves the way for the development of lactylation-focused strategies that could revolutionize glioblastoma treatment and improve outcomes for patients facing this devastating disease.

Subject of Research: Metabolic reprogramming through histone lactylation in glioblastoma, its association with tumor heterogeneity, immune evasion, and prognosis.

Article Title: Single-cell and spatial transcriptomics reveal lactylation-associated tumor cell clusters and define a prognostic risk model in glioblastoma

Article References:
Han, R., Chi, G., Sun, D. et al. Single-cell and spatial transcriptomics reveal lactylation-associated tumor cell clusters and define a prognostic risk model in glioblastoma. BMC Cancer (2025). https://doi.org/10.1186/s12885-025-15291-6

Image Credits: Scienmag.com

DOI: https://doi.org/10.1186/s12885-025-15291-6

Intratumoral heterogeneity is a hallmark of GBM, manifesting as a dynamic mosaic of genetically and phenotypically distinct neoplastic cell populations within the same lesion. Molecular targets such as EGFRvIII, once considered promising, reveal an unsettlingly patchy expression across tumor cells. This heterogeneity fosters immune evasion; antigen-negative clones often dominate at recurrence after an initial response to CAR-T therapy. Moreover, glioma stem cells (GSCs)—a subpopulation endowed with self-renewal capacity and notorious for therapeutic resistance—express distinct surface markers, including CD133 and CD44, that evade singular targeting strategies. These GSCs modulate antigen presentation pathways and contribute actively to immune suppression, thus escaping monolithic CAR-T approaches designed against singular antigens.

Beyond cellular heterogeneity, the GBM microenvironment compounds therapeutic challenges by erecting an immunologically suppressive fortress. The blood-brain barrier and blood-tumor barrier impose a stringent physical blockade, severely limiting the infiltration of systemically delivered CAR-T cells into the parenchymal tumor bed. Immunologically, the TME is enriched with tumor-associated macrophages, or glioma-associated macrophages (GAMs), that secrete immunosuppressive cytokines like IL-10 and TGF-β. Regulatory T cells (Tregs) further dampen effector immune responses. Engagement of immune checkpoints such as PD-1/PD-L1 within the niche induces functional exhaustion of T cells, including CAR-Ts, crippling their cytotoxic potential. The accumulation of metabolic byproducts like lactate and adenosine adds another layer of immunosuppressive milieu, impeding CAR-T cell metabolism and function.

In light of these formidable barriers, the CAR-T field is experiencing a paradigm shift from monofocal direct tumor lysis to multidimensional immune reprogramming approaches. Scientists are engineering sophisticated next-generation CAR constructs incorporating multi-target recognition to preempt antigen escape. Tandem CARs simultaneously targeting two antigens such as CD44 and CD133 have demonstrated enhanced tumor coverage. More innovative logic-gated CARs embed computational circuits that can enforce AND, OR, or IF-THEN gating. For example, SynNotch receptors enable conditional CAR expression only upon dual antigen engagement, refining specificity, and sparing healthy off-target tissues.

Complementing targeting complexity, switchable and universal CAR platforms (UniCARs) offer unprecedented flexibility. These modular systems decouple antigen recognition from T-cell activation by utilizing soluble adaptor molecules that bind distinct antigens and channel the CAR-T response. This decoupling permits clinicians to dynamically retarget the same engineered T cells against emerging tumor antigen profiles, critical in a disease fraught with clonal evolution such as GBM.

Another promising avenue is the exploitation of innate immune receptors in CAR designs. NKG2D-based CAR-T cells recognize stress-induced ligand families like MICA and MICB, commonly upregulated on malignant cells, providing a broad-spectrum detection mechanism. This inherent promiscuity targets heterogeneous tumor populations more effectively than single antigen-targeted CARs, thus circumventing the challenge of antigenic loss variants.

However, engineering CAR-T specificity is only part of the solution. Tackling the GBM immunosuppressive environment requires synergy with adjunctive therapies. Immune checkpoint blockade with PD-1/PD-L1 inhibitors can rejuvenate CAR-T populations compromised by exhaustion pathways. Genetic ablation of PD-1 in CAR-T cells or expression of dominant-negative PD-1 receptors have enhanced efficacy in preclinical GBM models, reawakening CAR-T cytotoxic potential.

Additionally, cytokine armoring endows CAR-T cells with self-sustaining proliferative and survival signals amid the hostile TME. Engineering CAR-T cells to secrete or respond to cytokines such as IL-7, IL-15, and IL-21 promotes their expansion and persistence, fostering a stem-like memory phenotype that is crucial for durable anti-tumor immunity. This bioengineering approach counters the nutrient deprivation and hypoxia that typically limit T-cell fitness within GBM.

Traditional modalities like temozolomide chemotherapy and radiotherapy are being strategically integrated with CAR-T therapy to potentiate immune responses. Temozolomide, beyond DNA alkylation, can deplete regulatory T cells, thereby alleviating immunosuppression and enhancing CAR-T efficacy. Radiotherapy induces immunogenic cell death, releasing tumor-associated antigens and damage-associated molecular patterns that promote dendritic cell activation and facilitate CAR-T priming, contributing to durable anti-tumor responses.

Given the formidable blood-brain barriers, novel routes of CAR-T delivery are under active exploration to optimize therapeutic concentrations at the tumor site while reducing systemic toxicity. Intratumoral and intraventricular administration bypass these barriers, ensuring direct CAR-T cell presence within the TME. These localized approaches have demonstrated promising safety profiles and enhanced anti-tumor activity in early-phase clinical investigations.

The sophistication of these efforts is underpinned by cutting-edge preclinical modeling techniques. Patient-derived glioblastoma organoids (GBOs) faithfully recapitulate the tumor’s heterogeneous cellular architecture and immunosuppressive stroma, providing a high-fidelity platform to screen CAR-T cell efficacy and optimize therapeutic design. The advent of single-cell RNA sequencing further refines this process by uncovering complex antigen co-expression patterns, exhaustion states, and cellular interactions within the TME, guiding precision engineering of multi-target CAR constructs and combination strategies.

Ultimately, the future of CAR-T therapy in glioblastoma hinges upon a holistic and integrated approach. Rather than solely focusing on direct tumor cell eradication, the next generation of immunotherapies aspires to reprogram and remodel the GBM microenvironment itself. Their goal is to convert an immunologically “cold” and formidable fortress into a vulnerable target accessible to durable immune attack. By harnessing combinatorial antigen targeting, checkpoint inhibition, cytokine support, advanced delivery mechanisms, and sophisticated preclinical tools, researchers aim to unlock the long-elusive potential of CAR-T therapy against this devastating brain tumor.

The convergence of molecular engineering, immunology, and translational science heralds a new frontier in neuro-oncology. Precision, adaptability, and multifaceted immune engagement promise to reshape the treatment landscape for GBM and fulfill the hope for durable remissions in a disease long deemed incurable. With sustained innovation and clinical integration, this emerging paradigm could ultimately break the barrier between promise and cure in glioblastoma treatment.

Subject of Research: Glioblastoma and CAR-T cell immunotherapy

Article Title: Dual Challenges and Innovative Strategies in Chimeric Antigen Receptor T-cell Therapy for Glioblastoma

News Publication Date: 30-Jul-2025

Web References:
https://www.xiahepublishing.com/journal/oncoladv
http://dx.doi.org/10.14218/OnA.2025.00014

Keywords: Glioblastoma, CAR-T therapy, antigen heterogeneity, tumor microenvironment, immune suppression, glioma stem cells, immune checkpoint blockade, cytokine armoring, blood-brain barrier, patient-derived organoids, single-cell RNA sequencing, next-generation CAR engineering