The creatine-phosphocreatine cycle serves as a crucial temporary energy buffering system in the brain, regulated by brain creatine kinase (CKB). CKB, responsible for reversible phosphorylation of creatine, plays a pivotal role in maintaining ATP levels. Alzheimer's disease (AD) has been linked to increased CKB oxidation and loss of its functions, although specific pathological processes and affected cell types remain unclear. In our study, cerebral cortex samples from AD, dementia with Lewy bodies (DLB), and age-matched controls were analyzed using antibody-based methods to quantify CKB levels and assess alterations associated with disease processes. Two independently validated antibodies exclusively labeled astrocytes in the human cerebral cortex. Immunofluorescence and Western blot analyses demonstrated a loss of CKB immunoreactivity correlated with increased plaque load, severity of tau pathology, and Lewy body pathology. Combining antibody-based and mass spectrometry (MS) approaches, we explored CKB availability in AD and DLB. Transcriptomics and targeted MS revealed unaltered total CKB levels. Reduced antibody binding levels, confirmed by Western blot under denatured conditions, suggested post-translational modifications that affect antibody binding. This aligns with altered efficiency at proteolytic cleavage sites indicated in the targeted MS experiment. These findings highlight that post-translational modifications extend to astrocytes, affecting their functions. CKB and the creatine-phosphocreatine cycle play crucial roles in energy buffering, securing constant ATP availability. Reduced ATP in astrocytes can disrupt ATP-dependent processes, such as the glutamate-glutamine cycle. Post-translational modifications in CKB and astrocyte dysfunction may disturb homeostasis, driving excitotoxicity in the AD brain. CKB and its activity emerge as potential biomarkers for monitoring early-stage energy deficits in AD.

Frozen brain tissue samples were homogenized with 3 mm steel beads using TissueLyser (Qiagen) in 7M Urea, 2M Thiourea and 2% sodium deoxycholate (SDC). Samples were centrifuged at 20,000×g for 30 min at 4°C and supernatants collected and mixed 1:4 with cold acetone and incubated over night at -20°C. Samples were centrifuged at 20,000×g for 30 min at 4°C and supernatants discarded. Pellets were washed two times with cold acetone and centrifuged at 20,000×g for 10 min at 4°C. Acetone excess was removed, and pellets were dried and dissolved in 200 µl 7M Urea, 2M Thiourea. Protein extract was diluted 8-times with 100mM tetraethylammonium bromide (TEAB) and protein concentrations measured using Bradford Protein Concentration Measurement (BioRad). Volumes corresponding to 100 µg of proteins were taken and mixed with a multiplex pool of heavy labelled protein standards - SIS PrESTs27. The mixtures were reduced in 10mM ditiotreitol (DTT) and incubated for 1 hour at 30°C at 500 rpm. Tubes were cooled down and 2-chloroacetamide (CAA) was added to a final concentration of 50 mM and incubated for 30 min in dark at ambient temperature. Pierce trypsin (Thermo Fischer Scientific) was added in 1:50 (enzyme: substrate) ratio and incubated for 16 h at 37°C at 500 rpm. Trifluoroacetic acid (TFA) was added to the final concentration of 0.5% to quench digestion and one fourth of the volume was taken to reverse-phase desalting performed as previously described by Rappsilber28. Purified samples were subjected to LC-MS/MS analysis using an online system of Ultimate 3000 (Thermo Fisher Scientific) LC connected to QExactive HF (Thermo Fisher Scientific) MS operating in parallel reaction monitoring (PRM) mode using an inclusion list with 3-minute isolation windows. Volume corresponding to 2 µg of digested protein content was loaded onto a trap column (PN 164535, Thermo Fisher Scientific) and washed for 5 min at 5 µL/min with 100% Solvent A (3% Acetonitrile (ACN), 0.1% formic acid (FA)). The peptides were separated on a 25-cm analytical column (PN ES902, Thermo Fisher Scientific) using a 90-minute gradient of 1 to 37% Solvent B (95% ACN, 0.1% FA) at 0.4 µL/min followed by a 13-min washout at 99% Solvent B and a 5-min equilibration. The MS PRM analysis was performed using a method with cycles consisting of an MS1 scan (60,000 resolution, AGC = 3e6, 300–1600 m/z, IT = 110 ms) followed by 20 PRM scans (60,000 resolution, AGC = 1e6, NCE = 27, 1.5 m/z isolation window, IT = 110 ms). Resulting raw files were imported to Skyline29 and ratios of heavy to light areas of all monitored endogenous and heavy-labelled SIS PrEST peptides extracted and processed using R (4.2.0).

Postmortem human brain tissues with well-documented clinical information and neuropathological evaluation from the superior frontal gyrus (SFG) or middle temporal gyrus (MTG) of 30 donors were obtained from the Netherlands Brain Bank (NBB). In this study, formalin-fixed and paraffin-embedded (FFPE) sections were examined to explore CKB-ir, while fresh-frozen tissue samples from the other hemisphere were used for WB and MS analysis. The SFG section of a non-demented middle-aged control (n = 1, male, aged = 56) was used to map base-line CKB distribution in the absence of plaques, tangles and Lewy bodies. We included 29 MTG samples across Alzheimer’s disease subjects (n=10, A1 to A10; Braak stage > 5, amyloid deposits stage > B; 5 women, 5 men; mean age 80.6 years), dementia with Lewy Body (n=10, B1 to B10; LB stage ≥ 5; 1 woman, 9 men; mean age 78.1 years), and non-demented age-matched controls (n=9, C1 to C9; AD Braak stage < 2, amyloid deposits stage ≤ B, LB stage < 2; 4 women, 5 men; mean age 83.1 years) for the comparison of experimental groups. The average mean postmortem delay for all these subjects was 5.56 h. All AD and DLB donors had a clinical history of progressive dementia and cognitive symptoms. Control donors were individuals without a history of dementia. The diagnosis was based on postmortem neuropathological and histochemical examination of the brains to determine the level of plaque load (amyloid deposits), tau pathology (Braak) and Lewy body (LB) stages. Additional information on the subjects is described in Table S1, including all available details on age, sex, postmortem delay, brain weight, APOE isoform type, neuropathological stage, and protein analysis methods. In the 4-stage amyloid classification, stage O: individuals exhibit no amyloid pathology; stage A: individuals experience mild pathology; stage B: individuals have moderate plaque load; stage C: individuals exhibit severe plaque load. Braak stages are determined by the distribution and severity of neurofibrillary tangle (NFT) pathology. Braak stage 0/1: - indicate NFT is confined mainly to the entorhinal region of the brain; Braak stage 6: indicate moderate to severe neocortical involvement of NFT. LB stages describe the distribution of α-synuclein in the brain. LB stage 0/1: people have no α-synuclein load; LB stage5/6: individuals experience severe α-synuclein load.