Periodic Reporting for period 4 - KupfferCellNiche (Determining the instructive tissue signals and the master transcription factors driving Kupffer cell differentiation)
Berichtszeitraum: 2022-01-01 bis 2022-12-31
This ERC CoG was called KupfferCellNiche. The Kupffer cells (KCs) are the main liver-resident macrophage population of the body. Previous work from our lab on mouse Kupffer cells demonstrated that these cells express a unique liver-specific transcriptome as compared to lung, brain or intestinal macrophages. The signals from the micro-environment of Kupffer cells (which we coined as the “KC niche”) that imprint this unique identity were unknown. Moreover, the precise identity of the cells constituting the KC niche were also unknown. The ERC CoG KupfferCellNiche project therefore had 3 main aims: (i) identify the cells representing the Kupffer cell (KC) niche, (ii) determine the specific cell-cell circuits that imprint the liver-specific identity in KCs, (iii) unravel the master transcription factor that drives KC development.
This ERC CoG project revolutionized the KC field. First, we demonstrated that Kupffer cells are long-lived cells that self-maintain for months in the liver and do not derive from circulating monocytes, as previously assumed. However, using our KC-DTR mice we showed that depletion of KCs triggers the activation of stellate cells that temporarily produce monocyte-attracting chemokines and recruit monocytes to the liver. These monocytes engraft in a single wave, adopt the transcriptional profile of KCs within a week and become long-lived cells (Bonnardel et al. Immunity 2019). Second, we demonstrated that LXRa is the master transcription factor driving KC development (Scott et al. Immunity 2018). Third, using correlated light and electron microscopy we found that KCs do not exclusively reside in the blood stream, as was proposed in all textbooks, but project an important part of their body across the endothelial barrier to be in direct cell-cell contact with stellate cells. The fact that each KC is paired with a stellate cell solved a 100-year-old mystery. In 1876, Dr. Karl von Kupffer first identified “Sternzellen” by gold chloride staining and later proposed that these cells were macrophages because he found cells phagocytosing inkt particles with a similar distribution in the liver. Later it became clear that these were two different cells: inkt-phagocytosing Kupffer cells now known to be macrophages and gold-chloride-stained stellate cells now known to be mesenchymal cells. Our work revealed that the reason that Karl von Kupffer confused both cells is that the stellate cell is an essential part of the KC niche and that both cells are in fact always found in pairs. Fourth, using NicheNet to predict cell-cell interactions, followed by in vitro and in vivo validations using our own CRE-specific mice, we found that LXRa expression in KCs is induced by three signals that synergise together: Notch-signal from LSECs, BMP-signal from stellate cells and lipids from hepatocytes and that this is completely conserved across seven species, including the human, the monkey, the pig, the hamster, the chicken and the zebrafish (Bonnardel et al. Immunity 2019, Guilliams et al. Cell 2022). Importantly, each signal alone is insufficient to imprint the KC identity. This means that a monocyte will only become a KC if it is integrated within the complete hepatic 4-cell module (reviewed in Guilliams et al. Immunity 2022). Fifth, our research provided formidable tools to the field: the first KC-DTR and KC-CRE mice, the transcriptome of KCs across seven species, conserved KC-specific surface markers that form the basis of robust gating and microscopy strategies to isolate KCs purely for functional analysis or to track them in clinical samples, and a public www.livercellatlas.org portal through which all the omics data are searchable and freely downloadable.
This ERC CoG project revolutionized the KC field. First, we demonstrated that Kupffer cells are long-lived cells that self-maintain for months in the liver and do not derive from circulating monocytes, as previously assumed. However, using our KC-DTR mice we showed that depletion of KCs triggers the activation of stellate cells that temporarily produce monocyte-attracting chemokines and recruit monocytes to the liver. These monocytes engraft in a single wave, adopt the transcriptional profile of KCs within a week and become long-lived cells (Bonnardel et al. Immunity 2019). Second, we demonstrated that LXRa is the master transcription factor driving KC development (Scott et al. Immunity 2018). Third, using correlated light and electron microscopy we found that KCs do not exclusively reside in the blood stream, as was proposed in all textbooks, but project an important part of their body across the endothelial barrier to be in direct cell-cell contact with stellate cells. The fact that each KC is paired with a stellate cell solved a 100-year-old mystery. In 1876, Dr. Karl von Kupffer first identified “Sternzellen” by gold chloride staining and later proposed that these cells were macrophages because he found cells phagocytosing inkt particles with a similar distribution in the liver. Later it became clear that these were two different cells: inkt-phagocytosing Kupffer cells now known to be macrophages and gold-chloride-stained stellate cells now known to be mesenchymal cells. Our work revealed that the reason that Karl von Kupffer confused both cells is that the stellate cell is an essential part of the KC niche and that both cells are in fact always found in pairs. Fourth, using NicheNet to predict cell-cell interactions, followed by in vitro and in vivo validations using our own CRE-specific mice, we found that LXRa expression in KCs is induced by three signals that synergise together: Notch-signal from LSECs, BMP-signal from stellate cells and lipids from hepatocytes and that this is completely conserved across seven species, including the human, the monkey, the pig, the hamster, the chicken and the zebrafish (Bonnardel et al. Immunity 2019, Guilliams et al. Cell 2022). Importantly, each signal alone is insufficient to imprint the KC identity. This means that a monocyte will only become a KC if it is integrated within the complete hepatic 4-cell module (reviewed in Guilliams et al. Immunity 2022). Fifth, our research provided formidable tools to the field: the first KC-DTR and KC-CRE mice, the transcriptome of KCs across seven species, conserved KC-specific surface markers that form the basis of robust gating and microscopy strategies to isolate KCs purely for functional analysis or to track them in clinical samples, and a public www.livercellatlas.org portal through which all the omics data are searchable and freely downloadable.
These were the main 3 aims of the grant:
AIM1: Determine the TxFs through which the niche imprints the KC identity
We found that Kupffer cells highly express the following transcription factors; ZEB2, LXRa and SPIC. We found that ZEB2 is required to maintain the tissue-specific identity of Kupffer cells and that loss of ZEB2 induces loss of the Kupffer cell population. A big part of the liver-specific signature of Kupffer cells is controlled by LXRa and loss of LXRa induces the loss of this signature and the disappearance from the liver (Scott et al. Immunity 2018).
AIM2: Map the crosstalk between KCs and niche cells during KC development
We discovered that the Kupffer cell niche is composed of Endothelial cells, Stellate cells and Hepatocytes. The main surprise is that we found that each Kupffer cell is paired with one stellate cell and has part of its body in the space of Disse where it touches the stellate cell and the hepatocyte (Bonnardel et al. Immunity 2019). This finding will strongly impact the liver field, as in all textbooks it was written that Kupffer cells only reside in the bloodstream. This means that most figures found in review articles and textbooks will need to be redrawn.
AIM3: Demonstrate which niche factors drive KC development
We found that the reason that Kupffer cells strongly interact with stellate cells and endothelial cells comes from the fact that these cells provide the instructive factors that drive Kupffer cell development (DLL4 from endothelial cells and BMP9 from stellate cells). Importantly, these factors are completely conserved across seven species (Guilliams et al. Cell 2022).
Global output of the ERC CoG project:
Importantly the macrophage niche concept was quickly picked up in other tissues. In fact, when we profiled the lungs of Covid patients, we identified that lung macrophages recruited during Covid lacked many genes induced by GM-CSF, one of the signals produced by the lung macrophage niche. We therefore decided to launch a clinical trial in Covid patients and treated them with inhaled GM-CSF to restore the homeostatic cell-cell circuits of the lung macrophage niche. This improved the oxygenation in the patients. We are therefore convinced that understanding how tissues wire the macrophages in healthy tissues will provide powerful tools to reprogram macrophages in diseased tissues and restore homeostasis.
AIM1: Determine the TxFs through which the niche imprints the KC identity
We found that Kupffer cells highly express the following transcription factors; ZEB2, LXRa and SPIC. We found that ZEB2 is required to maintain the tissue-specific identity of Kupffer cells and that loss of ZEB2 induces loss of the Kupffer cell population. A big part of the liver-specific signature of Kupffer cells is controlled by LXRa and loss of LXRa induces the loss of this signature and the disappearance from the liver (Scott et al. Immunity 2018).
AIM2: Map the crosstalk between KCs and niche cells during KC development
We discovered that the Kupffer cell niche is composed of Endothelial cells, Stellate cells and Hepatocytes. The main surprise is that we found that each Kupffer cell is paired with one stellate cell and has part of its body in the space of Disse where it touches the stellate cell and the hepatocyte (Bonnardel et al. Immunity 2019). This finding will strongly impact the liver field, as in all textbooks it was written that Kupffer cells only reside in the bloodstream. This means that most figures found in review articles and textbooks will need to be redrawn.
AIM3: Demonstrate which niche factors drive KC development
We found that the reason that Kupffer cells strongly interact with stellate cells and endothelial cells comes from the fact that these cells provide the instructive factors that drive Kupffer cell development (DLL4 from endothelial cells and BMP9 from stellate cells). Importantly, these factors are completely conserved across seven species (Guilliams et al. Cell 2022).
Global output of the ERC CoG project:
Importantly the macrophage niche concept was quickly picked up in other tissues. In fact, when we profiled the lungs of Covid patients, we identified that lung macrophages recruited during Covid lacked many genes induced by GM-CSF, one of the signals produced by the lung macrophage niche. We therefore decided to launch a clinical trial in Covid patients and treated them with inhaled GM-CSF to restore the homeostatic cell-cell circuits of the lung macrophage niche. This improved the oxygenation in the patients. We are therefore convinced that understanding how tissues wire the macrophages in healthy tissues will provide powerful tools to reprogram macrophages in diseased tissues and restore homeostasis.
For this project we clearly went beyond the state of the art and developed many novel technologies, including spatial transcriptomics on the human and mouse liver (Guilliams et al. Cell 2022), spatial proteomics on human and mouse liver (Guilliams et al. Cell 2022), combined RNA (whole genome) and Protein detection (100-plex)(Guilliams et al. Cell 2022), combined RNA (whole genome) and protein detection (200-plex) on human and mouse liver cells via CITE-seq. The techniques developed worked even better than what we had anticipated. We planned to do 30-plex flow cytometry but ended up developing a CITEseq pipeline that allowed to screen 100 antibodies in parallel, for example. All the data from our Liver Cell Atlas can be mined by any researcher through the public www.livercellatlas.org portal. We have more than 1000 individual IP-addresses visiting our website each week since the publication of our Atlas!