CoChat

Spi-1/PU.1 activates transcription through clustered DNA occupancy in erythroleukemia

Open access

The DOI matches its Crossref record; 45 of the 46 references it cites were found in a registry or an index; 11 of the 29 sources its claims cite were checked for support.

Abstract

Acute leukemias are characterized by deregulation of transcriptional networks that control the lineage specificity of gene expression. The aberrant overexpression of the Spi-1/PU.1 transcription factor leads to erythroleukemia. To determine how Spi-1 mechanistically influences the transcriptional program, we combined a ChIP-seq analysis with transcriptional profiling in cells from an erythroleukemic mouse model. We show that Spi-1 displays a selective DNA-binding that does not often cause transcriptional modulation. We report that Spi-1 controls transcriptional activation and repression partially through distinct Spi-1 recruitment to chromatin. We revealed several parameters impacting on Spi-1-mediated transcriptional activation. Gene activation is facilitated by Spi-1 occupancy close to transcriptional starting site of genes devoid of CGIs. Moreover, in those regions Spi-1 acts by binding to multiple motifs tightly clustered and with similar orientation. Finally, in contrast to the myeloid and lymphoid B cells in which Spi-1 exerts a physiological activity, in the erythroleukemic cells, lineage-specific cooperating factors do not play a prevalent role in Spi-1-mediated transcriptional activation. Thus, our work describes a new mechanism of gene activation through clustered site occupancy of Spi-1 particularly relevant in regard to the strong expression of Spi-1 in the erythroleukemic cells.

Overview

CoChat Summary

TL;DR

Combining ChIP-seq and expression profiling in a murine Spi-1/PU.1 erythroleukemia model, this study shows that Spi-1 DNA binding is selective but often not sufficient to change transcription. Activation is favored by Spi-1 occupancy near TSSs of CGI-free genes, with tightly clustered, same-orientation Spi-1 motifs and high Spi-1 recruitment.

Key findings

  • ChIP-seq identified 17,781 Spi-1 peaks and 15,946 RNA Pol II peaks; 83% of Spi-1 peaks fell within a gene-associated region versus 68% of control peaks, and the Spi-1 motif was found in 88% of Spi-1-bound regions. Spi-1 bound only 0.35% of mouse genome regions containing the Spi-1 motif, and only 8.1% (854/10,560) of Spi-1-occupied genes were transcriptionally modulated.
  • Spi-1-activated genes had closer binding: 60% contained Spi-1 peaks within 5 kb of the TSS, versus 19 kb for repressed genes and 47 kb for non-responsive genes; promoter and immediate-downstream enrichment was x2.3 and x2.8 for activated genes.
  • In promoter/immediate-downstream regions, Spi-1 occupancy of CGI-free genes was associated with more activation (14% versus 8% with CGI, P < 1.9 × 10−4), and activated genes had more Spi-1 motif hits in CGI-free promoters (mean 5.6 versus 3.9 with CGI, P = 8.1 × 10−5; versus 3.4 repressed and 3.8 NoResp). Motif pairs were mainly head-to-tail on the same strand, with 50% of such pairs within 80 bp.
  • In primary erythroleukemic cells, 73% of promoter/immediate-downstream Spi-1 sites were shared with cultured cells, and motif clustering plus higher peak height were also associated with activation. No prevalent lineage-specific cofactor was found; a Tal1::GATA1 motif appeared only in 10% of Spi-1-repressed genes lacking the Spi-1 motif (73/390).

Methods

  • ChIP-seq for Spi-1, IgG, RNA Pol II, and H3K36me3 was performed in cultured erythroleukemic cells from spi-1 transgenic mice and in primary spleen cells; libraries were sequenced on SOLiD, aligned to mm9, and peaks called with FindPeaks/MICSA.
  • Expression profiling used Affymetrix MOE430v2 microarrays (>39,000 transcripts) in control cells and doxycycline-inducible anti-spi-1 shRNA cells, with 23 Spi-1+ control arrays and 8 Spi-1− arrays; differential expression was assessed by ANOVA with adjusted P < 0.01 and validated by RT-qPCR.
  • Analyses included peak annotation relative to RefSeq TSS/TE, CpG island detection, motif discovery/enrichment (ChIPMunk, MEME, TOMTOM, PATSER), and statistical tests including Wilcoxon, χ2, and Kolmogorov–Smirnov.

Limitations

  • The paper states that too few peaks in some subregions/categories prevented separate subregion analysis, that further study is needed to determine whether GC-rich elements recruit specific proteins, and that complementary epigenetic analysis may be required to understand repression.
  • Evident from the text: multiple Spi-1 molecule occupancy is inferred from motif clustering and peak height, but direct simultaneous binding or homodimerization is not demonstrated; the authors note Spi-1 homodimerization has not been described.
  • Evident from the text: the proposed clustered-occupancy activation mechanism is correlative from occupancy and expression data; no functional perturbation of clustered motifs or CGI status is reported.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1093/nar/gks659

Cites

45 of 45 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1. Spi-1 and RNA Pol II exhibit preferential binding in the genome. (A) The definition of each genomic region is described. (B) The histograms represent the distribution of RNA Pol II peaks location relative to the nearest TSS. See the Methods section for details. χ2 test; Asterisk indicates P < 10−3. (C) The median peak height of RNA Pol II in the promoter or ImDown subregions (−1.5 kb + 2 kb) or score of H3K36me3 (y-axis) versus the gene expression levels measured by the median fluorescence using microarray data (x-axis). Each y-value was calculated for an interval of fluorescence. The solid curve is a fit with a polynomial of degree 4 by robust regression using an M estimator from the R package MASS and the ‘rlm’ function: http://cran.r-project.org/web/packages/MASS/MASS.pdf). (D) The histograms represent the distribution of the total Spi-1 and control peaks in the murine genome. See the ‘Materials and Methods’ section for details. χ2 test; Asterisk indicates P < 10−3. (E) Sequence logos of the PWM produced by ChIPMunk software using all 17 781 Spi-1 peak sequences generated using WebLogo are shown; the PWM threshold corresponds to a FDR of 10%.
Figure 1. Spi-1 and RNA Pol II exhibit preferential binding in the genome. (A) The definition of each genomic region is described. (B) The histograms represent the distribution of RNA Pol II peaks location relative to the nearest TSS. See the Methods section for details. χ2 test; Asterisk indicates P < 10−3. (C) The median peak height of RNA Pol II in the promoter or ImDown subregions (−1.5 kb + 2 kb) or score of H3K36me3 (y-axis) versus the gene expression levels measured by the median fluorescence using microarray data (x-axis). Each y-value was calculated for an interval of fluorescence. The solid curve is a fit with a polynomial of degree 4 by robust regression using an M estimator from the R package MASS and the ‘rlm’ function: http://cran.r-project.org/web/packages/MASS/MASS.pdf). (D) The histograms represent the distribution of the total Spi-1 and control peaks in the murine genome. See the ‘Materials and Methods’ section for details. χ2 test; Asterisk indicates P < 10−3. (E) Sequence logos of the PWM produced by ChIPMunk software using all 17 781 Spi-1 peak sequences generated using WebLogo are shown; the PWM threshold corresponds to a FDR of 10%.CC BY-NC · PubMed Central
Figure 2. Global expression profile of Spi-1-modulated genes. (A) shSpi-1-A2B (similar expression levels were observed in shSpi-1-A2C cells) and control cells were cultured with (26 h, 44 h and 96 h) or without dox (26 h and 44 h), and the extracted proteins were analyzed by western blotting using Spi-1 and adaptin antibodies. Two bands corresponding to Spi-1 were detected. (B) Ingenuity Pathway Analysis (IPA) of Spi-1 target genes. IPA was performed on Spi-1 activated and repressed genes as identified using expression microarrays. The 10 most significant molecular and cellular functions revealed are shown and -Log (Bonferroni–Holm P-value) are indicated. The orange bar represents the P-value threshold of 0.05 (-Log (P-value) = 1.3).
Figure 2. Global expression profile of Spi-1-modulated genes. (A) shSpi-1-A2B (similar expression levels were observed in shSpi-1-A2C cells) and control cells were cultured with (26 h, 44 h and 96 h) or without dox (26 h and 44 h), and the extracted proteins were analyzed by western blotting using Spi-1 and adaptin antibodies. Two bands corresponding to Spi-1 were detected. (B) Ingenuity Pathway Analysis (IPA) of Spi-1 target genes. IPA was performed on Spi-1 activated and repressed genes as identified using expression microarrays. The 10 most significant molecular and cellular functions revealed are shown and -Log (Bonferroni–Holm P-value) are indicated. The orange bar represents the P-value threshold of 0.05 (-Log (P-value) = 1.3).CC BY-NC · PubMed Central
Figure 3. Combination of ChIP-seq and gene expression analysis reveals a distinct Spi-1 peaks localization in activated and repressed genes. (A) Cumulative distribution of the distance between a gene TSS and the nearest Spi-1 or the control (Input and IgG) peaks for the different transcriptional response categories. (B) The histograms represent the enrichment of Spi-1-bound genes of each transcriptional category with respect to the total number of Spi-1-bound genes within each class of the subregion. Statistical analyses versus the total number of genes were performed using the χ2 test; Asterisk indicates P < 10−3; Double asterisk indicates P < 10−10. (C) Distribution of the distance from the Spi-1 and RNA Pol II peaks to the nearest gene TSS, as determined by the hierarchy model, in each transcriptional response category. The distributions on each side of the TSS within 50-bp windows are reported.
Figure 3. Combination of ChIP-seq and gene expression analysis reveals a distinct Spi-1 peaks localization in activated and repressed genes. (A) Cumulative distribution of the distance between a gene TSS and the nearest Spi-1 or the control (Input and IgG) peaks for the different transcriptional response categories. (B) The histograms represent the enrichment of Spi-1-bound genes of each transcriptional category with respect to the total number of Spi-1-bound genes within each class of the subregion. Statistical analyses versus the total number of genes were performed using the χ2 test; Asterisk indicates P < 10−3; Double asterisk indicates P < 10−10. (C) Distribution of the distance from the Spi-1 and RNA Pol II peaks to the nearest gene TSS, as determined by the hierarchy model, in each transcriptional response category. The distributions on each side of the TSS within 50-bp windows are reported.CC BY-NC · PubMed Central
Figure 4. Clustering of Spi-1 binding is associated with transcriptionally activated genes and occurs mainly in promoters devoid of CGIs. (A) Spi-1 occupancy in the promoter + ImDown subregions devoid of CGI increases the probability to positively influence gene transcription. The histograms represent the percentage of the repressed or activated genes when Spi-1 peaks are within promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs. (B) Distribution of the number of ChIPMunk-defined Spi-1 motif hits per Spi-1 peak in the promoter + ImDown, intragenic and enhancer subregions according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line inside the box plot indicates the median. The Wilcoxon test was used to statistically examine the differences in the distributions of the number of Spi-1-motif hits. P < 0.05 was considered statistically significant. (C) Distribution of the number of ChIPMunk-defined Spi-1-motif hits per Spi-1 peak sequence in the promoters + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs and according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line inside the box plot indicates the median. The Wilcoxon test was used to statistically examine the differences in the distributions of the number of Spi-1-motif hits. P < 0.05 was considered statistically significant.
Figure 4. Clustering of Spi-1 binding is associated with transcriptionally activated genes and occurs mainly in promoters devoid of CGIs. (A) Spi-1 occupancy in the promoter + ImDown subregions devoid of CGI increases the probability to positively influence gene transcription. The histograms represent the percentage of the repressed or activated genes when Spi-1 peaks are within promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs. (B) Distribution of the number of ChIPMunk-defined Spi-1 motif hits per Spi-1 peak in the promoter + ImDown, intragenic and enhancer subregions according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line inside the box plot indicates the median. The Wilcoxon test was used to statistically examine the differences in the distributions of the number of Spi-1-motif hits. P < 0.05 was considered statistically significant. (C) Distribution of the number of ChIPMunk-defined Spi-1-motif hits per Spi-1 peak sequence in the promoters + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs and according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line inside the box plot indicates the median. The Wilcoxon test was used to statistically examine the differences in the distributions of the number of Spi-1-motif hits. P < 0.05 was considered statistically significant.CC BY-NC · PubMed Central
Figure 5. Spi-1 motif hits distribution inside Spi-1 peaks respects rule of engagment. (A) Distribution of the Spi-1 peak heights in the promoter + ImDown subregions (all peaks) or peaks in the promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs and according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line within the box plot indicates the median. The Kolmogorov–Smirnov test was used to statistically examine the differences in the distribution of the Spi-1 peak height compared with the peaks in the NoResp group of genes; P < 0.05 was considered statistically significant. Distribution of the distances between the pairs of Spi-1 motif hits in (B) head-to-tail orientation (the same strand) or (C) in head-to-head or tail-to-tail orientation (opposite strands) for activated genes. The X-axis presents the length of the spacers separating pairs of Spi-1 motif hits. The Y-axis shows the fraction of sequences with at least one pair of Spi-1 motif hits separated by the selected spacer (see the Supplementary Methods for details).
Figure 5. Spi-1 motif hits distribution inside Spi-1 peaks respects rule of engagment. (A) Distribution of the Spi-1 peak heights in the promoter + ImDown subregions (all peaks) or peaks in the promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs and according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line within the box plot indicates the median. The Kolmogorov–Smirnov test was used to statistically examine the differences in the distribution of the Spi-1 peak height compared with the peaks in the NoResp group of genes; P < 0.05 was considered statistically significant. Distribution of the distances between the pairs of Spi-1 motif hits in (B) head-to-tail orientation (the same strand) or (C) in head-to-head or tail-to-tail orientation (opposite strands) for activated genes. The X-axis presents the length of the spacers separating pairs of Spi-1 motif hits. The Y-axis shows the fraction of sequences with at least one pair of Spi-1 motif hits separated by the selected spacer (see the Supplementary Methods for details).CC BY-NC · PubMed Central
Figure 6. Properties of the ChIP-seq peaks in primary erythroleukemic cells. (A) Distribution of the number of ChIPMunk-defined Spi-1-motif hits per Spi-1 peak sequence in the promoters + ImDown subregions (all peaks) or peaks in the promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs and according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line inside the box plot indicates the median. The Wilcoxon test was used to statistically examine the differences in the distributions of the number of Spi-1-motif hits. P < 0.05 was considered statistically significant. (B) Distribution of the Spi-1 peak heights in the promoter + ImDown subregions (all peaks) or peaks in the promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line within the box plot indicates the median. The Kolmogorov–Smirnov test was used to statistically examine the differences in the distribution of the Spi-1 peak height compared with the peaks in the NoResp group of genes; P < 0.05 was considered statistically significant.
Figure 6. Properties of the ChIP-seq peaks in primary erythroleukemic cells. (A) Distribution of the number of ChIPMunk-defined Spi-1-motif hits per Spi-1 peak sequence in the promoters + ImDown subregions (all peaks) or peaks in the promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs and according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line inside the box plot indicates the median. The Wilcoxon test was used to statistically examine the differences in the distributions of the number of Spi-1-motif hits. P < 0.05 was considered statistically significant. (B) Distribution of the Spi-1 peak heights in the promoter + ImDown subregions (all peaks) or peaks in the promoter + ImDown subregions that contain (with CGI) or not (w/o CGI) CGIs according to Spi-1 transcriptional activities. The red square indicates the mean, and the black line within the box plot indicates the median. The Kolmogorov–Smirnov test was used to statistically examine the differences in the distribution of the Spi-1 peak height compared with the peaks in the NoResp group of genes; P < 0.05 was considered statistically significant.CC BY-NC · PubMed Central

Verification

Claims the paper makes

8 Valid · 0 Not valid · 11 Not checked · 134 Own results and interpretations

114 more claims

Create a free account to see all 153 claims and to verify the 18 references still open.

See all claims — free account
  • It may state two separate claims as one. 25
  • Probably not a substantial claim. 13

Create a free account to see these claims.

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.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 45 references

Cited by

See all 28

Metadata

Authors
Ridinger-Saison, M., Boeva, V., Rimmelé, P., Kulakovskiy, I., Gallais, I., Levavasseur, B., Paccard, C., Legoix-Né, P., Morlé, F., Nicolas, A., Hupé, P., Barillot, E., Moreau-Gachelin, F., & Guillouf, C.
Affiliations
  • Institut Curie (FR)
  • Engelhardt Institute of Molecular Biology (RU)
Corresponding author
Guillouf, Christel
Venue
Nucleic Acids Research
Published
2012-07-11
Volume
40
Issue
18
Pages
8927-8941
ISSN
  • 0305-1048 (ISSN-L)
  • 1362-4962
OpenAlex
W1992430937
Cited by
28
References
46
Citations by year
2022: 1 · 2021: 5 · 2020: 2 · 2019: 3 · 2018: 1 · 2017: 1 · 2016: 1 · 2015: 3 · 2014: 4 · 2013: 5 · 2012: 1
FWCI
1.67
Type
journal article
Language
en
Open access
Locations
  • Nucleic Acids Research (publishedVersion, cc-by-nc)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • HAL (submittedVersion)
  • HighWire Press Open Archive (submittedVersion)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Chromatin Dynamics
  • Health Sciences › Medicine › Hematology › Acute Myeloid Leukemia Research
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Chromatin Remodeling and Cancer
Keywords
erythroleukemia, transcriptional activation, chromatin binding, transcriptional repression, CGI, transcription factor binding motifs, lineage-specific gene expression, acute leukemia, erythroid lineage
MeSH terms
Proto-Oncogene Protein Spi-1, Animals, Binding Sites, DNA, Leukemia, Erythroblastic, Acute, Mice, Transgenic, Proto-Oncogene Proteins, Transcriptional Activation, Trans-Activators, Gene Expression Regulation, Neoplastic, Genome, Sequence Analysis, DNA, CpG Islands, Gene Expression Profiling, Transcription Initiation Site, Cell Line, Tumor, Chromatin Immunoprecipitation, Regulatory Elements, Transcriptional, Mice, Nucleotide Motifs

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.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash).

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: Spi-1/PU.1 activates transcription through clustered DNA occupancy in erythroleukemia. https://app.cochat.ai/papers/10.1093/nar/gks659
The paper
Ridinger-Saison, M., Boeva, V., Rimmelé, P., Kulakovskiy, I., Gallais, I., Levavasseur, B., Paccard, C., Legoix-Né, P., Morlé, F., Nicolas, A., Hupé, P., Barillot, E., Moreau-Gachelin, F., & Guillouf, C. (2012). Spi-1/PU.1 activates transcription through clustered DNA occupancy in erythroleukemia. Nucleic Acids Research. https://doi.org/10.1093/nar/gks659

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.