Relationships of PGRN with sTREM2 in AD continuum and non-AD pathophysiology and their reciprocal roles in modulating amyloid pathology: two population-based study

Abstract

Progranulin (PGRN) and soluble triggering receptor expressed on myeloid cells-2 (sTREM2) are emerging biomarkers of Alzheimer's disease (AD). This study explores the roles of their interplay in modulating amyloid pathology. We analyzed data from 905 participants (mean age = 62.0) in the CABLE cohort and 973 participants (mean age = 73.1) in the ADNI, classified using the A/T/N biomarker framework. One-way ANOVA was used to assess whether cerebrospinal fluid (CSF) PGRN and sTREM2 differed across biomarker profiles and clinical stages. Multiple linear regression models and linear mixed-effects models were used to test the relationships among PGRN, sTREM2, and CSF Aβ_1-42 levels. Mediation analysis was used to explore the reciprocal relationships between sTREM2 and PGRN in influencing amyloid pathology. CSF proteomic and bioinformatic analyses were finally used to investigate the underlying biological mechanisms. In both cohorts, PGRN and sTREM2 were higher in individuals within the TN+ profile irrespective of the A status, and followed similar trajectory across different clinical and biomarker stage. CSF PGRN was associated with higher sTREM2 across AD continuum and non-AD pathophysiology. Bidirectional mediation was observed between PGRN (14.6% in CABLE, 15.6% in ADNI) and sTREM2 (29.7% in CABLE, 33.5% in ADNI) in modulating Aβ pathology (p < 0.0001). Proteomic analysis identified 1539 CSF proteins (Bonferroni-corrected p < 7.13 × 10^-6) simultaneously associated with PGRN, sTREM2, and Aβ_1-42. These proteins are mainly enriched in immune processes and neural plasticity. These findings suggest that the interplay between lysosome function and microglia-related neuroinflammation plays key roles in amyloid metabolism.

Fig. 1. CSF PGRN and sTREM2 within the framework combining A/T/N biomarker and clinical profile.

Scatter plot depicting the levels of CSF PGRN A, C and sTREM2 B, D for each of the four biomarker profiles, as defined by the A/T/N framework, coupled with clinical staging, as defined by CDR (for ADNI). The T (tau pathology) and N (neurodegeneration) group were merged. In ADNI, the CDR = 1 stage group includes some biomarker profiles containing insufficient number of subjects, which precludes conducting group comparison. They are still shown for sake of completeness. Each A/T/N biomarker profile is represented in a different color: A-/TN- are depicted in purple, A+/TN- in blue, A+/TN+ in grey, and A-/TN+ in orange. Solid bars represent the mean and the standard deviation (SD). P-values were assessed by a one-way ANCOVA adjusted for age, gender, APOE ε4 status and education. Significant P values after Bonferroni corrected post hoc pairwise comparisons are marked. A Aβ pathology biomarker status, T tau pathology biomarker status, N neurodegeneration biomarker status, AD Alzheimer’s disease, HC healthy controls, CDR clinical dementia rating, CSF cerebrospinal fluid.

Fig. 2. The relationships between CSF PGRN and CSF sTREM2.

PGRN exhibited positive correlations with sTREM2 in both cohorts A. The significance remained in each A/TN group (B for CABLE and C for ADNI). A Aβ pathology biomarker status, T tau pathology biomarker status, N neurodegeneration biomarker status, CSF cerebrospinal fluid.

Fig. 3. The associations of CSF PGRN and sTREM2 with amyloid pathology.

CSF PGRN was associated with higher levels of CSF Aβ_1–42 A. CSF sTREM2 was associated with higher levels of CSF Aβ_1–42 B. PGRN could increase CSF Aβ_1–42 partially by modulating CSF sTREM2. Mediation analyses revealed that CSF PGRN could increase CSF Aβ_1–42 partially by modulating CSF sTREM2, and CSF sTREM2 could also modulate the relationship between CSF PGRN and CSF Aβ_1–42 C. The CABLE data are depicted in blue and ADNI data in red. CSF cerebrospinal fluid.

Fig. 4. Proteins and underlying pathways underpinning the associations of CSF PGRN and sTREM2 with amyloid pathology.

After Bonferroni correction (p < 7.13 × 10⁻⁶), we identified 1539 CSF proteins consistently associated with CSF PGRN, sTREM2, and Aβ_1–42. Among these, 403 proteins were negatively correlated A, while 1136 proteins were positively correlated B. The bar charts on the right display the top five enriched pathways for these proteins.