Phosphorylated tau 181 and 217 are elevated in serum and muscle of patients with amyotrophic lateral sclerosis

Abstract

Blood phosphorylated (p)-tau 181 and p-tau 217 have been proposed as accurate biomarkers of Alzheimer's disease (AD) pathology. However, blood p-tau 181 is also elevated in amyotrophic lateral sclerosis (ALS) without a clearly identified source. We measured serum p-tau 181 and p-tau 217 in a multicentre cohort of ALS (n = 152), AD (n = 111) cases and disease controls (n = 99) recruited from four different centres. Further, we investigated the existence of both p-tau species using immunohistochemistry (IHC) and mass spectrometry (MS) in muscle biopsies of ALS cases (IHC: n = 13, MS: n = 5) and disease controls (IHC: n = 14, MS: n = 5) from one cohort. Serum p-tau 181 and p-tau 217 were higher in AD and ALS patients compared to disease controls. IHC and MS analyses revealed the presence of p-tau 181 and 217 in muscle biopsies from both ALS cases and disease controls, with ALS samples showing increased p-tau reactivity in atrophic muscle fibres. Blood p-tau species could potentially be used to diagnose both ALS and AD.

Fig. 1. Distribution of serum p-tau 181 and serum p-tau 217 in the diagnostic groups and biomarker diagnostic accuracies.

A Distribution of serum p-tau 181 in the total cohort of ALS [n = 152, including 16 cases with (A+) and 111 cases without AD co-pathology (A−)], AD cases (n = 111) and disease controls (DCo, n = 99). B Serum p-tau 181 in the diagnostic groups stratified according to the individual cohorts (Halle cohort: 63 ALS, 66 AD and 40 disease controls; Milan cohort: 67 ALS, 20 AD and 22 disease controls; Mannheim 1 cohort: 22 ALS and 12 disease controls; Mannheim 2 cohort: 25 AD and 25 controls). C Distribution of serum p-tau 217 in the total cohort of ALS [n = 152, including 16 cases with (A+) and 111 cases without AD co-pathology (A−)], AD cases (n = 111) and DCo (n = 99). D Serum p-tau 217 in the diagnostic groups stratified according to the individual cohorts (Halle cohort: 63 ALS, 66 AD and 40 disease controls; Milan cohort: 67 ALS, 20 AD and 22 disease controls; Mannheim 1 cohort: 22 ALS and 12 disease controls; Mannheim 2 cohort: 25 AD and 25 controls). Biomarker levels are reported on a logarithmic scale. Dots represent single data points. Horizontal lines represent the median values, the lower and upper lines correspond to the first and third quartiles, and the vertical line is the interquartile range. Biomarker differences between groups were assessed by Mann-Whitney or Kruskal–Wallis followed by Dunn’s post hoc test (adjustment for multiple comparisons). Two-sided p-values are reported. E Receiver Operating Characteristic (ROC) curve of serum p-tau 181 in the distinction between ALS, AD and DCo. F Receiver Operating Characteristic (ROC) curve of serum p-tau 217 in the distinction between ALS, AD and DCo. Areas under the curve and standard deviation are reported. AD Alzheimer’s disease, ALS amyotrophic lateral sclerosis, ALS A+ amyotrophic lateral sclerosis with AD co-pathology, ALS A- amyotrophic lateral sclerosis without AD co-pathology, AUC area under the curve, DCo disease controls, Mann 1 Mannheim 1 cohort, Mann 2 Mannheim 2 cohort, p-tau phosphorylated tau protein. Source data are provided as a Source Data file.

Fig. 2. Distribution of serum t-tau, serum NfL and CSF NfH in the diagnostic groups.

A Distribution of serum t-tau in the total cohort of ALS [n = 141, including 14 cases with (A+) and 104 cases without AD co-pathology (A-)], AD cases (n = 104) and disease controls (DCo, n = 90). B Serum t-tau in the diagnostic groups stratified according to the individual cohorts (Halle cohort: 58 ALS, 60 AD and 32 disease controls; Milan cohort: 66 ALS, 19 AD and 21 disease controls; Mannheim 1 cohort: 17 ALS and 12 disease controls; Mannheim 2 cohort: 25 AD and 25 controls). C Distribution of serum NfL in the total cohort of ALS [n = 152, including cases with (A+) and cases without AD co-pathology (A-)], AD cases (n = 111) and DCo (n = 99). D Serum NfL in the diagnostic groups stratified according to the individual cohorts (Halle cohort: 63 ALS, 66 AD and 40 disease controls; Milan cohort: 67 ALS, 20 AD and 22 disease controls; Mannheim 1 cohort: 22 ALS and 12 disease controls; Mannheim 2 cohort: 25 AD and 25 controls). E Distribution of CSF NfH in the total cohort of ALS [n = 152, including cases with (A+) and cases without AD co-pathology (A-)], AD cases (n = 111) and DCo (n = 99). F CSF NfH in the diagnostic groups stratified according to the individual cohorts (Halle cohort: 63 ALS, 66 AD and 40 disease controls; Milan cohort: 67 ALS, 20 AD and 22 disease controls; Mannheim 1 cohort: 22 ALS and 12 disease controls; Mannheim 2 cohort: 25 AD and 25 controls). Biomarker levels are reported on a logarithmic scale. Dots represent single data points. Horizontal lines represent the median values, the lower and upper lines correspond to the first and third quartiles, and the vertical line is the interquartile range. Biomarker differences between groups were assessed by Mann–Whitney or Kruskal–Wallis followed by Dunn’s post hoc test (adjustment for multiple comparisons). Two-sided p-values are reported. AD Alzheimer’s disease, ALS amyotrophic lateral sclerosis, ALS A+ amyotrophic lateral sclerosis with AD copathology, ALS A- amyotrophic lateral sclerosis without AD copathology, AUC area under the curve, CSF cerebrospinal fluid, DCo disease controls, Mann 1 Mannheim 1 cohort, Mann 2 Mannheim 2 cohort, NfH neurofilament heavy chain protein, NfL neurofilament light chain protein, t-tau total tau protein. Source data are provided as a Source Data file.

Fig. 3. P-tau 181 and p-tau 217 immunoreactivity in muscle biopsies.

Immunohistochemical findings from a disease control (subject no. 8, male, 45 years old, diagnose: myalgia and cramps, disease duration 24 months, biopsy from biceps brachii, further details in Supplementary Table 6) and a patient with ALS (subject no. 4, male, 75 years old, disease duration 29 months, definite ALS according to the revised El Escorial criteria, King’s stage 4, biopsy from vastus lateralis, further details in Supplementary Table 6) are shown. Muscle biopsies from disease controls (n = 14) showed normal-sized, predominantly hexagonal fibres with random variations in muscle fibre size and shape. Conversely, biopsies from patients with ALS (n = 13) showed neurogenic features, including grouped or even fascicular atrophy, with small-angled fibres (example of atrophic fibre marked with white arrowhead) and nuclear bags, often accompanied by compensatory hypertrophy of unaffected fibres (black asterisk). Both ALS (n = 13) and disease control (n = 14) muscle biopsies showed p-tau 181 and p-tau 217 immunoreactivity (scale bar: 50 µm) predominantly localised to the myonuclei (black arrowheads), sometimes extending into the perinuclear regions. In all ALS samples, we found increased sarcoplasmic reactivity to p-tau181 and p-tau 217 in atrophic muscle fibres (white arrowhead), whereas normal or hypertrophic fibres did not show increased sarcoplasmic reactivity (black asterisk).

Fig. 4. Mass spectrometry-based analysis of tau phosphorylation in muscle biopsies.

Exemplary spectra and for the identification of p-tau 181 (A) and p-tau 217 (B) in muscle extracts are shown. Both phosphosites were identified in all analysed ALS patients (n = 5) and disease controls (n = 5). C Additional phosphorylation sites were identified in all samples. Two sites, S437 and S438, were identified in a protein region that is not expressed in human CNS tau. Protein regions shown in purple-blue are expressed in adult CNS and other tissues. Protein regions shown in orange, but not in purple-blue, are not expressed in adult CNS. Positions are indicated as the site in the CNS-expressed isoform. Sites outside this isoform are given with respect to the full-length protein. CNS central nervous system, p-tau phosphorylated tau.