CoChat

3D microfluidic liver cultures as a physiological preclinical tool for hepatitis B virus infection

Open access
Ortega-Prieto, A. M., Skelton, J. K., Wai, S. N., Large, E., Lussignol, M., Vizcay-Barrena, G., Hughes, D., Fleck, R. A., Thursz, M., Catanese, M. T., Dorner, M. · 2018 · Nature Communications

The DOI matches its Crossref record; 56 of the 56 references it cites were found in a registry or an index.

Abstract

With more than 240 million people infected, hepatitis B virus (HBV) is a major health concern. The inability to mimic the complexity of the liver using cell lines and regular primary human hepatocyte (PHH) cultures pose significant limitations for studying host/pathogen interactions. Here, we describe a 3D microfluidic PHH system permissive to HBV infection, which can be maintained for at least 40 days. This system enables the recapitulation of all steps of the HBV life cycle, including the replication of patient-derived HBV and the maintenance of HBV cccDNA. We show that innate immune and cytokine responses following infection with HBV mimic those observed in HBV-infected patients, thus allowing the dissection of pathways important for immune evasion and validation of biomarkers. Additionally, we demonstrate that the co-culture of PHH with other non-parenchymal cells enables the identification of the cellular origin of immune effectors, thus providing a valuable preclinical platform for HBV research.

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/s41467-018-02969-8

Cites

56 of 56 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 3D PHH cultures form physiological hepatic microtissues for extended periods of time. a Schematic of the perfused bioreactor, in which media is recirculated via a pneumatically driven micro-pump, and the collagen-coated scaffold for cell adherence. Each plate consists of 12 individual bioreactors, in which media flow can be regulated. b Cell viability of PHH following seeding in 3D scaffolds. Viable cells are stained green with calcein AM while dead cells are stained red with ethidium homodimer-1 3 days post-seeding. c Kinetic of hepatic microtissue formation and comparison to 3D spheroid, static 2D PHH, and SACC PHH hepatocyte morphologies. d Immunofluorescence microscopy of albumin (green) and DAPI (blue) in 2D PHH, 3D spheroid, SACC PHH, and 3D PHH cultures after 14 days of culture. e Secretion of albumin from 2D PHH, 3D spheroid, SACC PHH, and 3D PHH cultures after 14 days of culture as determined by ELISA. All data shown are mean ± SD of three to six independent experiments. p-values calculated by two-way ANOVA with Bonferroni post-test. Scale bars: white (200 μm), gray (500 μm)
Fig. 1 3D PHH cultures form physiological hepatic microtissues for extended periods of time. a Schematic of the perfused bioreactor, in which media is recirculated via a pneumatically driven micro-pump, and the collagen-coated scaffold for cell adherence. Each plate consists of 12 individual bioreactors, in which media flow can be regulated. b Cell viability of PHH following seeding in 3D scaffolds. Viable cells are stained green with calcein AM while dead cells are stained red with ethidium homodimer-1 3 days post-seeding. c Kinetic of hepatic microtissue formation and comparison to 3D spheroid, static 2D PHH, and SACC PHH hepatocyte morphologies. d Immunofluorescence microscopy of albumin (green) and DAPI (blue) in 2D PHH, 3D spheroid, SACC PHH, and 3D PHH cultures after 14 days of culture. e Secretion of albumin from 2D PHH, 3D spheroid, SACC PHH, and 3D PHH cultures after 14 days of culture as determined by ELISA. All data shown are mean ± SD of three to six independent experiments. p-values calculated by two-way ANOVA with Bonferroni post-test. Scale bars: white (200 μm), gray (500 μm)CC BY · PubMed Central
Fig. 2 3D PHH cultures are metabolically stable and exhibit functional tissue formation. a Comparison of cytochrome P450 gene (HNMT, CYP2A6, SLC22A1, CYP2A13, and GSTP1) mRNA expression between static 2D PHH cultures, 3D spheroid cultures, SACC PHH cultures, and 3D PHH cultures following 14 days of culture as well as freshly thawed PHH and HepG2 cells. b Longitudinal comparison of the mRNA expression profile of HNMT, CYP2A6, ABCB11, SLC22A17, CYP2A13, and GSTP1 as well as HBV receptors (NTCP, ASGPR) in HepG2 cells and freshly thawed PHH as well as static 2D and 3D PHH cultures. c Determination of CYP3A activity in six different hepatocyte donors 7 days following seeding in 3D PHH cultures as determined by CYP3A-Glo assay. d CYP450 activity in 3D PHH following 7, 14, or 21 days of culture time as determined by quantification of metabolites of Tacrine (CYP1A2), Diclofenac (CYP2C9), and Midazolam (CYP3A4) by quadrupole linear ion trap mass spectrometry. e Stable maintenance of the bile canaliculi marker DPPIV/CD26 in 3D PHH cultures as determined by Luminex. f Functionality of bile canaliculi as determined by CDFDA staining. g, h Ultrastructural analysis of 3D PHH by transmission electron microscopy after 20 days of culture. (Filled arrows: bile canaliculi, open arrows: hepatic microvilli, asterisk: tight junctions, L: lipid droplet). i Immunofluorescence microscopy of tight junction and hepatocyte markers (ZO-1, CD81, ITGB1, Connexin 32) in 3D PHH cultures following 26 days of culture. All data shown are mean ± SD of three to six independent experiments. Scale bars: white (200 μm), gray (2 nm), black (500 nm)
Fig. 2 3D PHH cultures are metabolically stable and exhibit functional tissue formation. a Comparison of cytochrome P450 gene (HNMT, CYP2A6, SLC22A1, CYP2A13, and GSTP1) mRNA expression between static 2D PHH cultures, 3D spheroid cultures, SACC PHH cultures, and 3D PHH cultures following 14 days of culture as well as freshly thawed PHH and HepG2 cells. b Longitudinal comparison of the mRNA expression profile of HNMT, CYP2A6, ABCB11, SLC22A17, CYP2A13, and GSTP1 as well as HBV receptors (NTCP, ASGPR) in HepG2 cells and freshly thawed PHH as well as static 2D and 3D PHH cultures. c Determination of CYP3A activity in six different hepatocyte donors 7 days following seeding in 3D PHH cultures as determined by CYP3A-Glo assay. d CYP450 activity in 3D PHH following 7, 14, or 21 days of culture time as determined by quantification of metabolites of Tacrine (CYP1A2), Diclofenac (CYP2C9), and Midazolam (CYP3A4) by quadrupole linear ion trap mass spectrometry. e Stable maintenance of the bile canaliculi marker DPPIV/CD26 in 3D PHH cultures as determined by Luminex. f Functionality of bile canaliculi as determined by CDFDA staining. g, h Ultrastructural analysis of 3D PHH by transmission electron microscopy after 20 days of culture. (Filled arrows: bile canaliculi, open arrows: hepatic microvilli, asterisk: tight junctions, L: lipid droplet). i Immunofluorescence microscopy of tight junction and hepatocyte markers (ZO-1, CD81, ITGB1, Connexin 32) in 3D PHH cultures following 26 days of culture. All data shown are mean ± SD of three to six independent experiments. Scale bars: white (200 μm), gray (2 nm), black (500 nm)CC BY · PubMed Central
Fig. 3 3D PHH cultures are susceptible to long-term HBV infection. a, b Immunofluorescence detection of a HBsAg and HBcAg 10 days following infection of 3D cultures with patient-derived HBV (100 GE/cell) and b HBcAg in 3D PHH cultures, 3D spheroid cultures, and SACC PHH cultures 12 days following infection with HepDE19-derived sucrose-purified HBV (500 GE/cell). c Susceptibility of 3D PHH cultures compared to 3D spheroid cultures (left) and SACC PHH cultures (right) following infection with HepDE19-derived, sucrose-purified HBV at the indicated MOI via detection of secreted HBV DNA. d Longitudinal cumulative HBV DNA and HBsAg secretion following infection of 3D PHH cultures with the indicated MOI of sucrose gradient-purified HepDE19-derived HBV. e Blocking of HBV entry in 3D PHH cultures by pretreatment of cells with 5 U/mL heparinase III or myristoylated preS1 peptide (aa2-48, 1 μM) for 1 h prior to infection with 500 GE/cell sucrose-purified HepDE19-derived HBV. HBsAg secretion was determined 6 days postinfection. f–h Response of HBV-infected 3D PHH cultures to treatment with recombinant IFNα (1000 IU/mL) as determined by quantification of f cccDNA copies per cell using quantitative PCR of Hirt-extracted DNA, secretion of g HBV DNA, and h HBsAg at 21 days postinfection. IFNα treatment was initiated 2 days postinfection and maintained throughout the culture period. i Longitudinal cumulative HBsAg and HBV DNA secretion following infection of 3D PHH cultures with 200 GE/cell of patient-derived HBV. j HBeAg and HBV DNA secretion of HBV-infected 3D PHH cultures (100 GE/cell) in the presence of 1 μM Tenofovir alafenamide for 10 days. Treatment was initiated 2 days postinfection and was maintained throughout the culture period. All data shown are mean ± SD of three to six independent experiments. p-values calculated by two-way ANOVA with Bonferroni post-test. Scale bars: white (200 μm), gray (100 μm)
Fig. 3 3D PHH cultures are susceptible to long-term HBV infection. a, b Immunofluorescence detection of a HBsAg and HBcAg 10 days following infection of 3D cultures with patient-derived HBV (100 GE/cell) and b HBcAg in 3D PHH cultures, 3D spheroid cultures, and SACC PHH cultures 12 days following infection with HepDE19-derived sucrose-purified HBV (500 GE/cell). c Susceptibility of 3D PHH cultures compared to 3D spheroid cultures (left) and SACC PHH cultures (right) following infection with HepDE19-derived, sucrose-purified HBV at the indicated MOI via detection of secreted HBV DNA. d Longitudinal cumulative HBV DNA and HBsAg secretion following infection of 3D PHH cultures with the indicated MOI of sucrose gradient-purified HepDE19-derived HBV. e Blocking of HBV entry in 3D PHH cultures by pretreatment of cells with 5 U/mL heparinase III or myristoylated preS1 peptide (aa2-48, 1 μM) for 1 h prior to infection with 500 GE/cell sucrose-purified HepDE19-derived HBV. HBsAg secretion was determined 6 days postinfection. f–h Response of HBV-infected 3D PHH cultures to treatment with recombinant IFNα (1000 IU/mL) as determined by quantification of f cccDNA copies per cell using quantitative PCR of Hirt-extracted DNA, secretion of g HBV DNA, and h HBsAg at 21 days postinfection. IFNα treatment was initiated 2 days postinfection and maintained throughout the culture period. i Longitudinal cumulative HBsAg and HBV DNA secretion following infection of 3D PHH cultures with 200 GE/cell of patient-derived HBV. j HBeAg and HBV DNA secretion of HBV-infected 3D PHH cultures (100 GE/cell) in the presence of 1 μM Tenofovir alafenamide for 10 days. Treatment was initiated 2 days postinfection and was maintained throughout the culture period. All data shown are mean ± SD of three to six independent experiments. p-values calculated by two-way ANOVA with Bonferroni post-test. Scale bars: white (200 μm), gray (100 μm)CC BY · PubMed Central
Fig. 4 HBV subverts hepatocyte-intrinsic innate immune responses. a, b Analysis of the induction of p38, ERK1/2, JNK1/2/3, STAT2, and HSP60 phosphorylation following stimulation of uninfected or patient-derived HBV-infected 3D PHH cultures (100 GE/cell) with 1 μM myristoylated preS1 peptide for 24 h using phospho-kinase arrays. Cells were treated/infected 3 days post-seeding of PHH and phosphorylation was determined 24 h later. c–f Transcriptional profile of c type I and e III IFN and d, f ISG within the first 48 h after infection. g–k mRNA expression of g TLR1, h TLR2, i TLR8, j DDX58/RIG-I, and k IRF3 in 3D PHH cultures, either left uninfected, infected with 100 GE/cell patient-derived HBV, or HBV-infected cultures treated with 1000 IU/mL IFNα at 10 days postinfection. l Expression of ISG in 3D PHH cultures infected with patient-derived HBV (100 GE/cell), either untreated or treated with 1000 IU/mL IFNα at 10 days postinfection. m–o Inactivation of patient-derived HBV by 4% paraformaldehyde for 1 h, heat treatment at 98 degrees for 2 min, or exposure to UV irradiation for 30 min and resulting innate immune activation as measured by m HBsAg secretion 10 days after culture initiation and qPCR of n IFI27 and o IFITM3. All data shown are mean ± SD of three to six independent experiments (*p < 0.05, **p < 0.01, ***p < 0.001; two-way ANOVA with Bonferroni post-test)
Fig. 4 HBV subverts hepatocyte-intrinsic innate immune responses. a, b Analysis of the induction of p38, ERK1/2, JNK1/2/3, STAT2, and HSP60 phosphorylation following stimulation of uninfected or patient-derived HBV-infected 3D PHH cultures (100 GE/cell) with 1 μM myristoylated preS1 peptide for 24 h using phospho-kinase arrays. Cells were treated/infected 3 days post-seeding of PHH and phosphorylation was determined 24 h later. c–f Transcriptional profile of c type I and e III IFN and d, f ISG within the first 48 h after infection. g–k mRNA expression of g TLR1, h TLR2, i TLR8, j DDX58/RIG-I, and k IRF3 in 3D PHH cultures, either left uninfected, infected with 100 GE/cell patient-derived HBV, or HBV-infected cultures treated with 1000 IU/mL IFNα at 10 days postinfection. l Expression of ISG in 3D PHH cultures infected with patient-derived HBV (100 GE/cell), either untreated or treated with 1000 IU/mL IFNα at 10 days postinfection. m–o Inactivation of patient-derived HBV by 4% paraformaldehyde for 1 h, heat treatment at 98 degrees for 2 min, or exposure to UV irradiation for 30 min and resulting innate immune activation as measured by m HBsAg secretion 10 days after culture initiation and qPCR of n IFI27 and o IFITM3. All data shown are mean ± SD of three to six independent experiments (*p < 0.05, **p < 0.01, ***p < 0.001; two-way ANOVA with Bonferroni post-test)CC BY · PubMed Central
Fig. 5 HBV-infected 3D PHH cultures predict human host responses. a Cytokine XL antibody array of uninfected 3D PHH cultures and cultures infected with 100 GE/cell patient-derived HBV for 10 days. b–g Comparison of the longitudinal protein expression values of b IL-8, c MIP-3α, d CXCL10, e VEGF, f Serpin E1, and g MCP-1 between HBV-infected patients, healthy controls, and 3D PHH cultures infected with patient-derived HBV (100 GE/cell) as well as correlation of each cytokine with HBsAg (HBV-infected patients and infected 3D PHH cultures) and HBeAg (infected 3D PHH cultures) levels. Data shown are mean ± SD of three independent experiments. p-values calculated by two-way ANOVA with Bonferroni post-test
Fig. 5 HBV-infected 3D PHH cultures predict human host responses. a Cytokine XL antibody array of uninfected 3D PHH cultures and cultures infected with 100 GE/cell patient-derived HBV for 10 days. b–g Comparison of the longitudinal protein expression values of b IL-8, c MIP-3α, d CXCL10, e VEGF, f Serpin E1, and g MCP-1 between HBV-infected patients, healthy controls, and 3D PHH cultures infected with patient-derived HBV (100 GE/cell) as well as correlation of each cytokine with HBsAg (HBV-infected patients and infected 3D PHH cultures) and HBeAg (infected 3D PHH cultures) levels. Data shown are mean ± SD of three independent experiments. p-values calculated by two-way ANOVA with Bonferroni post-testCC BY · PubMed Central
Fig. 6 Kupffer cells do not alter HBV infection kinetics in 3D PHH cultures. a Cytokine XL antibody array of 3D PHH and 3D PHH/KC co-cultures (ratio 10:1) 13 days post-seeding. b Longitudinal osteopontin protein levels in 3D PHH and 3D PHH/KC co-cultures, determined by Luminex. c, d Response of 3D PHH and 3D PHH/KC co-cultures to 1 μg/mL LPS added 11 days post-seeding as determined by measurement of c IL-6 and d TNF-α 24 h post stimulation. e Acute phase response elicited by KC following infection with patient-derived HBV (100 GE/cell), determined by measurement of C-reactive protein using Luminex. f–k Comparative infection of f, h, i, k PHH and g, h, j, k PHH/KC co-cultures with patient-derived HBV (100 GE/cell) as determined by cumulative secretion of f, g HBsAg and i, j HBeAg and intracellular accumulation of h subgenomic HBV RNA and k pregenomic RNA. l, m Transcription of l type I and III interferon and m ISG following 10 days of infection of PHH or PHH/KC co-cultures with patient-derived HBV (100 GE/cell) in the absence or presence of 1000 IU/mL exogenous IFNα. Data shown are mean ± SD of four independent experiments
Fig. 6 Kupffer cells do not alter HBV infection kinetics in 3D PHH cultures. a Cytokine XL antibody array of 3D PHH and 3D PHH/KC co-cultures (ratio 10:1) 13 days post-seeding. b Longitudinal osteopontin protein levels in 3D PHH and 3D PHH/KC co-cultures, determined by Luminex. c, d Response of 3D PHH and 3D PHH/KC co-cultures to 1 μg/mL LPS added 11 days post-seeding as determined by measurement of c IL-6 and d TNF-α 24 h post stimulation. e Acute phase response elicited by KC following infection with patient-derived HBV (100 GE/cell), determined by measurement of C-reactive protein using Luminex. f–k Comparative infection of f, h, i, k PHH and g, h, j, k PHH/KC co-cultures with patient-derived HBV (100 GE/cell) as determined by cumulative secretion of f, g HBsAg and i, j HBeAg and intracellular accumulation of h subgenomic HBV RNA and k pregenomic RNA. l, m Transcription of l type I and III interferon and m ISG following 10 days of infection of PHH or PHH/KC co-cultures with patient-derived HBV (100 GE/cell) in the absence or presence of 1000 IU/mL exogenous IFNα. Data shown are mean ± SD of four independent experimentsCC BY · PubMed Central
Fig. 7 Exogenous signals are required for Kupffer cell responses to HBV infection. a Longitudinal secretion of TNF-α and IL-6 in 3D PHH cultures infected with patient-derived HBV (100 GE/cell). b, c Serum levels of b TNF-α and c IL-6 as well as correlation with serum HBsAg levels of HBV-infected patients and healthy controls. d–f Secretion of d TNF-α and e IL-6 as well as f HBsAg in 3D PHH cultures and 3D PHH/KC co-cultures infected with patient-derived HBV (100 GE/cell). Cultures were treated with 1 μg/mL LPS 8 days postinfection, TNF-α and IL-6 levels were quantified by Luminex 24 h post LPS stimulation, and HBsAg secretion was determined 48 h post LPS stimulation. Data shown are mean ± SD of three independent experiments (*p < 0.05, ***p < 0.001; two-way ANOVA with Bonferroni post-test)
Fig. 7 Exogenous signals are required for Kupffer cell responses to HBV infection. a Longitudinal secretion of TNF-α and IL-6 in 3D PHH cultures infected with patient-derived HBV (100 GE/cell). b, c Serum levels of b TNF-α and c IL-6 as well as correlation with serum HBsAg levels of HBV-infected patients and healthy controls. d–f Secretion of d TNF-α and e IL-6 as well as f HBsAg in 3D PHH cultures and 3D PHH/KC co-cultures infected with patient-derived HBV (100 GE/cell). Cultures were treated with 1 μg/mL LPS 8 days postinfection, TNF-α and IL-6 levels were quantified by Luminex 24 h post LPS stimulation, and HBsAg secretion was determined 48 h post LPS stimulation. Data shown are mean ± SD of three independent experiments (*p < 0.05, ***p < 0.001; two-way ANOVA with Bonferroni post-test)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 56 references

Cited by

See all 236

Metadata

Authors
Ortega-Prieto, A. M., Skelton, J. K., Wai, S. N., Large, E., Lussignol, M., Vizcay-Barrena, G., Hughes, D., Fleck, R. A., Thursz, M., Catanese, M. T., & Dorner, M.
Affiliations
  • Imperial College London (GB)
  • CN Bio Innovations (United Kingdom) (GB)
  • King's College London (GB)
Corresponding author
Dorner, M.
Venue
Nature Communications
Published
2018-02-14
Volume
9
Issue
1
Pages
682-682
ISSN
2041-1723 (ISSN-L)
OpenAlex
W2782129578
Cited by
236
References
56
Citations by year
2026: 12 · 2025: 15 · 2024: 30 · 2023: 26 · 2022: 29 · 2021: 51 · 2020: 35 · 2019: 27 · 2018: 11
FWCI
24.21
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Research Portal (King's College London) (submittedVersion, cc-by)
  • Repositorio Institucional de la Universidad de Málaga (University of Málaga) (submittedVersion, cc-by)
  • Spiral (Imperial College London) (submittedVersion, cc-by)
Funders
  • Wellcome Trust (104771, 104771/Z/14/Z)
  • National Institute for Health and Care Research
  • NIHR Imperial Biomedical Research Centre
  • H2020 European Research Council (637304)
Topics
  • Health Sciences › Medicine › Epidemiology › Hepatitis B Virus Studies
  • Health Sciences › Medicine › Hepatology › Liver physiology and pathology
  • Physical Sciences › Engineering › Biomedical Engineering › 3D Printing in Biomedical Research
Keywords
hepatitis B virus, primary human hepatocytes, HBV life cycle, HBV cccDNA, innate immune response, cytokine response, immune evasion, biomarkers, non-parenchymal cells, co-culture, host-pathogen interaction, preclinical models
MeSH terms
Adult, Aged, Animals, Cells, Cultured, Female, Hepatitis B, Hepatitis B virus, Humans, Infant, Kupffer Cells, Liver, Male, Middle Aged, Virus Replication, Reproducibility of Results, Coculture Techniques, Hepatocytes, NIH 3T3 Cells, Microfluidics, Cell Line, Tumor, Mice, Host-Pathogen Interactions, Hep G2 Cells

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: 3D microfluidic liver cultures as a physiological preclinical tool for hepatitis B virus infection. https://app.cochat.ai/papers/10.1038/s41467-018-02969-8
The paper
Ortega-Prieto, A. M., Skelton, J. K., Wai, S. N., Large, E., Lussignol, M., Vizcay-Barrena, G., Hughes, D., Fleck, R. A., Thursz, M., Catanese, M. T., & Dorner, M. (2018). 3D microfluidic liver cultures as a physiological preclinical tool for hepatitis B virus infection. Nature Communications. https://doi.org/10.1038/s41467-018-02969-8

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.