CoChat

Sirtuin-1 sensitive lysine-136 acetylation drives phase separation and pathological aggregation of TDP-43

Open access

The DOI matches its Crossref record; 55 of the 55 references it cites were found in a registry or an index.

Abstract

Trans-activation response DNA-binding protein of 43 kDa (TDP-43) regulates RNA processing and forms neuropathological aggregates in patients with amyotrophic lateral sclerosis and frontotemporal lobar degeneration. Investigating TDP-43 post-translational modifications, we discovered that K84 acetylation reduced nuclear import whereas K136 acetylation impaired RNA binding and splicing capabilities of TDP-43. Such failure of RNA interaction triggered TDP-43 phase separation mediated by the C-terminal low complexity domain, leading to the formation of insoluble aggregates with pathologically phosphorylated and ubiquitinated TDP-43. Introduction of acetyl-lysine at the identified sites via amber suppression confirmed the results from site-directed mutagenesis. K84-acetylated TDP-43 showed cytoplasmic mislocalization, and the aggregation propensity of K136-acetylated TDP-43 was confirmed. We generated antibodies selective for TDP-43 acetylated at these lysines, and found that sirtuin-1 can potently deacetylate K136-acetylated TDP-43 and reduce its aggregation propensity. Thus, distinct lysine acetylations modulate nuclear import, RNA binding and phase separation of TDP-43, suggesting regulatory mechanisms for TDP-43 pathogenesis.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-022-28822-7

Cites

55 of 55 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 TDP-43 is acetylated and its acetylation at K84 may affect its nuclear-cytoplasmic trafficking while acetylation at K136 causes a nuclear droplet-like distribution. a HEK293E cells were transfected with 6xHis-tagged TDP-43, which was purified with NiNTA beads and subjected to MS. The residues marked in the domain structure of TDP-43 were found to be acetylated (black arrows). K145 has been reported to be acetylated in the literature but we could not find it in our study (grey arrow). b Immunostaining of HEK293E cells transfected with 6xHis-tagged wtTDP-43, acetyl-dead (K84R, K136R and K145R) or acetyl-mimics (K84Q, K136Q and K145Q). Scale bar represents 10 μm. c Quantification of percentage of cells with 6xHis-TDP-43 protein distribution diffuse nuclear (blue bars), nuclear clumpy (grey bars), nuclear + cytosolic, diffuse (orange bars) and nuclear + cytosolic, clumpy (yellow bars). A total of 300 transfected cells from three independent experiments per group were classified. ***p < 0.001 as measured by Chi-squared test. Source data are provided as a Source Data file.
Fig. 1 TDP-43 is acetylated and its acetylation at K84 may affect its nuclear-cytoplasmic trafficking while acetylation at K136 causes a nuclear droplet-like distribution. a HEK293E cells were transfected with 6xHis-tagged TDP-43, which was purified with NiNTA beads and subjected to MS. The residues marked in the domain structure of TDP-43 were found to be acetylated (black arrows). K145 has been reported to be acetylated in the literature but we could not find it in our study (grey arrow). b Immunostaining of HEK293E cells transfected with 6xHis-tagged wtTDP-43, acetyl-dead (K84R, K136R and K145R) or acetyl-mimics (K84Q, K136Q and K145Q). Scale bar represents 10 μm. c Quantification of percentage of cells with 6xHis-TDP-43 protein distribution diffuse nuclear (blue bars), nuclear clumpy (grey bars), nuclear + cytosolic, diffuse (orange bars) and nuclear + cytosolic, clumpy (yellow bars). A total of 300 transfected cells from three independent experiments per group were classified. ***p < 0.001 as measured by Chi-squared test. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 2 K136 mutations reduce the solubility of TDP-43 and trigger its ubiquitination and pathological phosphorylation. a HEK293E shTDP-43 cells were transfected with 6xHis-tagged TDP-43 (wt, K136R and K136Q) or Flag-tagged FFLL TDP-43. RIPA soluble and insoluble fractions were separated and analyzed by Western blotting. N = 3 biologically independent samples. Source data are provided as a Source Data file. b ShTDP-43-HEK293E cells were transfected with 6xHis-tagged TDP-43 (wt, K136R and K136Q). Proteasome activity was inhibited by treatment with 20 µM MG-132 for 6 h, which caused wtTDP-43 ubiquitination (lane 3). 6xHis-tagged TDP-43 was purified with NiNTA beads and analyzed on Western blots. N = 3 biologically independent samples. Source data are provided as a Source Data file. c Transfected shTDP-43-HEK293E cells were double immunostained for TDP-43 and phosphorylated TDP-43 (S409/410). Scale bar represents 10 μm. N = 3 biological replicates.
Fig. 2 K136 mutations reduce the solubility of TDP-43 and trigger its ubiquitination and pathological phosphorylation. a HEK293E shTDP-43 cells were transfected with 6xHis-tagged TDP-43 (wt, K136R and K136Q) or Flag-tagged FFLL TDP-43. RIPA soluble and insoluble fractions were separated and analyzed by Western blotting. N = 3 biologically independent samples. Source data are provided as a Source Data file. b ShTDP-43-HEK293E cells were transfected with 6xHis-tagged TDP-43 (wt, K136R and K136Q). Proteasome activity was inhibited by treatment with 20 µM MG-132 for 6 h, which caused wtTDP-43 ubiquitination (lane 3). 6xHis-tagged TDP-43 was purified with NiNTA beads and analyzed on Western blots. N = 3 biologically independent samples. Source data are provided as a Source Data file. c Transfected shTDP-43-HEK293E cells were double immunostained for TDP-43 and phosphorylated TDP-43 (S409/410). Scale bar represents 10 μm. N = 3 biological replicates.CC BY · PubMed Central
Fig. 3 K136 mutations alter the RNA binding and splicing capabilities of TDP-43. a HEK293E shTDP-43 cells were transfected with an empty plasmid, wt or K136Q TDP-43. The lysates were then incubated with biotinylated (UC)12 or (UG)12 RNA oligomers. Protein-RNA mix was incubated with magnetic streptavidin beads. Bound proteins were eluted and analyzed by Western blotting. N = 3 biological replicates. Source data are provided as a Source Data file. b Increasing concentrations of 6xHis purified TDP-43 proteins were incubated with biotinylated (UG)12 oligomers. The resulting complexes were analyzed by a filter-binding assay. Membranes were incubated with HRP-coupled streptavidin. One of the [K136Q]TDP-43 data points illustrates overload problems at high protein concentration (>875 nM), precluding firm establishment of saturation binding curves. Source data are provided as a Source Data file. c Quantification of three replicates of filter-binding assay of wt (full line) and K136Q (dashed line) TDP-43 to biotinylated (UG)12 RNA oligomers. **p < 0.01. Data are presented as mean values ± SD. Unpaired, two-sided t-test was used for comparison. Source data are provided as a Source Data file. d Representative Western blot of the protein levels before and after the purification of native TDP-43 via NiNTA pulldown. N = 3 biological replicates. Source data are provided as a Source Data file. e HEK293E shTDP-43 cells were cotransfected with the specified TDP-43 constructs and a plasmid containing the CFTR minigene. RNA was extracted and splicing of CFTR exon 9 was assessed via rtPCR (upper panel). Protein levels are shown in the lower panel. N = 3 biologically independent samples. f Quantification of three replicates assessing effects of mutations at different lysines of TDP-43 on the splicing of CFTR exon 9. Data are presented as mean values ± SD. Unpaired, two-sided t-test was used for comparison. **p < 0.01. Source data are provided as a Source Data file.
Fig. 3 K136 mutations alter the RNA binding and splicing capabilities of TDP-43. a HEK293E shTDP-43 cells were transfected with an empty plasmid, wt or K136Q TDP-43. The lysates were then incubated with biotinylated (UC)12 or (UG)12 RNA oligomers. Protein-RNA mix was incubated with magnetic streptavidin beads. Bound proteins were eluted and analyzed by Western blotting. N = 3 biological replicates. Source data are provided as a Source Data file. b Increasing concentrations of 6xHis purified TDP-43 proteins were incubated with biotinylated (UG)12 oligomers. The resulting complexes were analyzed by a filter-binding assay. Membranes were incubated with HRP-coupled streptavidin. One of the [K136Q]TDP-43 data points illustrates overload problems at high protein concentration (>875 nM), precluding firm establishment of saturation binding curves. Source data are provided as a Source Data file. c Quantification of three replicates of filter-binding assay of wt (full line) and K136Q (dashed line) TDP-43 to biotinylated (UG)12 RNA oligomers. **p < 0.01. Data are presented as mean values ± SD. Unpaired, two-sided t-test was used for comparison. Source data are provided as a Source Data file. d Representative Western blot of the protein levels before and after the purification of native TDP-43 via NiNTA pulldown. N = 3 biological replicates. Source data are provided as a Source Data file. e HEK293E shTDP-43 cells were cotransfected with the specified TDP-43 constructs and a plasmid containing the CFTR minigene. RNA was extracted and splicing of CFTR exon 9 was assessed via rtPCR (upper panel). Protein levels are shown in the lower panel. N = 3 biologically independent samples. f Quantification of three replicates assessing effects of mutations at different lysines of TDP-43 on the splicing of CFTR exon 9. Data are presented as mean values ± SD. Unpaired, two-sided t-test was used for comparison. **p < 0.01. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 4 Amber suppression-mediated acetylation at K136 of TDP-43 prevents CFTR splicing, triggers the formation of nuclear droplets via its C-terminal domain. Acetylation at K84 reduces nuclear import. a HEK293E cells cotransfected with a plasmid containing AcRS and 4xtRNATAG (E451) and with TDP-43 with an amber stop codon at different positions and a C-terminal 6xHis tag. Media was changed to 5 mM acK-containing media 24 h before lysis. Protein lysates were analyzed via Western blot. N = 3 biological replicates. Source data are provided as a Source Data file. b Cells were cotransfected with the E451 plasmid and different C-terminally 6xHis-tagged TDP-43 constructs containing amber suppression codons. Cells were fixed after 24 h in the presence of 5 mM acK. Scale bar represents 10 µm. N = 3 biological replicates. c HEK293E shTDP-43 cells cotransfected with E451, CFTR minigene, and an empty plasmid, wt or K136TAG TDP-43. After 24 h in the presence of 5 mM acK, RNA was extracted and the splicing of CFTR exon 9 was assessed via rtPCR and visualized in an agarose gel. N = 3 biological replicates. Source data are provided as a Source Data file. d HEK293E shTDP-43 cells were cotransfected with E451 plasmid and different C-terminally tagged TDP-43 constructs. Lanes 3–5 have an amber stop codon at position 136 in addition to a mutated NLS, an amber stop codon at position 84 or a deletion of the C-terminus. Cells were treated with 5 mM acK for 24 h before lysis. N = 3 biological replicates. e HEK293E shTDP-43 cells were cotransfected with E451 and different C-terminally tagged TDP-43 constructs. Cells were fixed and immunolabeled after 24 h in the presence of 5 mM acK. Scale bar represents 10 µm. N = 3 biological replicates. f Quantification of percentage of cells with 6xHis-TDP-43 protein distribution diffuse nuclear (blue bars), nuclear clumpy (grey bars), nuclear + cytosolic, diffuse (orange bars) and nuclear + cytosolic, clumpy (yellow bars). A total of 250 cells per group cotransfected with E451 plasmid and different 6xHis C-terminally tagged TDP-43 constructs. Source data are provided as a Source Data file.
Fig. 4 Amber suppression-mediated acetylation at K136 of TDP-43 prevents CFTR splicing, triggers the formation of nuclear droplets via its C-terminal domain. Acetylation at K84 reduces nuclear import. a HEK293E cells cotransfected with a plasmid containing AcRS and 4xtRNATAG (E451) and with TDP-43 with an amber stop codon at different positions and a C-terminal 6xHis tag. Media was changed to 5 mM acK-containing media 24 h before lysis. Protein lysates were analyzed via Western blot. N = 3 biological replicates. Source data are provided as a Source Data file. b Cells were cotransfected with the E451 plasmid and different C-terminally 6xHis-tagged TDP-43 constructs containing amber suppression codons. Cells were fixed after 24 h in the presence of 5 mM acK. Scale bar represents 10 µm. N = 3 biological replicates. c HEK293E shTDP-43 cells cotransfected with E451, CFTR minigene, and an empty plasmid, wt or K136TAG TDP-43. After 24 h in the presence of 5 mM acK, RNA was extracted and the splicing of CFTR exon 9 was assessed via rtPCR and visualized in an agarose gel. N = 3 biological replicates. Source data are provided as a Source Data file. d HEK293E shTDP-43 cells were cotransfected with E451 plasmid and different C-terminally tagged TDP-43 constructs. Lanes 3–5 have an amber stop codon at position 136 in addition to a mutated NLS, an amber stop codon at position 84 or a deletion of the C-terminus. Cells were treated with 5 mM acK for 24 h before lysis. N = 3 biological replicates. e HEK293E shTDP-43 cells were cotransfected with E451 and different C-terminally tagged TDP-43 constructs. Cells were fixed and immunolabeled after 24 h in the presence of 5 mM acK. Scale bar represents 10 µm. N = 3 biological replicates. f Quantification of percentage of cells with 6xHis-TDP-43 protein distribution diffuse nuclear (blue bars), nuclear clumpy (grey bars), nuclear + cytosolic, diffuse (orange bars) and nuclear + cytosolic, clumpy (yellow bars). A total of 250 cells per group cotransfected with E451 plasmid and different 6xHis C-terminally tagged TDP-43 constructs. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 5 [K136Q]TDP-43 droplets have reduced mobility, growing in size and fusing over time. a Representative images of FRAP analysis of HEK293E cells transfected with C-terminally EGFP-tagged wt and K136Q TDP-43. Two different groups of [K136Q]TDP-43-EGFP droplets based on size were determined. Time in seconds. Scale Bar represents 5 µm. b Quantification of FRAP images (three cells quantified per group). The inclusions bodies are classified as follows: diffuse TDP-43 (circles), small inclusions (triangles) and big inclusions (squares). Source data are provided as a Source Data file. c Quantification of mobile fraction of FRAP analysis of wt and K136Q TDP-43 mutants. Three cells quantified per group. Data are presented as mean values ± SEM. Unpaired, two-sided t-test was used for comparison *p < 0.05. Source data are provided as a Source Data file. d Selected frames from green fluorescence live-cell imaging of cells expressing [K136Q]TDP-43-EGFP. Close-up of granules fusing (white arrows). Time in seconds. Scale bar represents 5 µm. N = 3 biological replicates. e Histogram of K136Q TDP-43 aggregate size in HEK293E fixed 24, 48 or 72 h after transfection. A total of 100 transfected cells analyzed for each time point. Different patterns represent the respective hours after transfection. Unpaired, two-sided t-test was used for comparison. Data are presented as mean values ± SEM. **p = 0.0075. Source data are provided as a Source Data file.
Fig. 5 [K136Q]TDP-43 droplets have reduced mobility, growing in size and fusing over time. a Representative images of FRAP analysis of HEK293E cells transfected with C-terminally EGFP-tagged wt and K136Q TDP-43. Two different groups of [K136Q]TDP-43-EGFP droplets based on size were determined. Time in seconds. Scale Bar represents 5 µm. b Quantification of FRAP images (three cells quantified per group). The inclusions bodies are classified as follows: diffuse TDP-43 (circles), small inclusions (triangles) and big inclusions (squares). Source data are provided as a Source Data file. c Quantification of mobile fraction of FRAP analysis of wt and K136Q TDP-43 mutants. Three cells quantified per group. Data are presented as mean values ± SEM. Unpaired, two-sided t-test was used for comparison *p < 0.05. Source data are provided as a Source Data file. d Selected frames from green fluorescence live-cell imaging of cells expressing [K136Q]TDP-43-EGFP. Close-up of granules fusing (white arrows). Time in seconds. Scale bar represents 5 µm. N = 3 biological replicates. e Histogram of K136Q TDP-43 aggregate size in HEK293E fixed 24, 48 or 72 h after transfection. A total of 100 transfected cells analyzed for each time point. Different patterns represent the respective hours after transfection. Unpaired, two-sided t-test was used for comparison. Data are presented as mean values ± SEM. **p = 0.0075. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 6 Antibodies against acetylated TDP-43 specifically recognize the modified forms. a, b HEK293E shTDP-43 cells were co transfected with E451 plasmid and different C-terminally 6xHis-tagged TDP-43 constructs. After 24 h in the presence of 5 mM of acK cells were lysed. Samples were Western probed with antibodies against acetylated K84 (clone 9G9) and K136 (clone 14D4), respectively. N = 3 biological replicates. Source data are provided as a Source Data file. c, d HEK293E shTDP-43 cells were cotransfected with E451 plasmid and different C-terminally 6xHis-tagged TDP-43 constructs. After 24 h in the presence of 5 mM of acK cells were fixed and immunostained for 6xHis or acetylated TDP-43. Scale bar represents 10 µm. N = 3 biological replicates.
Fig. 6 Antibodies against acetylated TDP-43 specifically recognize the modified forms. a, b HEK293E shTDP-43 cells were co transfected with E451 plasmid and different C-terminally 6xHis-tagged TDP-43 constructs. After 24 h in the presence of 5 mM of acK cells were lysed. Samples were Western probed with antibodies against acetylated K84 (clone 9G9) and K136 (clone 14D4), respectively. N = 3 biological replicates. Source data are provided as a Source Data file. c, d HEK293E shTDP-43 cells were cotransfected with E451 plasmid and different C-terminally 6xHis-tagged TDP-43 constructs. After 24 h in the presence of 5 mM of acK cells were fixed and immunostained for 6xHis or acetylated TDP-43. Scale bar represents 10 µm. N = 3 biological replicates.CC BY · PubMed Central
Fig. 7 Sirtuins 1 and 2 can deacetylate K136 of TDP-43 and prevent its aggregation. a HEK293E shTDP-43 cells were cotransfected with E451, K136TAG TDP-43 and SIRT1 or SIRT2. Twenty-four hours before lysis, 5 mM acK and DMSO, 1 µM or 10 µM Ex527 containing media was supplied to the cells. N = 3 biological replicates. Source data are provided as a Source Data file. b Stably amber suppressed shTDP-43-HEK293E cells were cotransfected with different sirtuin constructs together with K136TAG TDP-43. Cells were fixed after 24 h in the presence of 5 mM acK. Scale bar represents 10 µm. c Quantification of the number of cells with TDP-43 positive inclusions (6xHis staining). A total of 200 cells from 4 independent experiments were recorded per condition. Unpaired, two-sided t-test was used for comparison. Data are presented as mean values ± SEM. **p = 0.0083. Source data are provided as a Source Data file.
Fig. 7 Sirtuins 1 and 2 can deacetylate K136 of TDP-43 and prevent its aggregation. a HEK293E shTDP-43 cells were cotransfected with E451, K136TAG TDP-43 and SIRT1 or SIRT2. Twenty-four hours before lysis, 5 mM acK and DMSO, 1 µM or 10 µM Ex527 containing media was supplied to the cells. N = 3 biological replicates. Source data are provided as a Source Data file. b Stably amber suppressed shTDP-43-HEK293E cells were cotransfected with different sirtuin constructs together with K136TAG TDP-43. Cells were fixed after 24 h in the presence of 5 mM acK. Scale bar represents 10 µm. c Quantification of the number of cells with TDP-43 positive inclusions (6xHis staining). A total of 200 cells from 4 independent experiments were recorded per condition. Unpaired, two-sided t-test was used for comparison. Data are presented as mean values ± SEM. **p = 0.0083. Source data are provided as a Source Data file.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 55 references

Cited by

See all 97

Metadata

Authors
Garcia Morato, J., Hans, F., von Zweydorf, F., Feederle, R., Elsässer, S. J., Skodras, A. A., Gloeckner, C. J., Buratti, E., Neumann, M., & Kahle, P. J.
Affiliations
  • German Center for Neurodegenerative Diseases (DE)
  • Science for Life Laboratory (SE)
  • Hertie Institute for Clinical Brain Research (DE)
  • International Centre for Genetic Engineering and Biotechnology (IT)
Corresponding author
Kahle, Philipp J.
Venue
Nature Communications
Published
2022-03-09
Volume
13
Issue
1
Pages
1223-1223
ISSN
2041-1723 (ISSN-L)
OpenAlex
W4220753802
Cited by
94
References
54
Citations by year
2026: 13 · 2025: 26 · 2024: 24 · 2023: 22 · 2022: 7
FWCI
10.40
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • FreiDok plus (Universitätsbibliothek Freiburg) (submittedVersion)
  • Site cant be reached (submittedVersion, cc-by)
  • DZNE Pub (submittedVersion, cc-by)
Funders
  • NOMIS Stiftung
  • Deutsche Forschungsgemeinschaft (KA1675/3-2)
  • Deutsches Zentrum für Neurodegenerative Erkrankungen
Topics
  • Health Sciences › Medicine › Clinical Neurology › Amyotrophic Lateral Sclerosis Research
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Alzheimer's disease research and treatments
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Genetic Neurodegenerative Diseases
Keywords
TDP-43, lysine acetylation, SIRT1, phase separation, protein aggregation, amyotrophic lateral sclerosis, frontotemporal lobar degeneration, nuclear import, RNA binding, RNA splicing, low-complexity domains, post-translational modification, phosphorylation, ubiquitination, cytoplasmic mislocalization
MeSH terms
Acetylation, Amyotrophic Lateral Sclerosis, DNA-Binding Proteins, Humans, Lysine, Protein Processing, Post-Translational, RNA, Sirtuin 1, Protein Aggregation, Pathological

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Sirtuin-1 sensitive lysine-136 acetylation drives phase separation and pathological aggregation of TDP-43. https://app.cochat.ai/papers/10.1038/s41467-022-28822-7
The paper
Garcia Morato, J., Hans, F., von Zweydorf, F., Feederle, R., Elsässer, S. J., Skodras, A. A., Gloeckner, C. J., Buratti, E., Neumann, M., & Kahle, P. J. (2022). Sirtuin-1 sensitive lysine-136 acetylation drives phase separation and pathological aggregation of TDP-43. Nature Communications. https://doi.org/10.1038/s41467-022-28822-7

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.