Black communities are disproportionately affected by higher levels of inflammation and vascular factors, which may contribute to a higher rate of cognitive impairment. While high blood pressure is a known risk factor for cognitive decline, the relationship between diastolic blood pressure (DBP) and inflammatory pathways, and how antihypertensive medications modify this relationship, is understudied in older Black adults. This study included 305 Non-Hispanic Black adults (Mean Years of Age = 73.24 ± 6.06) who were from the Rush Memory and Aging Project and Minority Aging Research Study. Linear regression models were used to test the association of DBP with markers of chronic and acute inflammation measured with blood serum. We then re-ran the model including an interaction between antihypertensive medication and DBP. Covariates included years of age, sex/gender, and diabetes (presence vs. absence). Higher DBP was associated with lower chronic inflammation (β= -0.0131, SE=0.0041, p=0.002). Higher mean arterial pressure (MAP) was also associated with lower chronic inflammation (β= -0.0097, SE=0.0037, p=0.009). Antihypertensive medication was associated with higher acute inflammation (β=0.2574, SE=0.1125, p=0.023). These findings underscore the complexity of the relationship between vascular factors and inflammation in older Black adults. Future studies are needed to explore mechanisms by which lower DBP and MAP may affect chronic inflammation and brain health.

Alzheimer’s disease (AD) involves broad cellular and metabolic changes across the brain, including dysregulation of lipid pathways. Lipoprotein lipase (LPL) is a key enzyme in lipid metabolism and has been linked to AD risk and AD-relevant microglial biology. In peripheral tissues, GPIHBP1 is a capillary endothelial cell protein that transports LPL to the capillary lumen to enable triglyceride processing. Motivated by the connection between lipid metabolism and AD, this ongoing project tests whether GPIHBP1 expression is upregulated in AD brain, with a specific focus on oligodendrocytes. Using publicly available single-cell and single-nucleus RNA-seq datasets from GEO, we analyzed: one scRNA-seq AD vs. control dataset from hippocampal cortex (GSE175814), one snRNA-seq dataset spanning AD pathology stages (GSE147528; caudal entorhinal cortex), and one snRNA-seq AD vs control dataset (GSE138852; entorhinal cortex). In GSE175814, oligodendrocytes show higher normalized GPIHBP1 expression in AD than control, and a larger fraction of oligodendrocytes are GPIHBP1-positive in AD. In GSE138852, containing 6 AD and 6 control donors, oligodendrocytes again showed higher GPIHBP1 signal in AD than control. Across three independent datasets this signal is most consistent in oligodendrocytes, strengthening the case that GPIHBP1 upregulation may represent a reproducible AD-associated feature rather than a dataset-specific finding. This observation, to our knowledge, has not been previously highlighted in the AD single-cell literature.

Mesenchymal stem cell (MSC) based therapies represent a promising strategy for treating degenerative and inflammatory conditions through their immunomodulatory properties and multilineage potential. However, the large scale expansion required for clinical efficacy induces cellular senescence, severely impairing therapeutic potency. While senescence is driven by stressors such as DNA damage, oxidative stress, and oncogene activation, reversing this state remains challenging, as redundant signaling networks can compensate when a single pathway is inhibited. To address this complexity, we hypothesize that rejuvenating MSCs requires a synergistic multidrug approach. Yet identifying effective combinations is hindered by the vast combinatorial space and limited traditional screening throughput. Here, we present an integrated discovery platform coupling mechanistic language modeling with high throughput pharmacological assays. We use large language models to map MSC senescent pathway crosstalk and prioritize candidate drug combinations through in silico screening, then experimentally validate rejuvenation efficacy. Using this approach, we identified 15 novel drug combinations with an 80% hit rate and observed cell yield increases of up to 3.15 times relative to senescent controls. By integrating LLM guided prioritization with automated validation, this framework accelerates lead identification, reveals senescent signaling architecture, and supports future prediction of higher order three drug therapies for regenerative medicine.

Kikuchi Fujimoto Disease (KFD) is a highly uncommon, benign, and self-limiting lymphadenopathy. Currently, the originations of KFD are unclear, and it remains to be seen how emerging technology that probes the disease on a genetic level can reveal its origin. Because of its rarity, the disease carries a high misdiagnosis rate. A case study of a 20-year-old female illustrates the potential for near misdiagnosis and highlights the need for greater standardization in the treatment of lymphadenopathy as well as continued clinical education to prevent similar occurrences.

A new emerging treatment, called chimeric antigen receptor (CAR) T-cell immunotherapy, is under investigation for its potential therapeutic effects against osteosarcoma. However, this treatment is highly patient-specific and expensive, ranging up to $1 million per infusion. There is an urgent need to quantify the dynamics of T-cell kinetics to accurately predict an effective dosing schedule to minimize costs and maximize treatment efficacy for patients. CAR T-cell and tumor dynamics were modeled through a predator-prey model over time, examining various dosages and frequencies in a PK-PD model. Our findings suggest that the optimal concentration for efficacy in the cellular populations with regard to cost is significantly lower than the previously standard average value, which provides quantitative values for the next steps in designing cancer treatment.