Cerebrospinal Fluid Mitochondrial DNA in Rapid and Slow Progressive Forms of Alzheimer's Disease

Abstract

Alzheimer's type dementia (AD) exhibits clinical heterogeneity, as well as differences in disease progression, as a subset of patients with a clinical diagnosis of AD progresses more rapidly (rpAD) than the typical AD of slow progression (spAD). Previous findings indicate that low cerebrospinal fluid (CSF) content of cell-free mitochondrial DNA (cf-mtDNA) precedes clinical signs of AD. We have now investigated the relationship between cf-mtDNA and other biomarkers of AD to determine whether a particular biomarker profile underlies the different rates of AD progression. We measured the content of cf-mtDNA, beta-amyloid peptide 1-42 (Aβ), total tau protein (t-tau) and phosphorylated tau (p-tau) in the CSF from a cohort of 95 subjects consisting of 49 controls with a neurologic disorder without dementia, 30 patients with a clinical diagnosis of spAD and 16 patients with rpAD. We found that 37% of controls met at least one AD biomarker criteria, while 53% and 44% of subjects with spAD and rpAD, respectively, did not fulfill the two core AD biomarker criteria: high t-tau and low Aβ in CSF. In the whole cohort, patients with spAD, but not with rpAD, showed a statistically significant 44% decrease of cf-mtDNA in CSF compared to control. When the cohort included only subjects selected by Aβ and t-tau biomarker criteria, the spAD group showed a larger decrease of cf-mtDNA (69%), whereas in the rpAD group cf-mtDNA levels remained unaltered. In the whole cohort, the CSF levels of cf-mtDNA correlated positively with Aβ and negatively with p-tau. Moreover, the ratio between cf-mtDNA and p-tau increased the sensitivity and specificity of spAD diagnosis up to 93% and 94%, respectively, in the biomarker-selected cohort. These results show that the content of cf-mtDNA in CSF correlates with the earliest pathological markers of the disease, Aβ and p-tau, but not with the marker of neuronal damage t-tau. Moreover, these findings confirm that low CSF content of cf-mtDNA is a biomarker for the early detection of AD and support the hypothesis that low cf-mtDNA, together with low Aβ and high p-tau, constitute a distinctive CSF biomarker profile that differentiates spAD from other neurological disorders.

Keywords: Alzheimer’s disease; biomarker; cerebrospinal fluid; digital PCR; mitochondrial DNA.

Figure 1

CSF content of cf-mtDNA-85 in AD disease progression. The absolute number cf-mtDNA-85 copies was measured directly in CSF without nucleic acid extraction by droplet digital PCR (dPCR) in the presence of a hydrolysis probe using the mtDNA-85 primer pair, which targets a region of the cytochrome B gene producing an amplicon of 85 base pairs length. The CSF content of cf-mtDNA-85 from all subjects of the whole cohort is shown in (A) and from subjects selected using the indicated cut-off values of the core AD biomarkers Aβ and t-tau is shown in (B). Subjects were distributed in three different groups: ND-Ctrl, control group composed of patients with neurological diseases and with no clinical signs of AD type dementia; spAD, group composed of patients with clinical signs of AD type dementia of slow progression; rpAD: group composed of patients with clinical signs of AD of rapid progression. Numbers within squares show the number of patients in each group. Dots represent individual values. Bars represent mean ± 95% CI. ** Statistically significantly different p < 0.01, * statistically significantly different p < 0.05; n.s = statistically nonsignificantly different.

Figure 2

Relationship between cf-mtDNA-85 and CSF biomarkers of AD. (A,C,E): linear regression graphs and Pearson correlation (r) between CSF content values of cf-mtDNA-85 and Aβ (A), t-tau (C) or p-tau (E). Dotted lines represent 95% CI. Values are from all subjects of the entire cohort (n = 95). Dots represent individual values. All values were transformed to natural logarithm. n.s = nonsignificant. (B,D,F): bar graphs showing the CSF cf-mtDNA-85 content from subjects of the entire cohort segregated in two groups according to the cut-off value indicated for Aβ (B), t-tau (D) or p-tau (F). Numbers within squares show the number of patients in each group. Dots represent individual values. Bars represent mean ± 95% CI. ** Statistically significantly different, p < 0.01; * statistically significantly different, p < 0.05. n.s = statistically nonsignificantly different.

Figure 3

Sensitivity and specificity of cf-mtDNA-85/p-tau ratio. (A,D): bar graphs showing the ratio of cf-mtDNA-85/p-tau from each group, including all subjects of the entire cohort in (A) and from the groups of the AD biomarker-selected cohort in (D). ND-Ctrl, patients with neurological diseases and with no clinical signs of AD type dementia; spAD, patients with clinical signs of AD type dementia of slow progression; rpAD: patients with clinical signs of AD of rapid progression. Within squares is the number of patients in each group. Dots represent individual values. Bars represent mean ± 95% CI. BM- = ND-Ctrl patients without AD biomarkers in CSF. BM+ = spAD or AD patients with both low Aβ and high t-tau levels in CSF. (B–F) Receiving operating curve analyses (ROC) of the whole cohort (B,C) and the biomarker-selected cohort (D) and (F). Using the cut-off value of < 0.885, the sensitivity and specificity of cf-mtDNA-85/p-tau ratio are higher for patients with spAD in the biomarker-selected cohort (F). *** Statistically significantly different, p < 0.001; ** statistically significantly different, p < 0.01; * statistically significantly different p < 0.05; n.s = nonsignificantly different.