WHAT IS ALREADY KNOWN ON THIS TOPIC

Previous research has highlighted the contradictory functions of ACKR2 in tumor progression and metastasis, demonstrating its dual roles as both antitumoral and prometastatic. However, the specific contribution of ACKR2-expressing cell types to the immune response against tumors remained unclear, necessitating further investigation.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

Selective targeting of ACKR2 expression in lung capillaries could enhance lymphocyte infiltration, promoting the immune response against metastatic cancer and pathogens, thereby influencing therapeutic strategies aimed at combating these diseases.

Introduction

Chemokines, small chemoattractant cytokines, play a pivotal role in dictating leukocyte migration by interacting with a specific family of seven-transmembrane G-protein coupled receptors.1–3 Atypical chemokine receptors (ACKRs), a subfamily of chemokine receptors, have emerged as critical regulators of the chemokine system. Unlike canonical chemokine receptors mainly expressed on immune cells, ACKRs are present on diverse cell types, including blood and lymphatic endothelial cells (BECs and LECs, respectively). While ACKRs do not induce classical signaling for leukocyte migration, they activate β-arrestin-biased signaling which results in scavenging, transport, or intracellular depot of chemokines. The main physiological role of ACKRs is to act in a complementary way with canonical chemokine receptors by fine-tuning chemokine levels, enabling the creation of gradients essential for guiding leukocyte trafficking within tissues.4 5

One of the ACKRs, namely ACKR2, functions as a regulator of inflammation by binding and inducing the degradation of a wide range of inflammatory CC chemokines that bind the canonical chemokine receptors CCR1–CCR5.6 In humans, ACKR2 expression is detected in LECs in the skin and gut, as well as in trophoblast cells in the placenta. Additionally, low levels of ACKR2 expression were reported in hematopoietic cells.7–11

The investigation of ACKR2 protein expression in the mouse system has been hindered by the absence of specific antibodies. Consequently, data on its expression were primarily derived from mRNA analysis and indirect labeling methods.11 However, studies utilizing Ackr2 KO mice unveiled its central role in regulating the inflammatory response, resulting in exacerbated inflammatory reactions,12 that impact the correct development of adaptive immune response. The impairment in chemokine clearance resulted in lymphatic congestion and suppression of dendritic cell migration.13 14 Furthermore, ACKR2 plays a crucial role in resolving inflammatory responses12 and in maintaining a balance between protective and pathological immune responses in infections.15

In the context of tumors, ACKR2 exhibits a dual role, demonstrating both protumoral and antitumoral functions. It is a negative regulator of cancer-related inflammation, influencing the local availability of inflammatory chemokines that can facilitate tumor growth in skin carcinogenesis and colon cancer models.16 ACKR2 expression in Kaposi sarcoma spindle cells controls tumor progression by inhibiting inflammatory chemokines (CCL2, CCL5 and CCL3), leading to reduced macrophage infiltration and angiogenesis.17 These findings suggest that ACKR2 inhibits cancer proliferation through its chemokine scavenger activity, preventing the recruitment of protumoral leukocytes that support and enhance tumor growth.

Conversely, in preclinical models of melanoma and breast cancer, ACKR2-deficient mice exhibited protection against metastasis. This protective effect was attributed to the increased recruitment and activation of neutrophils and natural killer (NK) cells with antimetastatic properties.18 19

To better understand ACKR2 expression and delineate its role in cancer and inflammation, a reporter and conditional Ackr2 knockout (KO) mouse was generated. This revealed that the organ displaying the highest ACKR2 level is the lung, where it is expressed not only by LECs, as observed in most organs, but also by BECs, particularly by specialized endothelial cells of alveolar capillaries called aerocytes. Selective depletion of ACKR2 in endothelial cells resulted in protection against metastasis and altered leukocyte infiltration in murine melanoma and colon carcinoma lung metastasis. In an acute lung injury model, the lack of endothelial ACKR2 resulted in increased lymphocyte extravasation. Endothelial ACKR2 in the lung appears to specifically regulate leukocyte extravasation, and its selective deletion from endothelial cells resulted in increased lymphocyte lung infiltration.

ResultsACKR2 is expressed at the highest levels in murine lungs

To investigate the expression of ACKR2 in murine tissues, a floxed and reporter Ackr2 KO mouse was generated by homologous recombination in C57BL/6 ES cells by flanking the exon 2 (ENSMUSE00000368995) of the Ackr2 transcript (ENSMUST00000050327) and an inverted construct of Luciferase-P2A-tdTomato, with loxP and lox2272 sites. Cre-mediated recombination of the floxed region inverts the construct into the proper orientation and excises the exon (online supplemental figure S1A,B).

Through crossing Ackr2 floxed mice with mice carrying inducible Cre-recombinase under the ubiquitin C promoter (UBC-Cre-ERT2) and after tamoxifen treatment, Ackr2 KO and reporter mice for all tissues were obtained (Ubc Ackr2luc-tdT/luc-tdT hereafter referred as ACKR2ΔUbc). (online supplemental figure S1C,D). The recombination was confirmed through PCR on genomic DNA (gDNA; online supplemental figure S1E) and by qRT-PCR (online supplemental figure S1F).

Subsequently, the expression and distribution of ACKR2 in murine tissues were assessed by excising organs from ACKR2ΔUbc mice and their corresponding control littermates. Luciferase imaging revealed low expression of ACKR2 in the spleen, liver, stomach, ependymal fat, diaphragm, bone, small intestine, colon, testis, and trachea (figure 1A). ACKR2 was minimally detectable in all other analyzed organs (figure 1A). Conversely, high levels of ACKR2 expression were observed in the lung (figure 1A,B), which was further confirmed by evaluating TdTomato fluorescence intensity (figure 1C).

Figure 1

ACKR2 is highly expressed by pulmonary BECs. Ex vivo quantification of luciferase average radiance of organs of ACKR2ΔUbc (A) and ACKR2ΔProx1 (D) reporter mice. Representative images of luciferase signal emitted by the lungs of ACKR2ΔUbc (B) and ACKR2ΔProx1 mice (E) (bottom) and their relative control littermates (top). Region of interest (ROI) was bidimensionally designed on WT lung (top) to detract the radiance emitted by controls. The color bars show the Radiant Efficiency [(p/sec/cm2/sr)/μW/cm2]. TdTom signal from lungs of ACKR2ΔUbc (C) and ACKR2ΔProx1(F) mice (bottom) and their relative control littermates (top). Data shown are representative of at least three mice analyzed in three independent experiments. (G) Representative flow cytometry plot showing the identification of LEC, BEC, and fibroblast populations in the lung gated on viable CD45- cells with CD31 and Pdpn expression combination. (H) Frequency of ACKR2/TdTom+ cells among LECs, BECs, fibroblast and double negative (DN) cells from the lung of ACKR2ΔUbc mice. (I,J) Bright-field microscopy on pulmonary BECs from WT (I) and ACKR2ΔUbc (J) mice for the expression of CD31 and ACKR2/TdTom. (K) Overlaid flow cytometry histogram showing ACKR2/TdTom expression in LECs and BECs in ACKR2ΔUbc mice (red peak) compared with WT littermates (gray peak) in the indicated organs. Results are representative of at least three mice in three independent experiments. Data are represented as mean (SD). P value was generated using the unpaired t-test. *p<0.05, **p<0.01, ***p<0.001. ACKR, atypical chemokine receptors; BEC, blood endothelial cell; LEC, lymphatic endothelial cell; WT, wild type.

Given that ACKR2 expression was previously reported in LECs,8 Ackr2 floxed mice were crossed with mice expressing an inducible Cre-recombinase under the control of the lymphatic-specific promoter Prox1. This breeding strategy resulted in the generation of selective ACKR2 KO and reporter mice specifically for LECs (Prox1-ACKR2luc-tdT/luc-tdT hereafter referred as ACKR2ΔProx). Luciferase and fluorescence imaging revealed low ACKR2 expression in the organs of these mice (figure 1D) including the lung (figure 1E,F), suggesting that in this organ ACKR2 is expressed by cells outside the lymphatic compartment.

ACKR2 is expressed by LECs and by lung BECs

To identify the cell types expressing ACKR2 in the organs found positive by In Vivo Imaging System (IVIS), tissues taken from ACKR2ΔUbc mice were disaggregated and labeled with selected antibodies for flow cytometry analysis. As a positive control, placenta was disaggregated and ACKR2 was found expressed by CD45- cells9 (online supplemental figure S2A). ACKR2 expression was not detectable in CD45+ cells in any of the organs studied including the lymph nodes, spleen, and lung (online supplemental figure S2B–D). ACKR2 was not expressed by fibroblasts (CD45-CD31+PDPN-) (figure 1G) while it was expressed by LECs (CD45-CD31+PDPN+) in all the organs analyzed (figure 1H,K). Notably, ACKR2 was detected in the lungs in 40% of BECs (CD45-CD31+PDPN-; figure 1G,H,K). The presence of ACKR2 in lung BECs was further validated through imaging flow cytometry (figure 1I,J).

Figure 2

Lung ACKR2+ BECs exhibit an anti-inflammatory gene profile, sharing similarities with the gene profile of capillary type I or aCaps. (A) Heatmap showing 1087 differentially expressed genes between BEC ACKR2+ (green) and BEC ACKR2- (blue) (p value<0.05). Genes are clustered considering Pearson correlation. The gene expression scale is shown as Z-score. (B) Ingenuity pathway analysis (IPA) identified pathways significantly upregulated (Z-score>2) or downregulated (Z-score < −2) in ACKR2+ versus ACKR2- BECs. (C–D) Overlap analysis of DEGs expressed by BECs ACKR2+ (green) and BECs ACKR2neg (blue) with analysis of Goveia et al.20 (C) and Gillich et al. 22 (D). aCaps, aerocytes; ACKR, atypical chemokine receptors; BEC, blood endothelial cell; DEG, differentially expressed genes.

Lung ACKR2+ BECs express RNA signature similar to lung capillary ECs

To better define the subpopulation of vessels expressing ACKR2 in the lung, a transcriptome analysis was performed on isolated BECs obtained from the lung of ACKR2ΔUbc/flox mice. Cells were sorted into PDPN-CD31+ACKR2+ and PDPN-CD31+ACKR2- populations. Additionally, transcriptomic analysis was performed on FACS-sorted BECs from the lungs of ACKR2ΔUbc (sorting strategy reported in online supplemental figure S3A). As shown by principal component analysis (PCA; online supplemental figure S3D), BEC ACKR2+ and ACKR2- and ACKR2ΔUbc display different gene expression profiles.

Unsupervised hierarchical cluster analysis revealed 1087 differentially expressed genes (DEGs) (p<0.01) between ACKR2+ and ACKR2- BECs (figure 2A). Ingenuity pathway analysis (IPA) indicated that ACKR2+ BECs exhibited reduced expression levels of genes associated with the migration and activation of myeloid cells, as well as genes involved in the coagulation cascade (figure 2B). Furthermore, through overlap analysis with published single-cell data of murine ECs, it was found that ACKR2+ cells exhibited a gene signature resembling that of type I alveolar capillary ECs (figure 2C) or “aerocytes” (figure 2D) a capillary type unique to the lung, specialized for gas exchange and leukocyte trafficking.20–22

T-distributed stochastic neighbor embedding (tSNE) analysis of lung data sets of the EC atlas21 (http://www.lungendothelialcellatlas.com/) further validated that ACKR2 is predominantly expressed by capillary type I endothelial cells, characterized by higher levels of Ly6a (Sca1) and Sirpα expression while being negative for TEK (Tie2) expression (figure 3A–D). Through FACS analysis of lung BECs obtained from ACKR2ΔUbc/flox mice, it was observed that ACKR2 is indeed expressed by a subset of cells that are Sirpα+ and TEK- (figure 3E–J). Additionally, selective expression of ACKR2 by aerocytes was confirmed through analysis of the tabula muris lung dataset (online supplemental figure S4).

Figure 3

ACKR2 is expressed by type I lung capillaries/aCaps. (A) t-SNE plots of lung ECs clustering and (B–D) of the expression of indicated genes based on the RNA-seq dataseset.21 (E, F G and H) FACS analysis of TdTomato, TIE2 and SIRPα expression in BEC subpopulations from the heterozygous reporter mice Ackr2ΔUbc/flox. (I) Uniform Manifold Approximation and Projection (UMAP)analysis of clustering made by PhenoGraph of 3000 concatenated BEC subpopulation events/sample. (J) Heatmap of relative marker expression among UMAP clusters of BECs. (K–L) Immunofluorescence images in the lung tissue of ACKR2ΔUbc mice. Nuclei stained with 4',6-diamidin-2-fenilindolo (DAPI-blue); endothelial cells stained with anti-CD31 (red); endogenous TdTomato (light blue) and anti- Lyve-1 (green) (K) or SIRPα (L). Scale bar: 25 μm. aCaps, aerocytes; ACKR, atypical chemokine receptors; BEC, blood endothelial cell.

Figure 4

ACKR2 regulates the expression of IFN-inducible genes in lung BECs. (A) Unsupervised hierarchical cluster analysis of genes differentially expressed by lung BECs isolated from ACKR2ΔUbc and heterozygous reporter Ackr2ΔUbc/flox. (B) Heatmap of overlap analysis of IFN pathway EC signature20 with ACKR2ΔUbc and ACKR2+ BECs transcripts. Genes are clustered considering Pearson correlation. The gene expression scale is shown as Z-score. (C) Highly significant enriched gene sets (GSEA) between ACKR2ΔUbc BECs (KO on the left) and ACKR2+ BECs (on the right). False discovery rate (FDR<0.25) was taken as significant. Pathways associated with the greatest Normalized Enrichment Score are taken into consideration. (D) Biological processes significantly upregulated (red) or downregulated (blue) in ACKR2ΔUbc BECs compared with ACKR2+ BECs (IPA analysis, |z-score|>+2). ACKR, atypical chemokine receptors; BEC, blood endothelial cell; EC, endothelial cell; IFN, interferon; IPA, Ingenuity pathway analysis; KO, knockout.

Finally, ACKR2 expression was investigated using confocal microscopy on sections of lungs from ACKR2ΔUbc/flox mice. It was observed that ACKR2 was expressed by both LECs identified as CD31+ and Lyve1+, as well as by BECs (CD31+ and Lyve1-) (figure 3K). Moreover, ACKR2 expressed by BECs was found to colocalize with the aerocyte marker Sirpα (figure 3L). These findings conclusively demonstrate that in the lung, ACKR2 is expressed not only by LECs but also by a specific subset of vascular capillaries known as aerocytes or capillary type I endothelial cells.

ACKR2 negatively regulates interferon response of lung BECs

To understand if ACKR2 expression modulates the transcriptome of BECs, a comparison of the gene expression profiles between BECs ACKR2ΔUbc and BECs ACKR2+ was performed. Unsupervised hierarchical cluster analysis revealed 1501 DEGs between the two groups (figure 4A). IPA indicates that ACKR2ΔUbc BECs display activation of pathways related to T cell differentiation and development while the pathway “infection of mammalians” was found significantly inhibited compared with BEC ACKR2+ (figure 4D). Transcriptomic data were also analyzed using Gene Set Enrichment Analysis (GSEA) and transcripts of ACKR2ΔUbc BECs were associated with interferon-gamma response (figure 4C). This finding gained additional support through an overlap analysis involving single-cell data from murine endothelial cells, which revealed that ACKR2ΔUbc BECs exhibit an increase in the expression of interferon-stimulated genes (ISGs) (figure 4B). In summary, these results indicate that ACKR2 expression regulates the interferon response in BECs, potentially affecting lymphocyte differentiation and activation.

Selective inactivation of ACKR2 on endothelial cells protects mice from lung melanoma and colorectal carcinoma metastasis

To study the functional relevance of ACKR2 on ECs, floxed ACKR2 mice were crossed with Cdh5-CreERT2 mice obtaining ACKR2ΔCdh5 mice, in which the deletion of ACKR2 occurs only in ECs. ACKR2 expression in the lung was confirmed by luciferase activity (online supplemental figure S5A) and selective deletion was confirmed by qRT-PCR on lung biopsies (online supplemental figure S5B). To understand the role of ACKR2 in controlling lung metastasis, a syngeneic B16F10 metastasis melanoma model was used in ACKR2ΔUbc, ACKR2ΔCdh5, ACKR2ΔProx1, and relative control littermates. Consistent with previously published findings,18 ACKR2ΔUbc mice exhibited reduced lung tumor burden compared with wild-type (WT) mice (figure 5A). Interestingly, ACKR2ΔCdh5 mice showed partial protection, displaying fewer metastases than WT mice but more than ACKR2ΔUbc mice (figure 5B,D). Conversely, ACKR2ΔProx1 mice did not exhibit a significant difference in metastatic score compared with WT mice (figure 5C). The partial protection observed in ACKR2ΔCdh5 mice was further supported by RT-PCR analysis of the melanoma-specific transcript S100b (figure 5E). Moreover, ACKR2ΔCdh5 mice were also protected from lung metastasis when intravenously injected with the colorectal carcinoma (CRC) cell line MC38/SL4 (figure 5F,G).

Figure 5

Selective inactivation of ACKR2 on ECs protects mice from melanoma and CRC metastasis. (A–C) Melanoma lung metastasis in the indicated mice ACKR2ΔUbc (dark gray bar), ACKR2ΔCdh5 (red bar) and ACKR2ΔProx1 mice (white bar) expressed as ratio compared with their relative control littermates. (D) Representative images of excised lungs from mice melanoma metastasis-bearing mice. (E) qRT-PCR of the expression of S100b on Gapdh in ACKR2ΔUbc, ACKR2ΔCdh5 and control littermate metastatic lungs. (F) Metastatic ratio of MC38/SL4 in ACKR2ΔCdh5 lungs compared with their control littermates. (G) Representative images of excised lungs from WT and ACKR2ΔCdh5 MC38/SL4 metastasis-bearing mice. (H) Percentage of MC38-EGFP+ cells in the lung after 6 and 24 hours from intravenous injection in WT (blue dots) and ACKR2ΔCdh5 mice (red dots). (I) Representative flow cytometry dot plot on single, live cells, CD45 neg and EGFP pos from the murine lung stained with anti-CD98 antibody intravenously injected 2 min before the end of the experiments. Extravasated tumor cells are CD98 negative while cells in the capillaries are CD98 positive. (J) Quantification of MC38-EGFP+ cells in the lung of WT (blue bar) and ACKR2ΔCdh5 mice (red bar) 6 and 24 hours after intravenous injection. At least six mice/group were used per three independent experiments. Data are represented as mean (SD). P value was generated using the unpaired t-test. *P<0.05, **p<0.01, ***p<0.001. ACKR, atypical chemokine receptors; BEC, blood endothelial cell; EC, endothelial cell; WT, wild type.

ACKR2 limits lymphocyte extravasation and activation in the metastatic lung

To assess whether the absence of ACKR2 on BECs could hinder the dissemination of tumor cells into the lung, mice were intravenously injected with MC38-EGFP cells. Comparable engraftment of tumor cells into the lung was observed in both ACKR2ΔCdh5 and WT mice (figure 5H). To formally demonstrate tumor cell extravasation in the lung, the anti-CD98 antibody that labels MC38 cells23 was administered intravenously 2 min prior to lung tissue processing in order to label only intravascular cells (figure 5I). At 6 hours, approximately 20% of MC38-EGFP cells were observed in the interstitial space of the lung. By 24 hours, all MC38-EGFP cells had extravasated into the lung interstitium, with no observed differences between WT and ACKR2ΔUbc mice (figure 5J). To investigate whether variances in metastasis burden were associated with differences in immune responses, immune profiling of B16F10 metastatic lungs obtained from ACKR2ΔCdh5, ACKR2ΔUbc, and WT mice was performed by flow cytometry (gating strategy online supplemental figure S6A). No differences were observed between WT, ACKR2ΔCdh5 or ACKR2ΔUbc mice in the distribution of lung monocytes, neutrophils, NK cells, or B lymphocytes. On the contrary, there was a significant increase in T cells in ACKR2ΔCdh5 lungs compared with both WT and ACKR2ΔUbc mice (figure 6A,C). The enhancement of T cells in ACKR2ΔCdh5 mice was due to an increase of both CD4 and CD8 T lymphocytes (figure 6D). These results were further confirmed by multiple qPCR analyses performed on RNA isolated from the lungs of tumor-bearing mice . In ACKR2ΔCdh5 mice but not in ACKR2ΔUbc compared with WT mice, there was an increased expression of the T cell transcripts CD4 and CD8 (figure 6B). To further investigate the role of T cells in the different susceptibility to metastasis, analysis of the T cell activation was performed (online supplemental figure S5C). In ACKR2ΔCdh5, there was a significant increase of effector memory with a concomitant decrease of naïve CD8 T cells compared with WT littermates (figure 6E and online supplemental figure S5D), while no difference was found in the percentage of effector memory CD4 T cells (online supplemental figure S5E). Similar results were found in the lung of MC38/SL4 metastatic mice where the percentage of T cells in ACKR2ΔCdh5 lungs was higher compared with WT mice, driven by an increased frequency of CD8+T lymphocytes (figure 6F,G). In this model, while no difference was found in the frequency of CD8+ effector memory cells (online supplemental figure S5F,G), the ratio of lung-infiltrating CD3+ T lymphocytes to CD11b+ myeloid cells was significantly increased in ACKR2ΔCdh5 compared with WT mice (figure 6H). This parameter is predictive of a more effective immune response, as myeloid cells often exhibit immunosuppressive properties.24

Figure 6

Increased frequencies of T cells in metastatic lungs of ACKR2ΔCdh5 mice. (A) Immune infiltrate in B16F10 metastatic lungs of ACKR2ΔCdh5 (red), ACKR2ΔUbc (gray) and WT mice (blue) calculated by FACS analysis. (B) qPCR analysis of the indicated genes in the lung of ACKR2ΔUbc (gray bar), ACKR2ΔCdh5 (red bar) mice and their relative control littermates (blue bar). qPCR data are relative to Gapdh expression. (C) Frequency of CD3+ T cells on CD45+ in the B16F10 metastatic lungs of WT (blue bar), ACKR2ΔCdh5 (red bar) and ACKR2ΔUbc (gray bar) mice. (D) Frequency of CD4+ and CD8+ T cells on CD45+ in the B16F10 metastatic lungs of WT (blue bar) and ACKR2ΔCdh5 (red bar) mice. (E) Frequency of naïve, central memory (CM) and effector CD8+ T cells on CD45+ leukocytes in the B16F10 metastatic lungs of WT (blue bar) and ACKR2ΔCdh5 (red bar) mice. (F) Frequency of CD3+ T cells on CD45+ cells in the MC38/SL4 metastatic lungs of WT (blue bar) and ACKR2ΔCdh5 (red bar) mice. (G) Frequency of CD4+ and CD8+ T cells on CD45+ cells in the MC38/SL4 metastatic lungs of WT mice (blue bar) and ACKR2ΔCdh5 (red bar). (H) CD3/CD11b ratio in the lungs of ACKR2ΔCdH5 (red bar) and WT mice (blue bar). All the frequencies were calculated on data obtained by FACS analysis. (I) Chemokine concentration in the sera of WT (blue), ACKR2ΔCdh5 (red) and ACKR2ΔUbc (gray) mice after intravenous injection of MC38/SL4. At least six mice/group were used per three independent experiments. Bars represent mean±SD. P value was generated using the unpaired t-test. *P<0.05, **p<0.01. ACKR, atypical chemokine receptors; WT, wild type.

To investigate whether endothelial ACKR2 regulates leukocyte recruitment through chemokine scavenging activity, an analysis of chemokine concentrations in the sera of tumor-bearing mice was conducted. Compared with WT mice and ACKR2ΔUbc, ACKR2ΔCdh5 mice exhibited elevated levels of circulating CCL2 while no statistically significant differences were found for the other measured chemokines (figure 6I). These findings suggest that endothelial ACKR2 plays a role in modulating CCL2 levels in the blood of lung metastasis bearing mice, impacting leukocyte recruitment during tumor progression.

Selective inactivation of ACKR2 on ECs modulates lung lymphocyte extravasation in LPS-challenged mice

To demonstrate that ACKR2 expressed by lung ECs controls T lymphocyte extravasation WT, ACKR2ΔUbc and ACKR2ΔCdh5 mice were challenged intranasally with 10 μg LPS. After 4 hours, analysis of lung leukocytes was performed by FACS analysis (gating strategy online supplemental figure S6A). In both WT and ACKR2ΔCdh5 mice, LPS induced an increase in lung neutrophils while ACKR2ΔUbc lungs had a higher basal proportion of neutrophils that was not significantly increased by LPS (figure 7A). The percentage of monocytes was reduced after LPS in ACKR2ΔUbc and ACKR2ΔCdh5 mice. In addition, ACKR2ΔCdh5 displayed an increased frequency of T lymphocytes compared with WT and ACKR2ΔUbc mice (figure 7A). These results indicate that the lack of endothelial ACKR2 reduces monocyte while promotes T cell recruitment while having no effect on neutrophil recruitment in the inflamed lung.

Figure 7

Selective deletion of ACKR2 in ECs modulate leukocyte extravasation in the lungs of LPS-challenged mice. (A) Frequency of indicated leukocytes on CD45+ cells in ACKR2ΔUbc, ACKR2ΔCdh5 and WT lungs calculated by flow cytometry analysis in basal conditions (empty dots) and after 4 hours intranasal instillation of LPS (full dots) (B). Representative flow cytometry dot plot on single, live cells from the murine lung stained with anti-CD45 antibody intravenously injected 3 min before the end of the experiments. Extravasated leukocytes are CD45 negative while vascular leukocytes are positive CD45 antibody. Frequency of neutrophils (C) and T cells (D) in pulmonary and vascular compartment in ACKR2ΔUbc, ACKR2ΔCdh5 and WT mice. (E) Representative immunofluorescence images of CD3 cells infiltrating lungs in WT and LPS challenged mice. (F) Quantification of CD3+ cells in the indicated mice treated with phosphate-buffered saline (PBS) or LPS. At least three mice/group were used. Data are represented as mean (SEM). P value was generated using unpaired t-test . *P<0.05, **p<0.01, ***p<0.001. ACKR, atypical chemokine receptors; WT, wild type.

To formally demonstrate leukocyte extravasation in the lung, the CD45-PECy7 antibody was administered intravenous 2 min before lung processing25 (figure 7B). Neutrophils were equally distributed in the blood vessels and the interstitial space of the lung of WT, ACKR2ΔCdh5 and ACKR2ΔUbc mice (figure 7C) indicating that the increased frequency of neutrophils is accounted by both intravascular and interstitial neutrophils. Otherwise, an increased frequency of infiltrating T cells was found in the lung of ACKR2ΔCdh5 compared with WT and ACKR2Δubc mice demonstrating an increased extravasation of T cells in the lung of these mice (figure 7D). The increased infiltration of T cells in the lungs of ACKR2ΔCdh5 compared with WT LPS-treated mice was also confirmed by immunofluorescence staining (figure 7E,F). In this model, chemokine concentrations do not differ between WT, ACKR2ΔCdh5 and ACKR2ΔUbc mice (online supplemental figure S6B).

Discussion

Using a novel Ackr2-reporter and multiple conditional KO strains, this study has provided critical insights into the role of Ackr2 in leukocyte recruitment and immune response in the lung. First, ACKR2 was found selectively expressed at high levels by BECs in the lung. Second, through RNA sequencing, multiparametric flow cytometry, and confocal microscopy, we identified type I alveolar capillaries also called aerocytes, as the major pulmonary compartment with constitutive ACKR2 expression. Third, deletion of Ackr2 led to an increased expression of ISG by BECs. Lastly, the deletion of Ackr2 from ECs increased lymphocyte extravasation promoting an antitumoral immune response in the lung.

In both mice and humans, the information on ACKR2 expression across various organs has been severely limited by the lack of specific tools, and at present is limited to the expression by LECs8 and by trophoblasts in the placenta.9 This study, utilizing a reporter mouse, revealed the ubiquitous homeostatic expression of ACKR2 by LECs across all murine tissues. Unexpectedly, the lung emerged as the organ with the highest level of ACKR2 expression, where, intriguingly, it was found selectively expressed not only by LECs but also by a specialized capillary type called type I or aerocytes. This is a recently described type of capillaries that is unique to the lung and specialized for both gas exchange and the trafficking of leukocytes.20 22 The gene profile of Ackr2+ BEC significantly overlaps with that of alveolar capillaries type I and aerocytes (aCaps), (figure 2), indicating enrichment of ACKR2 expression in this specific lung capillary subtype. This result was further validated through multiparametric FACS and confocal microscopy analyses, confirming the expression of ACKR2 by aCaps.

The deletion of ACKR2 from ECs, but not selectively from LECs provided protection against lung metastasis. This protection was less pronounced than the one achieved through ubiquitous ACKR2 deletion, where neutrophils with antimetastatic activity played a key role.18 In this context, the results presented in this manuscript confirm previous findings on metastatic protection in ACKR2 KO mice, utilizing the newly developed Ackr2-reporter and KO strain that was generated from C57BL/6 ES cells. This guarantees tha