From Heterogeneity to Plasticity: Endothelial Dynamics in Lung Disease

ABSTRACT

Endothelial heterogeneity and plasticity play an important role in lung development, homeostasis, and pathology. In recent years, increasing evidence has demonstrated that endothelial dysfunction contributes to the progression of various lung diseases, such as ADRS, PF, PH, and lung developmental disorders. Therefore, targeting endothelial cells could hold promising therapeutic strategies for preventing disease development. Although significant advances in technology have revolutionized our understanding of endothelial heterogeneity and plasticity, effective and curative treatment options remain limited. Here, we discuss the molecular and functional diversity of lung endothelial cells and their critical role in maintaining lung homeostasis and in lung pathologies. We also briefly describe advanced technologies, such as single‐cell RNA sequencing and spatial transcriptomics, to uncover complex cell communication and underlying mechanisms. Furthermore, this review will identify future research questions for developing therapeutic approaches targeting lung endothelial cells.

1. Introduction

Endothelial cells (ECs) lining the interior of all blood vessels play a crucial role in various functions, including permeability, leukocyte transmigration, hemostasis, angiogenesis, regulation of vasomotor tone, and innate and acquired immunity. The endothelium exhibits distinct molecular and functional diversities in different organs or specific subsets of vascular beds or blood vessel types, which enhance substantial heterogeneity in the vascular endothelium [1]. Thus, understanding the molecular and functional diversity of ECs under physiological and pathological conditions has important implications.

Lungs possess a unique function for gas exchange with specialized structural features that are characterized by a thin alveolar‐capillary barrier. The alveolar‐capillary network not only provides an extensive surface area but also a dense and intricate network of capillaries. This architecture efficiently exchanges gases between the air and the blood [2]. As a highly vascularized tissue, the lung vasculatures are composed of a unique microenvironment with various specialized ECs, which comprise approximately 30% of the total cells in the lung [3]. Recently, single‐cell RNA sequencing (scRNA‐seq) has revolutionized our understanding of endothelial heterogeneity and plasticity in lung development and pathology. These understandings or discoveries could further develop novel therapeutic targets or treatments for lung repair and regeneration.

This review highlighted discoveries in endothelial heterogeneity and plasticity in lung health and disease, including a groundbreaking exploration of functionally distinct capillary ECs in the lungs. In addition, we discuss the application of advanced techniques, such as scRNA‐seq and spatial transcriptomics, in providing mechanistic or novel insights into lung EC biology. Identifying novel EC subtypes, molecular expression, and cell‐cell communications has implications for lung engineering and the discovery of novel therapeutic strategies for lung regeneration.

2. Endothelial Heterogeneity in Lung Health

2.1. Molecular and Functional Diversity

ECs are heterogeneous across different tissues and exhibit the complex structures and metabolic functions of each tissue and organ [4, 5, 6]. Lung ECs are categorized into two broad populations: vascular ECs (arteries, veins, and capillaries) and lymphatic ECs. All of which share the pan EC markers, such as CDH5, PECAM1, CLDN5, and transcription factor ERG [7, 8]. Advances in scRNA‐seq have revealed that different EC subpopulations display unique gene expression profiles with distinct functions (Table 1). While vascular ECs exhibit function in supplying and draining blood, lymphatic ECs, lining the form of lymphatic vessels, are important for absorbing interstitial fluid and carrying it as lymph to a lymph node or collecting vein [9, 10]. A previous holistic review provided a comprehensive understanding of the role of lymphatic ECs in contributing to lung homeostasis and disease [11]. This review focuses on vascular ECs (arterial, venous, and capillary ECs), particularly highlighting a novel discovery in distinct subsets of capillary ECs (Figure 1).

Table 1.

Representative marker genes of human and mouse lung EC subpopulations.

Lung EC subpopulations	Human markers	Mouse markers	References GEO
Pan ECs	PCAT19, GNG11, TIE1, CLDN5, SASH1, CDH5, ERG, RAMP2, PECAM1, CAVIN2	Cavin2, Calcrl, Cldn5, Tie1, Cdh5, Pecam1, Sash1, Ramp2, Erg, Cav1	[9] Human GEO publicly available datasets (GSE164829, GSE135893, GSE136831, and GSE133747) Murine GEO publicly available datasets (GSE133747, GSE129605, GSE132901, and GSE133992)
Arterial ECs	DKK2, IGFBP3, FBLN5, SERPINE2, CLDN10, GJA5, CXCL12, BMX, LTBP4, HEY1	Dkk2, Gja5, Cxcl12, Bmx, Efnb2, Fbln2, Gja4, Dll4, Hey1, Htra1
Aerocytes	SOSTDC1, EDNRB, HPGD, CYP3A5, PRKG1, TBX2, RCSD1, EDA, NCALD, S100A4	Car4, Emp2, Ednrb, Rcsd1, Prx, Tbx2, Fibin, Igfbp7, Itga3, Kdr
General capillary ECs	FCN3, IL7R, CD36, NRXN3, SLC6A4, GPIHBP1, ARHGAP18, IL18R1	Gpihbp1, Plcb1, Glp1r, Bmp6, Cd74, Aplnr, Adgrl3, Sntb1, Kit, Cd93
Pulmonary venous ECs	CPE, CLU, C7, PTGS1, EFEMP1, PKHD1L1, PDZRN4, DKK3, PLAT, HDAC9	Cpe, Slc6a2, Fgl2, Ptgs1, Vcam1, Nr2f2, Hdac9, Prcp, Ackr3, Rgs5, Vwf
Systemic venous ECs	COL15A1, ZNF385D, EBF1, TSHZ2, FLRT2, OLFM1, CPXM2, PLVAP, TPD52L1, PDE7B	None
Lymphatic ECs	MMRN1, CCL21, PROX1, PKHD1L1, SEMA3D, TFF3, TM4SF18, FLT4, LYVE1, PDPN	Mmrn1, Reln, Ccl21a, Prox1, Tbx1, Klhl4, Scn3a, Sema3d, Pdpn, Flt4

Figure 1.

Endothelial heterogeneity in pulmonary vasculature. A. Endothelial heterogeneity contributes to different zonation of pulmonary vasculature from artery, arteriole, capillary, venule, to vein. There are three main types of endothelial cells, including arterial ECs, venous ECs, and capillary ECs. Each EC population expressed specialized markers. Notably, capillary ECs identified two subtypes, such as aerocytes and gCaps. These two capillary subtypes have different functions in lungs. B. Alveolar‐capillary network shows cell heterogeneity communication. Aerocytes are localized in the thinnest regions with AT1 cells and express Car4 and Ednrb, a role for gas exchange, while gCaps are localized in thick regions with fibroblast and pericyte and express Gpihbp1 and Plvap, a role for maintenance and repair. EC, Endothelial cells; gCap, general capillary cells; AT2, Alveolar epithelial type 2 cells; AT1, Alveolar epithelial type 1 cells; RBC, Red Blood Cell.

Arterial ECs, which line the interior surface of arteries and arterioles, play a role in regulating hemostasis, coagulation, and inflammatory responses [7]. During pulmonary arterial development, Cxcl12 is highly expressed through the arterial endothelium but is absent from the venous endothelium. Genetic depletion of Cxcl12 leads to branching defects in the proximal and distal pulmonary artery and pulmonary vascular hypoplasia. Single‐cell transcriptomic analyses further identified two distinct arterial EC subpopulations: one characterized by high expression of Gkn3 and Gj5, and another enriched for Cxcl12 and Alox12. Functional enrichment analyses suggest that these populations exhibit different biological roles. Specifically, Gkn3+ arterial ECs are associated with structural organization, extracellular matrix regulation, and multicellular organism development, whereas Alox12+ arterial ECs are more strongly linked to vascular development. Importantly, this transcriptional heterogeneity corresponds to spatial zonation along the proximal‐distal axis during embryonic pulmonary artery development. Alox12 expression is enriched in distal arterial ECs, whereas Gkn3 expression predominates in proximal arterial ECs, indicating positional specification within the developing pulmonary arterial tree [12]. In the human lung, arterial ECs are characterized by the expression of EFNB2, SOX17, BMX, SEMA3G, HEY1, LTBP4, FBLN5, GJA5, and GJA4. To resist transmural pressure and shear stress, arterial ECs express genes encoding tight and gap junction proteins (CLDN10, GJA5, GJA4, and FBLIM1). They also produce extracellular matrix components (FBLN5, FBLN2, MGP, BGN, LTBP4, LTBP1, and FN1) and protease inhibitors (SERPINE2, CPAMP8), which contribute to vessel wall elasticity and structural integrity [9].

In contrast to arteries, veins are thin‐walled and subject to low shear forces, which makes the postcapillary venules the primary location for leukocyte extravasation. The human pulmonary vasculature comprises both pulmonary and bronchial circulations. The pulmonary circulation delivers deoxygenated blood from the heart to the alveoli, whereas the bronchial circulation provides oxygenated blood to lung structural cells [7]. Venous ECs were identified through the canonical transcription factor NR2F2 (COUP‐TFII, COUP transcription factor 2), as well as VCAM1, ACKR1, and SELP expressions [9]. During lung development, pulmonary venous ECs (Vwf, Vegfc, Prss23) emerge as early as embryonic day (E) 12 and maintain relatively stable cell numbers from E12 to postnatal day (P) 14 [13]. Lineage tracing combined with single‐cell RNA sequencing further demonstrated that pulmonary venous ECs undergo expansion during normal postnatal development, accompanied by increased Slc6a2 expression from P0 to P14 and P28 [14]. Recent studies have identified 2 distinct populations in the lung, including pulmonary venous ECs and systemic venous ECs, that differ in COL15A1 expression and anatomical location. Pulmonary venous ECs are characterized by the absence of COL15A1 expression and are localized in the lung parenchyma, whereas systemic venous ECs are COL15A1‐positive and are localized to the airways and visceral pleura [9, 15]

At the capillary level, capillary ECs account for the largest proportion of EC subclusters from the vascular bed, playing distinct functions with distinct subpopulations (Figure 1B) [6]. First, aerocytes are large, complex cells that are localized in the thinnest regions, where they are associated with alveolar type 1 (AT1) cells and express sequestration and adhesion genes, suggesting a role in gas exchange and leukocyte trafficking. Second, general capillary cells (gCap) are stem/progenitor cells localized in thick regions of the pulmonary stroma and are important for the maintenance and repair of alveolar capillary endothelial cells. Furthermore, gCap also expresses Edn1, Nos3, and Ptgis, suggesting a source of vasodilators and a regulator of vasomotor tone via interactions with pericytes. Of note, whereas mouse gCap preferentially expresses MHC class II genes involved in antigen presentation, suggesting a function in immune surveillance, these genes are expressed by human aerocytes [2]. Both cell types can regulate each other through their ligands and receptors, such as aerocytes ligands (Apln, Kitl) through cognate receptor of gCap (Aplnr, Kit), or gCap ligands (Edn1, Vegfa) with cognate receptor (Ednrb, Kdr) on aerocytes [2]. During lung organogenesis, scRNA‐seq has revealed that aerocytes (expressing Car4, Igfbp7, and Kitl) emerged at E18 and originated from the gCap population, then persisted in relatively stable numbers through to P14 [13]. With time, aerocytes begin to highly express Kdr (Vegf receptor), which is an important factor for aerocyte formation [13, 16].

2.2. Endothelial Cells Crosstalk With Other Lung Cells

Discovering complex communication among cell types is vital in tissue homeostasis and disease pathogenesis. Understanding the intricate crosstalk between ECs and other lung cells in the alveolar‐capillary unit is crucial for developing novel therapeutic interventions for lung diseases. Here, we describe three common lung crosstalk: EC‐alveolar epithelial cells, EC‐mural cells, and EC‐immune cells, as well as major receptor‐ligand interactions (Table 2).

Table 2.

Major receptor‐ligand interactions in cell‐cell signaling in the lung.

Ligand	Receptor	Source cells	Target cells	Effect	References
BDNF	NTRK2‐FL	AT1 cells	gCap	Cell survival, migration, proliferation	[17]
BDNF	NTRK2‐T1	AT1 cells	gCap	Cytoskeleton rearrangement, vascular permeability	[17]
Vegfa	Kdr	AT1	Aerocyte	Aerocyte specification and alveolar morphogenesis	[16]
Edn3	Ednrb	AT1/SMCs	Aerocyte	Angiogenesis	[2, 18]
Apln/Kitl	Aplnr/Kit	Aerocyte	gCap	Bidirectional signaling maintains capillary homeostasis	[2]
Edn1	Ednrb	gCap	Aerocyte	Vasomotor control	[2]
Edn1	Ednra	gCap	Fibroblast, Pericyte	Vasoconstriction	[2]
PGI2	Ptgir	gCap	Pericyte	Vasodilation	[2]
Cxcl12	Ackr3	gCap	gCap	Angiogenesis modulation	[2, 19]
Angpt1	Tek	Fibroblast, pericyte	gCap	Vascular stability, alveologenesis	[2, 20]
MHC class II	TCR antigen	gCap	CD4+ T cells	Antigen presentation	[2]
TxA2	Tbxa2r	Leukocyte	Aerocyte	Leukocyte trafficking	[2]
DLL4, JAG1, JAG2	NOTCH3	Arterial ECs	Pericyte/SMCs	Arterial EC specification, Mural cells proliferation and differentiation	[9]
CXCL12	CXCR4	Arterial ECs	Lymphoid cells	Immune cells homing, migration, and survival	[9]

Endothelial cell‐alveolar epithelial cell interactions. During lung development, epithelial and ECs interact to form the alveolar‐capillary network [21]. Similar to capillary endothelium, which includes aerocytes and gCap, alveolar epithelium consists of two distinct cell types, AT1 and alveolar type 2 (AT2) cells. AT1 cells and aerocytes are crucial for gas exchange in the lungs, while AT2 and gCap ECs are bifunctional stem/progenitor cells essential for alveolar‐capillary maintenance and repair [2]. Various signaling pathways facilitate crosstalk between epithelial cells and ECs, including VEGFA, FGF, and Wnt [16, 22, 23]. Importantly, aerocytes express cognate receptors (Ednrb and Kdr), suggesting that they interact with AT1‐expressing ligands (Vegfa and Edn3) [2]. Among all lung cell populations, AT1 cells exhibit the highest Vegfa expression, underscoring their central role in maintaining the alveolar vascular niche. AT1‐derived Vegfa is required for the specification and development of aerocytes, which characterizes an extended net‐like morphology and stimulates alveolar morphogenesis independent of myofibroblasts [16]. Furthermore, endothelial‐specific aryl hydrocarbon receptor activity plays a crucial role in the prominent crosstalk between the endothelial and epithelial cells, which prevents endothelial‐epithelial cell barrier disassembly in response to virus infection [24]. AT2 cell differentiation is vital for alveologenesis, which is associated with the contribution of ECs in releasing angiocrine factors. For example, ECs upregulated SPARCL1 expression, mediating NF‐κB signaling pathway in reprogramming AT2 cells from the intermediate state to mature cells [25]. Interestingly, a significant increase in gCap CXCL12 signaling changes cell‐cell interaction in the lung with advanced COPD. This ligand could interact with CXCR3 and CXCR4 [26]. The question is whether, along with these receptors (CXCR3 and CXCR4) expressing AT2 cells [27], the CXCL12‐CXCR axis might contribute to the crosstalk between gCap and AT2 cells in regulating COPD progression.

Endothelial cell‐mural cell interactions. The participation of mural cells (pericytes and vascular smooth muscle cells (VSMCs)) in established blood vessels is required for vascular stabilization and maturation. Whereas VSMCs are primarily found in arteries and veins, pericytes are associated with capillaries [28]. VSMCs are a continuous single or multilamellar SMC layer that surrounds the EC lining and play a vital function in regulating blood flow and stability via contractility. Unlike VSMCs, pericytes do not exhibit a continuous layer, feature direct contact with ECs, and play crucial roles in modulating the alveolar‐capillary membrane integrity [29]. Although advances in new genetic technologies and imaging techniques allow us to study the contributions of pericytes in lung pathobiology and cell interactions, significant challenges are still encountered in distinguishing pericytes from other mural cells regarding a unique single molecular marker for pericyte identification. Recently, the emergence of integration scRNA‐seq with spatial transcriptomics demonstrated that HIGD1B serves as a specific and superior cell marker for human and murine lung pericytes [30]. In addition, two distinct subtypes of pericytes have been identified, including type 1 pericytes (HIGD1B^high PDGFRβ^low) and type 2 pericytes (HIGD1B^low PDGFRβ^high). Type 1 pericytes are believed to be quiescent, playing a role in maintaining capillary homeostasis and preventing EC proliferation. In contrast, type 2 pericytes display multipotent properties and are thought to inhibit VSMC migration from arteries to capillaries [30]. However, further studies need to be conducted to validate pericytes' function and plasticity. Indeed, to maintain capillary homeostasis, communication between pericytes and ECs is essential through various signaling pathways, such as angiopoietin‐1/Tie2, PDGF‐β/PDGFRβ, TGF‐β, VEGF, Notch, and Wnt [20, 31]. For instance, angiopoietin‐1 is released by pericytes and consequently binds to Tie2/Tek, which localizes in ECs. This event leads to vascular stability and participates in alveologenesis [20, 32]. According to communication between VSMCs and ECs in the pulmonary vasculature, VSMCs and ECs are essential players in vessel diameter and blood pressure regulation [33]. Not only is paracrine signaling, such as nitric oxide production and endothelin‐1 (ET‐1), the key mechanism in contributing to this crosstalk, but also increasing evidence suggests the contribution of myoendothelial gap junctions and extracellular vesicles. These underlying mechanisms were comprehensively elucidated in the previous work [33]. Most recently, ECs released factors, including CXCL12, PDGF‐B, ET‐1, and MIF, to stimulate VSMC proliferation‐mediated pulmonary vascular remodeling through upregulation of Forkhead Box M1 (FoxM1) transcription factor [34]. Additionally, the relationship between BMPR2 and SOX17 via prostacyclin signaling further demonstrated the role of EC in modulating VSMC proliferation [35]. Further studies are needed to design effective pharmacotherapies that orchestrate the complexity of EC‐mural cell communication and to gain a deeper understanding of the interactions among distinct subpopulations, such as type 1 pericyte, type 2 pericyte, and gCap.

Endothelial cell‐immune cell interactions. In addition to epithelial and mural cells. ECs are involved in interactions with lung‐resident immune cells (macrophages, dendritic cells (DCs), innate lymphoid cells (ILCs)) as well as recruited immune cells (monocytes, neutrophils). Our previous review highlighted the intricate interplay between macrophages and ECs in acute lung injury (ALI), providing potential signaling pathways or therapeutic targets to modulate this crosstalk for preventing ALI/ acute respiratory distress syndrome (ARDS) [36]. Here, we present several recent progresses in understanding the communication between ECs and other immune cells in the lungs. Under homeostatic conditions, ECs function as a barrier, regulating immune cell recruitment, trafficking, activation, and infiltration. However, under pathological conditions, such as hypoxia, infection, and lung injury, ECs initiate signaling pathways that facilitate the appropriate infiltration of immune cells into the alveolar‐capillary compartment. For instance, arterial ECs serve as a ligand sender, CXCL12, which binds to CXCR4, facilitating homing, migration, and survival of lymphoid and dendritic cells [9]. In terms of capillary EC subpopulations, while aerocytes specifically express adhesion and leukocyte‐sequestration genes, suggesting that they primarily contribute to leukocyte trafficking, gCap expresses MHC class II components, proposing that they could activate CD4+ T cells via TCR antigen‐MHC‐II axis [2].

3. Endothelial Plasticity in Lung Health

3.1. Definitions and Mechanisms

Endothelial plasticity, a fundamental characteristic of ECs, refers to their ability to undergo reversible changes in phenotype and function in response to various environmental and physiological cues [37, 38]. In the lung, endothelial plasticity is important for lung development and also contributes to lung pathology. During embryo development, vasculogenesis involves the differentiation of EC progenitors to form a primitive plexus. Following lung vascularization, ECs differentiate into arterial, venous, or lymphatic fates. ECs acquire a specialized role throughout adult life depending on physiological or environmental conditions, such as various mechanical, inflammatory, and metabolic environments [1, 28, 37, 38]. Hence, the phenotypic plasticity of lung ECs is remarkably capable of giving rise to vessels with diverse morphological, functional, and molecular signatures. The current understanding of lung endothelial plasticity highlights the change in cell morphology, function, and underlying mechanisms in response to environmental and physiological stimuli, such as hypoxia, shear stress, and inflammatory signals. Importantly, endothelial‐to‐mesenchymal transition (EndoMT) is a cell transdifferentiation process characterized by ECs progressively changing their endothelial phenotype into a mesenchymal phenotype. ECs lose endothelial marker expressions, such as VE‐cadherin and CD31, and a concomitant increase of mesenchymal marker expressions, such as α‐SMA and Vimentin. This transition is a vital feature in the pathogenesis of chronic lung diseases, such as PF or PH.

3.2. Role in Angiogenesis and Tissue Repair

Endothelial plasticity is indispensable for the intricate process of angiogenesis, a process by which blood vessels expand through the formation of new vessels from pre‐existing ones. This can be further classified into sprouting angiogenesis and intussusceptive angiogenesis. Sprouting angiogenesis is characterized by EC sprouting with two distinct EC phenotypes: tip cells and stalk cells. In contrast, intussusceptive angiogenesis occurs rapidly without active cell proliferation, invasion, or increased endothelial permeability [39, 40]. During sprouting angiogenesis, tip cells guide the new vessel growth by promoting sprouting, invasion, and migration, while stalk cells follow, forming the lumen of the new vessel through elongation and proliferation [41]. ECs can adapt to different environments by modulating vascular endothelial growth factor receptor (VEGFR) levels. The fates of tip and stalk cells are interchangeable and regulated by VEGFR2/DLL4/NOTCH feedback mechanisms. Mechanistically, VEGFR2 signaling induces the expression of NOTCH signaling in adjacent stalk cells, which in turn laterally suppresses tip cell identity [42], suggesting that VEGF, a pro‐angiogenic cue, may serve a dual role in stimulating the initial differentiation of tip‐stalk cells [43]. At the microvasculature level, transcriptomic analysis of pulmonary ECs identified aerocytes not only enriched tip cell genes, such as VEGFR2 and NRP1, at homeostasis and after injury, but also expressed stalk cell genes, VEGFR1. In addition, gCap enriched several stalk cell transcripts, such as HES1, HEY1, and JAGGED1 [44]. Further studies need to be conducted on the role of aerocytes in angiogenesis, contributing to lung health.

4. Endothelial Heterogeneity and Plasticity in Lung Disease

4.1. Acute Lung Injury and Acute Respiratory Distress Syndrome (ARDS)

Disruption of the alveoli‐capillary network is a hallmark of acute lung injury (ALI) and its severe complications, acute respiratory distress syndrome (ARDS), which leads to leukocyte extravasation, edema formation, and hypoxemia. ECs are essential contributors to the maintenance of alveoli‐capillary homeostasis and ARDS pathobiology [44, 45]. In homeostatic conditions, ECs are quiescent, with anti‐inflammatory mechanisms, platelet disaggregation, and vasodilation. Upon exposure to inflammatory stimuli or infectious pathogens, initially damaged ECs occur, and following ECs regenerate to repair the lung microvascular structure [44, 46, 47]. For example, in an intratracheal instillation of the diphtheria toxin (DT)‐induced lung EC depletion model, gCap serves as the stem‐like cells expressing apelin and other progenitor cell markers such as Procr and Cd34, then transitioning to a strong proliferative phenotype to rapidly replenish these depleted ECs, including aerocytes, which exhibit Foxm1 and Mki67 expression [48]. Consistently, the emergence of highly proliferating ECs in ALI indicates not only high expression of proliferation markers, such as Mki67, Ccnb2, Cdk1, and Cdc20, but also expression of EC markers, such as Cd31, Gpihbp1, Plvap, and Cd34 [44]. Mechanistically, activating transcription factor 3 (Atf3)‐expressing gCap increases gene expressions associated with angiogenesis, endothelial development, and differentiation, enhancing proportion and contributing to regenerative and proliferative processes in pulmonary ECs after influenza A virus subtype H1N1 infection [49]. Notably, in response to viral infection, the discovery of injury‐induced capillary ECs (iCAP) exhibited higher expression of gCap markers (Gpihbp1, Kit), along with lower expression of aerocyte markers (Ednrb, Car4), expression of unique markers (Sparcl, Ntrk2), and upregulation of MHC class II and Ifngr1. Notably, iCAP state reprograms bidirectionally from gCap and aerocytes and is independent of the presence of Krt5+ epithelium [50]. However, underlying mechanisms in the differentiation of the intermediate cell states need to be further investigated. For example, the signaling pathways regulate iCAP formation and sustenance in injured areas and their capacity to differentiate into canonical gCap or aerocytes (Figure 2). Notably, although the role of aerocytes in ALI progression via endothelial regeneration has been given attention and gained more evidence, such as an increase in the cell population and alterations in gene transcriptions, their specialized role in gas exchange by connecting with the alveolar epithelium has not been fully explored. Particularly, during lung injury, an enhancement of a transitional Krt5+ epithelial cell is essential in epithelial regeneration for gas exchange. Hence, more studies need to be done to elucidate the endothelial‐epithelial regeneration process, which could hold promise in developing interventions to alleviate ALI progression. Furthermore, under viral infection, venous ECs exhibits stem cell‐like properties for lung regeneration and repair. Venous ECs have a capability of proliferation and differentiation into aerocytes and gCap during vascular regeneration [14]. Depletion of COUP‐TFII (Nr2f2) in ECs, a regulator of venous EC identity, reduced endothelial proliferation and prevented effective recovery after viral lung injury in vivo [46].

Figure 2.

Capillary endothelial cell heterogeneity in lung injury. In the context of lung injury, generally capillary (gCap) ECs and aerocytes undergo the differentiation intermediate state and persist for at least one year after injury, termed injury‐induced capillary ECs (iCAP). It is characterized by unique marker expressions, such as SPARCL1 and NTRK2, and MHC class II and IFNGR1 upregulation. Interestingly, the iCAP state is derived bidirectionally from gCap ECs and aerocytes. One distinct subgroup of iCAP acquires a gCap‐like gene expression profile, another acquires a gCap and aerocyte‐like gene expression profile. gCap, general capillary cells; iCAP, injury‐induced capillary. Gpihbp1, Glycosylphosphatidylinositol anchored high density lipoprotein binding protein 1; Plvap, Plasmalemma vesicle associated protein; Nr2f2, Nuclear receptor subfamily 2 group F member 2; Sparcl1, Secreted protein acidic and cysteine rich like 1; Car4, Carbonic anhydrase 4; Ednrb, Endothelin receptor type B.

4.2. Pulmonary Hypertension (PH)

Pulmonary hypertension (PH), a heterogenous cardiopulmonary disease, is defined as an increase in mean pulmonary arterial pressure (mPAP) greater than 20 mmHg. Pulmonary arterial wedge pressure (PAWP) and pulmonary vascular resistance (PVR) distinguish pre‐capillary PH (PAWP ≤ 15 mmHg, PVR > 2 Wood Units (WU)), isolated post‐capillary PH (PAWP > 15 mmHg, PVR ≤ 2 WU), and combined post‐ and pre‐capillary PH (PAWP > 15 mmHg, PVR > 2 WU) [51, 52]. It is characterized by a pulmonary vascular remodeling with a spectrum of structural and functional changes occurring primarily in the distal part of the pulmonary circulation, leading to increased pulmonary vascular resistance, subsequently affecting the structure and function of the right ventricle [52, 53]. Pulmonary vascular remodeling is the primary pathological mechanism underlying stenosis and occlusion of the pulmonary arteries in PH, accompanied not only by endothelial dysfunction and fibroblast and SMCs activation but also by loss of capillary ECs and by an exaggerated inflammatory cells infiltration [54]. ECs are involved in the initial stage of vascular remodeling leading to PH. In particular, EC apoptosis was observed in the initial stages of the disease, following hyperproliferative apoptosis‐resistant ECs that may contribute to plexiform lesions formation [55]. At a specific distinct endothelial population, scRNA‐seq analyses demonstrate that hypoxia induces an anti‐apoptotic, proliferative phenotype in arterial ECs, whereas capillary ECs undergo apoptosis [56]. Specifically, the emergence of a de‐differentiated EC population, which was characterized by loss of endothelial tight junction genes (Cdh5, Cldn5) and a typical gCap markers (Plvap, Aplnr), as well as upregulation of a hyper‐transcriptomic state characteristic of progenitor cells (Rpl) and antigen presentation and inflammation (Cd74, RT1‐Da), plays an role in promoting vascular inflammation. Furthermore, persistent and robust changes in transcriptional activity occurred in ‘activated’ arterial ECs (Tm4sf1+ ECs), a hyperproliferative apoptosis‐resistant EC phenotype. Tm4sf1+ ECs exhibited persistent dysregulated cell growth and reduced capacity for vascular repair and angiogenesis, suggesting a central role in arterial remodeling in PH progression [57]. An integration of an endothelial lineage‐tracing mouse with scRNA‐seq has been shown to reveal only a slightly higher proportion of venous ECs in PH but not in four other EC populations (arterial, aerocytes, gCap, and lymphatic ECs) [58]. Interestingly, the role of aerocytes in PH progression has been highlighted, involving the upregulation of numerous genes associated with angiogenesis, cell localization, and negative cell death regulation. A recent study revealed that the differentiation from gCap to Cxcl12+ arterial ECs leads to an increase in arterial ECs in distal PH lung via HIF‐2α/Notch4 signaling [59]. Furthermore, one specialized microvascular EC progenitor, c‐Kit+ ECs, represents a subset of gCap [2] and is important for stimulating neonatal pulmonary angiogenesis and mitigating hyperoxia‐induced alveolar simplification via Forkhead box F1 (Foxf1) signaling [60]. In the lung vascular remodeling context, three subgroups of c‐Kit+ ECs were identified: Notch1‐expressing, Nr2f2‐expressing, and Foxf1‐expressing c‐Kit+ ECs. Each subgroup represents the specification and development of pulmonary arteries, veins, and aerocytes, respectively. In particular, upregulation of Nr2f2 expression in PH promotes the differentiation of c‐Kit+ ECs into pulmonary venous ECs, thereby alleviating vascular remodeling [61]. Additionally, c‐Kit+ expressing Nr2f2, Notch1, and Foxf1 may also play crucial roles in vascular remodeling, angiogenesis, and microvascular repair [62, 63, 64, 65]. However, the function of capillary ECs, including aerocytes and gCap in PH, remains to be fully elucidated. Further investigation is needed, particularly regarding the role of aerocytes in angiogenesis and their contribution to PH progression.

In addition, EndoMT may be involved in vascular remodeling, contributing to PH progression. EndoMT accumulates in the subendothelial space with an upregulation of α‐SMA fibers and vimentin phosphorylation features. This accumulation is associated with an alteration in BMPR2 signaling in regulating the effect of TGF‐β in PH [66]. In contrast, using the scRNA‐seq dataset demonstrated an undetectable EndoMT cell population with low expression of EC markers (Cdh5 and Pecam1) and high expression of mesenchymal markers (Acta2 and Col1a1) [58]. The absence of EndoMT in another PH mouse study may be attributed to its potentially transient nature, emphasizing the need to investigate different stages of disease progression to gain a deeper understanding of EndoMT as well as endothelial heterogeneity in PH pathogenesis. Recent scRNA‐seq analysis in the SU5416‐chronic hypoxia (SuHx) rat model has identified a de‐differentiated EC population in early phase with very low activity of both the endothelial‐restricted transcription factor (Erg) and the Friend Leukemia Integration 1 transcription factor (Fli1), which is primed to undergo EndoMT to give rise to fibroblast phenotype [57].

4.3. Chronic Obstructive Pulmonary Disease (COPD)

COPD is a heterogeneous lung disease and the third most common cause of death worldwide. It is characterized by persistent inflammation and parenchymal destruction in the airway (bronchitis‐bronchiolitis) and alveoli (emphysema) [67, 68]. There are three common cell types contributing to COPD pathogenesis: ECs, epithelial cells, and macrophages. Recently, the role of ECs in COPD pathology has been given attention. This review focuses on how EC heterogeneity and plasticity contribute to COPD. In an elastase‐induced mouse COPD/emphysema model, loss of endothelial marker expression (Vegfr2 and Vegfa) and blood vessel development and angiogenesis contribute to the development of emphysema [69]. In contrast, by using single‐cell transcriptional technologies on lung sample of COPD with advanced emphysema, although advanced COPD did not involve in EC population alterations in both macro and microvascular, capillary EC gene expressions displayed an increase in inflammatory signaling, cellular stress response, and vesicular trafficking, especially aberrant gCap CXCL12 signaling, and a decrease in angiogenesis (ID1, ID3, and LDB2) and blood vessel development [26]. Consistently, observations in COPD and alpha‐1 antitrypsin (AAT) deficiency patients show that capillary ECs exhibit an inflammatory transcriptional expression, but not in COVID‐19 or pulmonary fibrosis diseases. Mechanistically, NTRK2 expression was highly upregulated in gCap in COPD and AAT compared to healthy lung control, proposing that highly expressed NTRK2 in gCap is a hallmark of emphysematous lung disease [50]. Furthermore, EC‐derived LRG1 has shown a potential correlation with COPD development and progression [69]. The question is whether from those markers or signaling, such as CXCL12, NTRK2, LRG1, any specific mechanisms or therapeutic targets could be potential diagnostic and treatment for COPD development.

4.4. Pulmonary Fibrosis (PF)

Pulmonary fibrosis (PF), a chronic and progressive interstitial lung disease, is the most common type of idiopathic interstitial pneumonia [70]. Its pathogenesis is considered a consequence of genetic susceptibility and a variety of environmental factors, such as smoking, microaspiration, viral infection, air pollution, occupational exposures, and mechanical strain. The injury of alveolar epithelium leads to aberrant epithelial‐fibroblast communications, promoting the extracellular matrix deposition and lung interstitial remodeling [70, 71]. Recently, the important role of vascular endothelium in PF has been highlighted [72, 73, 74, 75, 76, 77], suggesting that EC dysfunction may contribute to PF progression and be a promising therapeutic target in PF prevention. The characteristics of lung vasculature of human PF at single‐cell resolution have been shown to include an expansion of ectopic peribronchial EC population in the distal lung parenchyma, COL15A1‐expressing ECs [72]. Furthermore, an increase in the abundance of systemic venous and capillary EC populations is involved in increased migratory and fibrogenic properties in PF lung, whereas the proportion of aerocytes and gCap significantly decreased [72, 73]. Pulmonary venous ECs showed profound alterations when studying fibrotic aged mouse lungs by single‐cell transcriptomic analysis, which have uncovered several distinct subpopulations. Specifically, Cyp4b1+ and Cpe+ venous ECs correspond to a quiescent state characterized by homeostatic gene expression. In contrast, Slc6a6+ and Ackr1+ venous ECs represent an activated state that emerges in response to lung injury and fibrosis. Activated venous ECs exhibit transcriptional signatures associated with inflammation and chemokine bioavailability [15]. Notably, Ackr1+ venous ECs are enriched in aged fibrotic areas, suggesting a pathogenic profibrotic function and regulation of inflammatory responses. In a recent study, novel subpopulations have emerged during fibrosis formation and delayed in the aging condition which are capillary ECs overexpressing Lrg1. Lrg1 promotes alveolar regeneration and angiogenesis through Tgf‐β signaling pathways [78]. A time course of alveolar dysfunction illustrating the severity of alveolar remodeling in PF was investigated through spatial transcriptomics. These results suggest that alveolar remodeling by damaging the alveolar‐capillary structure followed by subepithelial fibroblast activation, then subsequent myeloid cell infiltration and proliferation [79]. Furthermore, the transition of ECs into a mesenchymal phenotype may be involved in PF progression [80, 81]. FOXF1, a transcription factor, high expression in capillary ECs, serves as an anti‐fibrotic factor that modulates EC function and myofibroblast activation [77]. In addition, overexpression of SREBP2 in ECs regulates the mesenchymal and fibrosis‐related genes, such as TGF‐β, Wnt, Snai1, and α‐SMA [74]. At present, two available antifibrotic agents, nintedanib and pirfenidone, slow PF progression, but are not curative and have side effects that limit their efficacy [82]. Given the understanding of PF pathology, especially the contribution of ECs and gene‐specific expressions in ECs, the development of therapies in protecting and stabilizing EC functions could hold promise as novel therapeutic strategies in PF prevention.

4.5. Lung Development Disorders

Defective lung vascular development can lead to severe lung developmental disorders, including alveolar capillary dysplasia (ACD) and bronchopulmonary dysplasia (BPD) [4]. ACD is a rare and typically lethal congenital condition characterized by a profound failure of alveolar capillary formation, abnormal vascular patterning, and impaired alignment between capillaries and the alveolar epithelium, often accompanied by misalignment of pulmonary veins [4, 23, 83]. BPD is a chronic lung disease that primarily affects premature infants and is characterized by impaired alveolarization and dysmorphic vascular development. Impaired angiogenesis, endothelial dysfunction, and disrupted endothelial‐epithelial crosstalk contribute to defective alveolar septation [4, 17]. There is growing evidence that capillary ECs are the key pathological driver and potential therapeutic target in BPD. Transcriptomic analyses reveal that dysregulated biological processes in capillary ECs were significantly associated with leukocyte migration, vasculogenesis, and cell‐substrate adhesion, highlighting their central role in suppression of angiogenesis. Notably, a marked expansion of NTRK2+ gCap has been observed in human BPD lungs. RNA velocity analysis and lineage tracing suggest that these NTRK2+ gCap arise from pre‐existing gCap populations rather than from aerocytes during pulmonary injury. By integrating multiomic profiling, spatial transcriptomics, and functional validation, recent studies have demonstrated that the balance between full‐length NTRK2 (NTRK2‐FL) and truncated NTRK2 (NTRK2‐T1) isoforms critically regulates the regenerative capacity of gCap, including angiogenesis and alveolar development. Mechanistically, NTRK2‐FL contains an intracellular kinase domain that activates MAPK/ERK and PI3K/AKT pathways, thereby promoting EC survival, migration, and proliferation. In contrast, NTRK2‐T1 lacks the kinase domain and engages RhoGTPase and calcium signaling, leading to cytoskeletal rearrangement and increased vascular permeability. During BPD progression, an imbalance characterized by downregulation of NTRK2‐FL and upregulation of NTRK2‐T1 contributes to persistent vascular defects and impaired alveolar‐capillary regeneration [17].

In addition, the majority of ACD with misalignment of pulmonary veins (ACDMPV) cases were caused by genic deletions and mutations in the FOXF1 gene [84]. FOXF1 is a transcription factor essential for lung branching morphogenesis and pulmonary vasculature development [85]. It is expressed in aerocytes and gCap, as well as in multiple mesenchymal cells, including pericytes and fibroblasts. Notably, FOXF1+KIT+ gCap cells were recently identified as endothelial progenitor cells capable of engrafting into the neonatal lung tissue of ACDMPV mice, thereby promoting the formation of new alveolar capillaries [65]. Although FOXF1 activity and Foxf1 mRNA are detected in other ECs and mesenchymal cells, the highest FOXF1 activity has been predicted in aerocytes [85]. In ACDMPV lungs, aerocytes are markedly reduced or absent, whereas COL15A1+ systemic venous ECs were significantly expanded. This imbalance is attributed to uncontrolled VEGFA signaling within cell‐cell interactions among AT1 cells, aerocytes, and systemic venous ECs. Furthermore, loss of FOXF1 in ACDMPV disrupted gene expression in gCap and reduced gCap progenitors, thereby impairing the differentiation or survival of aerocytes. Because aerocytes are critical for alveolar gas exchange, their deficiency ultimately leads to hypoxemia at birth [23].

5. Single‐Cell and Spatial Omics in Studying Lung Endothelium

Multi‐omics technologies capture a more holistic picture of a cell's biological state by measuring and integrating multiple types of biological data, such as genomics, epigenomics, transcriptomics, and proteomics, from the same cell [86]. The advent and rapid development of scRNA‐seq have revolutionized the field of endothelial heterogeneity and molecular alterations in the lung, providing unprecedented resolution at the level of gene expression [2, 5, 9, 13]. For example, during early stages of lung development, single‐cell analyses identified the characterization of endothelial heterogeneity in various distinct EC subtypes with specific markers, such as arterial ECs (Vwf, Cxcl12, and Pcsk5), venous ECs (Vwf, Vegfc, and Prss23), gCap (Gpihbp1 and Kit), proliferating gCap (Mki67 and Gpihbp1), aerocytes (Car4, Igfbp7, and Kitl), and lymphatic ECs (Flt4 and Ccl21a) [13]. Thus, applying single‐cell omics technologies is a powerful approach for providing deeper insights into EC development, plasticity, and lung heterogeneity (Table 3).

Table 3.

Single‐cell RNA sequencing studies of lung endothelial cell heterogeneity in lung diseases.

Diseases	Species	Endothelial phenotypes	Markers	Characteristics	Reference
Acute lung injury/ARDS	Mouse	Venous ECs	Slc6a2, Bst1	Venous EC increase proliferation and differentiation to capillary ECs (gCaps and aerocytes)	[14]
	Mouse	Injury‐induced capillary ECs (iCAPs)	High expression (Gpihbp1 and Kit), low expression (Ednrb and Car4), Sparcl1, Ntrk2, MHC class II, Ifngr1	iCAP state arises bidirectionally from gCap and aerocytes after infection, significant increase in lung injury, and persists for at least a year	[50]
	Mouse	Ntrk2+ ECs	Ntrk2, Plvap	Ntrk2+ ECs derived from gCap ECs and upregulated in respond to injury	[87]
Pulmonary hypertension	Mouse	c‐Kit+ ECs	Kit, Pecam1, Eng, Cdh5, Ly6a, Cd34, Kdr, Klf4	c‐Kit+ ECs increase population and differentiate into pulmonary venous ECs and aerocytes	[61]
	Mouse	Capillary, venous, arterial, lymphatic ECs, Proliferating, Sftp+ ECs	gCaps (Nrp1, Sema3c), aerocytes (Car4), venous (Vwf, Prss23), arterial (Cxcl12, Mgp), lymphatic (Ccl21a, Prox1), proliferating (Top2a), Sftp+ ECs (Sftp protein gene)	Higher proportion of venous ECs in PH, Upregulation of MHC‐II and CD74 in arterial and gCap ECs Aerocytes related to apoptotic, pro‐migratory and pro‐angiogenic genes	[58]
	Mouse	gCap ECs	Gpihbp1, Plvap	gCaps adopt arterial ECs characteristic in PH	[59]
	Rat	‘Activated’ arterial ECs, lymphatic ECs. gCap EGR, ‘differentiated’ ECs	gCap‐ERG (Atf3, Egr1, Fos, Zfp36, Btg2), ‘activated’ arterial ECs (Tm4sf1, Fn1, Amd1, Cxcl12, Serpine1), lymphatic ECs (Ccl21, Mmrn1, Lrg1, Nrp2, Igfbpl1), ‘differentiated’ ECs (Sftpc, Scgb1a1, Hba‐a2, Scgb3a1)	An early and sustained increase in ‘activated’ arterial and ‘differentiated’ EC populations Later expansion in arterial and lymphatic ECs ‘Differentiated’ ECs loss endothelial identity (Cdh5, Cldn5) and increase ribosomal proteins (Rpl), antigen presentation and inflammation (Cd74, RT1‐Da), and EndMT related genes (Zeb2, Nr2f2, Tpt1, Crip1, and S100a4) ‘Activated’ arterial ECs (Tm4sf1+ ECs) exhibit a growth dysregulated state contributing to arterial remodeling	[57]
	Rat	Endothelial arterial type1, endothelial arterial type 2, endothelial capillary cells	Endothelial arterial type 1 (Nostrin), Endothelial arterial type 2 (Tm4sf1), endothelial capillary cells (Car4)	Tm4sf1+ ECs may have stem/progenitor potential, exhibit a distinct transcriptomic signature enriched for angiogenesis and Cxcl12 signaling	[88]
COPD	Human	Capillary ECs	Aerocytes (CA4, HPGD), gCap ECs (FCN3, NOSTRIN)	No significant differences in the proportion of ECs in advanced COPD Capillary ECs related to inflammatory signaling and cellular stress Upregulation of genes in aerocytes (TNFRSF10D, IRF1), gCap ECs (TNFAIP3, IFI6, IL6) CXCL12+ gCap ECs increased in COPD	[26]
	Human	gCap ECs	NTRK2	NTRK2 was significantly upregulated in gCap ECs in emphysematous diseases	[50]
Pulmonary fibrosis	Human	Peri‐bronchial ECs	COL15A1	Increase the population in the distal IPF parenchyma	[72]
	Human	gCap ECs, Systemic venous and capillary ECs	gCap ECs (NOSTRIN, FCN3, BTNL9, CX3CL1, ICAM1), systemic venous ECs (COL15A1, SPRY1, ZNF385D, POSTN), systemic capillary ECs (COL15A1, INSR, ZNF385, ZNF385D)	A decrease in population of gCaps and an increase in population of systemic venous and capillary ECs	[73]
	Mouse	Capillary ECs	Lrg1+ gCaps (Lrg1, Aplnr), Lrg1+ aerocytes (Lrg1, Ednrb)	Capillary ECs reprogram into a pro‐angiogenic phenotype expressing Lrg1 in the fibrosis model	[78]
	Mouse	Pulmonary venous ECs	Bst1, Slc6a2, and Amigo2 quiescent venous ECs (Cyp4b1 and Cpe), activated venous ECs (Slc6a6 and Ackr1)	Slc6a6+ activated venous ECs generated from pre‐existing Cyp4b1+ post‐capillary venules. Ackr1+ venous ECs exhibited inflammatory and angiogenic features and were generated from pre‐existing Fabp4+ venous ECs	[15]
ACDMPV	Human	Aerocyte, gCap, Systemic venous ECs	Aerocyte (HPGD, EDNRB), gCap (FOXF1, KIT), systemic venous ECs (COL15A1, ABCB1)	Reduced or absent aerocytes and loss of gCap function Expansion of systemic venous ECs	[23]
BPD	Human and mouse	NTRK2+ gCap	CLIC5, NTRK2	Emerge NTRK2+ gCap in BPD Imbalance of NTRK2‐FL and NTRK2‐T1 leads to impaired vascular and alveolar development	[17]

Recent advances in spatial transcriptomics enable high‐resolution mapping of gene expression directly within tissue sections, overcoming the loss of anatomical context inherent in dissociation‐based methods. For instance, the Xenium platform utilizes multiplexed fluorescent in situ hybridization (FISH) to visualize the spatial distribution of mRNA molecules in the lung of a PH mouse model, identifying multiple cell populations – including ECs, SMCs, fibroblasts, macrophages, pericytes, AT1, and AT2 cells – along with their precise spatial localization [89]. Furthermore, high‐resolution spatial transcriptomics have also revealed abnormal changes in cell proportions and cell‐cell interactions in disease states, such as expansion of arterial ECs and reduction of gCap and aerocytes in PH [89]. Integration of scRNA‐seq and spatial transcriptomics technologies enables mapping gene signatures within tissue context, providing crucial insights into the spatial organization of different EC types and their responses in specific regions of the lung, particularly in diseased states.

6. Therapeutic Implications

6.1. Targeting Endothelial Dysfunction in Lung Disease

Given the critical role of endothelial dysfunction in the pathogenesis of many lung diseases, targeting the endothelium has emerged as a promising therapeutic strategy. Novel therapeutic approaches should be investigated to restore normal endothelial function in conditions such as ARDS, PH, COPD, and PF. For example, an enrichment of EC phenotypes involving the systemic vasculature (systemic venous ECs and systemic capillary ECs) in progressive PF may represent an interesting anti‐fibrotic target. Furthermore, both pulmonary venous ECs and gCap have been played a role as progenitor stem cells, targeting on promoting these cells differentiation during injury or underlying mechanisms on particular endothelial regeneration could open more avenues of inquiry.

Despite significant advances, no curative treatment options are currently available for lung disease. Cell‐based therapies have emerged as promising therapeutic interventions in promoting damaged tissue regeneration. Besides mesenchymal stromal cells (MSCs), endothelial progenitor cells (EPCs) therapy is being explored, which has been investigated in regenerative medicine and potential treatment for a variety of lung diseases, including COPD, IPF, and PH [90, 91]. There were two types of EPCs: multipotent EPCs (c‐Kit+ Aplnr+ Foxf1+) in the peripheral pulmonary microcirculation that differentiate into capillary, arterial, and venous EC, and endothelial colony‐forming cells (Sca1+ Prorc+) in the proximal pulmonary microcirculation that produces only arterial and venous cells [92]. Embryonic stem cells (ESCs)‐derived multipotent EPCs could engraft into the pulmonary vasculature, promote blood vessel formation, and improve oxygenation, leading to the prevention of ACDMPV [93]. Mechanistically, multipotent EPCs promote neonatal lung angiogenesis via FOXF1‐mediated BMP9/ACVRL1 activation [94]. Notably, EPCs have been evaluated for the safety and potential benefit in PH patients, as indicated in a phase 1 clinical study (PHACeT trial). PH patients administered with EPCs overexpressing eNOS showed well‐tolerated, short‐term hemodynamic improvements and sustained increases in exercise capacity [95]. However, further studies are needed to investigate cell dosing, efficacy, and timing of administration for EPC‐based gene therapies.

6.2. Modulating Endothelial Plasticity

Lung disease is one of the leading causes of death worldwide. Thus, strategies aimed at promoting beneficial endothelial plasticity, such as enhancing endothelial repair and regeneration or inhibiting the EndoMT process as maladaptive plasticity, could hold potential strategies for lung repair and regeneration. Furthermore, understanding or discovering the underlying mechanisms or signaling pathways in regulating endothelial plasticity could enhance therapeutic targets for lung diseases. Endothelial repair and regeneration in damaged tissue is a process involving both the proliferation of existing ECs and the differentiation of ECs or circulating EPCs. During lung injury, gCap acts as an endothelial stem‐ and progenitor‐like cell, contributing to vascular regeneration and repair via its proliferation and differentiation to unique regenerative cells (co‐expression of Procr and apelin) or aerocytes to compensate for the loss of numbers of cells. Generating adoptive cell therapy of regenerative cells in repairing damaged lungs could be a promising therapeutic approach. In addition, venous ECs are capable of proliferating, differentiating, and acquiring capillary ECs upon lung injury. Thus, venous ECs also contribute to the repair of the capillary ECs [14]. However, the molecular mechanisms and signaling pathways for venous EC differentiation need to be further investigated.

EndoMT has been implicated in lung disease pathogenesis, contributing to fibrosis and vascular remodeling in PF and PH. Mechanistically, EndoMT could be modulated by TGF‐β, BMP, Wingless/Integrated, or Notch signaling pathways [96, 97]. There are several therapeutic agents that have been investigated targeting EndoMT in models of PF and PH, including Ponatinib, Imatinib, Salvianolic acid, and Vildagliptin [96]. However, further studies need to be done to determine whether EndoMT occurs in patients with PF and PH. Translating basic research findings on targeting EndoMT with negative regulators to clinical applications should be evaluated.

7. Future Perspectives and Open Questions

ECs are heterogeneous in the lung, playing distinct roles in distinct subpopulations. Thanks to the development of cutting‐edge technologies, such as scRNA‐seq, spatial transcriptomics, and lineage tracing, researchers have unraveled the diversity and heterogeneity of the endothelium within lung biology in terms of development, homeostasis, and pathologies. This will further open possibilities in discovering therapeutic agents and interventions for lung repair and regeneration.

The discovery of specialized EC subpopulations in the lung could enhance the complexity of cell communication during lung development and homeostasis. Therefore, further studies need to be done to explore how the heterogeneous communication between ECs and other cell types within lung tissue, such as epithelial, mural cells, and immune cells, maintains the alveolar‐capillary structure. For example, along with the novel identification of pericyte subtypes and the location of gCap and pericytes, the important function of pericytes and gCap in PH is accumulating. Furthermore, understanding the function of perivascular macrophages in endothelial regeneration could yield promising therapeutic strategies targeting lung barrier integrity and immune modulation in lung diseases.

Given that many chronic lung diseases are more prevalent in older adults, investigating how the aging process affects the lung endothelium could provide critical insights into disease pathogenesis and progression. For example, investigating a mechanistic understanding of how endothelial heterogeneity and plasticity contribute to maladaptive mechanisms in age‐related endothelial dysfunction. The application of multi‐omics technologies could be important for shedding light on specialized EC types and the underlying mechanisms of age‐related lung diseases.

Based on the current findings and understanding of endothelial heterogeneity and plasticity, this will allow us to further develop therapeutic interventions for endothelial dysfunction in lung diseases. Focusing on targeting specialized ECs rather than pan‐ECs could hold a promising approach to discover effective treatments for lung repair and regeneration. Furthermore, the application of multi‐omics approaches to human tissue samples is crucial to move the promising findings from preclinical models into effective clinical applications that can benefit patients with lung diseases.

8. Conclusions

Endothelial heterogeneity and plasticity are fundamental characteristics of the lung endothelium that are crucial for maintaining lung health and contributing to disease pathogenesis. The remarkable diversity in structure and function among ECs across different regions of the pulmonary vasculature reflects their specialized roles in processes such as gas exchange, regulation of vasomotor tone, vasodilation, leukocyte migration, and immune surveillance. Especially, EC dysfunction is a common feature of a variety of lung diseases. Recent advances have revealed lung EC heterogeneity within both the macrovascular and microvascular systems. It has improved our understanding of the complex interplay between cell types in the lung during development, homeostasis, and disease. With the continued use of cutting‐edge technologies, such as scRNA‐seq, spatial transcriptomics, intravital imaging, lung organoids, and microfluidics, we can enhance our understanding of complex cell communication, cell diversity, and underlying mechanisms within lung development, maintenance, and pathogenesis. We will then discover novel therapeutic strategies and interventions in preventing lung disease progression.

Author Contributions

Van Dung Nguyen: writing – original draft, literature search, figures and table preparation. Bisheng Zhou: conceptualization, writing – original draft, writing – review and editing, literature survey, supervision, funding acquisition. All authors have reviewed, contributed, and approved of this manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.