Decoding the multiple functions of ZBP1 in the mechanism of sepsis-induced acute lung injury

Abstract

Sepsis-induced acute lung injury (ALI), characterized by severe hypoxemia and pulmonary leakage, remains a leading cause of mortality in intensive care units. The exacerbation of ALI during sepsis is largely attributed to uncontrolled inflammatory responses and endothelial dysfunction. Emerging evidence suggests an important role of Z-DNA binding protein 1 (ZBP1) as a sensor in innate immune to drive inflammatory signaling and cell death during infections. However, the role of ZBP1 in sepsis-induced ALI has yet to be defined. We utilized ZBP1 knockout mice and combined single-cell RNA sequencing with experimental validation to investigate ZBP1’s roles in the regulation of macrophages and lung endothelial cells during sepsis. We demonstrate that in sepsis, ZBP1 deficiency in macrophages reduces mitochondrial damage and inhibits glycolysis, thereby altering the metabolic status of macrophages. Consequently, this metabolic shift leads to a reduction in the differentiation of macrophages into pro-inflammatory states and decreases macrophage pyroptosis triggered by activation of the NLRP3 inflammasome. These changes significantly weaken the inflammatory signaling pathways between macrophages and endothelial cells and alleviate endothelial dysfunction and cellular damage. These findings reveal important roles for ZBP1 in mediating multiple pathological processes involved in sepsis-induced ALI by modulating the functional states of macrophages and endothelial cells, thereby highlighting its potential as a promising therapeutic target.

Full text

communications biology Article
https://doi.org/10.1038/s42003-024-07072-x
Decoding the multiple functions of ZBP1 in
the mechanism of sepsis-induced acute
lung injury
Check for updates
Ting Gong 1,2,7 ,YuFu 1,3,7,QingdeWang 1, Patricia A. Loughran1,YuehuaLi
1, Timothy R. Billiar 1,4,
Zongmei Wen3, Youtan Liu2& Jie Fan 1,4,5,6
Sepsis-induced acute lung injury (ALI), characterized by severe hypoxemia and pulmonary leakage,
remains a leading cause of mortality in intensive care units. The exacerbation of ALI during sepsis is
largely attributed to uncontrolled inflammatory responses and endothelial dysfunction. Emerging
evidence suggests an important role of Z-DNA binding protein 1 (ZBP1) as a sensor in innate immune
to drive inflammatory signaling and cell death during infections. However, the role of ZBP1 in sepsis-
induced ALI has yet to be defined. We utilized ZBP1 knockout mice and combined single-cell RNA
sequencing with experimental validation to investigate ZBP1’s roles in the regulation of macrophages
and lung endothelial cells during sepsis. We demonstrate that in sepsis, ZBP1 deficiency in
macrophages reduces mitochondrial damage and inhibits glycolysis, thereby altering the metabolic
status of macrophages. Consequently, this metabolic shift leads to a reduction in the differentiation of
macrophages into pro-inflammatory states and decreases macrophage pyroptosis triggered by
activation of the NLRP3 inflammasome. These changes significantly weaken the inflammatory
signaling pathways between macrophages and endothelial cells and alleviate endothelial dysfunction
and cellular damage. These findings reveal important roles for ZBP1 in mediating multiple pathological
processes involved in sepsis-induced ALI by modulating the functional states of macrophages and
endothelial cells, thereby highlighting its potential as a promising therapeutic target.
Acute Lung Injury (ALI) and acute respiratory distress syndrome (ARDS)
are clinical syndromes characterized by severe hypoxemia and pulmonary
leakage, posing not only a threat to patient life but also imposing significant
psychological and economic burdens on patients and their families1.Despite
medical advancements over the past few decades, such as implementing low
tidal volume ventilation strategies to minimize ventilator-associated lung
injury and the use of anti-inflammatory medications like dexamethasone,
these interventions have not significantly improved patient survival rates2,3.
Currently, the mortality rates for ALI/ARDS remain alarmingly high, ran-
ging from 30% to 40%, and account for approximately 10% of all deaths in
intensive care units (ICUs) globally4.
Sepsis is a leading cause of ALI/ARDS, characterized by an uncon-
trolled inflammatory response and endothelial barrier dysfunction5.During
sepsis, the extensive release of inflammatory mediators precipitates a
systemic inflammatory response syndrome (SIRS), which subsequently
triggers ALI/ARDS6. In septic ALI, the dysfunction of pulmonary vascular
endothelial cells (PVECs) is characterized by increased vascular perme-
ability and the disruption of the alveolar-capillary barrier. This disruption
leads to pulmonary edema and compromised gas exchange, resulting in life-
threatening hypoxemia and respiratory failure7. Therefore, the development
of innovative therapeutic strategies targeting septic ALI is critical. Such
strategies should focus on controlling the inflammatory response and pre-
serving endothelial barrier function to reduce mortality rates.
Single-cell RNA sequencing (scRNA-seq) is a cutting-edge technology
that allows us to decode cellular heterogeneity, differentiation, and inter-
cellular communication during the pathogenesis of sepsis. Evidence indi-
cates that many genes expressed in macrophages and endothelial cells (ECs)
are clustered, enabling the identification of gene transcription and
1Department of Surgery, University of Pittsburgh School of Medicine, Pittsburgh, 15213, USA. 2Department of Anesthesiology, Shenzhen Hospital, Southern
Medical University, Shenzhen, 518000, China. 3Department of Anesthesiology, Shanghai Pulmonary Hospital, School of Medicine, Tongji University, Shanghai,
200433, China. 4Department of Immunology, University of Pittsburgh School of Medicine, Pittsburgh, 15213, USA. 5Research and Development, Veterans Affairs
Pittsburgh Healthcare System, Pittsburgh, PA, 15240, USA. 6McGowan Institute for Regenerative Medicine, University of Pittsburgh, Pittsburgh, PA, 15219, USA.
7
These authors contributed equally: Ting Gong, Yu Fu. e-mail: tinggong@pitt.edu;jif7@pitt.edu
Communications Biology | (2024) 7:1361 1
1234567890():,;
1234567890():,;
Content courtesy of Springer Nature, terms of use apply. Rights reserved

pathological changes in local tissues during the progression of sepsis8,9.In
this study, we applied scRNA-seq to decode the roles of Z-DNA binding
protein 1 (ZBP1) in the mechanism of sepsis-induced ALI.
ZBP1 has been reported as a sensor of double-stranded DNA and RNA
helices, adopting the unusual left-handed Z-conformations known as
Z-DNA and Z-RNA10. Recent studies have identified ZBP1 as a crucial
upstream regulator in the pathways of cell death and pro-inflammatory
signaling11,12. However, the precise role and underlying mechanisms of
ZBP1 in the context of sepsis-induced ALI remain poorly understood. In
this study, we used ZBP1 knockout (Zbp1−/−
) mice to investigate the role of
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 2
Content courtesy of Springer Nature, terms of use apply. Rights reserved

ZBP1 in sepsis-induced ALI. Utilizing a combination of scRNA-seq,
immunofluorescence, and protein blot analysis, we observed an upregula-
tion of ZBP1 expression across various cellular components of the lung
during the pathological process of sepsis-induced ALI. Our comprehensive
analysis of scRNA-seq data delineates the role of ZBP1 in the progression of
sepsis-induced ALI, demonstrating its regulatory influence on various
pulmonary cellular components, including macrophages and endothelial
cells. The absence of ZBP1 provides a protective effect against lung tissue
damage in sepsis-induced ALI, primarily manifesting as reduced inflam-
matory responses and preserved vascular integrity. These findings highlight
ZBP1 as a potential therapeutic target for sepsis-induced ALI and under-
score the need for further investigation into its regulatory mechanisms and
therapeutic implications.
Results
scRNA-seq reveals alterations in different lung cells
following sepsis
A comprehensive illustration outlines the experimental workflow beginning
with the induction of the cecal ligation and puncture (CLP) sepsis model in
mice. The measurement process includes lung tissue extraction, isolation of
single cells, and their analysis through next-generation sequencing using the
10x Genomics platform, followed by detailed bioinformatic analysis and
subsequent validation of the findings (Supplementary Fig. 1A).
We utilized UMAP to identify cell clusters based on the expression of a
panel of known cell type-specific marker genes. Through this bioinformatic
approach, we were able to delineate ten major somatic cell types present in
the lung tissue of wild-type mice subjected to either sham (WT Sham) or
CLP (WT CLP). These identified cell populations included epithelial cells,
mononuclear phagocytes (MPC), endothelial cells, alveolar type 2 cells
(AT2), epithelial progenitor cells, fibroblasts, mesothelial cells, macrophages,
pericytes, and lymphatic endothelial cells. The UMAP plots display distinct
cellular groupings in both WT Sham and WT CLP groups, demonstrating
the diversity of cell types and the impact of septic injury on the lung cellular
microenvironment. Dot plot analysis further confirmed the presence and
relative expression levels of key markers within these cell types, providing a
comprehensive cellular atlas of the septic lungs (Supplementary Fig. 1B, C).
In the intercellular interaction network diagrams, we observed con-
nections and changes in the interactions among single-cell subpopulations
in the lungs of septic mice (Supplementary Fig. 1D, E). Furthermore, the
analysis of enriched signaling pathways reveals significant upregulation of
pro-inflammatory pathways, including TNF, IL6, and IL1, in CLP mice.
Conversely, in sham mice, pathways such as WNT and VEGF are notably
enhanced (Supplementary Fig. 1F).
To further analyze the changes in intercellular interaction pathways
during sepsis, we examined the probabilities of macrophage receptor-ligand
pairs. We observed a significant enhancement in both paracrine (macro-
phages affecting other cells) and autocrine (macrophages affecting them-
selves) pathways in sepsis. This was notably marked by the secretion of TNF
and its binding to receptors such as Tnf - Tnfrsf1b, Tnf - Tnfrsf1a, and Spp1
-(Itga4+Itgb1) (Supplementary Fig. 2A–D). Subsequently, we conducted a
detailed analysis of macrophage subgroups through dimensionality reduc-
tion, clustering, and subgrouping of macrophages (Supplementary Fig. 2E).
scMetabolism software13 was utilized to assess the activity of the gly-
colysis pathway in macrophages, with AUCell14 employed for scoring. We
observed a significant enhancement of glycolysis/gluconeogenesis meta-
bolism in macrophages of CLP mice (Supplementary Fig. 3A, B). Further-
more, we found that glycolysis-related genes HIF1-alpha, LDHA, and
Slc2a1 were significantly upregulated in the CLP group (Supplementary
Fig. 3C–E). Additionally, the inflammatory response andhypoxiapathway
activity scores were markedly higher in CLP mice compared to sham mice
(Supplementary Fig. 3F–I). Moreover, by referring to the functional gene
sets for macrophages summarized by Bischoff 15, the analysis of M1 and M2
pathway activation scores revealed that the M1 macrophage scores were
significantly elevated in the CLP mice (Supplementary Fig. 3J).
These results suggest that in sepsis, macrophages predominantly
engage in hypoxia and glycolysis pathways, undergo metabolic repro-
gramming towards a pro-inflammatory M1 polarization, and release
inflammatory mediators that affect the function of other cell types.
ZBP1 expression increases in lung cells in response to sepsis
Lung histology demonstrated cellular infiltration,edema,andalveolarwall
thickening in the CLP group (Fig. 1A). The severity of lung injury in the CLP
group was further quantified with scoring of these histopathological features
(Fig. 1B). The bronchoalveolar lavage fluid (BALF) from the CLP group
exhibited a marked increase in IL-6 levels (Fig. 1C) and neutrophil exuda-
tion (Fig. 1D).
Differences in gene expression between WT CLP and WT Sham groups
were delineated in a volcano plot, using a cutoff of |log
2
FC| = 1 and P=0.05.
Genes surpassing this threshold of log
2
FC > 1 and P<0.05wereclassified as
upregulated in the WT CLP group. Among these, the expressions of key
inflammatory and chemotactic factors, such as Cxcl2, Ccl5, Ccl4, Il1b, and
Cxcl10, were notably elevated in the WT CLP group, suggesting active
inflammatory signaling (Fig. 1E). To further probe into the biological sig-
nificance of these differentially regulated genes, pathway enrichment ana-
lysis was conducted. This analysis brought to light a significant enrichment
within the cytokine-receptor interaction and the TNF signaling pathway
(Fig. 1F), elucidating the molecular mechanisms at play during ALI.
Moreover, the distribution and expression levels of ZBP1 mRNA
across different cell populations were elucidated, revealing a marked
increase in the CLP group (Fig. 1G).Thiselevationwasvisuallycaptured
and quantified through UMAP analysis, depicting widespread upregulation
of ZBP1 mRNA expression (Fig. 1H). Immunofluorescence staining cor-
roborated these findings, showing pronounced expression of ZBP1 in the
lung tissue cells of CLP mice, thus aligning with the transcriptional data
(Fig. 1I). Further analyses using Western blot and qPCR to examine ZBP1
protein levels and mRNA revealed significant increase in ZBP1 post-CLP,
peaking at the 24 h postoperative interval (Fig. 1J, K).
The increased expression of ZBP1 in macrophages was further vali-
dated in the macrophages isolated from the BALF of septic mice. While
ZBP1 was primarily in nuclear and minimally expressed in the macrophages
from sham group, it exhibited a substantial cytoplasmic presence with
significantly elevated expression levels in the macrophages from CLP group
(Fig. 1L). The data reveal significant changes in ZBP1 expression and
intracellular translocation in the lung following sepsis-induced ALI.
ZBP1 is involved in sepsis-induced lung injury and mortality
The pivotal role of ZBP1 in sepsis-induced ALI was shown through a
comparative analysis of Zbp1−/−and WT mice following CLP. Lung
Fig. 1 | ZBP1 expression increases in lung cells in response to sepsis.
ARepresentative H&E-stained images of mouse lung tissues 24 h after CLP and
sham procedures. Scale bars, 50 μm. BLung injury was quantitatively assessed
through the scoring of histopathological features. CELISA detection of IL-6 levels in
BALF from CLP mice. DNeutrophil exudation counts in BALF. EDifferential gene
expression between WT CLP and WT Sham groups is presented in a volcano plot.
Genes upregulated in the WT CLP group are shown as red dots, while downregulated
genes are shown as blue dots. FEnrichment analysis of upregulated genes in the WT
CLP group identified significant KEGG pathways, with the y-axis representing
pathway items and the x-axis displaying the enrichment score. GViolin plots from
single-cell analysis display the expression levels of ZBP1 mRNA in different cell
types between the two groups. HUMAP plots represent the distribution of ZBP1
mRNA expression across all analyzed cells, with color intensity corresponding to
expression level. IImmunofluorescence detection of ZBP1 expression in lung tissue
24 h post-CLP. JWestern blot and qPCR (K) analysis were used to examine ZBP1
protein and mRNA expression levels in lung tissue at 6, 12, 24, and 48 h post-CLP
surgery. LLocalization and expression of ZBP1(green) in macrophages from BALF
of septic mice detected by IF. Scale bar, 20 μm. All results are based on three
replicates, and data are presented as the mean ± SD, *P< 0.05,
**P< 0.01, ***P< 0.001.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 3
Content courtesy of Springer Nature, terms of use apply. Rights reserved

histology demonstrated that in Zbp1−/−mice the lungs exhibited less cellular
infiltration, reduced edema, and thinner alveolar walls in comparison to WT
mice, indicating attenuated lung damage in the absence of ZBP1 (Fig. 2A).
The lung injury scores, derived from the pathological examination of these
tissues, were considerably lower in the Zbp1−/−mice than that in the WT
group in sepsis, which underscores the role of ZBP1 in sepsis-induced lung
injury (Fig. 2B).
Neutrophil counts in BALF showed a decreased cellular infiltration in
the Zbp1−/−mice following CLP, suggesting a role of ZBP1 in promoting
sepsis-induced cell infiltration into the lungs (Fig. 2C). The BALF protein
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 4
Content courtesy of Springer Nature, terms of use apply. Rights reserved

content, a marker of vascular permeability, was also significantly lower in
the Zbp1−/−mice (Fig. 2D). Additionally, the Zbp1−/−mice showed reduced
Evans blue dye extravasation in the lungs compared to the control group
following CLP (Fig. 2E).
Pro-inflammatory cytokines in the BALF were quantified using ELISA,
revealing that the levels of IL-6, IL-1β,andTNF-αweremarkedly reduced in
the Zbp1−/−mice, further suggesting a mitigated inflammatory response in
the absence of ZBP1 (Fig. 2F). In addition, immunofluorescence staining for
the neutrophil marker Ly6G showed a reduced neutrophil infiltration in the
lungs of Zbp1−/−mice 24 h post-CLP (Fig. 2G). Post-CLP survival outcomes
were significantly improved in Zbp1−/−mice, as depicted by Kaplan-Meier
survival curves. Over the course of 96 h following the procedure, Zbp1−/−
mice demonstrated a pronounced survival advantage over the WT mice
(Fig. 2H). Furthermore, monitoring of core body temperature changes post-
CLP indicated a more stable physiological response in Zbp1−/−mice com-
paredtotheWTgroup(Fig.2I).
Serum lactate levels, measured at 24 h post-CLP, were substantially
lower in Zbp1−/−mice, which is indicative of enhanced lactate clearance and/
or lactate creation. This metabolic parameter correlates with the improved
survival seen in these animals, suggesting a better overall outcome in the
Zbp1−/−group (Fig. 2J). To further evaluate the impact of ZBP1 on organ
function, we conducted additional experiments measuring amino-
transferases and creatinine levels in the blood samples of these mice. Our
findings show that ZBP1 deficiency mitigates CLP-induced liver and kidney
damage (Supplementary Fig. 4).
Collectively, these results support the hypothesis that ZBP1 plays a
detrimental role in the progression of sepsis-induced ALI, and its absence
can lead to a significant reduction in lung injury, better preservation of
vascular integrity, decreased inflammation, enhanced metabolic function,
and overall improved survival.
scRNA-seq reveals the role of ZBP1 in altering transcriptional
profiles of various lung cells following sepsis
To elucidate the role of ZBP1 in sepsis-induced ALI, we collected lung cells
from WT and Zbp1−/−mice that had undergone either sham or CLP
operations. A cohort of 14 samples was subjected to this scrutiny, and
following stringent quality control measures (Supplementary Fig. 5). We
conducted a thorough scRNA-seq data analysis to dissect the cellular and
molecular intricacies resulting from the deficiency of ZBP1 (Fig. 3A).
Western blot analysis of lung tissues confirmed the knockout of ZBP1 in
Zbp1−/−mice and showed an absence of ZBP1 expression post-CLP as
opposed to the elevated levels observed in the WT mice, which substantiates
the genetic manipulation undertaken in this study (Fig. 3B).
By UMAP analysis, we mapped the intricate cellular landscape of the
lung tissue, identifying and counting various cell populations (Fig. 3C).
Annotations for cell subpopulations are consistent with those presented in
Supplementary Fig. S1C. Our quantitative assessment across the different
experimental setups revealed no significant disparities in the relative
abundanceofeachcelltype(Fig.3D).
Next, we observed shifts in the distribution frequencies of different
lung cells (Fig. 3E). Further analysis exposed a substantial number of dif-
ferentially expressed genes (DEGs) across cell types, highlighting several
populations that exhibited marked transcriptional alterations when the WT
and Zbp1−/−mice were contrasted following CLP (Fig. 3F). The integrity of
the cell typing was affirmed by a box plot illustrating cell purity (Fig. 3G).
These scRNA-seq data reveal a distinct transcriptional landscape eli-
cited by ZBP1 knockout in lung cells following sepsis and suggest an
important role for ZBP1 in transcriptional reprogramming across various
lung cells, which may critically affect the cellular response to sepsis and
influence the outcomes of ALI.
ZBP1 regulates alveolar macrophage differentiation in sepsis
WT and Zbp1−/−mice were subjected to CLP model. t-SNE-based dimen-
sionality reduction and clustering analysis identified three distinct macro-
phage subpopulations (Macrophage C1, Macrophage C2, and Macrophage
C3) across the WT sham, WT CLP , Zbp1−/−Sham, and Zbp1−/−CLP groups
(Fig. 4A). These findings are augmented by a heatmap that delineates the
differential expression of the top 15 marker genes within these subgroups,
providing a molecular signature of each cluster (Fig. 4B). M1 and M2
polarization states were quantitatively assessed, with violin plots depicting
the gene set activity scores that distinguish between the pro-inflammatory
and anti-inflammatory macrophage phenotypes within the subpopula-
tions (Fig. 4C).
Through the trajectory analysis, constructed using the Slingshot
algorithm16, we visualized the differentiation pathways of macrophage
subsets, indicating a potential reprogramming of macrophage lineage
commitment upon ZBP1 knockout (Fig. 4D). CytoTRACE17 analysis
showed the differentiation probability scores, which range from 0 to 1,
offering a nuanced view into the relative differentiation states of macro-
phage subpopulations, where scores nearer to 0 are associated with a higher
state of differentiation (Fig. 4E). The two-dimensional t-SNE scatter plot
demonstrated the dispersion of macrophage subpopulations, visually sup-
porting the data obtained from the trajectory and CytoTRACE analyses
(Fig. 4F). Box plots showcasing CytoTRACE scores elucidate the disparity in
differentiation states across macrophage subsets, with Macrophage C3
emerging as the most differentiated macrophage subpopulation (Fig. 4G).
Expression dynamics of the top nine genes with the highest correlation
to the differentiation trajectory endpoints were illustrated in line plots,
shedding light on the gene expression changes that may underpin macro-
phage identity and function in this model of ALI (Fig. 4H). The heatmap of
the top 30 genes with the highest correlation to differentiation endpoints
complemented these findings by offering a broader perspective of the
molecular factors involved in macrophage differentiation (Fig. 4I).
Finally, the cell percentage chart provided a visual summary of the
different cell cluster distributions among the groups, highlighting the overall
impact of ZBP1 knockout on macrophage heterogeneity in the inflamma-
tory milieu of sepsis-induced ALI (Fig. 4J). In conclusion, these data suggest
that ZBP1 is a key regulator of macrophage phenotype and function, with its
knockout leading to significant changes in macrophage differentiation
patterns.
ZBP1 regulates macrophage metabolic and inflammatory status
in sepsis
We have delineated distinct macrophage populations, uncovering the
impact of ZBP1 disruption on macrophage diversity (Fig. 5A). We observed
a marked upregulation of pro-inflammatory genes such as iNOS, TNF, IL6,
and SPP1 in the WT CLP group, which was notably mitigated in the Zbp1−/−
mice (Fig. 5B, C and Supplementary Fig. 6A).
The flow cytometry data further confirmed these findings, indicating a
notable decrease in pro-inflammatory SPP1+cells within the ZBP1
Fig. 2 | ZBP1 is involved in sepsis-induced lung injury and mortality. A H&E
staining of lung sections from WT and Zbp1−/−mice 24 h post-sham or CLP surgery
(n= 5). Scale bar, 50 μm. BLung injury scores were obtained from the pathological
assessment of lung tissues. CBALF neutrophil counts were quantitated in each
group of mice. DProtein levels in BALF, indicative of vascular permeability, were
assessed using BCA assay. EEvaluation of lung tissue damage and vascular leakage
was performed using Evans blue dye extravasation, with quantification illustrating
statistically significant differences (n= 5). FLevels of pro-inflammatory cytokines
IL-6, IL-1β, and TNF-αin BALF were quantified using ELISA.
GImmunofluorescence staining for neutrophil marker Ly6G in lung tissues 24 h
post-CLP, with nuclei stained blue and Ly6G in red. The number of neutrophils per
high power field (HPF) was quantified (n= 5). HKaplan-Meier survival curves
illustrate the survival probability of WT and Zbp1−/−mice subjected to sham or CLP
over 96 h (n= 10). IPost-CLP core body temperature variations were monitored
(n= 5). JSerum lactate levels were determined 24 h post-CLP using a lactate mea-
surement kit (n= 5). Data are represented as mean ± SD, *P< 0.05,
**P< 0.01, ***P< 0.001.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 5
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Fig. 3 | scRNA-seq reveals the role of ZBP1 in altering transcriptional profiles of
various lung cells following sepsis. A Schematic diagram of sample collection to
single-cell sequencing analysis. BWestern blot analysis for ZBP1 in lung tissues of WT
and Zbp1−/−mice at 24 h post-CLPand sham procedures.CTheUMAPplotrevealsthe
heterogeneity of cell populations within lung tissue, with distinct clusters representing
ten different cell types as identified by single-cell analysis. DThe stacked bar chart
depicts the proportional distribution of cell types within each experimental group.
EViolin plots compare the prevalence of cell populations, capturing the variability in cell
type frequencies across samples. FThe bar graph enumerates the differentia llyexpres sed
genes (DEGs) across cell types, contrasting the transcr iptional profiles between WT and
Zbp1−/−mice post-CLP. GBox plots demonstrate the cell-type specificity, confirming
the cell purity within the identified populations. Data are presented as the mean ± SD,
*P< 0.05, **P< 0.01, ***P< 0.001.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 6
Content courtesy of Springer Nature, terms of use apply. Rights reserved

knockout landscape (Fig. 5D). Subsequent analysis of the M1 and M2
polarization scores revealed a tempered pro-inflammatory activation in the
macrophages of the Zbp1−/−groups (Fig. 5E, Supplementary Fig. 6B, C).
This trend continued with the percentage of iNOS+cells in primary mac-
rophages showing a significant increase in the WT CLP group, while ZBP1
knockout reversed this trend, signaling a departure from the classical pro-
inflammatory M1 phenotype in sepsis-induced ALI (Fig. 5F, G).
A characteristic feature of M1 activation is its distinct metabolic state,
divergent from that of resting macrophages. Cellular metabolic repro-
gramming is vital for macrophage activation and function. M1
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved

macrophages increase glucose consumption and lactate release while
reducing the oxygen consumption rate18.Ourfindings propose that ZBP1
promotes a shift towards an inflammatory M1 phenotype through meta-
bolic reprogramming. To investigate metabolic pathways, we quantified
activities across oxidative phosphorylation, glycolysis, and ROS pathways
and uncovered a metabolic reconfiguration in Zbp1−/−macrophages
(Fig. 5H, Supplementary Fig. 6D–G). This was further exemplified by ele-
vated ROS levels, highlighting the metabolic shifts propelled by ZBP1
knockout (Fig. 5I).
Furthermore, we determined the energy state of macrophages by
measuring ATP content and demonstrated that ZBP1 knockout alleviated
the reduction in cellular energy caused by CLP (Fig. 5J). The electron
microscopy analysis of mitochondria illustrated how the structural integrity
of these organelles corresponds with metabolic changes. Intact mitochon-
drial architecture was observed in sham conditions, while CLP resulted in
mitochondrial disruption, which was ameliorated by ZBP1 knockout
(Fig. 5K). The mitochondrial structural and functional alterations form a
crucial component of cellular metabolic reprogramming, with compro-
mised mitochondria driving cells to rely more heavily on anaerobic glyco-
lysis for energy production19.
The regulatory mechanisms of metabolic reprogramming involve a
plethora of signaling pathways and regulators, among which HIF1αhas
been identified as a key mediator of monocyte metabolic reprogramming
during sepsis20. At the transcriptomic level, we evaluated the expression of
genes, including HIF1-alpha, LDHA, and Slc2a1, revealing alterations in
mRNA levels indicative of metabolic adaptation following ZBP1 knockout
(Fig. 5L). The modulation of lactate production, a crucial gauge of glycolytic
flow, was apparent in the Zbp1−/−groups (Fig. 5M). Our findings reveal that
ZBP1 knockout diminishes the CLP-induced upregulation of HIF-1αand
LDHA protein expression (Fig. 5N).
These findings present an important regulatory influence of ZBP1 in
macrophage metabolism and inflammation.
ZBP1 regulates macrophage NLRP3 inflammasome activation
and pyroptosis
The analysis of differential gene expression in macrophages indicates that
ZBP1 influences the transcriptional landscape under septic stress (Fig. 6A).
The enriched KEGG pathways among upregulated genes suggested that
ZBP1 knockout significantly impacted biological processes related to
inflammation and cell death (Fig. 6B).
The distribution of NLRP3 expression across macrophage sub-
populations points to a direct correlation between ZBP1 knockout and
inflammasome regulation (Fig. 6C). The expression profiles of pyroptosis-
associated markers revealed increased levels of IL-8, NLRP3, AIM2, IL-1β,
and caspase-1 in the WT CLP group compared to those in the Zbp1−/−CLP
group, suggesting that the absence of ZBP1 dampens the inflammatory and
pyroptotic response (Fig. 6D). Activity scores for apoptosis pathways across
macrophage populations showed the apoptotic potentialwithin each group
(Fig. 6E). We observed the co-localization of ZBP1 with NLRP3 within lung
macrophages post-CLP, which may trigger the pyroptotic cascade (Fig. 6F).
This interaction is further confirmed by a proximity ligation assay (PLA),
revealing physical proximity between ZBP1 and NLRP3 in situ, suggesting a
potential mechanistic link in inflammasome activation (Fig. 6G).
The expression levels of pro-caspase-1 and its activated form, as well as
pro-gasdermin D and its activated form, reveal the proteolytic processing
events that culminate in pyroptosis in primary lung macrophages from the
WT CLP group (Fig. 6H). Furthermore, the presence of ASC specks points
to the assembly of the inflammasome complex, further validating the acti-
vation of pyroptotic pathways in these cells (Fig. 6I). Moreover, the iden-
tification of pyroptotic cells through TUNEL positivity in macrophages
underscores the active execution of cell death pathways, highlighting cells
undergoing pyroptosis, particularly in WT CLP macrophages as compared
to those in Zbp1−/−CLP macrophages (Fig. 6J).
These results demonstrate a critical role of ZBP1 in modulating the
activation of the NLRP3 inflammasome and subsequent macrophage
pyroptosis.
scRNA-seq reveals macrophages are a dominating regulator in
the lung cellular networks in sepsis
CellChat analysis was conducted to investigate the complexities of intercellular
interactions. We observed the role of ZBP1 in regulating cell-cell interactions.
In Zbp1−/−mice, we found significant decreases in the strength of the cell-cell
interactions (Fig. 7A, B). Particularly, the interaction strength between mac-
rophages and endothelial cells was diminished in the Zbp1−/−group as
compared to the WT group (Fig. 7C). Analysis of TNF and IL-1 signaling
pathways, which elucidated how the macrophage interactions with other cell
types were redistributed, demonstrated that macrophages are a dominating
regulator in controlling lung inflammation during injury (Fig. 7D, E).
Further pathway analysis revealed that the IL6 pathways were sub-
stantially enriched in the WT CLP group, whereas the ZBP1 knockout led to
an upregulation of VEGF pathways and a downregulation of IL6 signaling
(Fig. 7F). In addition, the SPP1 signaling pathway was substantially
downregulated in the Zbp1−/−group (Fig. 7G, H). Studies suggested that
SPP1, a cytokine influential in modulating immune responses, acts through
various receptors including integrins and CD44, initiating signaling cas-
cades that affect gene expression, cytoskeletal reorganization, and cell
survival21,22.
The t-SNE plots showed differential SPP1 mRNA distribution between
the WT and Zbp1−/−groups, indicating transcriptional changes in macro-
phage function due to ZBP1 knockout. Notably, there was a pronounced
reductioninSPP1expressioninthelungcellsinZbp1−/−animals (Fig. 7I, J).
This reduction likely contributes to mitigating macrophage-driven
inflammation.
Overall, inflammatory signaling pathways in macrophages from
Zbp1−/−mice were significantly attenuated in sepsis. They underscore the
pivotal role of macrophages, particularly in relation to endothelial cells, as
essential mediators of the cellular response to sepsis and point towards
macrophage signaling pathways as potential therapeutic targets to alleviate
the impact of acute lung injury.
ZBP1 is involved in sepsis-induced endothelial cell damage and
dysfunction
In sepsis-induced ALI, PVEC dysfunction results in increased vascular
permeability and disruption of the alveolar-capillary barrier, leading to
pulmonary edema and impaired gas exchange23,24. Our study highlights the
crucial role of ZBP1 in maintaining endothelial cell integrity. Based on
Fig. 4 | ZBP1 regulates alveolar macrophage differentiation in sepsis. A t-SNE-
based dimensionality reduction and clustering of macrophages from WT Sham, WT
CLP, Zbp1−/−Sham, and Zbp1−/−CLP groups reveal three macrophage sub-
populations: Macrophage C1, Macrophage C2, and Macrophage C3. BThe heatmap
illustrates differential gene expression with the top 15 genes highlighted within the
distinct macrophage subgroups. CViolin plots depict the M1 and M2 gene set
activity scores across the three macrophage subpopulations. DTrajectory analysis of
macrophage differentiation among subgroups, constructed using Slingshot, is
visualized in a scatter plot, with arrows denoting the direction of different iation and
developmental progression. ECytoTRACE analysis yields a scatter plot showing the
differentiation probability scores of macrophage subpopulations, ranging, and
scaled from 0 to 1, predicting relative cellular differentiation states. Scores closer to 0
indicate higher differentiation, whereas scores approaching 1 denote a less differ-
entiated state. FA t-SNE scatter plot visualizes the two-dimensional dispersion of
macrophage subpopulations. GBox plots reveal the variation in differentiation as
quantified by CytoTRACE scores among the macrophage subtypes. HLine plots
illustrate the expression changes of the top 9 genes with the highest correlation to the
starting and ending points of the differentiation pathway. IThe heatmap showcases
the expression levels of the top 30 genes with the highest correlation to the differ-
entiation trajectory endpoints. JThe cell percentage chart depicts the distribution of
different cell clusters among the groups. Data are presented as the mean ± SD,
*P< 0.05, **P< 0.01, ***P< 0.001 .
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 8
Content courtesy of Springer Nature, terms of use apply. Rights reserved

significant differences in gene expression, we have categorized the endo-
thelial cell landscape into three distinct subpopulations (Fig. 8A, B).
We observed a notable increase in the inflammatory endothelial cell E3
subpopulation with high expression of Lcn2 and Saa3 in the CLP group.
However, this increase was significantly reduced in the Zbp1−/−CLP group
(Fig. 8C). Moreover, compared to WT mice, the Zbp1−/−mice exhibited a
significant downregulation in the expression levels of inflammatory cyto-
kines and cell adhesion molecules, including Icam1, Vcam1, Nfkbia, Ccl2,
and Il6, underscoring the modulatory effects of ZBP1 on inflammatory gene
expression in response to sepsis (Fig. 8D, E).
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Consistently, in the WT group, sepsis decreased the protein expression of
endothelial junction proteins VE-cadherin and claudin 5 and increased the
expression of Icam1 and Vcam1. In contrast, ZBP1 knockout mitigated the
CLP-induced decrease in VE-cadherin and claudin 5 expression and the
increase in Icam1 and Vcam1 (Fig. 8F).ThesresultssuggestthatZBP1con-
tributes to endothelial barrier dysfunction by reducing endothelial adhesion
and tight junction proteins while increasing adhesion molecule expression.
Furthermore, we observed reduced endothelial cell apoptosis in the
Zbp1−/−CLP group by quantitating Annexin V/PI staining endothelial cells
(Fig. 8G). Lung tissue analysis provided further support for these findings, as
evidenced by fewer TUNEL-positive endothelial cells in Zbp1−/−CLP mice
compared to WT CLP mice (Fig. 8H).
These findings demonstrate that ZBP1 involves in sepsis-induced
endothelial cell damage and dysfunction, positioning ZBP1 as a therapeutic
target to alleviate endothelial injury in inflammatory and septic conditions.
Discussion
The current study using Zbp1−/−mice revealed that ZBP1 plays an
important role in promoting the development and progression of ALI fol-
lowing systemic inflammatory response in sepsis. ZBP1 induces mito-
chondrial damage and glycolysis, which modulates macrophage metabolic
status. This, in turn, increases the differentiation of macrophages into pro-
inflammatory states. Furthermore, it induces the activation of the NLRP3
inflammasome in macrophage and subsequent pyroptosis. These actions
collectively enhance the inflammatory signaling pathways between mac-
rophages and other cells. Additionally, ZBP1 also plays a role in inducing
endothelial cell dysfunction, which causes endothelium damage and
increased permeability. Taken together, these multiple functions of
ZBP1 significantly contribute to the severity of sepsis-induced ALI and
higher mortality (Fig. 9).
The results underscore the complex interplay between cellular com-
ponents within the lungs during sepsis. Subsequently, we observed elevated
expression of ZBP1 across multiple cell types, suggesting ZBP1 is an
important regulator in the inflammatory response25. Spontaneous activation
of ZBP1 leads to the necroptotic cell death of keratinocytes or intestinal
epithelial cells, which results in sterile autoinflammation26,27.Importantly,in
sham condition, there was no difference in the lung status between WT and
Zbp1−/−mice. However, in the context of sepsis, ZBP1 knockout significantly
improved outcomes by reducing the inflammatory factors, alleviating
endothelial dysfunction, and enhancing the survival rates of the mice.
The modulation of immune cell metabolism in sepsis impacts their
immune functions19. M1 macrophages, noted for their pro-inflammatory
functions, enhance glucose uptake and lactate secretion while reducing
oxygen consumption18,28. Our study utilized single-cell transcriptomics to
reveal a notable downregulation of oxidative phosphorylation (OXPHOS)
and an enhancement of glycolysis in macrophages during sepsis, which was
accompanied by an increase in ROS production. Notably, this increase in
ROS was substantially reduced in Zbp1−/−mice.
Mitochondrial dysfunction shifts energy production from OXPHOS to
anaerobic glycolysis to satisfy ATP demands29. The inhibition of the mito-
chondrial respiratory chain is linked to an increase in ROS production,
underlining the relationship between OXPHOS suppression and ROS
generation30. Metabolic reprogramming is part of the mitochondrial stress
response, allowing a damaged mitochondrial network to recover by
switching from oxidative metabolism to aerobic glycolysis, thereby
increasing the production of ATP and NADPH for energy and antioxidant
defense31,32.Themetabolicprofile of immune cells is crucial for determining
their inflammatory phenotype. In this study, we observed significant
metabolic reprogramming in lung macrophages during the septic response,
contributing to the progression of ALI. Similarly, research suggests that
liver-resident Kupffer cells also undergo metabolic reprogramming under
stress conditions, characterized by alterations in mitochondrial function or
lactate production, which may shift their phenotype toward inflammation
regulation and influence the progression of liver injury33,34.
OurstudyfoundthatZBP1knockoutsignificantly alleviated mito-
chondrial dysfunction and ROS accumulation in macrophages during
sepsis. Research shows that ZBP1 acts as an innate immune sensor for
mitochondrial genome instability, collaborating with cGAS to maintain
IFN-I signaling, which in turn promotes mitochondrial dysfunction and
cardiac injury11. Overall, our results provide new evidence for the role of
ZBP1 in regulating macrophage metabolic reprogramming through mito-
chondrial modulation in sepsis.
HIF-1αcritically regulates glycolysis by upregulating the transcription
of glycolytic enzymes and membrane transport proteins, thereby enhancing
glucose flux and glycolysis35. The products of glycolysis further regulate
HIF-1αactivity and metabolic reprogramming, which in turn promotes a
pro-inflammatory phenotype in macrophages36.Inthisstudy,wefoundthat
ZBP1 deficiency reduces lactate production and glyc olysis induced by sepsis,
as well as the elevated expression and transcriptional activity of HIF-1α.The
increase in glycolysis and the accumulation of succinate enhance the pro-
duction of IL-1β37. Consequently, the reduced release of inflammatory
cytokinessuchasTNF-αand IL-1βfollowing inflammasome activation in
ZBP1-deficient macrophages may be influenced by a combination of gly-
colytic and mitochondrial signaling pathways.
It has been reported that ZBP1 on mitochondria, potentially a key
downstream event of telomere stress signaling, leads to MAVS activation38.
MAVS facilitates the recruitment of NLRP3 to mitochondria, thereby
promoting the production of IL-1βand the activation of the NLRP3
inflammasome39. Studies have shown that upon pathogenic stimulation,
ZBP1, as a specific sensor, activates the NLRP3 inflammasome40,41.Our
findings suggest that the absence of ZBP1 dampens inflammatory and
pyroptotic responses, as indicated by lower expression levels of pyroptosis-
associated markers, including IL-8, NLRP3, AIM2, IL-1β, and caspase-1 in
the Zbp1−/−mice. Additionally, the interaction between ZBP1 and NLRP3
in lung macrophages following CLP points to a potential mechanism trig-
gering the pyroptotic cascade. Our study underscores the critical role of
ZBP1 in modulating inflammasome activity and pyroptosis.
The significant amelioration of endothelial cell damage and dysfunc-
tion in Zbp1−/−mice indicates the importance of ZBP1 in damaging
endothelial stability and alveolar-capillary barrier integrity during sepsis.
The ZBP1 knockout mitigated changes in endothelial junction integrity
caused by sepsis, particularly preserving the expression of VE-cadherin and
Claudin-5, while reducing the expression of adhesion molecules. Recent
studies have identified ZBP1 as an innate sensor during infection, regulating
cell death, inflammasome activation, and pro-inflammatory responses12,42.
Our findings indicate that the improvement in endothelial cell damage
Fig. 5 | ZBP1 regulates macrophage metabolic and inflammatory status in sepsis.
At-SNE visualization displays the macrophage subpopulations across four experi-
mental groups. BThe expression distribution of the genes Spp1 and Nos2 within the
macrophage subpopulations is illustrated. CSingle-cell analysis reveals the mRNA
expression levels of inflammatory markers, including iNOS, TNF, IL6, and SPP1,
across different macrophage groups. DAlveolar macrophages isolated from each
mouse group were analyzed by flow cytometry to determine the percentage of SPP1+
cells (n= 3). EBox plots show M1 scoring within macrophages of each group.
FiNOS+cells percentage in primary macrophages from each mouse group were
detected by flow cytometry (n= 5). GIF staining was used to visualize F4/80 (red)
and iNOS (green) in lung tissues 24 h post-CLP (n= 5). White arrowheads point to
F4/80+iNOS+macrophages. HViolin plots present the activity scores for oxidative
phosphorylation, glycolysis, and ROS pathways in macrophages of each group.
ILevels of ROS were measured in macrophages via flow cytometry (n= 5). JRelative
amounts of intracellular ATP in WT and Zbp1−/−macrophages. KThe structure of
mitochondria was evaluated in macrophages using transmission electron micro-
scopy, with black arrows indicating mitochondria (n= 5). LThe mRNA expression
levels of HIF1-alpha, LDHA, and Slc2a1 in single-cell transcriptomes. MLactate
levels in primary macrophages from each group were measured with a lactate assay
kit. NProtein expression levels of HIF1-a and LDHA in macrophages were assessed
by Western blot. All results are based on three independent experiments. Data are
represented as mean ± SD, *P< 0.05, **P< 0.01, ***P< 0.001.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved

mechanisms in sepsis by ZBP1 knockout is due to reduced programmed cell
death, leading to less cellular damage.
On the other hand, we found that ZBP1 knockout reduces the upre-
gulation of inflammatory genes such as Icam1,Nfkbia,Ccl2,andIl6in
endothelial cells during sepsis. This regulation suggests that ZBP1 is
involved in the inflammatory cascade during inflammation12,andour
results indicate that ZBP1 knockout may weaken the inflammatory sig-
naling pathways between macrophages and endothelial cells, potentially by
reducing the release of inflammatory factors, thereby mitigating the
inflammatory cascade.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 11
Content courtesy of Springer Nature, terms of use apply. Rights reserved

However, the current study has limitations. Our research mainly
focused on macrophages and endothelial cells,and future studies areneeded
to determine the specific role of ZBP1 in other cell types involved in
inflammation and disease pathogenesis. Additionally, although we identi-
fied that ZBP1 knockout reduces mitochondrial damage and observed
potential changes in glycolysis, glycolysis was not the main pathway
explored in this study. Future studies should include experiments using
glycolysis inhibitors, such as metformin, to further clarify the role of gly-
colysis in ZBP1-mediated effects.
In conclusion, our single-cell transcriptomic analysis of acute lung
injury in sepsis provides a foundational dataset for studying lung
damage in sepsis. Our study shows that ZBP1 knockout reduces mito-
chondrial damage and inhibits glycolysis, leading to altered macrophage
metabolism and reduced differentiation into pro-inflammatory states.
ZBP1 knockout also diminishes macrophage pyroptosis by inhibiting
NLRP3 inflammasome activation, weakening inflammatory signaling
across cells. Additionally, ZBP1 knockout helps alleviate endothelial
dysfunction and cellular damage, offering potential therapeutic impli-
cations for managing sepsis-related inflammation. Further under-
standing of the interactions between ZBP1, metabolic reprogramming,
mitochondrial homeostasis, and immune response through single-cell
transcriptomics and validation may bring new insights into the patho-
genesis of sepsis.
Materials and methods
Animal strains
The experimental protocols involving animals were rigorously reviewed and
received approval from the Institutional Animal Care and Use Committees at
the University of Pittsburgh and the VA Pittsburgh Healthcare System. The
University of Pittsburgh’s animal protocol number is IS0002501524045015,
and the VA Pittsburgh Healthcare System’s protocol number is 1617201.
C57BL/6 wild-type (WT) mice were sourced from Jackson Laboratories. ZBP1
knockout (Zbp1−/−
) mice were obtained from the University of Pittsburgh.
Cecal ligation and puncture model
For the procedure, mice received an intraperitoneal injection of 50 mg/kg
ketamine and 5 mg/kg xylazine for anesthesia. Following a 1.5 cm abdom-
inal incision, the cecum was externalized, securely ligated with 4-0 silk
sutures, and punctured once using a 22-gauge needle to create a through-
and-through puncture. The abdominal incision was then sutured closed
with 4-0 silk, and the cecum was repositioned internally. Post-surgery, the
mice were monitored for mortality at 6-h intervals during survival studies.
For specific experiments, mice were euthanized 24 h post-CLP to collect
blood, BALF, and lung tissues.
Apoptosis analysis
To evaluate cell apoptosis, we followed the procedures outlined in the
Annexin V-FITC/PI Cell Apoptosis Detection Kit (BD Biosciences, East
Rutherford, NJ, USA). Cells were treated with Annexin V-FITC binding
solution and propidium iodide (PI). Data from these assays were ana-
lyzed using FlowJo software (version 10.0.7, Tree Star, Inc., Ash-
land, OR, USA).
Western blot analysis
Protein extraction was performed using RIPA lysis buffer supplemented
with protease inhibitors (Sigma), and protein concentration was determined
using a BCA protein assay kit (Thermo). We loaded 30 μgoftotalprotein
into each lane and separated them on 10% SDS-PAGE, then transferred
onto PVDF membranes (Millipore, Billerica, MA, USA). The membranes
were blocked using 5% non-fat milk for 1 h at room temperature, followed
by overnight incubation with specific primary antibodiesat4°C.Thepri-
mary antibodies included CASP1, GSDMD, ZBP1, VE-cadherin, Claudin 5,
VCAM1, ICAM1, HIF1a, LDHA, β-actin, and GAPDH. Details on these
antibodies are provided in Supplementary Table 1. After incubation with
HRP-conjugated secondary antibodies, protein bands were visualized using
enhanced chemiluminescence (ECL) reagents (Merck Millipore). Images of
the bands were captured using a ChemiDoc imaging system (Bio-Rad).
Hematoxylin and eosin (H&E) staining and lung injury scoring
Lung samples were fixed in 4% paraformaldehyde, embedded in paraffin,
and then sectioned into 5-μm slices. These sections underwent H&E
staining using standard histopathological techniques. Observations of his-
topathological changes were made using a light microscope. The extent of
lung injury, which included criteria such as atelectasis, alveolar and inter-
stitial inflammation, hemorrhage, edema, necrosis, and overdistension, was
assessed in six sections from the lower lobes. The scoring was as follows: 0
indicated no injury; 1 indicated injury to 25% of the field; 2 to 50%; 3 to 75%;
and 4 indicated diffuse injury throughout the field. Independent patholo-
gists, who were blinded to the experimental groups, conducted the lung
injury evaluations.
Immunofluorescence staining
Cells or tissues are fixed at room temperature in 4% paraformaldehyde for
30 min, permeabilized with 0.1% Triton X-100 for 5 min, and then blocked
with 5% BSA at room temperature for another 30 min. Overnight incuba-
tion at 4 °C with primary antibodies against Ly6G, CD31, F4/80, iNOS,
NLRP3, and ZBP1 follows. The next day, the slides are incubated with
fluorescently labeled secondary antibodies (1:100) in the dark for 1 h. Cell
nuclei are stained with DAPI for 5 min. Samples are then observed and
imaged under a Nikon A1R confocal microscope.
Proximity ligation assay (PLA)
For the PLA, cells are seeded in confocal dishes, washed with PBS, and fixed
with 4% formaldehyde for 15 min. Blocking is done with 5% BSA for
30 min, followed by overnight incubation at 4 °C with antibodies against
ZBP1 and NLRP3. Probe incubation, ligation, and amplification steps are
performed according to the manufacturer’s instructions (Duolink Detection
Kit, DUO92102-1KT, Sigma-Aldrich). Nuclei are stained with DAPI and
slides are mounted. PLA samples are imaged and analyzed using a 60x
objective on a Nikon A1R confocal microscope.
Lung microvascular permeability assessment
Lung microvascular permeability is evaluated using Evans blue dye extra-
vasation. Thirty minutes before euthanasia, mice receive an intravenous
injection of Evans blue dye (20 mg/kg, Sigma). Lungs are perf used with PBS to
Fig. 6 | ZBP1 regulates macrophage NLRP3 inflammasome activation and pyr-
optosis. A Volcano plot highlighting differential gene expression in macrophages
between WT CLP and Zbp1−/−CLP mice. Upregulated genes are denoted in red,
while downregulated genes are shown in blue. BBubble chart illustrates enriched
KEGG pathways among upregulated differentially expressed genes. Ct-SNE
visualization displays the expression distribution of NLRP3 across macrophage
subpopulations in the four groups. The expression level is represented by the
intensity of red color, with 0 values not displayed. Dsingle-cell analysis showing the
mRNA expression levels of pyroptosis-related genes, including IL-8, NLRP3, AIM2,
IL-1β, and caspase-1. EViolin plots present the activity scores for apoptosis path-
ways in macrophages from each group. FConfocal microscopy was used to observe
the co-localization of ZBP1 (green) and NLRP3 (red) in lung macrophages from
both CLP and sham groups (n= 5). Co-localization analysis was performed using
ImageJ software. Scale bar, 20 μm. GProximity ligation assay (PLA) employed
specific antibodies against ZBP1 and NLRP3, with DAPI staining indicating nuclei.
Scale bars, 20 μm. HWestern blot analysis assessed the protein expression levels of
pro-caspase-1 (P45) and its activated form (P20), as well as pro-gasdermin D
(GSDMD, P53) and its activated form (P30), in primary lung macrophages from
each group. IPyroptosis was evaluated by detecting ASC specks using IF. JIF
staining detected F4/80 (red) and TUNEL (green) in lung tissues 24 h post-CLP
(n= 5). White arrowheads highlight macrophages positive for both TUNEL and
F4/80. Data are represented as mean ± SD, *P< 0.05, **P< 0.01, *** P< 0.001.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved

remove intravascular dye and then harvested. The lung tissue is homogenized
and incubated overnight at 37 °C with PBS containing 16.7% formamide. The
homogenate is filtered through a 70-μm mesh and plated in a 96-well plate.
Absorbance at 620 nm and 740 nm is measured, and the amount of extra-
vasatedEvansbluedyeiscalculatedbasedonastandardcurve
43.
Single-cell suspension preparation, library construction, and
sequencing
Twenty-four hours post-surgery, all mice were euthanized, and lung tissues
were swiftly excised. Under sterile conditions, tissues were washed twice
with ice-cold PBS supplemented with 0.04% BSA. Using sterile surgical
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved

scissors, the clean tissues were carefully minced into approximately 0.5 mm3
fragments and then placed in freshly prepared digestion solution. The
digestion was conducted at 37 °C for 30 min, with intermittent stirring every
10 min. The resulting cell suspension was filtered twice through a BD 70-μm
cell strainer, followed by centrifugation at 4 °C at 400×gfor 5 min. The pellet
was resuspended in an appropriate volume of medium, mixed with an equal
volume of red blood cell lysis buffer, and incubated at 4 °C for 5 min.
Afterward, the sample was centrifuged at 400×gfor 5 min, and the super-
natant was discarded. The pellet was washed once with medium and cen-
trifuged again at 400×gfor 5 min. After discarding the supernatant, the
sample was resuspended in 100 μl of medium. The freshly prepared single-
cell suspension was adjusted to a concentration of 30,000 cells per sample.
Library construction followed the manufacturer’s instructions for the 10x
Genomics Chromium Next GEM Single Cell 3’Reagent Kit v3.1 (Catalog
No. 000388). The libraries were sequenced using high-throughput
sequencing on the NovaSeq 6000 platform.
Single-cell data processing and cell identification
We utilized Cell Ranger software (version 5.0.0) from 10x Genomics to
demarcate cell barcodes and assign sequence reads to the corresponding
genomic and transcriptomic profiles. The output was a matrix representing
gene counts across cells. Quality control was enforced using critical
metrics: each cell was required to express between 500 and 7000 unique
genes (nFeature_RNA), have fewer than 50,000 total RNA counts
(nCount_RNA), and contain less than 5% mitochondrial genes.
For data analysis, we adopted the Seurat pipeline (version 5.0.3),
employing dimensionality reduction and unsupervised clustering for
quality checks. Data normalization was performed using NormalizeData
and ScaleData functions from Seurat44, selecting 2,000 highly variable genes
for detailed examination via FindVariableFeatures, and employing 19
principal components for further dimensionality reduction.
To mitigate batch effects, we applied the Harmony algorithm (version
0.1)45. Cluster dimensionality was adjusted using “dims = 1:19, resolution =
0.5”parameters in RunUMAP and FindClusters. Initial cell type prediction for
each cluster was conducted using scMayoMap software (version 0.1.0)46.For
refinement, the top 50 differentially expressed genes from each cluster were
analyzed using CellMarker 2.0, PanglaoDB, and the ACT database, enhancing
the accuracy of cell type assignments. In terms of marker gene presentation, we
focused on genes that showed cluster-specificexpressionfromthetop50
DEGs of each cluster, ensuring a balance between the breadth of candidate
genes and the focus on those most differentially expressed, thus likely to be
biologically significant and reliable markers of specific cell types or states.
ROGUE index
To assess the heterogeneity within different clusters, we employed the
ROGUE (version 1.0)47,whichquantifies the diversity of cell states or types
in each cluster. The ROGUE index was calculated using this package for
each cluster identified in our single-cell RNA sequencing data. Subse-
quently, we performed a statistical significance assessment of the observed
ROGUE values between the groups of interest.
M1 and M2 functional scores in macrophages
Using the AddModuleScore function on our single-cell RNA sequencing
dataset, we quantified the M1 and M2 functional scores for each macro-
phage cell, indicative of pro-inflammatory and anti-inflammatory activities,
respectively. These scores represent the relative expression levels of genes
linked to these specific functional modules. A higher score denotes greater
enrichment of the corresponding functional state. We based our selection of
gene sets for M1 and M2 modules on the macrophage functional gene set15,
which encompasses established markers and signature genes characterizing
these distinct functional states.
Metabolic pathway analysis
To evaluate differences in metabolic pathways among various cell groups,
we downloaded 98 metabolism-related pathways from the GSEA database
and calculated the pathway scores for distinct cell clusters using the
AddModuleScore function in Seurat. For identifying significantly altered
metabolic pathways, we applied the Wilcoxon test (P <0.05)tocompare
pathways between Zbp1−/−and WT mice post-CLP. Bonferroni correction
was used to adjust the p-values for multiple comparisons.
Pathway enrichment analysis
To unravel the functional mechanisms and biological pathways linked to
different cell clusters, we conducted Gene Ontology (GO) and Kyoto
Encyclopedia of Genes and Genomes (KEGG) pathway enrichment ana-
lyses using Metascape. For each cell cluster, we identified the top 150 dif-
ferentially expressed genes (DEGs) based on their natural logarithmic fold
change (logFC), with a threshold of logFC > 0.25. These DEGs are the most
notably upregulated genes within each cluster relative to the general
population. Our focus on these top DEGs was aimed at pinpointing the
crucial functional pathways and biological processes distinct to each cell
cluster.
Inference of cellular differentiation trajectories
To elucidate potential lineage differentiation and developmental trajectories
within our single-cell RNA sequencing dataset, we utilized two com-
plementary computational approaches: CytoTRACE (version 0.3.3)17 and
Slingshot (version 2.0.0)16. By computing a CytoTRACE score for each cell,
we can determine its relative position along a differentiation trajectory. Cells
with higher CytoTRACE scores are typically less differentiated, suggesting a
proximity to the start of the differentiation process, whereas cells with lower
scores are more differentiated, indicative of nearing the trajectory’send.
Fig. 7 | scRNA-seq reveals macrophages are a dominating regulator in the lung
cellular networks in sepsis. A Network diagrams demonstrate the differences in the
number of ligand-receptor pairs (left) and communication probabilities (right)
between cellsubpopulations within WT CLP and Zbp1−/−CLP groups. Theperipheral
solid circlesrepresent various cell subpopulations, withthe circle size correspondingto
the number of ligand-receptor pairs. Blue lines indicate stronger communication in
the WT CLP group, while red lines denote stronger communication in the Zbp1−/−
CLP group. The line thickness reflects the magnitude of communication change.
BHeatmaps depict the differential number of interactions (left) and interaction
strengths (right) among all cell subpopulations between the two groups. The y-axis
represents the ligand-expressingcells, the x-axis the receptor-expressing cells, and the
color scale the difference in communication probability. Bars at the top and right
represent the cumulative differences in communication probability along each axis.
CGraphical representations of macrophage interactions as ligand cellsaffecting other
cell subpopulations. Thesize of each circle denotesthe number of ligand-receptorpairs
within that subpopulation, and the width of the lines represents the probability of
communication, with thicker lines indicating higher probabilities. DHierarchical
plots of the TNF signaling pathway network and (E) IL1 signaling pathway network
show autocrine and paracrine signaling interactions within specified cell subpopula-
tions and the remaining cell subpopulations, respectively. Each colored circle repre-
sents a cell subpopulation, solid circles for ligand cells, and hollow circles for receptor
cells. The line thickness reflects the cell communication probability. FBar charts
illustratethe differences in enriched signaling pathways betweenthe Zbp1−/−CLP and
WT CLP groups. Pathways on the y-axis are colored red for significantly stronger
communication in the WT CLP group. Blue pathways indicate stronger commu-
nication in the Zbp1−/−CLP group. GA bubble plot represents the probability of
macrophage interactions as ligand cells with othercell types between the Zbp1−/−CLP
and WT CLP groups. The x-axis labels the cell pairs, differentiated by color (blue for
Zbp1−/−CLP, red for WT CLP), and the y-axis represents ligand-receptor pairs. HA
bubble plot visualizes the probability differences in macrophage-to-macrophage
ligand-receptor pairs between the Zbp1−/−CLP and WT CLP groups. It-SNE plots
show the distribution of SPP1 mRNA expression between the groups,with zero values
in gray and a gradient indicating the expression level. JWestern blot analysis was
performedto assess the protein expressionof SPP1 in macrophages from WT CLP and
Zbp1−/−CLP mice. All results are based on three measurements. Data are represented
as mean ± SD, *P<0.05,**P< 0.01, ***P<0.001.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 14
Content courtesy of Springer Nature, terms of use apply. Rights reserved

This method provides insights into the spectrum of cellular states and
potential lineage connections within our dataset.
To further explore the dynamic changes in gene expression and the
potential branching patterns of cellular differentiation, we applied Slingshot,
ahighlyflexible and robust method for inferring pseudotime trajectories.
Slingshot utilizes a novel approach based on principal curves and graph
theory to construct continuous developmental trajectories that capture the
progression of cells through different states or lineages.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 15
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Cell-cell communication and interactions
To explore the complex network of cell-cell communication among the
distinct cell populations identified in our single-cell RNA sequencing data,
we utilized the CellChat package (version 1.5.0)48. Initially, we created a
CellChat object using the normalized gene expression matrix along with cell
meta-data, which included cell type labels for each cell. This object forms the
basis for all subsequent analyses and stores essential information for
deducing cell-cell interactions. We then implemented the CellChat pipeline
to identify potential ligand-receptor interactions among the cell popula-
tions. CellChat uses a curated database of known ligand-receptor pairs,
analyzing their expression patterns across cell types to predict intercellular
communication networks. It quantifies communication probabilities and
signaling strengths for each pair, offering a detailed measure of the inter-
actions’likelihood and intensity.
Statistical analysis
All data were processed using R Studio (version 4.3.3) or GraphPad Prism
(version 9.5). The results are presented as the mean ± SD, based on data
from at least three independent experiments. Survival curves were generated
using the Kaplan-Meier method and differences were evaluated with the
log-rank test. Statistical significance was established at a threshold of
P< 0.05. For comparative analysis, Student’st-test was employed for two-
group comparisons, while one-way ANOVA was used for analyses invol-
ving multiple groups.
Reporting summary
Further information on research design is available in the Nature Portfolio
Reporting Summary linked to this article.
Data availability
The source data for the graphs and charts in the main figures are provided in
Supplementary Data 1–4. Uncropped original blot/gel images from the
main figures are available in Supplementary Fig. 7. The single-cell RNA
sequencing data have been deposited in the GSE278767. All other data can
be obtained from the corresponding author upon reasonable request.
Received: 27 July 2024; Accepted: 14 October 2024;
References
1. Meyer, N. J., Gattinoni, L. & Calfee, C. S. Acute respiratory distress
syndrome. Lancet 398, 622–637 (2021).
2. Tomazini, B. M. et al. Effect of dexamethasone on days alive and
ventilator-free in patients with moderate or severe acute respiratory
distress syndrome and COVID-19: the CoDEX randomized clinical
trial. JAMA 324, 1307–1316 (2020).
3. Karalapillai, D. et al. Effect of intraoperative low tidal volume vs
conventional tidal volume on postoperative pulmonary complications
in patients undergoing major surgery: a randomized clinical trial.
JAMA 324, 848–858 (2020).
4. Matthay, M. A. et al. Acute respiratory distress syndrome. Nat. Rev.
Dis. Prim. 5, 18 (2019).
5. Park, I. et al. Neutrophils disturb pulmonary microcirculation in sepsis-
induced acute lung injury. Eur. Respir. J. 53, 1800786 (2019).
6. Alsabani, M. et al. Reduction of NETosis by targeting CXCR1/2
reduces thrombosis, lung injury, and mortality in experimental human
and murine sepsis. Br. J. Anaesth. 128, 283–293 (2022).
7. Li, Z. et al. BMX represses thrombin-PAR1-mediated endothelial
permeability and vascular leakage during early sepsis. Circ. Res. 126,
471–485 (2020).
8. Sun, X. et al. Transcriptional switch of hepatocytes initiates
macrophage recruitment and T-cell suppression in endotoxemia. J.
Hepatol. 77, 436–452 (2022).
9. Godoy-Tena, G. et al. Epigenetic and transcriptomic reprogramming
in monocytes of severe COVID-19 patients reflects alterations in
myeloid differentiation and the influence of inflammatory cytokines.
Genome Med. 14, 134 (2022).
10. Yuan, F. et al. Z-DNA binding protein 1 promotes heatstroke-induced
cell death. Science 376, 609–615 (2022).
11. Lei, Y. et al. Cooperative sensing of mitochondrial DNA by ZBP1
and cGAS promotes cardiotoxicity. Cell 186, 3013–3032.e3022
(2023).
12. Jiao, H. et al. Z-nucleic-acid sensing triggers ZBP1-dependent
necroptosis and inflammation. Nature 580, 391–395 (2020).
13. Wu, Y. et al. Spatiotemporal immune landscape of colorectal cancer
liver metastasis at single-cell level. Cancer Discov. 12, 134–153
(2022).
Fig. 8 | ZBP1 is involved in sepsis-induced endothelial cell damage and dys-
function. A t-SNE analyses demonstrate dimensionality reduction and distinct
clustering among endothelial cells from the WT Sham, WT CLP, Zbp1−/−Sham, and
Zbp1−/−CLP groups, revealing three subpopulations. BThe heatmap details the
expression levels of the top 15 marker genes within each endothelial cell subgroup,
with red signaling higher expression and blue indicating reduced expression.
Ct-SNE visualizations underscore distinct patterns of subpopulation clustering
within the four groups. DThe volcano plot of the endothelial subpopulations
between Zbp1−/−CLP and WT CLP groups features genes with differential
expression; upregulated genes are marked by red dots and downregulated genes by
blue. ESingle-cell analysis shows a notable increase in mRNA levels of Icam1 and
Vcam1 in the WT CLP groups, which are diminished in the Zbp1−/−mice. FWestern
blot analysis depicts the differential protein expression of VE-cadherin, claudin 5,
Icam1, and Vcam1 in primary mouse endothelial cells, with GAPDH serving as a
loading control. GEndothelial cell apoptosis is assessed by Annexin V/PI staining
through flow cytometry (n= 3). HImmunofluorescence staining for CD31 (red) and
TUNEL (green) in lung sections post-CLP at 24 h (n= 5). Data are represented as
mean ± SD, *P< 0.05, **P< 0.01, ***P< 0.001.
MACROPHAGE
NLRP3 inflammasome activation
Pyroptosis
Inflammatory response
TNFD, IL-1E, IL-6
Macrophage differentiation
Damage and dysfunction
ENDOTHELIAL CELL
Mitochondrial dysfunction, Glycolysis
Proinflammation
M1 macrophage
ZBP1
Sepsis-
induced
ALI and
mortality
Fig. 9 | Summary model. The current study using Zbp1−/−mice revealed that ZBP1
plays an important role in promoting the development and progression of ALI
following systemic inflammatory response in sepsis. ZBP1 induces mitochondrial
damage and glycolysis, which modulates macrophage metabolic status. This, in turn,
increases the differentiation of macrophages into pro-inflammatory states. Fur-
thermore, it induces the activation of the NLRP3 inflammasome in macrophage and
subsequent pyroptosis. These actions collectively enhance the inflammatory sig-
naling pathways between macrophages and other cells. Additionally, ZBP1 also
plays a role in inducing endothelial cell dysfunction, which causes endothelium
damage and increased permeability. Taken together, these multiple functions of
ZBP1 significantly contribute to the severity of sepsis-induced ALI and higher
mortality.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 16
Content courtesy of Springer Nature, terms of use apply. Rights reserved

14. Aibar, S. et al. SCENIC: single-cell regulatory network inference and
clustering. Nat. Methods 14, 1083–1086 (2017).
15. Bischoff, P. et al. Single-cell RNA sequencing reveals distinct tumor
microenvironmental patterns in lung adenocarcinoma. Oncogene 40,
6748–6758 (2021).
16. Street, K. et al. Slingshot: cell lineage and pseudotime inference for
single-cell transcriptomics. BMC Genomics 19, 477 (2018).
17. Gulati, G. S. et al. Single-cell transcriptional diversity is a hallmark of
developmental potential. Science 367, 405–411 (2020).
18. Pearce, E. L. & Pearce, E. J. Metabolic pathways in immune cell
activation and quiescence. Immunity 38, 633–643 (2013).
19. Cheng, S. C. et al. Broad defects in the energy metabolism of
leukocytes underlie immunoparalysis in sepsis. Nat. Immunol. 17,
406–413 (2016).
20. Shalova, I. N. et al. Human monocytes undergo functional re-
programming during sepsis mediated by hypoxia-inducible factor-1α.
Immunity 42, 484–498 (2015).
21. Bill, R. et al. CXCL9:SPP1 macrophage polarity identifies a network of
cellular programs that control human cancers. Science 381,
515–524 (2023).
22. Hulsmans, M. et al. Recruited macrophages elicit atrial fibrillation.
Science 381, 231–239 (2023).
23. Yang, J. et al. Hemorrhagic shock primes for lung vascular endothelial
cell pyroptosis: role in pulmonary inflammation following LPS. Cell
Death Dis. 7, e2363 (2016).
24. Xiang, M. et al. Hemorrhagic shock activates lung endothelial reduced
nicotinamide adenine dinucleotide phosphate (NADPH) oxidase via
neutrophil NADPH oxidase. Am. J. Respir. Cell Mol. Biol. 44,
333–340 (2011).
25. de Reuver, R. et al. ADAR1 prevents autoinflammation by suppressing
spontaneous ZBP1 activation. Nature 607, 784–789 (2022).
26. Devos, M. et al. Sensing of endogenous nucleic acids by ZBP1
induces keratinocyte necroptosis and skin inflammation. J. Exp. Med.
217, e20191913 (2020).
27. Schwarzer, R., Jiao, H., Wachsmuth, L., Tresch, A. & Pasparakis, M.
FADD and caspase-8 regulate gut homeostasis and inflammation by
controlling MLKL- and GSDMD-mediated death of intestinal epithelial
cells. Immunity 52, 978–993.e976 (2020).
28. Liu, L. et al. Proinflammatory signal suppresses proliferation and shifts
macrophage metabolism from Myc-dependent to HIF1α-dependent.
Proc. Natl. Acad. Sci. USA 113, 1564–1569 (2016).
29. Owen, A. M. et al. Chronic muscle weakness and mitochondrial
dysfunction in the absence of sustained atrophy in a preclinical sepsis
model. eLife 8, e49920 (2019).
30. West, A. P. et al. TLR signalling augments macrophage bactericidal
activity through mitochondrial ROS. Nature 472, 476–480 (2011).
31. Timblin, G. A. et al. Mitohormesis reprogrammes macrophage
metabolism to enforce tolerance. Nat. Metab. 3, 618–635 (2021).
32. Nargund, A. M., Fiorese, C. J., Pellegrino, M. W., Deng, P. & Haynes, C.
M. Mitochondrial and nuclear accumulation of the transcription factor
ATFS-1 promotes OXPHOS recovery during the UPR(mt). Mol. Cell
58, 123–133 (2015).
33. Zhang, H. et al. Pre-operative exercise therapy triggers anti-
inflammatory trained immunity of Kupffer cells through metabolic
reprogramming. Nat. Metab. 3, 843–858 (2021).
34. Blériot, C. et al. A subset of Kupffer cells regulates metabolism
through the expression of CD36. Immunity 54, 2101–2116.e2106
(2021).
35. Kim, J. W., Tchernyshyov, I., Semenza, G. L. & Dang, C. V. HIF-1-
mediated expression of pyruvate dehydrogenase kinase: a metabolic
switch required for cellular adaptation to hypoxia. Cell Metab. 3,
177–185 (2006).
36. Palsson-McDermott, E. M. et al. Pyruvate kinase M2 regulates Hif-1α
activity and IL-1βinduction and is a critical determinant of the Warburg
effect in LPS-activated macrophages. Cell Metab. 21,65–80 (2015).
37. Tannahill, G. M. et al. Succinate is an inflammatory signal that induces
IL-1βthrough HIF-1α.Nature 496, 238–242 (2013).
38. Nassour, J. et al. Telomere-to-mitochondria signalling by ZBP1
mediates replicative crisis. Nature 614, 767–773 (2023).
39. Subramanian,N.,Natarajan,K.,Clatworthy,M.R.,Wang,Z.&
Germain, R. N. The adaptor MAVS promotes NLRP3 mitochondrial
localization and inflammasome activation. Cell 153,348–361
(2013).
40. Kuriakose, T. et al. ZBP1/DAI is an innate sensor of influenza virus
triggering the NLRP3 inflammasome and programmed cell death
pathways. Sci. Immunol. 1, aag2045 (2016).
41. Zheng, M., Karki, R., Vogel, P. & Kanneganti, T. D. Caspase-6 is a key
regulator of innate immunity, inflammasome activation, and host
defense. Cell 181, 674–687.e613 (2020).
42. Lee, S. et al. AIM2 forms a complex with pyrin and ZBP1 to drive
PANoptosis and host defence. Nature 597, 415–419 (2021).
43. Ulland, T. K. et al. Nlrp12 mutation causes C57BL/6J strain-specific
defect in neutrophil recruitment. Nat. Commun. 7, 13180 (2016).
44. Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell
184, 3573–3587.e3529 (2021).
45. Korsunsky, I. et al. Fast, sensitive and accurate integration of single-
cell data with Harmony. Nat. Methods 16, 1289–1296 (2019).
46. Yang, L. et al. Single-cell Mayo Map (scMayoMap): an easy-to-use
tool for cell type annotation in single-cell RNA-sequencing data
analysis. BMC Biol. 21, 223 (2023).
47. Liu, B. et al. An entropy-based metric for assessing the purity of single
cell populations. Nat. Commun. 11, 3155 (2020).
48. Jin, S. et al. Inference and analysis of cell-cell communication using
CellChat. Nat. Commun. 12, 1088 (2021).
Acknowledgements
We thank Chaowei Shang, Ph.D., Director of Microscopy Facility at the
University of Pittsburgh for assistance with confocal microscopy and
immunofluorescence. This work was supported by the US National
Institutes of Health Grant R01-HL-139547(J.F.) and R21AI185275 (J.F.), US
Department of Veterans Affairs Award 1I01BX004838 (J.F.) and
IK6BX006297 (J.F.).
Author contributions
All authors contributed to the study conception and design. Material and
animal preparation, data collection, and analysis were performed by T.G.,
Y.F., Q.W., Y.L., and P.A.L.; T.G., Z.W., T.R.B., Y.L., and J.F. planned the
project and conceived the experiments. T.G., Y.L., and J.F. wrote the
manuscript. All authors approved the final manuscript.
Competing interests
The authors declare no competing interests.
Additional information
Supplementary information The online version contains
supplementary material available at
https://doi.org/10.1038/s42003-024-07072-x.
Correspondence and requests for materials should be addressed to
Ting Gong or Jie Fan.
Peer review information Communications Biology thanks the anonymous
reviewers for their contribution to the peer review of this work. Primary
handling editors: Connie Wong and Joao Valente.
Reprints and permissions information is available at
http://www.nature.com/reprints
Publisher’sNoteSpringer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 17
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Open Access This article is licensed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License,
which permits any non-commercial use, sharing, distribution and
reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative
Commons licence, and indicate if you modified the licensed material. You
do not have permission under this licence to share adapted material
derived from this article or parts of it. The images or other third party
material in this article are included in the article’s Creative Commons
licence, unless indicated otherwise in a credit line to the material.If material
is not included in the article’s Creative Commons licence and your intended
use is not permitted by statutory regulation or exceeds the permitted use,
you will need to obtain permission directly from the copyright holder. To
view a copy of this licence, visit http://creativecommons.org/licenses/by-
nc-nd/4.0/.
© The Author(s) 2024
https://doi.org/10.1038/s42003-024-07072-x Article
Communications Biology | (2024) 7:1361 18
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com