Research Highlights

CAR Treg Cells Mediate Linked Suppression and Infectious Tolerance in Islet Transplantation in Mice Wardell CM, Fung VCW, Chen E, et al. Sci Transl Med. 2025;17(812):eadp6519. doi:10.1126/scitranslmed.adp6519 Since the discovery of regulatory T cells (Treg)—defined by the expression of CD251 and the transcription factor forkhead box (FOXP3)2,3—this leukocyte subset has been exemplified as indispensable for preventing deleterious immune responses4 and was most recently recognized by the 2025 Nobel Prize in Physiology or Medicine. Leveraging the suppressive potential of Treg for therapeutic purposes—including allograft rejection—has been a focus in the transplant field, culminating in their use in the ONE study.5 Recent advances in T-cell biology,6 including the development of chimeric antigen receptor (CAR)-engineered Treg, can provide unprecedented targeting to prespecified antigens and tissues. Treg possess several properties, including infectious tolerance,7 which enables them to moderate effector T-cell activity or phenotypically convert into Treg to reinforce a tolerant state. HLA A2 is a common allele expressed across multiple ethnic groups, but it is rarely the only antigenic mismatch in a transplanted organ. Wardell et al8 tested the concept that A2-CAR Treg could induce tolerance to antigens, applying this to an islet transplant model. Treg were generated from the spleen and lymph nodes of nonobese diabetic (NOD) mice. Retroviruses encoding either Thy1.1 or both A2-CAR and Thy1.1 were used for transduction to generate polyclonal or antigen-specific A2-CAR Treg. After coculturing with HLA-A2– or HLA-A2+ NOD islets for 3 d, only antigen-specific A2-CAR Treg increased expression of CD25 and production of transforming growth factor beta in response to HLA-A2+ NOD islets. A2-CAR Treg also proved effective in vitro against antigen-presenting cells, a key mechanism linked to bystander suppression. A2-CAR Treg were cocultured with HLA-A2– or HLA-A2+ NOD splenic CD11c+ dendritic cells (DCs). Expressions of costimulatory molecules (CD80 and CD86) and the NOD major histocompatibility complex class II molecule I-Ag7 were significantly reduced on HLA-A2+ DCs relative to HLA-A2– cells (Figure 1A). Polyclonal Treg demonstrated similar effects on DCs regardless of HLA-A2 expression. For further validation, A2-CAR Treg were tested against diabetogenic BDC2.5 CD4+ (effector) T cells (which are diabetogenic by recognizing a portion of the pro-insulin C-peptide). HLA-A2+ NOD DCs were cocultured with polyclonal or A2-CAR Treg, BDC2.5 T cells, and the cognate peptide. Proliferation was suppressed by A2-CAR Treg in a dose-dependent manner, significantly more so than by polyclonal Treg. These data indicate that A2-CAR Treg could tolerize HLA-A2+ DCs.FIGURE 1.: Mechanisms of action of A2-CAR Treg. A, In vitro suppression of antigen-presenting cell (APC) activation. B, Protection against hyperglycemia in NOD scid mice receiving an HLA-A2 islet transplant. C, Evidence of linked suppression. ACE, anterior chamber of the eye; CAR, chimeric antigen receptor; DC, dendritic cell; FOXP3, forkhead box P3; NOD, nonobese diabetic; scid, severe combined immunodeficiency; Teff, effector T cell; Treg, regulatory T cell. Created with Biorender.The next step was to test the efficacy of A2-CAR Treg toward autoreactive effector T cells. HLA-A2+ NOD islets were transplanted into the anterior chamber of the eye in NOD-severe combined immunodeficiency-gamma (NSG) mice, with engraftment occurring in 2 weeks. Diabetogenic BDC2.5 T cells were injected and hyperglycemia developed within 7 days. Protection against hyperglycemia was conferred to 33% of mice coinjected with (3× as many) polyclonal Treg, compared with 82% of mice coinjected with the same number of A2-CAR Treg (Figure 1B). Furthermore, A2-CAR Treg engrafted at higher numbers (with increased Ki67 expression), demonstrating lower systemic interferon-gamma and circulating BDC2.5 T cells compared with untreated mice. An increase in FOXP3+ BDC2.5 T cells was observed only in A2-CAR Treg-treated mice, suggesting the emergence of a tolerogenic balance between effector and Treg. Evaluation of transplanted eyes and endogenous pancreas (2 endogenous sources of insulin) demonstrated tracking of A2-CAR Treg to the graft, with lower BDC2.5 T-cell engraftment. Decreased effector T-cell infiltration was seen in the pancreas (lacking HLA-A2), with expansion of FOXP3+ BDC2.5 Treg. Enucleation (and removal of the HLA-A2 graft) did not change euglycemic status, with stable numbers of circulating FOXP3-expressing A2-CAR Treg, indicative of linked suppression. Finally, to explore whether infectious tolerance was induced, diphtheria toxin–receptor A2-CAR Treg were generated and injected into immunodeficiency-gamma mice with engrafted HLA-A2+ NOD.Rag1−/− islets (no passenger T/B cells). Enucleation and diphtheria toxin–based elimination of A2-CAR Treg did not result in loss of glycemic control. BDC2.5 Treg remained expanded and the ratio of BDC2.5 T effector to the total number of FOXP3+ cells in the blood was stable. These data support the notion of developed infectious tolerance. The development of Treg holds promise to redefine transplant immunotherapy and illuminate new pathways for tolerance induction. This mechanistic study provides an interesting insight into the in vivo function of Treg in the context of infectious tolerance and linked suppression. Comprehensive Human Proteome Profiles Across a 50-y Lifespan Reveal Aging Trajectories and Signatures Ding Y, Zuo Y, Zhang B, et al. Cell. 2025;188(20):5763–5784.e26. doi:10.1016/j.cell.2025.06.047 Aging is inherently associated with a decline in organ function, and this phenomenon is related to disrupted proteostasis—the proteomic network that regulates protein synthesis, modification, transport, and degradation to maintain cellular function. Temporal trends in transplantation demonstrate that the most common age range for kidney donation is 50–64 y in Australia,1 United States,2 and United Kingdom.3 The impact of increasing donor age on allograft outcomes has been well documented for decades for liver,4,5 kidney,6 pancreas,7 heart,8 and lung.9 Although this encapsulates a rather reductionist view of factors that determine graft (and patient) survival—particularly as recipient age has also increased—our molecular understanding of aging is rather limited to amplified senescence,10 impaired mitochondrial integrity,11 and altered antigen-presenting cell function,12 remaining relatively unexplored in human tissue. Ding et al13 addressed the lack of data by analyzing aging trajectories of different organs, generating an encyclopedic catalogue of the proteome across sex and age (14–68 y). The team identified >12 000 proteins, profiling them across cellular location and functional specificity and pairing these data with transcriptomics (Figure 1). This showed limited correlation between transcription and translation, which declined with age, and decoupling likely impacts aging tissue function and maintenance. Gene ontology analysis demonstrated enrichment in complement activation, inflammatory responses, and cell adhesion, but downregulation of mRNA processing and protein-RNA complex assembly. Both proteomic analysis and immunohistochemistry demonstrated accumulation of serum amyloid proteins and concurrent activation of immune and complement pathways as part of an “aging protein community.” These were variably upregulated across organs, although sex-based proteomic differences were less with age, suggesting conservation of an aging signature. Downregulated differentially expressed proteins included cellular transport processes and protein maturation, mitochondrial biogenesis, and concurrent epigenetic instability (including regulation of DNA methylation and chromatin remodeling), which mirrors the physiological decline in homeostatic properties.FIGURE 1.: The human proteome across the lifespan. A, A comprehensive aging atlas generated from proteomic and transcriptomic data across age, sex, and organs. B, Differentially expressed proteins. GAS6, growth arrest-specific 6. Created with Biorender.Approximately 40% of differentially expressed proteins were shared across all tissues. This precipitated the development of an in silico “hallmark assessment system” of aging, characterized by loss of heterochromatin and DNA repair, cell-cycle arrest, and inflammation. This was supported by tissue-based analysis of aging tissues using established biomarkers, including elevated p21Cip1 expression, reduced heterochromatin mark H3K9me3, increased proinflammatory cytokine interleukin-6, and tissue fibrosis indicated by collagen IV. The authors also developed a novel method for calculating biological age across tissues, devising a proteomic age clock that effectively predicts age regardless of heterogeneity. Metalloproteinase inhibitor 3 was identified as a contributor to the tissue clock across nine distinct organs. The same technique was used to predict transition points across the aging spectrum, with heightened humoral immune response and deregulated RNA splicing proteins most prominent around 50 years. Aorta, spleen, and adrenal glands demonstrated the most marked proteomic changes—even at age 30 years—and intensification across all organs that occurred 15–25 years after that. The aorta showed the greatest sensitivity to aging. An age-related tissue-based secretome was also generated, consistent with the senescence-associated secretory phenotype associated with aging and pathophysiological organ function. The inflammatory chemokine stromal cell-derived factor-1 (CXCL12) was upregulated in most organs. Mapping of receptor-ligand interactions across the secretome revealed interorgan communication, primarily driven by the aorta, secondary lymphoid organs (spleen, lymph nodes), and endocrine organs (adrenals). These effects appeared to be driven by chemokines and Ephrin family members involved in regulating cell migration, tissue boundary formation, and vascular development. These data support the notion that arteries are a key focus of systemic aging—both in terms of sensitivity to age-related proteomic changes and amplification of aging signals. This was highlighted by the senolytic cytokine (senokine) growth arrest-specific 6, derived from and secreted by the aorta. In vitro, growth arrest-specific 6 produced a senescence-associated secretory phenotype reaction in human vascular smooth muscle cells, and administration of only 4 doses of the recombinant protein to 8-mo-old mice reduced grip strength, decreased wire-hang endurance, and compromised balance and coordination. This was accompanied by elevated vascular p21cip expression, inflammation, and oxidative stress. The accelerated aging seen at around 50 years confirms previous noninvasive clocks,14,15 and focuses on the vascular bed as the primary organ that can further accelerate systemic decline. It complements established aging biomarkers (eg, telomeres) and recent plasma-based analyses that correlate with physical function15 and mortality.16 How can this knowledge be leveraged for transplantation? Proteostasis influences immunogenicity, regenerative potential, and susceptibility to ischemia/reperfusion injury. Understanding aging offers opportunities for organ donor risk stratification, prediction of donor organ viability, and geroprotective intervention—pharmacological or cell-based preconditioning therapies that block/reverse aging.