Scientific Programme

Background: Apolipoprotein E (apoE) is a lipid transporter in the central nervous system (CNS), and is the strongest genetic factor for late-onset Alzheimer’s disease (AD). Although apoE is produced by various cell types in the CNS, including astrocytes, microglia and pericytes, it is unclear whether apoE lipoproteins are a homogeneous population or if their molecular composition varies depending on their cellular origin. There is emerging evidence to suggest that cell-type-specific apoE may contribute differently to AD-related processes, including amyloid deposition, neuroinflammation, synaptic maintenance, and blood–brain barrier function. We hypothesized that apoE lipoproteins possess distinct lipid and protein signatures determined by their cellular origin, which are reflected in human cerebrospinal fluid (CSF).

Methods: Human induced pluripotent stem cells (iPSCs) were differentiated into astrocytes, microglia and pericytes. After validating cell identity, apoE-containing lipoproteins secreted into media were isolated using apoE immunoprecipitation and characterized using lipidomics and proteomics using LC-MS/MS. ApoE particles immunoprecipitated from human CSF were used as in vivo reference. 

Results: ApoE particles secreted by astrocytes, microglia and pericytes exhibited distinct lipidomic and proteomic profiles, demonstrating that apoE lipoproteins are highly cell-type specific. Astrocyte-derived particles were enriched in sphingomyelins and proteins such as clusterin (CLU), whereas pericyte-derived particles showed enrichment of ether-linked phospholipids and proteins associated with antioxidant activity. In contrast, the lipidation of microglial apoE particles was markedly lower, but had higher levels of proteins involved in lipid metabolism and amyloid processing. Notably, one-third of molecules detected in cell-secreted apoE were unique, highlighting the substantial molecular specialisation that exists between cell types. The majority of the lipids and proteins identified in cell-derived apoE particles were also detected in CSF-apoE particles.

Significance:These findings establish that CNS apoE lipoproteins are not homogeneous but comprise cell-type-specific particles with unique lipid and protein signatures. Detecting these molecular signatures in human CSF demonstrates their physiological relevance. This work provides a framework for investigating how cell-specific apoE populations contribute to CNS function and AD pathogenesis.

Background and objective: Treatment monitoring in congenital adrenal hyperplasia (CAH) remains challenging as glucocorticoid dosing must be balanced against the clinical and biochemical phenotype of over or under treatment. CAH due to 21-hydroxylase deficiency (21OHD), is routinely diagnosed through neonatal screening with conventional 17a-hydroxyprogesterone (17OHP) measured as the single steroid biomarker. Additional adrenal-derived steroids have shown promise for diagnosis, including 21-deoxycortisol (21dF), 21-deoxycortisone (21dE) and 11-ketotestosterone (11KT), representing metabolites produced through alternative glucocorticoid and androgen pathways. Notably, 11KT, an 11-oxy androgen, also hints at a potential biomarker in treatment monitoring. These unconventional steroids could therefore serve as reliable predictors of treatment outcome, warranting their assessment in disease tracking.

Methods: Serum steroid profiling of 137 visits (71 patients with genetically confirmed 21OHD; age IQR 6-15 years; 44% female) were analyzed using liquid chromatography high-resolution mass spectrometry as part of a prospective observational cohort study. Visits were divided into optimal treatment (n=85) and undertreatment (n=52) based on standardized clinical and biochemical assessment. Odd ratios were calculated using mixed-effects logistic regression (adjusted for age, sex, daily hydrocortisone-equivalent dose, CAH subtype, BMI SDS and pubertal status), and discriminative metabolites were identified using sparse partial least squares-discriminative analysis (sPLS-DA).

Results: The strongest associations (p = 0.01) were observed for 11-oxy androgens, 21dF, 21dE and 11-deoxycortisol, where higher concentrations increased the odds of being undertreated versus optimally treated. The sPLS-DA, including 25 steroids, showed good discrimination between treatment groups with an AUROC of 0.79. In component 1, 17OHP, progesterone, 21dE, 16a-hydroxyprogesterone, 21dF, 11-deoxycorticosterone and corticosterone classified undertreatment; whereas in component 2, dehydroepiandrosterone sulfate classified optimal treatment. Including clinical variables resulted in a marginal increase in classification (AUROC 0.82).

Significance: These results underscore that serum steroid profiling can be used to predict treatment outcome, even without considering additional clinical variables and support the inclusion of unconventional steroids in the biochemical work-up of 21OHD to improve treatment monitoring.

Background: Inherited disorders of energy metabolism, particularly mitochondrial diseases, remain challenging to diagnose because clinical, metabolic or genetic findings are not always conclusive. Functional investigations often require invasive and labor-intensive procedures such as muscle or skin biopsies, limiting rapid clinical decision-making.

Objective: We therefore explored whether stable isotope tracing in peripheral blood mononuclear cells (PBMCs) obtained from routine EDTA blood samples could provide a minimally invasive alternative to characterize metabolic dysfunction.

Methods: Isolated PBMCs were exposed to mitochondrial inhibitors, including rotenone to model complex I dysfunction, in combination with ¹³C-labeled substrates to assess metabolic fluxes. Isotopologue distributions of metabolites such as pyruvate, lactate, and citrate were quantified by liquid chromatography-mass spectrometry to target glycolysis, and the oxidative and reductive carboxylation within the tricarboxylic acid cycle.

Results: Inhibition of complex I with rotenone, using [U-¹³C]-glutamine as tracer, increased the M+5/M+4 citrate ratio indicating enhanced reductive carboxylation. In parallel, tracing with [1-¹³C]-glucose showed an increased M+1/M lactate ratio, consistent with elevated glycolytic flux.

Relevance: This non-invasive ¹³C-isotope tracing approach has the potential to identify characteristic signatures of metabolic alterations in inherited disorders of energy metabolism. It may improve diagnostic precision and support development of targeted therapeutic strategies.

Background: Quantification of free platinum (Pt) is critical for therapeutic drug monitoring (TDM) of platinum-based chemotherapeutics, as this fraction is closely associated with pharmacological activity and toxicity. However, it is subject to time-dependent redistribution and protein binding, making preanalytical conditions a key determinant of result reliability.

Objective: This study aimed to evaluate the temporal stability of free platinum in plasma and to define preanalytical conditions for accurate and reproducible measurement.

Methods: A blank pooled human plasma matrix was spiked with a defined concentration of platinum standard to simulate circulating Pt. Samples were stored under controlled conditions on the bench at room temperature (RT, approx. 22°C), at +4°C, and -20°C, and analyzed at multiple time points (0, 1, 2, 3, 5, 24, 48, 72, and 144 h). At each time point, the free platinum fraction was isolated using a validated ethanol precipitation protocol and all naturally occurring stable isotopes quantified by inductively coupled plasma mass spectrometry (ICP-MS). Changes in measured free platinum concentrations over time were assessed to evaluate redistribution kinetics and stability.

Results: A time- and temperature-dependent decline in measurable free platinum was observed during the first 48 h, consistent with progressive binding to plasma proteins or adsorption processes. This decrease was slower at +4°C and almost absent at -20°C. For RT, the nadir was followed by a gradual, virtually linear increase until original levels were reached at 144 h. Variability between the differently stored samples increased with storage time, indicating reduced reliability of delayed processing if stored at RT or +4° C before. The observed changes were consistent across Pt isotopes, except for the lowly abundant ¹⁹⁰Pt and ¹⁹²Pt, which displayed higher variability due to lower signal, probably masking the kinetics.

Conclusion: These findings demonstrate that the free platinum fraction is highly sensitive to preanalytical storage duration. Handling conditions critically influence the accurate determination of free platinum in plasma. Immediate sample processing (within 3 h of collection), strict control and documentation of the interval between blood collection and preparation are essential for valid results. Standardization of sample handling protocols is therefore mandatory for reliable implementation of free platinum measurements in clinical TDM workflows.