Unusually Rapid Isomerization of Aspartic Acid in Tau

Abstract

Spontaneous chemical modifications in long-lived proteins can potentially change protein structure in ways that impact proteostasis and cellular health. For example, isomerization of aspartic acid interferes with protein turnover and is anticorrelated with cognitive acuity in Alzheimer's disease. However, few isomerization rates have been determined for Asp residues in intact proteins. To remedy this deficiency, we used protein extracts from SH-SY5Y neuroblastoma cells as a source of a complex, brain-relevant proteome with no baseline isomerization. Cell lysates were aged in vitro to generate isomers, and extracted proteins were analyzed by data-independent acquisition (DIA) liquid chromatography-mass spectrometry (LC-MS). Although no Asp isomers were detected at Day 0, isomerization increased across time and was quantifiable for 105 proteins by Day 50. Data analysis revealed that isomerization rate is influenced by both primary sequence and secondary structure, suggesting that steric hindrance and backbone rigidity modulate isomerization. Additionally, we examined lysates extracted under gentle conditions to preserve protein complexes and found that protein-protein interactions often slow isomerization. Base catalysis was explored as a means to accelerate Asp isomerization due to findings of accelerated asparagine deamidation. However, no substantial rate enhancement was found for isomerization, suggesting fundamental differences in acid-base chemistry. With an enhanced understanding of Asp isomerization in proteins in general, we next sought to better understand Asp isomerization in tau. In vitro aging of monomeric and aggregated recombinant tau revealed that tau isomerizes significantly faster than any similar protein within our dataset, which is likely related to its correlation with cognition in Alzheimer's disease.

Keywords: Alzheimer’s disease; aging; aspartic acid; data-independent acquisition; isomerization; liquid chromatography-mass spectrometry.

Figure 1.

Skyline chromatograms for the peptide ALVDGPCTQVR from the protein RL14 across 5 in-vitro aging time points. Each colored line corresponds a fragment ion associated with the precursor in the spectral library. The first isomer peak emerges at a different retention time on day 5, with all four peaks being clearly visible by Day 50.

Figure 2.

Impact of primary sequence on isomerization. a) Number of detected isomers and b) percent isomerization binned by the residue C-terminal to Asp. c) Number of detected isomers and d) percent isomerization binned by the residue N-terminal to Asp. Error bars for b) and d) represent one standard deviation.

Figure 3.

Crowding of the C-terminal amide. (a) percent isomerization compared to the number of atoms within 4 Å of the C-terminal amide involved in succinimide formation (b) percent isomerization compared to the number of atoms within 4 Å of the C-terminal amide involved in succinimide formation, binned by DG or non-DG isomer. (c) Box-and-whisker plot showing percent isomerization binned by presence of absence of hydrogen bonding (d) Ratio of the number of isomers where the hydrogen of the C-terminal amide has hydrogen bonding interactions to the total number of isomers.

Figure 4.

Impact of secondary structure on isomerization. (a) Number of detected Asp isomers sorted by secondary structure element (b) percent isomerization binned by both primary sequence and secondary structure element (c) percent isomerization compared to the number of atoms within 4 Å of the C-terminal amide involved in succinimide formation, binned by secondary structure element.

Figure 5.

Impact of Native lysis buffer on isomerization. (a) Isomerization at each timepoint in RIPA and Native buffers for the peptide ATAVVDGAFK from PRDX2, error bars correspond to biological replicates (n=3). (b) Difference in percent isomerization (delta) across timepoints for ATAVVDGAFK. (c) Isomerization at each timepoint in RIPA and Native buffers for TALIHDGLAR from RS12, error bars correspond to biological replicates (n=3). (d) Difference in percent isomerization (delta) across timepoints for TALIHDGLAR (e) Before-and-after plot comparing percent isomerization at Day 50 in RIPA and Native lysis buffers for isomerized peptides. Green and purple lines signify decreases and increases, respectively, in percent isomerization greater than two standard deviations of the biological triplicate error. (f) Number of isomers with significant changes to percent isomerization value when comparing RIPA and Native lysis buffers.

Figure 6.

Evaluation of base catalysis and effects of pH on Asp isomerization. (a) Comparisons between “Accelerated” and standard lysis buffers for RIPA (orange) and Native (green) lysis buffers, where red lines correspond to changes in percent isomerization with a difference greater than two standard deviations from triplicate percent isomerization calculations. (b) Bar plot showing whether observed differences led to no significant change (orange), significant decreases (green), or significant increases (purple) for both RIPA and Native lysis buffers. (c-d) Impact of ammonium hydroxide on Asn deamidation at Day 50, shown as in (a) and (b).

Figure 7.

Progression of isomerization for the peptide TDHGAEIVYK from recombinant tau. (a) Protein aged only in PBS buffer is shown in blue, while protein in aggregating conditions is shown in orange. Red-dotted line corresponds to 50 days of aging. Error bars are derived from technical replicates. (b) percent isomerization compared to the length of the protein in amino acids, with DG isomers in blue, non-DG isomers in red, and the average %isomerization for non-DG isomers in the red, dashed line. Percent isomerization at Day 50 for TDHGAEIVYK from tau shown, with monomeric tau in green and aggregated tau in yellow.

Scheme 1.

Mechanism of Aspartic Acid Isomerization

Scheme 2.

Experimental procedure for in vitro aging of cell lysates and preparation for LC-MS

Scheme 3.

Examples of potential conformations for deamidation site residue pairs, with the reaction site allowing succinimide ring formation highlighted in yellow.