Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index.

Lysosomal trapping of palbociclib and its functional implications

Open access
Llanos, Susana, Megias, Diego, Blanco-Aparicio, Carmen, Hernández-Encinas, Elena, Rovira, Miguel, Pietrocola, Federico, Serrano, Manuel · 2019 · Oncogene

Abstract

Palbociclib is a selective inhibitor of cyclin-dependent kinases 4 and 6 (CDK4/6) approved for the treatment of some cancers. The main mechanism of action of palbociclib is to induce cell cycle arrest and senescence on responsive cells. Here, we report that palbociclib concentrates in intracellular acidic vesicles, where it can be readily observed due to its intrinsic fluorescence, and it is released from these vesicles upon dilution or washing out of the extracellular medium. This reversible storage of drugs into acidic vesicles is generally known as lysosomal trapping and, based on this, we uncover novel properties of palbociclib. In particular, a short exposure of cells to palbociclib is sufficient to produce a stable cell-cycle arrest and long-term senescence. Moreover, after washing out the drug, palbociclib-treated cells release the drug to the medium and this conditioned medium is active on susceptible cells. Interestingly, cancer cells resistant to palbociclib also accumulate and release the drug producing paracrine senescence on susceptible cells. Finally, other lysosomotropic drugs, such as chloroquine, interfere with the accumulation of palbociclib into lysosomes, thereby reducing the minimal dose of palbociclib required for cell-cycle arrest and senescence. In summary, lysosomal trapping explains the prolonged temporal activity of palbociclib, the paracrine activity of exposed cells, and the cooperation with lysosomotropic drugs. These are important features that may help to improve the therapeutic dosing and efficacy of palbociclib. Finally, two other clinically approved CDK4/6 inhibitors, ribociclib and abemaciclib, present a similar behavior as palbociclib, suggesting that lysosomal trapping is a property common to all three clinically-approved CDK4/6 inhibitors.

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/s41388-019-0695-8

Cites

50 of 50 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Palbociclib is reversibly stored into acidic vesicles. a (Top images) SA-βgal staining of SK-Mel-103 cells treated with different senescence-inducing stimuli (bleomycin 12 mUnits/ml, doxorubicin 10 nM, nutlin 10 µM, palbociclib 1 µM) for 7 days. The drugs were added only once and the media were not changed for the length of the treatment. (Bottom images) Confocal images of live control and senescent cells (excitation 405 mm/emission 500–550 nm). b Representative flow cytometry histogram plots of the indicated cells after treatment with bleomycin (12 mUnits/ml), doxorubicin (10 nM for cancer cells or 250 nM for IMR90 cells), and palbociclib (5 μM) during 7 days. Cells were excited with a 405 nm violet laser and the intensity of the emitted fluorescence was detected at 550 nm. c Confocal images illustrating the co-localization of lysotracker red and palbociclib-fluorescence in SK-Mel-103 cells treated with palbociclib 1 µM for 7 days. To improve palbociclib detection cells were incubated with 4 µM palbociclib 1 h prior to image acquisition. d Analysis of palbociclib-fluorescence in live SK-Mel-103 cells treated with palbociclib 1 µM for 7 days and then seeded in flow chamber slides in the presence 4 µM palbociclib. 24 h after seeding, chamber slides were flowed with medium containing or not palbociclib 4 µM and pictures were taken every 10 min. The graph represents the average and standard deviation of palbociclib-fluorescence in 10 individual cells normalized to time 0
Fig. 1 Palbociclib is reversibly stored into acidic vesicles. a (Top images) SA-βgal staining of SK-Mel-103 cells treated with different senescence-inducing stimuli (bleomycin 12 mUnits/ml, doxorubicin 10 nM, nutlin 10 µM, palbociclib 1 µM) for 7 days. The drugs were added only once and the media were not changed for the length of the treatment. (Bottom images) Confocal images of live control and senescent cells (excitation 405 mm/emission 500–550 nm). b Representative flow cytometry histogram plots of the indicated cells after treatment with bleomycin (12 mUnits/ml), doxorubicin (10 nM for cancer cells or 250 nM for IMR90 cells), and palbociclib (5 μM) during 7 days. Cells were excited with a 405 nm violet laser and the intensity of the emitted fluorescence was detected at 550 nm. c Confocal images illustrating the co-localization of lysotracker red and palbociclib-fluorescence in SK-Mel-103 cells treated with palbociclib 1 µM for 7 days. To improve palbociclib detection cells were incubated with 4 µM palbociclib 1 h prior to image acquisition. d Analysis of palbociclib-fluorescence in live SK-Mel-103 cells treated with palbociclib 1 µM for 7 days and then seeded in flow chamber slides in the presence 4 µM palbociclib. 24 h after seeding, chamber slides were flowed with medium containing or not palbociclib 4 µM and pictures were taken every 10 min. The graph represents the average and standard deviation of palbociclib-fluorescence in 10 individual cells normalized to time 0CC BY · PubMed Central
Fig. 2 Short- and long-term effects of palbociclib on lysosomal function. a Confocal images of acridine orange-stained SK-Mel-103 after 1 h treatment with the indicated compounds (palbociclib 4 µM, chloroquine 50 µM, bafilomycin 40 nM). b Western blot depicting the levels of the autophagy marker p62 and the lysosomal marker LAMP-1 in SK-Mel-103 cells treated with the indicated concentrations of palbociclib for 24 h, or with the indicated compounds (palbociclib 1 µM, doxorubicin 10 nM, nutlin 10 µM) for 7 days. All the drugs were added once and the media were not changed for the duration of the treatment. Lysates from cells treated with 5 µM chloroquine for 48 h were included as control for autophagy inhibition. c Confocal images of acridine orange signal in control and palbociclib-treated SK-Mel-103 cells. d Palbociclib-fluorescence signal in non-senescent and senescent cells: SK-Mel-103 cells were treated for 7 days with the indicated senescence-inducing drugs (palbociclib 1 µM, bleomycin 12 mUnits/ml, doxorubicin 10 nM, nutlin 10 µM). The drugs were added only once and the culture media were not changed for the length of the treatment. Subsequently, control (non-senescent) and senescent cells were incubated in the absence (untreated) or presence of 4 μM palbociclib and lysotracker red for 1 h prior to confocal microscopy
Fig. 2 Short- and long-term effects of palbociclib on lysosomal function. a Confocal images of acridine orange-stained SK-Mel-103 after 1 h treatment with the indicated compounds (palbociclib 4 µM, chloroquine 50 µM, bafilomycin 40 nM). b Western blot depicting the levels of the autophagy marker p62 and the lysosomal marker LAMP-1 in SK-Mel-103 cells treated with the indicated concentrations of palbociclib for 24 h, or with the indicated compounds (palbociclib 1 µM, doxorubicin 10 nM, nutlin 10 µM) for 7 days. All the drugs were added once and the media were not changed for the duration of the treatment. Lysates from cells treated with 5 µM chloroquine for 48 h were included as control for autophagy inhibition. c Confocal images of acridine orange signal in control and palbociclib-treated SK-Mel-103 cells. d Palbociclib-fluorescence signal in non-senescent and senescent cells: SK-Mel-103 cells were treated for 7 days with the indicated senescence-inducing drugs (palbociclib 1 µM, bleomycin 12 mUnits/ml, doxorubicin 10 nM, nutlin 10 µM). The drugs were added only once and the culture media were not changed for the length of the treatment. Subsequently, control (non-senescent) and senescent cells were incubated in the absence (untreated) or presence of 4 μM palbociclib and lysotracker red for 1 h prior to confocal microscopyCC BY · PubMed Central
Fig. 3 Lysosomal trapping contributes to the long-term effects palbociclib. a, b SK-Mel-103 cells were treated with different concentrations of palbociclib for 24 h. Following palbociclib treatment, cells were washed with PBS and added fresh media lacking palbociclib. This procedure was repeated for the indicated number of times. Between successive changes of media, cultures were resting for 30 min. a The SA-βgal activity of the cultures was assessed 7 days after palbociclib withdrawal. b The percentage of senescent cells was quantified 7 days after palbociclib withdrawal by flow cytometry. Control and senescent cells were differentiated on the basis of their scatter patterns (FCS-A vs SSC-A) (see Figures S3b,c). A total of two independent experiments were performed, one is shown in this panel and the other one is shown with the FACS profiles in the Supplementary Information (Figure S3d). c SK-Mel-103 cells (1000 cells) were seeded in 35 mm dishes and treated with the indicated concentrations of palbociclib. 24 h later, cells were washed, as in a, for the specified number of times. Colonies were visualized by crystal violet staining 10 days after palbociclib withdrawal
Fig. 3 Lysosomal trapping contributes to the long-term effects palbociclib. a, b SK-Mel-103 cells were treated with different concentrations of palbociclib for 24 h. Following palbociclib treatment, cells were washed with PBS and added fresh media lacking palbociclib. This procedure was repeated for the indicated number of times. Between successive changes of media, cultures were resting for 30 min. a The SA-βgal activity of the cultures was assessed 7 days after palbociclib withdrawal. b The percentage of senescent cells was quantified 7 days after palbociclib withdrawal by flow cytometry. Control and senescent cells were differentiated on the basis of their scatter patterns (FCS-A vs SSC-A) (see Figures S3b,c). A total of two independent experiments were performed, one is shown in this panel and the other one is shown with the FACS profiles in the Supplementary Information (Figure S3d). c SK-Mel-103 cells (1000 cells) were seeded in 35 mm dishes and treated with the indicated concentrations of palbociclib. 24 h later, cells were washed, as in a, for the specified number of times. Colonies were visualized by crystal violet staining 10 days after palbociclib withdrawalCC BY · PubMed Central
Fig. 4 Paracrine senescence by release of lysosome-trapped palbociclib. a Senescent SK-Mel-103 cells (exposed to palbociclib 1 µM for 7 days) where counted and seeded in 10 cm dishes (600,000 cells per dish) in the presence of 4 µM palbociclib. 24 h later cells washed with PBS and added fresh media lacking palbociclib. This procedure was repeated every hour for 4 consecutive hours. Cell pellets (P) and conditioned media (M) were collected as illustrated in the diagram and the palbociclib content was assessed by mass spectrometry. Control pellets (P0) and media (M0) were collected prior to the first wash and residual media (R) was collected immediately after washing the dishes for the first time. The values correspond to a representative experiment from a total of two independent experiments. b Donor SK-Mel-103 cells were treated with the indicated senescence-inducing stimuli, namely, palbociclib (1 µM), bleomycin (12 mUnits/ml), doxorubicin (10 nM), nutlin (10 µM), and γ-irradiation (10 Gy). The different drugs were added only once and the culture media were not changed for the length of the treatment (7 days). To obtain conditioned media, the senescent cells were washed with PBS and incubated with fresh media for 3 additional days. The media were then collected, filtrated with 0.45 μM filters, diluted 1:2 with fresh media, and added to the recipient cells. Recipient SK-Mel-103 were treated with the conditioned media of donor cells for 7 days, with no media changes, and then stained for SA-βgal activity
Fig. 4 Paracrine senescence by release of lysosome-trapped palbociclib. a Senescent SK-Mel-103 cells (exposed to palbociclib 1 µM for 7 days) where counted and seeded in 10 cm dishes (600,000 cells per dish) in the presence of 4 µM palbociclib. 24 h later cells washed with PBS and added fresh media lacking palbociclib. This procedure was repeated every hour for 4 consecutive hours. Cell pellets (P) and conditioned media (M) were collected as illustrated in the diagram and the palbociclib content was assessed by mass spectrometry. Control pellets (P0) and media (M0) were collected prior to the first wash and residual media (R) was collected immediately after washing the dishes for the first time. The values correspond to a representative experiment from a total of two independent experiments. b Donor SK-Mel-103 cells were treated with the indicated senescence-inducing stimuli, namely, palbociclib (1 µM), bleomycin (12 mUnits/ml), doxorubicin (10 nM), nutlin (10 µM), and γ-irradiation (10 Gy). The different drugs were added only once and the culture media were not changed for the length of the treatment (7 days). To obtain conditioned media, the senescent cells were washed with PBS and incubated with fresh media for 3 additional days. The media were then collected, filtrated with 0.45 μM filters, diluted 1:2 with fresh media, and added to the recipient cells. Recipient SK-Mel-103 were treated with the conditioned media of donor cells for 7 days, with no media changes, and then stained for SA-βgal activityCC BY · PubMed Central
Fig. 5 Paracrine senescence can be mediated by palbociclib non-responsive donor cells. a Donor SK-Mel-103 and Saos2 cells were treated with palbociclib (4 µM), for 7 days. The drug was added only once and the culture media were not changed for the length of the treatment. To obtain conditioned media, SK-Mel-103 and Saos2 cells were tripsinized and identical numbers of both cell lines were seeded in new dishes in the presence of palbociclib. 24 h later the cells were washed with PBS and incubated with fresh media for 3 additional days. The media were then collected, filtrated with 0.45 μM filters, diluted 1:2 with fresh media, and added to the recipient cells. Recipient SK-Mel-103 were treated with the conditioned media for 7 days, with no media changes, and then stained for SA-βgal activity. b Mass spectrometry quantification of palbociclib concentration in the conditioned media of the indicated donor cells treated as in a. The graph represents the average values and SD of three independent experiments. c Donor palbociclib responsive (MCF7, T47D, and MDA-MB-415) and non-responsive (HCC1937 and MDA-MB-415) human breast cancer cell lines were treated with palbociclib 1 μM for 7 days. The drug was added only once and the culture media were not changed for the length of the treatment. Following the treatment, the cells were tripsinized and seeded in new dishes to be stained for SA-βgal activity (top images) and analyzed by confocal imaging to evaluate the presence of palbociclib (middle images). Prior to image acquisition cells were incubated with 4 µM palbociclib 1 h to improve palbociclib detection. In parallel, conditioned media from identical number of palbociclib-treated cells were obtained as in a and added to recipient SK-Mel-103 for 7 days, with no media changes, prior to assessing SA-βgal activity (bottom images)
Fig. 5 Paracrine senescence can be mediated by palbociclib non-responsive donor cells. a Donor SK-Mel-103 and Saos2 cells were treated with palbociclib (4 µM), for 7 days. The drug was added only once and the culture media were not changed for the length of the treatment. To obtain conditioned media, SK-Mel-103 and Saos2 cells were tripsinized and identical numbers of both cell lines were seeded in new dishes in the presence of palbociclib. 24 h later the cells were washed with PBS and incubated with fresh media for 3 additional days. The media were then collected, filtrated with 0.45 μM filters, diluted 1:2 with fresh media, and added to the recipient cells. Recipient SK-Mel-103 were treated with the conditioned media for 7 days, with no media changes, and then stained for SA-βgal activity. b Mass spectrometry quantification of palbociclib concentration in the conditioned media of the indicated donor cells treated as in a. The graph represents the average values and SD of three independent experiments. c Donor palbociclib responsive (MCF7, T47D, and MDA-MB-415) and non-responsive (HCC1937 and MDA-MB-415) human breast cancer cell lines were treated with palbociclib 1 μM for 7 days. The drug was added only once and the culture media were not changed for the length of the treatment. Following the treatment, the cells were tripsinized and seeded in new dishes to be stained for SA-βgal activity (top images) and analyzed by confocal imaging to evaluate the presence of palbociclib (middle images). Prior to image acquisition cells were incubated with 4 µM palbociclib 1 h to improve palbociclib detection. In parallel, conditioned media from identical number of palbociclib-treated cells were obtained as in a and added to recipient SK-Mel-103 for 7 days, with no media changes, prior to assessing SA-βgal activity (bottom images)CC BY · PubMed Central
Fig. 6 Chloroquine potentiates palbociclib effects by reducing its lysosomal entrapment. a Confocal images of the lysotracker red and palbociclib fluorescence of SK-Mel-103 cells treated with 2 μM palbociclib and/or 50 µM chloroquine for 1 h. b Flow cytometry analysis (FCS-A vs SSC-A) of the percentage of senescent SK-Mel-103 cells incubated with the indicated concentrations of palbociclib in the presence or absence of 5 μM chloroquine for 7 days. c, d SK-Mel-103 cells were incubated for 7 days with the indicated concentrations of palbociclib in the presence or absence of 5 μM chloroquine. c Following the treatment, identical number of cells were re-seeded in 96-well plates and viable cells were measured at different days to analyze their proliferative capability. Values correspond to a single experiment with four biological replicas. Each condition was normalized to its respective signal at day 0. d In addition, 1000 cells from each treatment were re-seeded in 35 cm dishes and 10 days later colony growth was assessed by crystal violet staining. e SK-Mel-103 cells were co-incubated for one hour with 0.125 μM palbociclib and the indicated concentrations of chloroquine. The protein levels of the CDK4 substrate FOXM1 were subsequently analyzed by Western blot
Fig. 6 Chloroquine potentiates palbociclib effects by reducing its lysosomal entrapment. a Confocal images of the lysotracker red and palbociclib fluorescence of SK-Mel-103 cells treated with 2 μM palbociclib and/or 50 µM chloroquine for 1 h. b Flow cytometry analysis (FCS-A vs SSC-A) of the percentage of senescent SK-Mel-103 cells incubated with the indicated concentrations of palbociclib in the presence or absence of 5 μM chloroquine for 7 days. c, d SK-Mel-103 cells were incubated for 7 days with the indicated concentrations of palbociclib in the presence or absence of 5 μM chloroquine. c Following the treatment, identical number of cells were re-seeded in 96-well plates and viable cells were measured at different days to analyze their proliferative capability. Values correspond to a single experiment with four biological replicas. Each condition was normalized to its respective signal at day 0. d In addition, 1000 cells from each treatment were re-seeded in 35 cm dishes and 10 days later colony growth was assessed by crystal violet staining. e SK-Mel-103 cells were co-incubated for one hour with 0.125 μM palbociclib and the indicated concentrations of chloroquine. The protein levels of the CDK4 substrate FOXM1 were subsequently analyzed by Western blotCC BY · PubMed Central
Fig. 7 Lysosomal trapping is a feature common to all CDK4 inhibitors. a Donor SK-Mel-103 cells were treated with 1 μM palbociclib, 1 μM abemaciclib or 1 μM ribociclib for 7 days. The drugs were added only once and the culture media were not changed for the length of the treatment. Following treatments, cells were tripsinized and identical numbers were seeded in new dishes in the presence of the drugs. 24 h later the dishes were washed with PBS and incubated with fresh media for 3 additional days. The media were then collected, filtrated with 0.45 μM filters, diluted 1:2 with fresh media, and added to the recipient cells. Recipient SK-Mel-103 were incubated with the conditioned media for 7 days, with no media changes, and then stained for SA-βgal activity. b The reversible storage of palbociclib into lysosomes (left) has three main implications: (top right) it prolongs the availability of the drug (long-term activity) and induces paracrine effects (paracrine activity); (bottom right) the presence of another lysosomotropic drug, reduces the accumulation of palbociclib into lysosomes and this increases its cytoplasmic concentration thereby potentiating its intracellular activity (enhanced activity) as well as its extracellular release (paracrine activity)
Fig. 7 Lysosomal trapping is a feature common to all CDK4 inhibitors. a Donor SK-Mel-103 cells were treated with 1 μM palbociclib, 1 μM abemaciclib or 1 μM ribociclib for 7 days. The drugs were added only once and the culture media were not changed for the length of the treatment. Following treatments, cells were tripsinized and identical numbers were seeded in new dishes in the presence of the drugs. 24 h later the dishes were washed with PBS and incubated with fresh media for 3 additional days. The media were then collected, filtrated with 0.45 μM filters, diluted 1:2 with fresh media, and added to the recipient cells. Recipient SK-Mel-103 were incubated with the conditioned media for 7 days, with no media changes, and then stained for SA-βgal activity. b The reversible storage of palbociclib into lysosomes (left) has three main implications: (top right) it prolongs the availability of the drug (long-term activity) and induces paracrine effects (paracrine activity); (bottom right) the presence of another lysosomotropic drug, reduces the accumulation of palbociclib into lysosomes and this increases its cytoplasmic concentration thereby potentiating its intracellular activity (enhanced activity) as well as its extracellular release (paracrine activity)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 50 references

Cited by

See all 96

Metadata

Authors
Llanos, S., Megias, D., Blanco-Aparicio, C., Hernández-Encinas, E., Rovira, M., Pietrocola, F., & Serrano, M.
Affiliations
  • Spanish National Cancer Research Centre (ES)
  • Institute for Research in Biomedicine (ES)
  • Institució Catalana de Recerca i Estudis Avançats (ES)
Corresponding author
  • Llanos, Susana
  • Serrano, Manuel
Venue
Oncogene
Published
2019-01-28
Volume
38
Issue
20
Pages
3886-3902
ISSN
  • 0950-9232 (ISSN-L)
  • 1476-5594
OpenAlex
W2914014134
Cited by
96
References
50
Citations by year
2026: 15 · 2025: 14 · 2024: 11 · 2023: 10 · 2022: 18 · 2021: 15 · 2020: 8 · 2019: 4
FWCI
8.27
Type
journal article
Language
en
Open access
Licence
  • cc-by
  • creativecommons.org/licenses/by/4.0 (tdm)
Locations
  • Oncogene (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • Institute for Research in Biomedicine
  • Banco Santander
  • Fundación Bancaria Caixa d'Estalvis i Pensions de Barcelona
  • Santander Universities
Topics
  • Health Sciences › Medicine › Oncology › Advanced Breast Cancer Therapies
  • Physical Sciences › Materials Science › Biomaterials › Nanoparticle-Based Drug Delivery
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cell Biology › Autophagy in Disease and Therapy
Keywords
palbociclib, lysosomal trapping, CDK4/6 inhibitors, lysosomotropic agents, cell cycle arrest, senescence, paracrine senescence, drug release, chloroquine, ribociclib, abemaciclib, drug resistance, cancer therapy
MeSH terms
Acridine Orange, Aminopyridines, Antineoplastic Agents, Benzimidazoles, Chloroquine, Fluorescent Dyes, Humans, Lysosomes, Piperazines, Purines, Pyridines, Cytokines, Cellular Senescence, Cell Line, Tumor, Protein Kinase Inhibitors, Cyclin-Dependent Kinase 4, Lysosomal Membrane Proteins, Lysosomal-Associated Membrane Protein 1

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: Lysosomal trapping of palbociclib and its functional implications. https://app.cochat.ai/papers/10.1038/s41388-019-0695-8
The paper
Llanos, S., Megias, D., Blanco-Aparicio, C., Hernández-Encinas, E., Rovira, M., Pietrocola, F., & Serrano, M. (2019). Lysosomal trapping of palbociclib and its functional implications. Oncogene. https://doi.org/10.1038/s41388-019-0695-8

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.