Research Highlights

Immune System Adaptation During Gender-affirming Testosterone Treatment Lakshmikarth T, Consiglio C, Sardh F, et al. Nature. 2024;633:155 Individuals who are genetically women and have successfully undergone testosterone therapy to transition to transgender men may respond differently to environmental infections and/or inflammatory stimuli compared with genetic women not undergoing testosterone therapy (cisgender women).1 In the current study,2 Lakshmikanth et al longitudinally followed 23 transgender men (genetically women who underwent gender-affirming testosterone therapy between the age of 18–27 y) and sampled their peripheral blood at baseline, as well as 3 and 12 mo after testosterone treatment. The goal was to study changes in plasma proteins, immune cell phenotypes, and functional immune cell responses to testosterone treatment. After confirming the expected rise of bioavailable testosterone levels and the corresponding decline in estradiol levels in the serum, the authors used bulk RNA sequencing of longitudinal blood samples to reveal an overall decrease in transcripts from the type 1 interferon-α (IFN-I) response pathways, with a reciprocal upregulation of transcripts from the tumor necrosis factor (TNF) signaling pathways. This finding supports a previously unrecognized role of testosterone in calibrating IFN-I versus TNF cross-regulation. Using 50-marker mass cytometry of peripheral blood mononuclear cells, the authors determined that testosterone treatment resulted in a contraction of plasmacytoid dendritic cells (DCs), CD8+ mucosa-associated invariant T cells, and CD24+CD8+ central memory T cells. Interestingly, after testosterone treatment, plasmacytoid DCs exhibited an attenuated IFN-I response following Toll like receptor (TLR) 7/8 stimulation in vitro. This attenuation correlated with an upregulation of suppressor of cytokine signaling 3 and 1 and a downregulation of IRF7 following testosterone therapy. A similar shift in IFN-I versus TNF cross-regulation was also observed in monocytes after testosterone therapy in response to TLR7/8 or TLR4 stimulation ex vivo. In this case, an upregulation of the monocyte surface receptor signaling lymphocyte activation molecule family 7 was induced by testosterone and correlated with enhanced TNF production in monocytes. Signaling lymphocyte activation molecule family 7 has been recently described as potentiating TNF responses through an autocrine loop in myeloid cells. Importantly, these altered immune responses were largely driven by elevated testosterone levels and, to a lesser extent, by decreased estrogen levels. Epigenetically, longitudinal single-cell assay for transposes-accessible chromatin with sequencing of peripheral blood mononuclear cells from individuals undergoing testosterone therapy revealed increased transcription factor activity in the canonical nuclear factor-kappa B pathway in T cells and natural killer cells but not in B cells or myeloid/DCs, and no increased activity in the JACK-STAT, AP-1, or MAPK pathways across all immune cells. Importantly, these findings were corroborated by studies involving cisgender male and female controls, further highlighting the effect of testosterone therapy in converting a genetic female immune cell phenotype to a phenotypically male one, characterized by a reduced IFN-I/TNF response to immune stimulation. This timely study reveals clinically relevant adaptation in the immune system during testosterone therapy in the transgender male population. These adaptations directly influence the host’s response to inflammatory stimuli, including infections, as well as T-cell exhaustion during chronic antigen stimulation.3 Consequently, these findings have significant implications in transplantation, as transplant recipients are highly susceptible to opportunistic infections and subsequent inflammation, yet T-cell exhaustion may be beneficial for allograft acceptance and long-term survival. Further understanding of the hormonal control of alloimmune responses may provide opportunities for immunosuppression tailoring or even therapeutic control. Proteomic Aging Clock Predicts Mortality and Risk of Common Age-related Diseases in Diverse Populations Argentieri MA, Xiao S, Bennett D, et al. Nat Med. 2024;30:2450 Chronological age is an imperfect surrogate for biological age. Biological age is a major determinant of common chronic diseases and mortality. Therefore, an accurate measure of biological age—a molecular age “clock”—would significantly advance our ability to prognosticate individual risks for chronic diseases and premature death, as well as prophylactically target disease processes. Previous attempts have primarily focused on DNA methylation (DNAm) as an age-dependent measure of epigenetic modification.1 It is now recognized that actual protein levels may provide more direct mechanistic and functional insights into aging biology. Consequently, a “proteomic age clock” that is generalizable across large and ethnically diverse populations is highly desirable. In the current study,2 Argentieri et al developed a proteomic age clock by leveraging a variety of biobanks. They initially used the UK Biobank (UKB), consisting of 45 441 participants from the United Kingdom, assigning 70% to training and 30% to test subsets to devise the proteomic age clock (ProtAge). This model was later validated in 2 different cohorts: the China Kadoorie Biobank with 3977 participants and the Finnish biobank with 1990 participants. The initial machine learning methods were trained to predict chronological age. In this process, they identified the gradient boosting method (LightGBM) as the best machine learning approach. They identified 204 aging-related proteins (APs) to predict chronological age and demonstrated stability over time of association between these APs and age. They then calculated the “ProtAgeGap”—the difference between biological age defined by the ProtAge and chronological age—which was consistent across all 3 cohorts studied. Furthermore, they defined a 20-protein panel (ProtAge20) that achieved 95% of the performance of the 204-protein ProtAge. The 20 proteins of the ProtAge20 were ELN, COL6A3, CDCP1, PODXL2, LTBP2, SCARF2, ENG, CXCL17, LECT2, GDF15, FSHB, AGRP, ACRV1, EDA2R, PTPRR, KLK3, KLK7, GFAP, and NEFL. Key aging processes that these proteins are involved in can be categorized as cell adhesion and extracellular matrix interactions, immune response and inflammation, hormone regulation and reproduction, cell signaling, protease activity and enzymatic function, regulation of body weight and energy balance, neuronal structure and function, and development and differentiation. In validating the association between ProtAgeGap (using 204 APs) and aging-related physiological and cognitive function, the authors found that an increasing ProgAgeGap was associated with increasing frailty index, declining physical measures of well-being, and cognitive function, as well as increasing biological measures of aging. The performance of ProtAgeGap20 was similar to ProtAgeGap. Importantly, the top decile of ProtAgeGap correlated with the highest cumulative incidence rates of osteoarthritis, all-cause mortality, ischemic heart disease, type 2 diabetes, and chronic kidney disease—which are all common age-related pathologies. Similarly, ProtAgeGap was associated with the cumulative incidence of 4 cancers (esophageal, lung, non-Hodgkin lymphoma, and prostate) in Cox multivariable models. It is important to note that proteins identified in this study showed very little overlap with corresponding genes previously identified by the leading DNAm clocks or targets from previous proteomic aging studies. This was explained by a possible lack of strong correlation between age-related changes in gene expression (the functional consequence of DNAm) and age-related protein expression, as well as the significantly larger study population of the current study. With a clear demonstration of generalizability using 2 additional independent samples (China Kadoorie Biobank and Finnish biobank) with very different participant genetic backgrounds, this study provides evidence that plasma proteomics can be a powerful tool for measuring biological age and quantifying common age-related diseases. Applying this to transplantation, this tool may provide a more objective measure of recipient appropriateness and readiness in pretransplant evaluation. Additionally, measurement of ProtAgeGap under the influence of immunosuppression may have the potential to further predict posttransplant well-being, for example, in terms of the development of age-related complications such as neoplasia, and even recipient survival.