CoChat

Inhibiting tumor cell-intrinsic UBA6 by inosine augments tumor immunogenicity

Open access
He, Baokun, Zhang, Lei, Jiang, Li, Yu, Liang, Li, Qin, Tian, Xiangjun, He, Jing-Quan, Zeng, Ling, Yang, Yuqin, Wang, Chaoran, Wei, Yuhan, Jiang, Xiaoyue, Li, Jing, Ge, Xiaolu, Gu, Qisheng, Li, Ji-Kun, Wu, Di, Sadler, Anthony, Yu, Di, Xu, Dakang, Gao, Yue, Yuan, Xiangliang · 2021 · Nature Communications

Preprint — published as 10.1038/s41467-022-33116-z

The DOI matches its Crossref record; 53 of the 53 references it cites were found in a registry or an index.

Abstract

Metabolic alteration influences cancer immunity. However, the role and mechanism of metabolic adaption on immune checkpoint blockade (ICB) responses remains ill-defined. Here, to identify metabolites that modulate ICB sensitivity, metabolomic profiling in mouse tumor models and cancer patients treated with ICB was performed. We identified that metabolite inosine was associated with ICB sensitivity in mice and humans, and overcame ICB resistance in several mouse tumor models. Notably, inosine sensitized tumor cells to T cell-mediated cytotoxicity by amplifying tumor-intrinsic immunogenicity. Chemical proteomics further identified that inosine directly bound and inhibited ubiquitin-activating enzyme UBA6. Tumor UBA6 loss augmented tumor immunogenicity and substituted the synergistic effect of inosine in combination with ICB. Clinically, tumor UBA6 expression negatively correlated with ICB response in cancer patients. Thus, we reveal an unappreciated function of inosine on tumor-intrinsic immunogenicity and provide UBA6 as a candidate target for immunotherapy.

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 · 10.21203/rs.3.rs-620390/v1

Cites

53 of 53 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Inosine overcomes resistance to immunotherapy by inflaming tumour immune microenvironment. a Volcano plot showing the hazard ratios (HR) of high/low levels of serum metabolites (n = 202, represented as points, high and low are stratified by the median of each metabolite) in CheckMate 025 renal cell carcinoma (RCC) patients with nivolumab (anti-PD1 Ab) treatment (n = 392). Overall survival is used as a Cox proportional hazards model. A cutoff at P = 0.01 is shown as a horizontal line and HR (high/low) = 1 is shown as a vertical line (HR: high/low <1 indicates that a high level of a metabolite is a benefit for ICB patients). b Kaplan–Meier plot of overall survival in RCC patients with nivolumab (anti-PD1 Ab) (High, n = 196, mean OS = 33; Low, n = 195, mean OS = 21) or everolimus (mTOR inhibitor) (High, n = 174, mean OS = 20; Low, n = 174, mean OS = 20) grouped by the inosine level at the median of baseline level. c–e Tumour volume and survival analysis of B16-F0 (c, n = 5), B16-GMCSF (d, n = 5), or 4T1 (e, n = 6) tumour-bearing mice treated with IgG2a (Ctrl), 400 mg/kg of Inosine (Ino), anti-CTLA4 + anti-PD1 (ICB) or ICB + Inosine (Ino + ICB) treatment. f tSNE plot of single-cell RNA sequencing of CD45+ immune cells from 4T1 tumours treated with Ctrl (n = 16199 cells) or Ino+ICB treatment (n = 9842 cells). g The frequency of gp70-specific CD8+ T cells in 4T1 tumour with Ctrl (n = 6), Ino (n = 6), ICB (n = 6), or Ino+ICB (n = 8) treatment for 15 days. Data are presented as Mean ± s.e.m. Statistical significance was determined by one-way ANOVA and Tukey test for multiple comparisons (tumour sized of c–e, g) or log-rank (Mantel–Cox) test (survival analysis of b–e). Source data are provided as a Source Data file.
Fig. 1 Inosine overcomes resistance to immunotherapy by inflaming tumour immune microenvironment. a Volcano plot showing the hazard ratios (HR) of high/low levels of serum metabolites (n = 202, represented as points, high and low are stratified by the median of each metabolite) in CheckMate 025 renal cell carcinoma (RCC) patients with nivolumab (anti-PD1 Ab) treatment (n = 392). Overall survival is used as a Cox proportional hazards model. A cutoff at P = 0.01 is shown as a horizontal line and HR (high/low) = 1 is shown as a vertical line (HR: high/low <1 indicates that a high level of a metabolite is a benefit for ICB patients). b Kaplan–Meier plot of overall survival in RCC patients with nivolumab (anti-PD1 Ab) (High, n = 196, mean OS = 33; Low, n = 195, mean OS = 21) or everolimus (mTOR inhibitor) (High, n = 174, mean OS = 20; Low, n = 174, mean OS = 20) grouped by the inosine level at the median of baseline level. c–e Tumour volume and survival analysis of B16-F0 (c, n = 5), B16-GMCSF (d, n = 5), or 4T1 (e, n = 6) tumour-bearing mice treated with IgG2a (Ctrl), 400 mg/kg of Inosine (Ino), anti-CTLA4 + anti-PD1 (ICB) or ICB + Inosine (Ino + ICB) treatment. f tSNE plot of single-cell RNA sequencing of CD45+ immune cells from 4T1 tumours treated with Ctrl (n = 16199 cells) or Ino+ICB treatment (n = 9842 cells). g The frequency of gp70-specific CD8+ T cells in 4T1 tumour with Ctrl (n = 6), Ino (n = 6), ICB (n = 6), or Ino+ICB (n = 8) treatment for 15 days. Data are presented as Mean ± s.e.m. Statistical significance was determined by one-way ANOVA and Tukey test for multiple comparisons (tumour sized of c–e, g) or log-rank (Mantel–Cox) test (survival analysis of b–e). Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 2 Inosine sensitises tumour cells to T-cell-mediated killing by modulating tumour immunogenicity. a Experimental strategy to evaluate the ability of inosine to enhance T-cell-mediated tumour killing by modulating T cells. b The relative cell viability of B16-GMCSF-OVA cells was shown. OT-1 T cells were pretreated with indicated concentrations of inosine or vehicle for 24 h, then co-cultured with B16-GMCSF-OVA tumour cells at a 2:1 E: T (Effector: T cells, Target: tumour cells) ratio for 48 h (n = 5). c The relative cell viability of 4T1 cells after incubation with activated CD8+ T cells pretreated with the indicated concentration of inosine at a 5:1 E: T ratio for 48 h (n = 5). d Experimental strategy to evaluate the ability of inosine to enhance T-cell-mediated tumour killing by modulating Tumour cells. e The relative cell viability of B16-GMCSF-OVA cells (left) and 4T1 cells (right) with Ctrl or inosine treatment following the method in a (n = 5). f Experimental strategy to evaluate the direct effect of inosine on tumour cells. g The relative cell viability (left) and apoptosis (right) of B16-GMCSF cells following inosine treatment at indicated concentrations for 48 h in vitro (n = 5). h The relative cell viability (left) and apoptosis (right) of 4T1 cells following inosine treatment at indicated concentrations for 48 h in vitro (n = 5). i Representative flow analysis (left panel) and quantifying (left panel) the intensity of cell surface MHC-I expression in B16-GMCSF cells treated with vehicle and inosine (100 μM) upon IFN-γ (20 ng/ml) treatment (n = 3). j, k Selective represented antigen processing/presentation and interferon-responsive gene expression in B16-GMCSF (j) and 4T1 (k) and tumour cells treated with inosine at indicated concentrations (n = 3). Data are presented as Mean ± s.e.m. Statistical significance was determined by Two-sided Student’s t test (i) and one-way ANOVA and Tukey test for multiple comparisons (b, c, e, g, h). NS no significant. Source data are provided as a Source Data file.
Fig. 2 Inosine sensitises tumour cells to T-cell-mediated killing by modulating tumour immunogenicity. a Experimental strategy to evaluate the ability of inosine to enhance T-cell-mediated tumour killing by modulating T cells. b The relative cell viability of B16-GMCSF-OVA cells was shown. OT-1 T cells were pretreated with indicated concentrations of inosine or vehicle for 24 h, then co-cultured with B16-GMCSF-OVA tumour cells at a 2:1 E: T (Effector: T cells, Target: tumour cells) ratio for 48 h (n = 5). c The relative cell viability of 4T1 cells after incubation with activated CD8+ T cells pretreated with the indicated concentration of inosine at a 5:1 E: T ratio for 48 h (n = 5). d Experimental strategy to evaluate the ability of inosine to enhance T-cell-mediated tumour killing by modulating Tumour cells. e The relative cell viability of B16-GMCSF-OVA cells (left) and 4T1 cells (right) with Ctrl or inosine treatment following the method in a (n = 5). f Experimental strategy to evaluate the direct effect of inosine on tumour cells. g The relative cell viability (left) and apoptosis (right) of B16-GMCSF cells following inosine treatment at indicated concentrations for 48 h in vitro (n = 5). h The relative cell viability (left) and apoptosis (right) of 4T1 cells following inosine treatment at indicated concentrations for 48 h in vitro (n = 5). i Representative flow analysis (left panel) and quantifying (left panel) the intensity of cell surface MHC-I expression in B16-GMCSF cells treated with vehicle and inosine (100 μM) upon IFN-γ (20 ng/ml) treatment (n = 3). j, k Selective represented antigen processing/presentation and interferon-responsive gene expression in B16-GMCSF (j) and 4T1 (k) and tumour cells treated with inosine at indicated concentrations (n = 3). Data are presented as Mean ± s.e.m. Statistical significance was determined by Two-sided Student’s t test (i) and one-way ANOVA and Tukey test for multiple comparisons (b, c, e, g, h). NS no significant. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 3 Inosine directly inhibits tumour UBA6 to sensitise T-cell-mediated killing. a Scheme of chemical proteomics for target identification. b Volcano plots of LiP-SMap experiments with inosine treatment. c Heat-map of the top 23 proteins changed with inosine treatment identified by LiP-SMap. d The effect of deletion of the top 23 genes in e on OT-1 T-cell-mediated tumour cell killing. e The effect of inosine on the interaction between UBA6 and USE1 in HEK293 cells. Data were representative of two independent experiments (n = 2). f The effect of inosine on USE1~S~FAT10 thioester in vitro. Data were representative of two independent experiments (n = 2). g The relative cell viability of WT (sgCtrl), Uba6-null (sgUba6) 4T1, or B16-GMCSF-OVA tumour cells with vehicle (Ctrl) or inosine (100 µM) following the method in a (n = 5). Data are presented as Mean ± s.e.m. Statistical significance was determined by one-way ANOVA and Tukey test for multiple comparisons (g). The P values in b and d were based on log-rank tests. NS no significant. Source data are provided as a Source Data file.
Fig. 3 Inosine directly inhibits tumour UBA6 to sensitise T-cell-mediated killing. a Scheme of chemical proteomics for target identification. b Volcano plots of LiP-SMap experiments with inosine treatment. c Heat-map of the top 23 proteins changed with inosine treatment identified by LiP-SMap. d The effect of deletion of the top 23 genes in e on OT-1 T-cell-mediated tumour cell killing. e The effect of inosine on the interaction between UBA6 and USE1 in HEK293 cells. Data were representative of two independent experiments (n = 2). f The effect of inosine on USE1~S~FAT10 thioester in vitro. Data were representative of two independent experiments (n = 2). g The relative cell viability of WT (sgCtrl), Uba6-null (sgUba6) 4T1, or B16-GMCSF-OVA tumour cells with vehicle (Ctrl) or inosine (100 µM) following the method in a (n = 5). Data are presented as Mean ± s.e.m. Statistical significance was determined by one-way ANOVA and Tukey test for multiple comparisons (g). The P values in b and d were based on log-rank tests. NS no significant. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 4 Inosine and genetic inhibition of UBA6 stimulate tumour cell-intrinsic immune response signalling. a Volcano plots of upstream regulator analysis of UBA6-dependent genes in 4T1 tumour cells by IPA. b Top-ranked GO terms in the transcriptome of Uba6-null 4T1 tumour cells. c The upregulated GSEA signatures in the Uba6-null 4T1 tumour. d Heat-map of proteins for differential signalling pathways (Red upregulated, blue downregulated). The expression level of these proteins in sgCtrl and sgUba6 4T1 cells is measured by the whole proteomics (n = 3). e The relative mRNA expression of IFNγ response-related genes in sgCtrl and sgUba6 4T1 tumour cells without or with inosine (100 µM) treatment for 48 h (n = 3). f Cell viability of Uba6-null and control 4T1 or B16-GMCSF tumour cells following stimulation with 10 ng/ml TNFα + 10 ng/ml IFNγ treatment for 48 h (n = 4). Data are presented as Mean ± s.e.m. Statistical significance was determined by two-sided Student’s t test (e, f). The P values in a and b were based on log-rank tests. NS no significant. Raw RNA-seq data is available in the GEO database with accession number GSE210225. For the remaining data, source data are provided in a Source Data file.
Fig. 4 Inosine and genetic inhibition of UBA6 stimulate tumour cell-intrinsic immune response signalling. a Volcano plots of upstream regulator analysis of UBA6-dependent genes in 4T1 tumour cells by IPA. b Top-ranked GO terms in the transcriptome of Uba6-null 4T1 tumour cells. c The upregulated GSEA signatures in the Uba6-null 4T1 tumour. d Heat-map of proteins for differential signalling pathways (Red upregulated, blue downregulated). The expression level of these proteins in sgCtrl and sgUba6 4T1 cells is measured by the whole proteomics (n = 3). e The relative mRNA expression of IFNγ response-related genes in sgCtrl and sgUba6 4T1 tumour cells without or with inosine (100 µM) treatment for 48 h (n = 3). f Cell viability of Uba6-null and control 4T1 or B16-GMCSF tumour cells following stimulation with 10 ng/ml TNFα + 10 ng/ml IFNγ treatment for 48 h (n = 4). Data are presented as Mean ± s.e.m. Statistical significance was determined by two-sided Student’s t test (e, f). The P values in a and b were based on log-rank tests. NS no significant. Raw RNA-seq data is available in the GEO database with accession number GSE210225. For the remaining data, source data are provided in a Source Data file.CC BY · PubMed Central
Fig. 5 UBA6 deletion substitutes the effect of inosine on antitumour immunity in vivo. a Tumour volume and survival analysis of sgCtrl and sgUba6 B16-GMCSF tumours in NSG, wild-type (WT) mice (n = 8). b Tumour volume and survival analysis of sgCtrl and sgUba6 B16-GMCSF tumours in WT mice with Ctrl, ICB, or ICB + Ino treatment (n = 8). c Tumour volume and survival analysis of sgCtrl and sgUba6 4T1 tumour-bearing NSG mice (left) or wild-type (WT) mice (right) (n = 10). d Tumour volume and survival analysis of sgCtrl and sgUba6 4T1 tumour-bearing WT mice with Ctrl, ICB, or ICB + Ino treatment (n = 10). Data are presented as mean ± SEM. Statistical significance was determined by ANOVA (tumour volume of a–d) or log-rank (Mantel–Cox) test (survival analysis of a–d). NS no significant. Source data are provided as a Source Data file.
Fig. 5 UBA6 deletion substitutes the effect of inosine on antitumour immunity in vivo. a Tumour volume and survival analysis of sgCtrl and sgUba6 B16-GMCSF tumours in NSG, wild-type (WT) mice (n = 8). b Tumour volume and survival analysis of sgCtrl and sgUba6 B16-GMCSF tumours in WT mice with Ctrl, ICB, or ICB + Ino treatment (n = 8). c Tumour volume and survival analysis of sgCtrl and sgUba6 4T1 tumour-bearing NSG mice (left) or wild-type (WT) mice (right) (n = 10). d Tumour volume and survival analysis of sgCtrl and sgUba6 4T1 tumour-bearing WT mice with Ctrl, ICB, or ICB + Ino treatment (n = 10). Data are presented as mean ± SEM. Statistical significance was determined by ANOVA (tumour volume of a–d) or log-rank (Mantel–Cox) test (survival analysis of a–d). NS no significant. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 6 UBA6 in tumour cells predicts patient outcomes to immunotherapy. a The association between CTL and OS of melanoma patients with distinct UBA6 levels. b Kaplan–Meier plots of PFS of melanoma patients with anti-CTLA4 (n = 42) and OS of melanoma patients with anti-PD1 (n = 47) based on UBA6 level. c Representative UBA6 protein staining tumour sections (top: 100×, bottom: 400×) (left) and CT scans (right) of lung cancer patients with anti-PD1 treatment. CT scans of tumours (top) and mediastinal lymph nodes (bottom) of patient 1 and left pleural effusion (top) and pericardial effusion (bottom) of patient 2 are highlighted by red arrows. Data were representative of three independent experiments (n = 3). d Waterfall plot depicting the responses to anti-PD1 treatment by the best change in the sum of target lesions, in comparison to baseline, in cancer patients with low UBA6 (n = 16) or high UBA expression (n = 6). Every bar represents one patient and the colours correspond to response to anti-PD1 treatment (PR partial response, SD stable disease, PD progressive disease). Dotted black lines indicate the response as described by RECIST1.1. e Pie charts of response fractions for each group of patients with UBA6-low and UBA6-high expression in tumour cells. Data are presented as Mean ± s.e.m. Statistical significance was determined by a log-rank test (a, b). Source data are provided as a Source Data file.
Fig. 6 UBA6 in tumour cells predicts patient outcomes to immunotherapy. a The association between CTL and OS of melanoma patients with distinct UBA6 levels. b Kaplan–Meier plots of PFS of melanoma patients with anti-CTLA4 (n = 42) and OS of melanoma patients with anti-PD1 (n = 47) based on UBA6 level. c Representative UBA6 protein staining tumour sections (top: 100×, bottom: 400×) (left) and CT scans (right) of lung cancer patients with anti-PD1 treatment. CT scans of tumours (top) and mediastinal lymph nodes (bottom) of patient 1 and left pleural effusion (top) and pericardial effusion (bottom) of patient 2 are highlighted by red arrows. Data were representative of three independent experiments (n = 3). d Waterfall plot depicting the responses to anti-PD1 treatment by the best change in the sum of target lesions, in comparison to baseline, in cancer patients with low UBA6 (n = 16) or high UBA expression (n = 6). Every bar represents one patient and the colours correspond to response to anti-PD1 treatment (PR partial response, SD stable disease, PD progressive disease). Dotted black lines indicate the response as described by RECIST1.1. e Pie charts of response fractions for each group of patients with UBA6-low and UBA6-high expression in tumour cells. Data are presented as Mean ± s.e.m. Statistical significance was determined by a log-rank test (a, b). Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 7 A summary model linking inosine and immunotherapy responses. The metabolic imbalance, especially inosine, is associated with immunotherapy responses in mice and humans. Inosine overcomes tumour-intrinsic resistance to immunotherapy by inhibiting UBA6 and increasing tumour immunogenicity in tumours with UBA6-high expression.
Fig. 7 A summary model linking inosine and immunotherapy responses. The metabolic imbalance, especially inosine, is associated with immunotherapy responses in mice and humans. Inosine overcomes tumour-intrinsic resistance to immunotherapy by inhibiting UBA6 and increasing tumour immunogenicity in tumours with UBA6-high expression.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 53 references

Cited by

See all 82

Metadata

Authors
He, B., Zhang, L., Jiang, L., Yu, L., Li, Q., Tian, X., He, J.-Q., Zeng, L., Yang, Y., Wang, C., Wei, Y., Jiang, X., Li, J., Ge, X., Gu, Q., Li, J.-K., Wu, D., Sadler, A., Yu, D., Xu, D., Gao, Y., & Yuan, X.
Affiliations
  • Guangzhou University of Chinese Medicine (CN)
  • Shanghai Jiao Tong University (CN)
  • Capital Medical University (CN)
  • The University of Texas MD Anderson Cancer Center (US)
  • Peking University
  • First People's Hospital Affiliated to Shanghai Jiaotong University
  • University of North Carolina at Chapel Hill (US)
  • Hudson Institute of Medical Research (AU)
  • The University of Queensland (AU)
  • Tianjin University of Traditional Chinese Medicine
Corresponding author
  • He, Baokun
  • Yu, Liang
  • Gao, Yue
  • Yuan, Xiangliang
Venue
Nature Communications
Published
2021-07-12
Volume
13
Issue
1
Pages
5413-5413
ISSN
2041-1723 (ISSN-L)
OpenAlex
W4296038319
Cited by
82
References
49
Citations by year
2026: 19 · 2025: 29 · 2024: 19 · 2023: 14 · 2022: 1
FWCI
6.58
Type
journal article
Language
en
Open access
Locations
  • Nature Communications (publishedVersion, cc-by)
  • Research Square (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Digital Commons-TMC (Texas Medical Center) (submittedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Natural Science Foundation of China (81402148, 81871274, 2019YFC1604900, 81974256, 81630102, 81772525)
  • National Key Research and Development Program of China (2019YFC1604900)
Topics
  • Health Sciences › Medicine › Oncology › Cancer Immunotherapy and Biomarkers
  • Life Sciences › Immunology and Microbiology › Immunology › Immunotherapy and Immune Responses
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Physiology › Adenosine and Purinergic Signaling
Keywords
inosine, UBA6, immune checkpoint blockade, tumor immunogenicity, melanoma, breast cancer, purine metabolism, anti-PD-1 therapy, CTLA-4 blockade, T cell cytotoxicity, tumor microenvironment, cancer metabolism, immunotherapy sensitivity
MeSH terms
Immune Checkpoint Inhibitors, Animals, Combined Modality Therapy, Humans, Inosine, Melanoma, Radioimmunotherapy, Ubiquitin-Activating Enzymes, Mice, Tumor Microenvironment

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: Inhibiting tumor cell-intrinsic UBA6 by inosine augments tumor immunogenicity. https://app.cochat.ai/papers/10.21203/rs.3.rs-620390/v1
The paper
He, B., Zhang, L., Jiang, L., Yu, L., Li, Q., Tian, X., He, J.-Q., Zeng, L., Yang, Y., Wang, C., Wei, Y., Jiang, X., Li, J., Ge, X., Gu, Q., Li, J.-K., Wu, D., Sadler, A., Yu, D., . . . Yuan, X. (2021). Inhibiting tumor cell-intrinsic UBA6 by inosine augments tumor immunogenicity. Nature Communications. https://doi.org/10.21203/rs.3.rs-620390/v1

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.