Elevation of CSF Sortilin Following Subarachnoid Hemorrhage in Patients and Experimental Model Rats.

Subarachnoid hemorrhage (SAH) can cause acute neuronal injury and chronic neurocognitive deficits; biomarkers reflecting its associated neuronal injury are of potential prognostic value. Sortilin, a member of the vacuolar protein sorting 10p (Vps10p) family, is enriched in neurons and is likely involved in neurodegenerative diseases. Here, we explored sortilin in the cerebrospinal fluid (CSF) as a potential biomarker for early neuronal injury after SAH. Sortilin levels in the CSF of SAH patients (n = 11) and controls (n = 6) were analyzed by immunoblot. SAH rats surviving 3-72 h (h) were evaluated neurologically, with their brain and CSF samples examined histologically and biochemically. Sortilin protein ~100 kDa was detected in the CSF from SAH patients only, with its levels correlated to Hunt-Hess scale. Rats in the SAH groups showed poorer Garcia score and beam balancing capability than sham controls. Sortilin ~100 kDa was detectable in the CSF of the SAH, but not sham, animals. Levels of sortilin ~100 kDa and fragments ~40 kDa in cortical lysates were elevated in the SAH relative to control rats. Levels of cortical glial fibrillary acidic protein (GFAP) were also elevated in the SAH rats. In immunohistochemistry, the pattern of sortilin labeling in the brain was largely comparable between the SAH and control rats, whereas an increased astrocytic GFAP immunolabeling was evident in the former. Together, these results suggest that SAH can cause an early and remarkable rise of sortilin products in CSF, likely reflecting neuronal change. Sortilin could be further explored as a potential biomarker in some brain disorders.

Soluble sortilin is present in excess and positively correlates with progranulin in CSF of aging individuals.

Mutations in progranulin are a major cause of frontotemporal lobe degeneration (FTLD). Hence, plasma progranulin is an attractive biomarker in FTLD but poorly reflects levels in cerebrospinal fluid (CSF), suggesting tissue-specific regulation of progranulin levels. Sortilin was recently identified as a progranulin scavenger receptor that destines it for lysosomal degradation. Proteolysis or alternative splicing generates soluble sortilin variants that retain progranulin binding and potentially functions as a decoy receptor. In the present study, we analyzed soluble sortilin and progranulin in plasma and CSF in 341 aging individuals. We found that soluble sortilin exists in CSF in ten-fold molar excess compared to progranulin and observed a highly significant positive correlation between soluble sortilin and progranulin levels in CSF but not in plasma. However, carriers of the minor allele of SNP rs646776 in SORT1 encoding sortilin displayed significantly increased soluble sortilin and reduced progranulin specifically in plasma but not in CSF. Taken together, our findings suggest that soluble sortilin may affect progranulin levels in both a tissue-specific and genotype-dependent manner.

Amyloid cascade and tau pathology cerebrospinal fluid markers in mild cognitive impairment with regards to Alzheimer's disease cerebral metabolic signature.

BACKGROUND: Biomarker relationships in early stages of Alzheimer's disease (AD) are elusive. Cerebrospinal fluid (CSF) levels of amyloid-ß 1-42 (Aß42) and total tau (tTau) as well as 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) contribute to help unravel AD pathology. Furthermore, peptides related to amyloid-ß protein precursor (AßPP) processing [e.g., soluble AßPPα and ß (sAßPPα and sAßPPß, respectively); sortilin-related receptor with A-type repeats (SORL1, also called LR11 or SORLA)] are factors crucially implicated in the formation of pathological hallmarks of AD. OBJECTIVE: To unveil differences in CSF concentrations of Aß42, sAßPPα, sAßPPß, tTau, and SORL1 between patients with mild cognitive impairment (MCI) who were categorized according to expert interpretation of FDG scans. METHODS: PET results were classified as suggesting high likelihood for AD (MCI-AD high), intermediate likelihood for AD (MCI-AD intermediate), or little likelihood for AD (MCI-AD unlikely). An AD dementia group was also included. Differences between the groups were tested by Kruskal- Wallis test, Mann-Whitney test, or χ2. Provided statistically significant differences were detected, multiple linear regression models were employed. RESULTS: Aß42 levels in patients with MCI-AD high (n = 15) were lower compared to MCI-AD intermediate (n = 18) and MCI-AD unlikely patients (n = 25) (p = 0.002), while they did not differ from patients with AD dementia (n = 17). The regression model revealed a significant impact of the metabolic pattern on Aß42 concentrations. SORL1, tTau, sAßPPα, and sAßPPß concentrations did not differ between the groups. CONCLUSION: These findings point to linkages between plaque pathology and glucose cerebral hypometabolism.