Role of the CXCL8-CXCR1/2 Axis in Cancer and Inflammatory Diseases

Abstract

The chemokine receptors CXCR1/2 and their ligand CXCL8 are essential for the activation and trafficking of inflammatory mediators as well as tumor progression and metastasis. The CXCL8-CXCR1/2 signaling axis is involved in the pathogenesis of several diseases including chronic obstructive pulmonary diseases (COPD), asthma, cystic fibrosis and cancer. Interaction between CXCL8 secreted by select cancer cells and CXCR1/2 in the tumor microenvironment is critical for cancer progression and metastasis. The CXCL8-CXCR1/2 axis may play an important role in tumor progression and metastasis by regulating cancer stem cell (CSC) proliferation and self-renewal. During the past two decades, several small-molecule CXCR1/2 inhibitors, CXCL8 releasing inhibitors, and neutralizing antibodies against CXCL8 and CXCR1/2 have been reported. As single agents, such inhibitors are expected to be efficacious in various inflammatory diseases. Several preclinical studies suggest that combination of CXCR1/2 inhibitors along with other targeted therapies, chemotherapies, and immunotherapy may be effective in treating select cancers. Currently, several of these inhibitors are in advanced clinical trials for COPD, asthma, and metastatic breast cancer. In this review, we provide a comprehensive analysis of the role of the CXCL8-CXCR1/2 axis and select genes co-expressed in this pathway in disease progression. We also discuss the latest progress in developing small-molecule drugs targeting this pathway.

Introduction

Chemokines and their cognate receptors play an essential role in the immune system by mediating the activation and trafficking of immune cells during innate and adaptive responses. Chemokines are also involved in hematopoiesis and development by directing and mobilizing precursor cells to sites of maturation ¹. Chemokines are small (6-14 kDa) secreted proteins that contain four cysteine residues that are essential for their structural integrity. Arrangement of these four cysteine residues is used to group chemokines into four different classes: CC, CXC, XC, and CX3C ². To date 50 chemokines and 20 chemokine receptors have been identified with CC and CXC being the two major classes of chemokines (Table 1) ³. CXC is further subdivided into ELR+ or ELR- denoting CXC chemokines that contain or lack the three amino acid motif (Glu-Leu-Arg) that precedes the first cysteine residue on the N-terminus ⁴. A phylogenetic tree of the chemokine receptors based on their sequence homology reveals similarity among the family members (Figure 1) ⁵.

Table 1.

Chemokine receptors and their corresponding antagonists

Class	Receptors	Ligand	Cell Expression	Antagonists
CCR	CCR1	CCL3, 5, 7, 8, 13, 15, 16, 23	Monocytes, immature dendritic cells (DCs), T cells, PMNs, eosinophils, mesangial cells, platelets	CP-481,715 (arthritis); MLN3897 (arthritis); BX471 (multiple sclerosis); AZD-4818 (COPD)
	CCR2	CCL2, 7, 8, 12, 13	Monocytes, immature DCs, basophils, PMNs, T cells, natural killer (NK) cells, endothelial cells, fibroblasts	MLN 1202 (MS, RA, atherosclerosis); INCB8696 (MS, lupus); CCX140 (MS); PF-4136309 (pain); MK-0812 (rheumatoid arthritis, multiple sclerosis)
	CCR3	CCL5, 7, 8, 11, 13, 14, 15, 24, 26	Eosinophils, basophils, T cells, DCs, platelets, mast cells	TPI ASM8 (asthma); KW-0761 (cancer); 776994 (asthma, allergic rhinitis); DPC-168 (asthma); GW766944 (asthma)
	CCR4	CCL17, 22	Immature DCs, basophils, T cells (Th2 T-cells), platelets	KW-0761 (lymphoma)
	CCR5	CCL3, 4, 5, 8, 11, 13, 14, 20	T-cells (Th1 cells), immature DCs, monocytes, NK cells, thymocytes	Maraviroc (approved for HIV, rheumatoid arthritis); Vicriviroc (HIV); Aplaviroc (HIV, potential toxicity); INCB9471 (HIV); Pro 140 (HIV); CCR5mAb004 (HIV); TBR-652 (HIV); Cenicriviroc (HIV)
	CCR6	CCL20	Immature DCs, T cells, B cells	None reported
	CCR7	CCL19, 21	Naïve and memory T cells Mature DCs, T cells, B cells	None reported
	CCR8	CCL1, 4, 16	Monocytes, B cells, T cells (Th2 cells), thymocytes	AZ084
	CCR9	CCL25	T cells, thymocytes, DCs, macrophages	CCX-282 (IBD, Crohn's disease); CCX8037; CCX282-B (IBD); GSK-1605786 (Crohn's disease)
	CCR10	CCL27, 28	T cells, melanocytes, dermal endothelia, dermal fibroblasts, Langerhans cells, astrocytes	None reported
CXCR	CXCR1	CXCL6, 8	PMNs, monocytes, astrocytes, endothelia, mast cells	SCH527123 (COPD); Reparixin (reperfusion injury)
	CXCR2	CXCL1, 2, 3, 5, 6, 7, 8	PMNs, monocytes, eosinophils, endothelia, mast cells	SCH527123 (COPD); Reparixin (reperfusion injury); SB656933 (COPD, cystic fibrosis); AZD5069 (neutrophil function); GSK1325756 (pulmonary disease)
	CXCR3	CXCL9, 10, 11	T cells, B cells, NK cells, mesangial cells, smooth muscle cells, endothelia	T-487/AMG-487 (psoriasis)
	CXCR4	CXCL12	Hematopoietic progenitors, T cells, immature DCs, monocytes, B cells, PMNs, platelets, astrocytes, endothelia	Plerixafor (multiple myeloma, NHL); BKT-140 (multiple myeloma); AMD 3100 (myelokathexis); AMD11070 (HIV); MSX-122 (cancer)
	CXCR5	CXCL13	T cells, B cells, astrocytes	None reported
	CXCR6	CXCL16	Memory T cells	None reported
	CXCR7	CXCL12		None reported
XCR	XCR1	XCL1, XCL2	T cells	None reported
CX3CR	CX3CR1	CX3CL1	PMNs, monocytes, NK cells, T cells, astrocytes	None reported
	Duffy	CXCL1, 7, 8, CCL1, 5	Red blood cells, endothelia	None reported
	D6	CCL2, 4, 5, 8, 13, 14, 15	B cells	None reported

Table adapted from references 1, 3, 6, 8-10.

Figure 1.

Phylogenetic tree of chemokine receptors. The phylogenetic tree was generated from sequence homology using the tools in the GPCR database (gpcrdb.org).

Given the vital roles of chemokines in the immune system and during inflammatory responses, a number of chemokines are involved in diseases as diverse as HIV, arthritis, multiple sclerosis (MS), chronic obstructive pulmonary diseases (COPD), lupus, pain, asthma, inflammatory bowel diseases (IBD), Crohn's disease, reperfusion injury (RI), cancer, and cystic fibrosis (CF) ⁶. The high significance of chemokine receptors for these diseases has led to the intense development of small-molecule inhibitors (Table 1). Currently, Maraviroc, a CCR5 antagonist used for HIV-1 infection, and Plerixafor, a CXCR4 antagonist, used as a hematopoietic stem cell mobilizer in patients with non-Hodgkin lymphoma and multiple myeloma, are the only two FDA approved chemokine receptor inhibitors ⁷. In this review we provide a comprehensive overview of the CXCL8-CXCR1/2 axis and its role in the pathogenesis of various diseases including inflammation and cancer.

CXCL8

CXCL8 is one of the first and most intensively studied chemokines acting as a pro-inflammatory chemokine. In the late 1980s, Peveri et al. found that LPS-stimulated blood monocytes produced a secretory protein (neutrophil activating factor, NAF) that stimulated neutrophil exocytosis (granule release) and oxidative burst (superoxide and hydrogen peroxide production) that appeared to be mediated by cell surface receptors ¹¹. NAF was the first chemokine to be purified and sequenced in 1987 and was later named as interleukin-8 (IL8) and CXCL8 upon identification of additional chemokines ^12-14.

Structural features of CXCL8

The crystal structure of CXCL8 was first solved in 1991. CXCL8 is a 72-amino acid peptide that has three antiparallel β-strands and one α-helix made up of C-terminal residues 57-72 ¹⁵. The structure is stabilized by two disulfide bonds formed by Cys7-Cys34 and Cys9-Cys50 (Figure 2A). The two cysteine residues (Cys7 and Cys9) are separated by one residue (Gln8), hence CXCL8 is classified as CXC chemokine. The crystal structure also revealed that CXCL8 exists as a dimer stabilized by hydrogen bonds between the first β-strand. CXCL8 monomer is the high affinity ligand for CXCR1, but both monomer and dimer bind CXCR2 with similar affinities. Interactions between CXCR1 N-terminal site-1 and CXCL8 play important role for higher affinity of the receptor towards CXCL8 monomer than dimer. These differences in the affinities of CXCR1 and CXCR2 towards CXCL8 monomer versus dimer may lead to different signaling outcome ^16-18. Scanning mutagenesis on CXCL8, in which the first 15 amino acids of CXCL8 were individually mutated to alanine, revealed the critical N-terminal motif, Glu-Leu-Arg (ELR). Mutants E4A, L5A, or R6A were inactive in receptor activation assays and showed reduced affinity to its receptors in competitive binding assays ^{4, 19}.

Figure 2.

Structures of CXCL8 and its receptors CXCR1/2. A. CXCL8 protein structure: The dimer structure was generated from the Protein Data Bank code PDB ID 1ICW. Two monomers are colored as green and cyan. The ELR motif is colored as yellow and CXC motif is in blue. B. CXCR2 2D structural domains: The N-terminus (extracellular face) of CXCR2 is critical for ligand binding and specificity. Transmembrane domain 4 and extracellular loop 2 is also important for ligand binding. The G-protein couples to the C-terminus (cytoplasmic face) of CXCR2 and involves intracellular loop 3. Several proteins, such as G protein coupled receptor kinase 6 (GRK6), vasodilator-stimulated phosphoprotein (VASP), β-arrestin1/2, adaptor protein-2 (AP-2), protein phosphatase 2A (PP2A) also associate with the C-terminus of CXCR2 to mediate different signaling cascades. C. CXCR1/2 3D structure: CXCR1 solid state NMR structure (PDB: 2LNL, green) aligned with CXCR2 homology model (blue) ²⁰. The predicted binding mode of the CXCR2 antagonist SCH527123 (brown) in the allosteric site located inside the CXCR2 receptor ²⁰.

CXCL8 expression and secretion

CXCL8 is secreted by different cell types including blood monocytes, alveolar macrophages, fibroblasts, endothelial cells, and epithelial cells ^{11, 21-23}. CXCL8 is virtually undetectable in unstimulated cells. CXCL8 expression is stimulated by various cytokines (interleukin-1, interleukin-6, CXCL12, and TNFα), hypoxia, reactive oxygen species (ROS), bacterial particles and other environmental stresses, and mediated by transcription factors, NF-κB and activator protein-1 (AP-1) ^{24, 25}. This stimulation leads to 10 to 100-fold up-regulation of CXCL8 expression ²⁶. The signaling cascades that stimulate the production of CXCL8 are depicted in Figure 3.

Figure 3.

CXCL8-CXCR1/2 signaling cascades and receptor recycling. CXCL8 binding to CXCR1/2 activates several G-protein-mediated signaling cascades. Receptor activation immediately leads to the dissociation of the Gαi subunit from the βγ subunits, and subsequently activates growth and stress kinases such as ERK1/2, JNK1, and p38. G-protein activation also induces rapid intracellular Ca²⁺ mobilization released from the endoplasmic reticulum (ER) and inhibition of adenylyl cyclase resulting in decreased cyclic AMP production. CXCR1/2 activation also leads to receptor phosphorylation on the C-terminus by GRK2/6 and recruitment of β-arrestin1/2 to mediate receptor internalization. Internalized receptors are either recycled back to the cell surface or routed to lysosomes for degradation.

The combination of at least three different mechanisms leads to the up-regulation of CXCL8 expression: a) de-repression of the CXCL8 gene promoter, b) trans-activation of CXCL8 expression by NF-κB and JNK pathways, and c) CXCL8 mRNA stabilization by the p38 MAPK pathway ²⁷. In unstimulated cells, the CXCL8 gene promoter is repressed as a consequence of three events: a) NF-κB-repressing factor (NRF) binds to the negative regulatory element (NRE) blocking the NF-κB binding site, b) octamer-1 (OCT-1) binds to the complementary strand of the promoter gene in the opposite direction of the C/EBP binding site, and c) histone deacetylase 1 (HDAC-1) induces the deacetylation of histone proteins ^27-29. In the presence of a stimulus, such as IL1 or TNFα, the activated p65 subunit of NF-κB translocates to the nucleus and binds to the DNA. C/EBP binds to the promoter by replacing OCT-1, followed by the recruitment of CREB-binding protein (CBP)/p300 resulting in histone hyperacetylation and chromatin remodeling. As a result, the CXCL8 promoter is de-repressed. AP-1 and NF-κB proteins are phosphorylated in a signal-dependent manner and trans-activate gene transcription of CXCL8 as well as other anti-apoptotic genes ²⁷.

There are several other genes that play a critical role in the transcription of CXCL8. Prostaglandin E2 (PGE2) induces de-repression of the CXCL8 gene promoter through concurrent association of site-specific DNA demethylation and histone H3 hyperacetylation ³⁰.

IκB (inhibitor of NF-κB) kinase (IKK) regulates CXCL8 production by trans-activating NF-kB transcription factor. IκB blocks the nuclear localization signal (NLS) of NF-κB proteins and thus inactivates NF-κB. IKK, a complex comprising IKK alpha and/or IKK beta and two molecules of NEMO, phosphorylates IκB. The phosphorylated IκB then is degraded by the proteasome allowing NF-κB subunits p65 and p50 to translocate into the nucleus ³¹. Reduced NF-κB-DNA binding activities as well as decreased CXCL8 production were observed in IκB-beta-stable transfectants of HONE1 and IκB-beta-infected HK1 cells compared to vector control. These observations reinforce the importance of IKK in CXCL8 expression ³².

The PI3K-AKT signaling pathway induces CXCL8 expression in human lung epithelial cells by activating IKK and NF-κB proteins ³³. The transcription factor AP-1, which trans-activates CXCL8 transcription, is activated by mitogen-activated protein kinases (MAPK), such as c-JUN kinase (JNK1) and ERK. JNK1 phosphorylates c-JUN, which then translocates into the nucleus and together with c-Fos, binds to the c-JUN promoter region of DNA to form the AP-1 transcription factor and promotes CXCL8 expression ^{27, 34}.

Twist-related protein 1 (TWIST1) also plays an important role in NF-κB-induced CXCL8 expression. The TWIST1 carboxy-terminal WR (Trp-Arg) domain mediates the formation of the TWIST1-p65 complex and activates the transcriptional activity of NF-κB by increasing the DNA binding affinity of p65 to the CXCL8 promoter and thus, enhances the expression of CXCL8 ³⁵.

CXCL8 production is affected by the stability of the CXCL8 mRNA, which is rapidly degraded by AU-rich cis-elements (ARE) contained in its 3′ untranslated region. p38 MAPK cascades play an important role in the post-translational regulation of CXCL8 expression by inhibiting ARE-based mRNA degradation. Dual specificity mitogen-activated protein kinase kinase 6 (MKK6) selectively activates p38 MAPK by phosphorylating a threonine and a tyrosine in the activation loop ^{27, 36}. MAP kinase-activated protein kinase 2 (MK2), which is downstream target of MKK6-p38 MAPK, actively contributes to the stabilization of CXCL8 mRNA. The dominant‐negative mutant of MK2 interferes with MKK6-induced CXCL8 mRNA stabilization ³⁶.

The CXCL8 Receptors: CXCR1 and CXCR2

CXCL8 mediates its signals via extracellular binding to two G protein-coupled receptors, C-X-C chemokine receptor type 1 (CXCR1) and C-X-C chemokine receptor type 2 (CXCR2). The two receptors share 76% sequence homology with each other and bind to CXCL8 with similar affinity (K_d ≈ 4 nM) ^37-39. The major differences between the two receptors occur in the second extracellular loop, the C-terminal (intracellular) and the N-terminal (extracellular) regions ^{37, 40}. CXCR2 interacts with all other ELR+ chemokines (CXCL1-3, 5-7) with high affinity, but CXCR1 only weakly binds to other ELR+ chemokines ⁴¹.

Activation of CXCR1 and CXCR2

Since the discovery of CXCR1 in 1991, a number of studies have been performed to characterize CXCR1 and CXCR2 receptor signaling and regulation. These studies were mainly carried out in neutrophils, or HEK293 and RBL-2H3 cells over-expressing CXCR1 and/or CXCR2. Upon chemokine binding, CXCR1/2 couples to pertussis toxin-sensitive G-protein via physical interaction with the Gα_i subunit to regulate several signaling cascades that mediate neutrophil chemotaxis and activation (Figure 3) ⁴². Activation of CXCR1/2 induces dissociation of the receptor with the G-protein and release of the Gβγ subunits from the Gα subunit. Release of the Gβγ subunits activates phospholipase C (PLC, β-2 isoform) and results in calcium mobilization from the endoplasmic reticulum to cytosol and protein kinase C (PKC) activation, which is critical for neutrophil chemotaxis (Figure 3) ^{43, 44}.

CXCL8 also induces rapid and transient phosphorylation of extracellular signal related kinases (ERK1/2) and phosphatidylinositide 3-kinase (PI3K)/Akt in human neutrophils ^45-48. ERK1/2 is a component of the Ras-Raf-MEK-ERK signaling cascade ⁴⁹. However, the role of CXCL8-mediated ERK1/2 activation in neutrophils migration remains unclear. Xythalis et al. showed MEK inhibitor PD098059 blocked CXCL8-induced neutrophil chemotaxis, while other studies showed no effects of PD098059 on CXCL8-induced neutrophil chemotaxis ^45-47. Inhibition of PI3K with small-molecule inhibitor LY294002 significantly reduced CXCL8-mediated cell migration in human neutrophils and L1.2 cells over-expressing CXCR2 ^{45, 50}.

CXCL8 signaling activates members of RhoGTPase family and thus, induces activation of protein kinases such as Src and focal adhesion kinase (FAK). Activation of FAK and Src by CXCL8 signaling resulted in increased cellular proliferation and motility ⁵¹. In CXCR1- and CXCR2-RBL (rat basophil leukemia) transfected cells, CXCL8 induces FAK phosphorylation and re-localization. It also induces actin and β-tubulin re-localization to promote cell spreading and motility that is directly correlated with the CXCL8-induced migratory response ^{52, 53}. FAK regulates cell motility by directing processes involved in cell spreading, attachment, and detachment ⁵⁴. LIM and SH3 protein 1 (LASP-1) directly associates with chemokine receptors (CXCR1-4), and its association is critical for chemotaxis, suggesting that LASP-1 may serve as an adaptor protein that connects chemokine receptors to components of the cytoskeleton ⁵⁵. CXCR2 also regulates other key regulators of actin polymerization such as Rac-GTPases (small monomeric GTPases) ^{56, 57}.

The uncoupling of the Gα subunit from CXCR2 upon ligand activation inhibits the enzyme adenylyl cyclase (AC) that converts ATP to cyclic AMP, and results in decreased intracellular cyclic AMP concentrations ⁴². To explore the effects of CXCL8 and CXCL1 on cyclic AMP levels, Hall et al. stimulated CXCR1/2-overexpressing CHO cells with forskolin (AC activator) in the presence of CXCL8 or CXCL1. Both chemokines dose-dependently inhibited CXCR2-mediated forskolin-induced cyclic AMP accumulation, while only CXCL8 inhibited CXCR1-mediated forskolin-induced cyclic AMP accumulation ⁵⁸.

Though CXCR1 and CXCR2 induce cell migration and granule release in neutrophils through similar pathways, phospholipase D (PLD) activation is exclusively mediated by CXCR1 ^59-61. PLD converts phosphatidylcholine to phosphatidic acid and choline. Phosphatidic acid activates NAPDH oxidase and subsequent superoxide anion production and thus, stimulates oxidative burst in neutrophils ⁶². CXCR1, but not CXCR2, in neutrophils significantly induced superoxide anion production, suggesting that CXCR1 is essential for CXCL8-mediated oxidative burst ⁶³.

Regulation of CXCR1 and CXCR2

The G-protein signaling of CXCR1/2 is tightly regulated and quickly desensitized to prevent constitutive signaling. Receptor desensitization is regulated by several mechanisms, including receptor phosphorylation/β-arrestin1/2-recruitment, AP-2 adaptor protein association, and receptor cross-desensitization.

Homologous desensitization (agonist-dependent)

Upon ligand stimulation, CXCR1/2 is phosphorylated by G-protein-coupled receptor kinases (GRKs) and associates with β-arrestin1/2 and AP-2 to promote dynamin- and clathrin-mediated receptor internalization ^64-67. CXCR2 internalization occurs at a faster rate and at lower ligand concentrations than CXCR1, suggesting differential regulation of receptor signaling ^{68, 69}. CXCR1 and CXCR2 are regulated by different GRKs. GRK2 mainly phosphorylates CXCR1, while GRK6 mediates CXCR2 phosphorylation ⁶⁷.

Receptor phosphorylation recruits β-arrestin1/2 to the receptor to terminate G-protein signaling via two distinct mechanisms. β-arrestin1/2 association inhibits G-protein coupling of receptors and recruits the endocytic machinery such as clathrin and AP-2 to mediate receptor internalization and sequestration ⁷⁰. For CXCR2 (but not CXCR1), β-arrestin1/2 may not be absolutely necessary for receptor internalization. Receptor internalization, though reduced, was still observed in phosphorylation-deficient CXCR2 (truncated C-terminal or GRK knockout) and β-arrestin-2 deficient cells, suggesting that receptor internalization is also mediated through alternative phosphorylation-independent mechanisms ^{64, 67, 71-74}. Phosphorylation or β-arrestin-2 deficiency also exhibits enhanced G-protein signaling resulting in ROS generation that induces cell death ^{73, 74}. Unlike β-arrestin1/2, receptor association with AP-2 does not require phosphorylation and AP-2 receptor association is required for CXCR1/2 internalization ⁷². AP-2 is a critical adaptor protein that directly links membrane-bound receptors to the clathrin lattice during endocytosis ^75-77.

Heterologous desensitization (agonist-independent)

CXCR1/2 is also regulated by the activation of other receptors (heterologous desensitization). CXCR1 and CXCR2 are cross-phosphorylated and desensitized to CXCL8 by receptors for N-formylated peptides (fMLP) or complement cleavage product C5a ^{60, 78}. fMLP and C5a are strong chemoattractants for leukocytes that mediate chemotaxis and leukocyte activation ^{79, 80}. CXCR1, but not CXCR2, also cross-phosphorylates and desensitizes fMLP and C5a receptors when these receptors are co-expressed together in RBL-2H3 cells ⁶⁰. However, C-terminal truncated CXCR2 is able to activate PLD and cross-phosphorylate and desensitize fMLP and C5a receptors, suggesting that PLD activation determines the ability of CXCR1/2 to regulate other receptors ⁶⁰.

Receptor trans-activation

In endothelial cells, CXCL8 activation of CXCR1/2 leads to an interaction between CXCR1/2 receptors and vascular endothelial growth factor receptor 2 (VEGFR2) and trans-activates VEGFR2 via receptor phosphorylation mediated by Src kinases. VEGFR2 trans-activation is required for CXCL8-induced endothelial cell permeability ⁸¹. The CXCL8-CXCR1/2 axis also stimulates VEGFR2 activation by inducing the transcription of VEGF in endothelial cells via the NFκB pathway ⁸². The CXCL8-CXCR2 axis also trans-activates epithelial growth factor receptor (EGFR) via receptor phosphorylation to mediate endothelial cell migration and capillary tube formation ^{57, 83}. CXCL8 stimulates expression of integrin α_vβ₃, which plays a key role in endothelial cell survival and tumor migration during angiogenesis ⁸⁴.

Structural features of CXCR1 and CXCR2

The difficulty of purifying membrane-bound receptors and the inherent flexibility of GPCRs has hindered their crystallization. Currently, the structures of about fifteen human GPCRs, including chemokine receptors CXCR4, CCR5 and CXCR1 are available ⁸⁵. The GPCR Network is implementing high throughput structure determination pipelines to characterize 15-25 representative human GPCRs within the next few years ⁸⁶. A three dimensional structure of CXCR1 (PDB code: 2LNL) was solved by NMR spectroscopy ⁸⁷. Figures 2B and 2C depict two- and three-dimensional structures of CXCR1/2 receptors.

The N-terminus of CXCR1/2

A number of CXCR1 and CXCR2 chimera, receptor truncation, and single amino acid point mutation studies have revealed several essential structural features that are critical for receptor binding, activation, and regulation. The N-terminus, transmembrane domain 4 (TM4) and extracellular loop 2 (ECL2) of CXCR1/2 are critical for ligand binding and specificity (Figure 2B) ^88-91 as well as for determining the rate of receptor internalization ⁹². In CXCR1, two disulfide bonds between Cys30 and Cys277 (connecting N-terminus to TM7) and Cys110 and Cys187 (connecting TM3 to ECL2) are important for ligand binding. Charged residues within the extracellular loops and transmembrane helices (Asp85 of TM2, Lys117 of TM3, and Asp288 and Glu291 of TM7) of CXCR1 are important for ligand binding and receptor signal transduction ⁸⁷. CXCR2 ligands bind to overlapping but distinct sites on CXCR2 and affinities of ligands do not necessarily correlate with potency on receptor activation. This suggests that sites of receptor binding and activation are distinct ⁹³. An anti-CXCR1 monoclonal antibody, 7D9, binds to the region of first 45 residues of the receptor and inhibits chemotaxis without affecting ligand binding, further supporting the notion that ligand binding and receptor activation are distinct ⁹¹.

The C-terminus of CXCR1/2

The C-terminus of CXCR1/2 regulates receptor phosphorylation, internalization, G-protein coupling and association with other cytoplasmic proteins. CXCR1 intracellular loop 3 (ICL3) that projects into the cytoplasm is important for G-protein coupling, calcium mobilization, and chemotaxis of the cell ⁸⁷. Also several amino acid residues of CXCR1, such as Ser132, Asp134 of TM3 and Met241, Phe251 of TM6 play critical roles in G-protein coupling and receptor activation ⁹⁴. Truncation of the C-terminus of CXCR1/2 impairs receptor phosphorylation, β-arrestin1/2 association, and internalization, as well as enhanced G-protein signaling (calcium release) and reduced chemotaxis ^{66, 69, 72, 95}. Reduced chemotaxis without receptor internalization suggests that G-protein signaling negatively regulates chemotaxis and not receptor internalization ⁶⁶. Alanine point mutations show that serine residues 342 and 346-348 are involved in receptor desensitization and sequestration but not receptor phosphorylation, suggesting other hydroxylated residues may be involved in receptor phosphorylation ⁹⁶. However, this is in conflict with studies showing involvement of serine 346-348 in agonist-induced phosphorylation in primary cultured cells isolated from mice ⁹⁷.

The C-terminus of CXCR2 is also a binding site for several adaptor proteins that regulate receptor desensitization and endocytosis. β-arrestin1/2 associates with CXCR1/2 upon receptor phosphorylation and mediate the recruitment of endocytic components (clathrin and dynamin) ^64-66. AP-2 also binds the LLKIL motif on CXCR2 and regulates receptor internalization and sequestration in HEK293 cells ⁷². C-terminal deletion of CXCR2 also reduces G-protein activation (measured by GTPγS exchange), suggesting it is also involved in G-protein coupling ⁹⁸. The third intracellular loop of CXCR2 is also involved in G-protein coupling and signaling ⁹⁹. Protein phosphatase 2A (PP2A), a serine/threonine phosphatase, also directly associates with CXCR2 on KFRHGL motif of the C-terminus independent of receptor phosphorylation and mediates receptor dephosphorylation and receptor recycling ⁷¹. CXCL8 stimulates phosphorylation of vasodilator-stimulated phosphoprotein (VASP) via PKA- and PKC-mediated signaling pathways and promotes the association of VASP with the C-terminus of CXCR2. VASP association is critical for CXCR2-mediated chemotaxis and polarization ¹⁰⁰.

Differential functions of CXCR1 and CXCR2

Both receptors mediate common GPCR signaling pathways and cellular functions such as calcium release, activation of Ras/MAPK and PI3K signaling cascades, as well as receptor internalization and chemotaxis. However, CXCR1 but not CXCR2 was shown to activate PLD and subsequently mediate ROS generation and oxidative burst in neutrophils ^{59, 61, 78}.

The receptor desensitization rate is also different between the two receptors. CXCR2 is internalized more rapidly and at lower ligand concentrations than CXCR1 ^{68, 69}. It is also recycled back to the surface at a much slower rate than CXCR1. In studies with CXCR2 mutants, where the C-terminus was truncated and receptor internalization was impaired, CXCR2 was able to activate PLD and mediate CXCL8-mediated superoxide anion production ⁶⁰. This is also corroborated with studies that inhibit the internalization mechanism of CXCR2 (cell lines with β-arrestin1/2 deficiency or dominant negative dynamin), which show similar increase in ROS production. This suggests that the functional differences between CXCR1 and CXCR2 may be regulated by the duration of the signal ⁶⁰.

Higher ligand concentrations are required for receptor desensitization than receptor activation. And the fact that CXCR2 can interact with all ELR+ chemokines suggests that CXCR2 may play a more important role in chemotaxis than CXCR1 ^{101, 102}.

In endothelial cells, CXCR1 and/or CXCR2 knockdown with shRNAs showed that both receptors were critical for CXCL8-mediated endothelial cell proliferation, survival, migration, invasion, tube formation and angiogenesis, which corroborates previous studies performed with antibodies against CXCR1/2 and in vivo studies with CXCR2^-/- mice ^{103, 104}. Interestingly, double knockdown of CXCR1 and CXCR2 did not show additive effects on endothelial cells, suggesting the knockout of either receptor is sufficient to alter CXCL8-mediated angiogenesis ¹⁰⁵.

Role of CXCL8-CXCR1/2 axis in infection

Inflammation is a defense mechanism that can be triggered by infection and tissue damage ¹⁰⁶. The CXCL8-CXCR1/2 axis recruits neutrophils at the site of infection and induces a neutrophil oxidative burst and a granule release to eliminate inflammatory stimulus and increase bacterial clearance (Figure 4) ^{101, 102}. Thus, this axis protects the host from further infection and tissue damage ¹⁰⁷. Disruption in the CXCL8-CXCR1/2 axis could severely affect the host's immune mechanisms against infection and even may lead to fatality. Impaired neutrophil recruitment often leads to a decrease in bacterial clearance and reduced survival rate in the experimental infectious disease models ¹⁰⁸.

Figure 4.

CXCR1 and CXCR2 mediate neutrophil recruitment during infection. In the presence of a microbial infection, macrophages at the site of infection begin to secrete CXCL8 to attract CXCR1/2-expressing neutrophils to the site of infection. Since CXCR2 is more sensitive to low ligand concentrations, CXCR2 is believed to play a more important role at recruiting neutrophils to the site of infection, whereas CXCR1 mediates oxidative burst and granule release to combat the microbes at the site of infection.

CXCR1/2 knockout studies

A number of CXCR2 knockout mice studies have been performed to further elucidate various roles of CXCR2. In general, most of these studies show that CXCR2 knockout mice are healthy. However, they do exhibit impaired wound healing and angiogenesis, increased susceptibility to pathogens, and decreased pathogen clearance due to reduced neutrophil recruitment ^109-117. Hyperoxia-induced neutrophil infiltration is significantly diminished in CXCR2^-/- mice, protecting them from liver injury as compared with the CXCR2^+/+ mice. Similar results from attenuation of hyperoxia-induced neutrophil infiltration and protection from liver injury were observed when normal mice were treated with an anti-CXCR2 antibody ^{118, 119}. In a separate study, CXCR2 knockout mice exhibited neurological defects including decreased spinal cord white matter area and reduced myelin sheath thickness ¹²⁰. These mice also had enlarged lymph nodes and spleen due to increased B-cells and neutrophils, suggesting that CXCR2 plays a role in B-cell and neutrophil expansion and development ¹²¹. CXCR2^-/- mice were resistant to cuprizone-induced demyelination and the transfer of CXCR2-positive neutrophils made mice susceptible to demyelination as they were before ¹²². CXCR2 knockout mice blocked LPS-induced neutrophil recruitment into their cerebral microvessels ¹²³.

CXCR2 is also involved in neutrophil trafficking from the bone marrow during development ¹²⁴. Lastly, CXCR2 knockouts were less susceptible to spontaneous tumorigenesis including melanoma, prostate and renal cancer ^125-130. The knockout studies in mice suggest that CXCR1 is important for embryonic oligodendrocyte precursor migration in developing spinal cord ¹³¹. All previously mentioned knockout studies prove the importance of CXCR2 in inflammatory diseases related to neutrophil infiltration as well as in tumorigenesis and metastasis. Therefore, blocking CXCR2 signaling could potentially be a novel therapy for these diseases.

Genes Implicated in the CXCL8-CXCR1/2 Signaling Pathway

Our bioinformatics analysis reveals important roles for the expression of CXCL8 and CXCR1/2 genes in tumor cell proliferation, migration, and activation of the inflammatory system. Protein-protein interaction analysis connects the CXCL8-CXCR1/2 axis with other cytokines through physical interactions, coexpression, pathway knowledge, and automated text-mining (STRING Database, v.10) (Figure 5) ¹³². We observed a high correlation between CXCL8 and other cytokines (e.g. CXCL1, CXCL2, CXCL3, LIF, IL1A, and IL1B) from the analysis of the Cancer Cell Line Encyclopedia (CCLE, Broad Institute) (Figure 6A) ¹³³. Interestingly, we observed significantly different patterns for CXCR1 and CXCR2 gene expression (Figures 6B, 6C). Table 2 summarizes a set of select genes involved in the CXCL8-CXCR1/2 signaling axis. The Oncomine gene expression analysis (cancerous vs. normal) of those genes (supplementary Figures S1-S6) reveals that CXCL8 as well as some of its correlated genes, such as CXCL1, CXCL2, and CXCL3 are highly expressed in cancerous tissue, particularly in colorectal cancers ¹³⁴. Targeting these genes by small-molecule drugs could be an efficient way of manipulating the CXCL8-CXCR1/2 signaling pathways.

Figure 5.

Genes implicated in the CXCL8-CXCR1/2 axis. Protein-protein interaction plots generated from STRING version 10 database using A. CXCL8 (IL8), B. CXCR1 and C. CXCR2 as queries. Pictures are top 20 genes connected to CXCL8 or CXCR1/2 either through physical interactions (experiments), co-expression, text mining, neighborhood on the genome, gene-fusion, database, co-occurrence and homology.

Figure 6.

Gene expression heat map for CXCL8 (IL8), CXCR1/2 co-expressed genes from the CCLE as well as co-expression correlation. A. Gene expression heat map and correlations for CXCL8 (IL8) co-expressed genes. The top 20 genes are listed. B. Gene expression heat map and correlations for CXCR1 co-expressed genes. C. Gene expression heat map and correlations for CXCR2 co-expressed genes.

Table 2.

Functions of select genes associated with the CXCL8-CXCR1/2 axis

Genes	Functions
RELA	The activated p65 subunit (RELA) of NF-κB translocates to the nucleus and binds to the DNA. As a result, CXCL8 promoter is derepressed and CXCL8 gene expression is induced [27].
DARC	CXCL8 is one of the natural ligands for duffy antigen receptor for chemokines (DARC). The receptor is involved in regulating angiogenesis in endothelial cells [135].
CXCL12	CXCL8 expression is upregulated by CXCL12-CXCR4 axis in several cell types, such as human mast cells, endothelial cells and leukemia cells [136, 137]. The CXCL8-CXCR2 axis is activated by the CXCL12-CXCR4 axis in breast cancer cells [137].
IL1B	Pro-inflammatory cytokine interleukin-1β (IL-1β) induces expression of CXCL7, which is a pharmacological ligand of CXCR1 and CXCR2 and promotes tumor cell proliferation [138]. Blockade of CXCR1/2 receptors by reparixin protects mice from cerebral damage in a model of middle cerebral artery occlusion and reperfusion by reducing IL-1β levels and PMN recruitment [139].
IL1A	Interleukin-1α (IL-1 α) activates NF-κB and AP-1-induced CXCL8 expression in head and neck squamous cell carcinomas [140].
JUN and FOS	Transcription factor AP-1 is homo or heterodimer of c-Jun and c-Fos, which are trans-activated by CXCL8 signaling [27].
CXCL1	CXCL1 binds to CXCR2 with an EC50 of 5 nM [41]. Like IL8, CXCL1 is also overexpressed in various cancers including colorectal cancer (source: Oncomine, Supplementary Figure S2).
PPBP (CXCL7)	CXCL7, also known as NAP-2, binds CXCR2 with high affinity (EC50 = 7 nM) [41]. Overexpression of CXCL7 and CXCR2 in liver metastases from colon cancer patients are correlated with shorter overall and disease-free survival [141].
CXCL12	Tumor-derived paracrine CXCL8 signaling induced expression and secretion of CXCL12 from stromal cells in prostate cancer and thus, augmented invasion of PTEN-deficient prostate cancer cells [142].
ARRB1	Phosphorylated CXCR2 recruits β-arrestin1/2 (ARRB1) and components of endocytosis such as clathrin and dynamin to mediate receptor internalization [64].
GRK6	GRK6 phosphorylates CXCR2 and negatively regulates receptor sensitization, internalization and chemotaxis, thus affecting cell signaling and angiogenesis [67].
VASP	Vasodilator-stimulated phosphoprotein (VASP) is phosphorylated by PKC and PKA signaling, interacts with the C-terminus of CXCR2, and plays a critical role in CXCR2-mediated chemotaxis and polarization [100].
GTF3A (AP2)	Adaptin 2 (AP2) binds CXCR2 and plays an important role in receptor internalization. AP2 interacts with LLKIL motif in the carboxyl terminus of CXCR2 and helps internalization of the receptor and receptor-mediated chemotaxis in HEK293 cells [72].
PP2A	Protein phosphatase 2A (PP2A) binds CXCR2 on the C-terminus independent of receptor phosphorylation and mediate receptor dephosphorylation and receptor recycling [71].
MMP9	Overexpression of MMP9 and CXCL8 correlates with poor prognosis of bladder cancer. High-grade tumors express significantly higher levels of MMP9 and CXCL8 compared to low-grade tumors [143].
EGFR	The CXCL8-CXCR2 axis trans-activates EGFR via receptor phosphorylation to mediate endothelial cell migration and tube formation [57, 83].

The CXCL8-CXCR1/2 Axis in Inflammatory Diseases

Since CXCL8 is a critical component of inflammation-mediated processes, aberrant regulation of CXCL8 and its receptors has been implicated in a number of inflammatory-mediated diseases that include cystic fibrosis, chronic obstructive pulmonary disorder, asthma, psoriasis, rheumatoid arthritis, and inflammatory bowel diseases ^{26, 144-150}. It is also involved in tumorigenesis of various cancers such as lung, colon, prostate, pancreatic, breast, ovarian, and melanoma (Table 3).

Table 3.

Important roles of CXCL chemokines and CXCR1/2 in cancer

Cancer type	Summary of findings
Lung Cancer (NSCLC)	IL-1β stimulated more CXC chemokine secretion in A549 cells than in human tracheobronchial epithelium cells via CREB and NF-κB activation [229]. Lewis lung carcinoma (LLC) cells transduced with human IL-1β exhibited increased tumor growth, which was inhibited by CXCR2 antibodies [270]. CXCL8 stimulated epithelial cell proliferation (A549 and NCI-H292) via EGFR trans-activation involving the MAPK pathways [271]. CXCL8 stimulated H460 and MOR/P (NSCLC cell lines) cell proliferation via CXCR1 but not CXCR2 [272]. CXCR2-/- mice implanted with LLC primary tumors in heterotopic and orthotopic models showed reduced tumor growth and vascular density as well as reduced spontaneous metastases [273]. Inhibition of CXCR2 with antibodies impeded the progression of premalignant alveolar lesions in mice with KRAS mutations known to develop lung adenocarcinoma [274]. Inhibition of CXCR2 with AZ10397767 reduced neutrophil infiltration in A549 tumor spheroids and primary tumors in mice [275]. CXCR2 antibodies inhibited SNAIL-mediated tumor burden in orthotopic and heterotopic lung cancer mouse models [276]. Depletion of CXCR2 via shRNA knockdown in a highly metastatic murine adenocarcinoma cell line with Kras/p53 mutant reduced tumor invasion and metastasis in in vitro and in vivo orthotopic syngeneic mouse models [277]. In lung adenocarcinoma, CXCR2 was a poor prognostic marker and promoted invasion and metastasis of tumors [277]. CXCL5 was one of the main drivers of the CXCL8-CXCR2 ligand axis in adenocarcinomas as it was the most upregulated gene in that cluster [277]. Single nucleotide polymorphisms (SNPs) in CXCR2 were associated with CXCR2 expression, signaling and susceptibility to lung cancer [278].
Colorectal Cancer	CXCL8 and CXCR2 were upregulated in colorectal tumor samples (n=8) [230]. CXCL8, CXCR1, and CXCR2 expression were higher in metastatic colon cancer cell lines (KM12C and KM12L4) than in Caco-2 cells. CXCL8 also induced cell proliferation which was attenuated by neutralizing antibodies to CXCR1/2 or CXCL8 [279]. CXCL1 expression was higher in primary colon adenocarcinoma than in normal colon epithelium. Inhibition of CXCL1 by siRNA reduced proliferation and increased apoptosis [231]. Primary colorectal cancer samples expressed CXCL1 and its expression was associated with tumor size and stage, metastasis, and patient survival; colon cancer cell lines also express CXCR2, and CXCL1 stimulation increased their invasiveness [280]. Over-expression of CXCL8 via stable transfection in human colon cancer cells (HCT116 and Caco-2) enhanced cell proliferation, migration, invasion, and resistance to oxaliplatin. CXCL8 over-expressing cells also formed larger tumors with increased microvessel density in xenograft models [281]. Single nucleotide polymorphisms (SNPs) in CXCL8, CXCR1, and CXCR2 were associated with colon and rectal cancer risk [282]. Immunodeficient mice expressing human CXCL8 on the skin had enhanced human and mouse colon cancer tumor growth, angiogenesis, and metastases to the lung and liver. Conversely, CXCR2 knockout mice exhibited reduced tumor growth and angiogenesis, and increased necrosis [283]. CXCL8 was an independent prognostic marker for colon cancer. CXCL8 expression was upregulated in colon cancer and its level was increased with disease progression and metastasis [284]. CXCL8 expression significantly correlated with expression of αvβ6 integrin. The CXCL8-CXCR1/2 axis enhanced migration of colorectal carcinoma cells by increasing αvβ6 integrin expression [285]. SCH527123 inhibited human colon cancer liver metastases in a mouse xenograft model; however, it had no effect on tumor growth [248]. SCH527123 inhibited colon cancer cell (HCT116 and Caco-2) proliferation, migration, and invasion, and increased apoptosis. It also reduced tumor growth and angiogenesis as well as improved oxaliplatin treatment in mice xenograft studies [247]. The CXCL8-CXCR2 axis played an important role in chemoresistance of HCT116 cells [286]. CXCL2-CXCR2 axis helps in the recruitment of tumor-associated neutrophils and thus, regulated colitis-associated colon cancer in mice [287].
Breast Cancer	Increased copy numbers of CXCL1/2 genes contributed to higher expression of CXCL1/2 in invasive breast tumors. CXCL1/2 participated in a paracrine loop involving the tumor microenvironment and cancer cells to enhance chemoresistance and metastasis in breast tumors [232]. Thrombin stimulated CXCL1 expression and secretion in tumor and endothelial cells. Antibodies against CXCL1 inhibited thrombin-induced angiogenesis (endothelial tube formation). Depletion of CXCL1 via shRNA in 4T1 cells reduced tumor growth, angiogenesis, and metastasis [288]. CXCL7 and CXCR2 expression were higher in malignant (MCF10CA1a.c11) than in premalignant (MCF10AT) cells. Premalignant cells transfected with CXCL7 showed increased invasiveness, which was attenuated by a CXCL7 antibody [233]. Activation of the fibroblast growth factor receptor (FGFR) in epithelial breast cancer cells led to downregulation of the TGFβ/SMAD3 pathways in tumor-associated macrophages, which is associated with increased expression of CXCL chemokines. These chemokines also stimulated breast epithelial cancer cell invasiveness which was inhibited by SB225002 (CXCR2 inhibitor) [289]. Mesenchymal stem cells produced CXCL1 and CXCL5 and recruited mammary cancer cells, facilitating bone metastasis. This process was inhibited by antibodies against CXCL1, CXCL5, and CXCR2 as well as by SB265610 [290]. CXCR2 knockdown via shRNA in metastatic murine mammary tumor cell lines (C166, 4T1) reduced cell invasion, but did not alter cell proliferation. Implantation of these cells into an orthotopic mouse model showed that CXCR2 knockdown reduced spontaneous lung metastasis by 40% compared to control. These shRNA knockdown cells also enhanced cytotoxicity of doxorubicin and paclitaxel in in vitro and in vivo mice models [256, 291]. CXCR1 blockade with CXCR1 antibodies or reparixin depleted breast cancer stem cells in HCC1954, MDA-MB-453 and MDA-MB-231 cell lines. Reparaxin also retarded tumor growth and metastasis in xenograft studies [292, 293]. CXCL8 induced activation of EGFR/HER2 signaling pathways mediated by SRC, PI3K, and MEK in breast cancer stem cells from metastatic and invasive breast cancers derived from human patients. CXCL8 also enhanced colony formation ability of these cells. Inhibition of CXCR1/2 with SCH563705 inhibited colony formation and improved the efficacy of lapatinib (tyrosine kinase inhibitor) [255]. CXCL8 levels were increased in breast cancer patients compared to healthy volunteers and the level was associated with the stage of the disease [294]. CXCL8 levels were significantly upregulated in breast cancer patients having bone metastasis compared with patients lacking bone metastasis. There was also a signicant correlation between plasma CXCL8 levels and bone resorption in breast cancer patients [295]. Higher expressions of CXCR2 ligands CXCL1, CXCL3, CXCL5 and CXCL7 were observered in drug resistance breast cancer cells which exhibited delayed tumor growth, but higher metastatic potential in mouse xenograft model [296].
Prostate Cancer	Oxaliplatin increased NF-κB activity and the transcription of CXCL1, CXCL8, and CXCR2. CXCR2 antagonist AZ10397767 inhibited oxaliplatin-induced NF-κB activity and increased oxaliplatin-induced apoptosis in androgen-independent prostate cancer resistant to chemotherapy [249]. 5-FU increased CXCL8 secretion and CXCR1 and CXCR2 gene expression in PC3 cells. AZ10397767 increased 5-FU cytotoxicity and apoptosis [250]. TRAMP (tumor adenocarcinoma of the mouse prostate)/CXCR2-/- mice were smaller than CXCR2 wild-type mice and had reduced angiogenesis [129]. CXCL1 and CXCL8 increased PC3 invasion and adhesion to laminin, while CXCR2 antibodies inhibited CXCL8-induced cell invasion [297]. High-producing and low-producing CXCL8 clones of PC3 cells were isolated and injected into the prostate of nude mice. Tumors with high-producing CXCL8 showed increased growth, vascularization, and lymph node metastasis compared to low-producing CXCL8 tumors [234]. Hypoxia induced CXCL8, CXCR1, and CXCR2 expression in PC3 cells via HIF-1 and NF-κB transcriptional activity. CXCR1/2 siRNA enhanced etoposide-induced cell death in hypoxic PC3 cells [252]. CXCR2 inhibition with AZ10397767 and NF-κB inhibition with BAY11-7082 enhanced ansamycin cytotoxicity in PC3 cells but not DU145 cells [253]. CXCL8 upregulated cFLIP (caspase 8 inhibitor) expression and pretreatment with AZ10397767 inhibited CXCL8-induced cFLIP expression in LnCAP and PC3 cells. It also sensitized PC3 cells to TRAIL treatment (TRAIL induce CXCL8 expression in PC3 and LnCAP cells) [254]. PTEN repression via siRNA and shRNA increased CXCL8, CXCR1 and CXCR2 expression in PCa cells. CXCL8 depletion via siRNA decreased cell viability in PTEN deficient cells through G1 cell cycle arrest and apoptosis [298]. Tumor-derived CXCL8 enhanced secretion of cytokines CCL2 and CXCL12 from stromal cells and augmented proliferation and invasion of PTEN deficient prostate cancer cells [142]. CXCL8 serum level was correlated with increasing grade of metastatic prostate cancer compared to healthy volunteers [237]. CXCL8 induced cyclin D1 translation via Akt and activation of translational components in PC3 and DU145 cells [299]. The highly metastatic PC-3M-LN4 cells overexpress CXCL8 compared to PC-3P cells. Knockdown of PC-3M-LN4 cells with antisense CXCL8 cDNA reduced MMP-9 expression, collagenase activity, and invasion in vitro and in an in vivo orthotopic model. Conversely, upregulation of CXCL8 in PC-3P cells had the opposite effect [300]. CXCL1 and CXCL2, derived from bone marrow adipocytes, accelarated osteolysis and promoted metastasis of prostate cancer to bone [301].
Ovarian Cancer	CXCR2 shRNA knockdown in ovarian cancer cells (T29Gro-1, T29H, and SKOV3) inhibited tumor growth and arrested cells in G0/G1 phase by regulating cell cycle modulators. CXCR2 also induced apoptosis and angiogenesis. CXCR2 expression was correlated with poor overall survival for ovarian cancer [302]. Matrix metalloprotease-1 (MMP-1) activation of protease-activated receptor-1 (PAR1) induced CXCL8, CXCL1, and CCL1 secretion and stimulated endothelial cell proliferation, tube formation, and migration. These activities were attenuated by CXCR1/2 inhibition with X1/2pal-i3 (cell-penetrating pepducin that targets third intracellular loop of CXCR1/2), which also reduced tumor growth in mice and MMP-1-mediated angiogenesis [303]. CXCL8 suppressed TRAIL-medaited OVCAR3 apoptosis via downregulation of death receptors [304]. CXCL1 enhanced epithelial ovarian cancer cell (SKOV3 and OVCAR-3) growth and trans-activated EGFR via EGF release which involved the ERK1/2 signaling pathway [305]. Ovarian cancer patients with the A/A or A/T genotype for the CXCL8 T-251A gene polymorphism were less responsive to cyclophosphamide and bevacizumab treatment than patients with the T/T genotype. The A/A genotype was associated with increased CXCL8 production [306]. Paclitaxel induced CXCL8 promoter activation in ovarian cancer through the activation of both AP-1 and NF-κB. CXCL8 inhibition by antibodies stimulated tumor growth via recruitment of neutrophils to tumor site [307-309]. Stably CXCR2 transfected SKOV3 cells had a faster proliferation rate compared to cells transfected with empty vector. CXCR2 positive cells potentiated EGFR trans-activation, which led to AKT signaling regulated by NF-κB activation. As a result CXCR2 ligands CXCL1/2 were upregulated [310]. CXC chemokines and CXCR2 expression was elevated in sorafenib resistant ovarian tumors comapred to responsive tumors. CXCR2 inhibitor in combination with sorafenib exhibited synergistic inhibition of tumor cell growth [311].
Melanoma	Low tumorigenecity melanoma cell line A375P overexpressing CXCR1 or CXCR2 had enhanced in vivo tumor growth which was associated with increased microvessel density and reduced apoptosis [312]. CXCL8 serum level was associated with patient response to dacarbazine, cisplatin, and vindesine with or without DVP/IFN-2/IL-2 chemotherapy/immunochemotherapy [313]. Expression of CXCR2 and CXCL8 were correlated with melanoma tumor grade [314]. Melanoma cell lines secrete CXCL8 and express CXCR1 and CXCR2. CXCL8 stimulation of melanoma cells enhanced cell proliferation, migration, and invasion, which was reversed with inhibition of CXCR2 with neutralizing antibodies [315]. CXCL8 overexpression in A375P cells or CXCL8 knockdown in A375SM cells showed that CXCL8 regulated cell proliferation, migration, invasion, and colony formation. CXCL8 overexpression was associated with enhanced tumor growth and lung metastasis in vivo [235]. CXCR1 and/or CXCR2 knockdown in A375-SM cells via shRNA inhibited cell proliferation, migration, and invasion in vitro, reduced tumor growth and mircovessel density, and increased apoptosis in nude mice compared to control cells. Similar in vitro and in vivo results were obtained with CXCR1/2 inhibition with small molecule antagonists, SCH-479833 and SCH-527123 [316, 317].
Pancreatic cancer	Capan-1 cells expressed CXCL1, CXCL8, and CXCR2. CXCL1 and CXCL8 antibodies inhibited Capan-1 growth [318]. BxPC3 cells secreted CXCL3, CXCL5 and CXCL8. The supernatant from BxPC3 cells induced neovascularization in a corneal micropocket assay, which was impaired in the presence of CXCR2 antibodies. CXCL5 and CXCL8 were overexpressed in pancreatic cancer tissue samples [236]. ELR+ chemokines were elevated in exocrine pancreatic secretions from pancreatic cancer patients. Pancreatic cancer cell lines (BxPC3, Colo-357, and Panc-28) also expressed more ELR+ chemokines than normal pancreatic ductal epithelial cell line. Supernatants from pancreatic cancer cell lines stimulated HUVEC tube formation, which was attenuated by CXCR2 antibodies. These results were replicated in an orthotopic mouse model [319]. CXCR2 knockout mice with orthotopic and heterotopic pancreatic cancer tumors had impaired mobilization of bone marrow derived endothelial progenitor cells associated with reduced tumor angiogenesis and tumor growth [246]. K-Ras4BG12V transformed human pancreatic duct epithelial cells show enhanced secretion of CXC chemokines and VEGF via MEK1/2 and c-Jun pathways. When these cells were co-cultured with HUVECs, they enhanced HUVEC tube formation and invasiveness which was inhibited by CXCR2 antagonist, SB225002, or VEGF antibody [320]. CXCL5 expression correlated with clinical stage and shorter patient survival in pancreatic cancer. CXCL5 siRNA knockdown inhibited tumor growth in pancreatic cancer xenograft mouse model [238]. CXCR1 expression was positively correlated with lymph node metastasis and poor survival rate in patients with pancreatic ductal adenocarcinoma (PDAC) [321]. CXCR2 formed a macromolecular complex with NHERF1 and PLCβ3 in pancreatic cancer cells. The CXCR2-NHERF1-PLCβ3 complex regulated CXCR2 signaling activity and played important role in tumor progression and invasion [322]. CXCR2 expression is upregulated in human pancreactic ductal adenocarinoma, particularly in neutrophils/myeloid derived suppressor cells and assocoiated with tumorigenesis, metastasis as well as poor prognosis and survival. CXCR2 inhibition significantly enhanced sensitivity to anti-PD1 therapy and improved survival time in mice bearing pancreatic tumors [323-325]. Increased expression of CXCR2 and its ligand was observed in KRAS(G12D) mutation-bearing PDAC cells. Knocking down of CXCR2 or CXCR2 antagonists selectively inhibited in vitro and in vivo tumor cell proliferation as well as altered KRAS protein levels [326].
Liver cancer	CXCL8 along with CXCR1/2 receptors played important role in invasion, angiogenesis and metastasis of different solid tumors including liver cancer [327]. CXCL5 and CXCL8, both bind CXCR2, were significantly overexpressed in liver cancer cells, HCCLM3, with high movement capacity as compared with HepG2 cells with low movement capacity or normal liver L02 cells [328]. Treatment with siRNAs against CXCL5 suppressed tumor growth, proliferation, migration and invasion in liver cancer [329].
Bladder cancer	CXCL5 along with CXCR2 promoted migration and invasion of tumor cells in bladder cancer [330]. CXCL8 expression was increased significantly in monomethylarsenous acid [MMA(III)]-induced malignant transformation of urothelial (UROtsa) cells. Internalization of CXCR1 was increased in those malignant cells [331].
Other cancers	CXCR2 expression is significantly correlated with high grade, advanced stage metastasis as well as shorter overall survival in patients with renal cell carcinoma. Immunohistochemical analysis using CXCR2 could be a positive prognostic marker for renal cell carcinoma [332]. CXCR2-CXCL2 interaction in the tumor microenvironment plays an important role in tumor progression and metastasis in hepatocellular carcinoma (HCC) [333]. CXCR2-CXCL1-axis regulated infiltration of neutrophils, which were enriched in the peripheral stromal cells and associated with reduced recurrence free as well as overall survival of HCC patients [334]. CXCR2 signaling is strongly activated in glioblastoma microenvironment and responsible for tumor neovascularization and metastasis as well as tumor recurrence. Blocking the CXCR2 signaling by anti-CXCR2 antibody or CXCR2 inhibitor suppressed glioma growth and cell migration [335, 336]. CXCR2 is a poor prognostic marker in gastric cancer patients [337]. IL-1β transactivated EGFR via the CXCL1-CXR2 axis by increasing CXCL1 expression in oral squamous cell carcinoma [338]. CXCR2 expression along with postoperative complications affect recurrence free as well as overall survival of patients with esophageal cancer [339]. CXCL8-CXCR1/2 axis plays important role in the head and neck squamous cell carcinoma (HNSCC). IL-8 siRNA inhibited proliferation of HNSCC cells [340].

Chronic obstructive pulmonary disorder

COPD is a leading cause of morbidity and mortality in developed countries and is characterized by progressive and irreversible airflow obstruction caused by fibrosis and narrowing of small airways, and destruction of alveolar attachments (emphysema), which are heavily mediated by neutrophils and lymphocytes ^{151, 152}. CXCL8 contributes to the pathogenesis of COPD through several mechanisms. CXCL8 and other chemokines secreted by lung macrophages orchestrate the trafficking of polymorphonuclear neutrophils (PMN) to the lungs in response to external stimuli (cigarette smoke, air pollutants) ^{153, 154}. CXCL8 also stimulates the airway epithelium, causing it to contract and increase its permeability to inflammatory cells ¹¹⁵. Protease secretion from the accumulation of neutrophils and other inflammatory cells leads to sustained and extensive tissue damage ^{115, 154, 155}. The concentrations of CXCL8 in sputum and bronchoalveolar lavage (BAL) are higher in patients with COPD than healthy volunteers and correlate with increased neutrophil accumulation ^156-160. p53 induces plasminogen activator inhibitor-1 (PAI-1) expression levels in alveolar epithelial cells and enhances the expression of CXCL1, CXCL2 and CXCR2 in BAL during chronic cigarette smoke exposure ¹⁶¹. Additionally, neutralizing CXCL8 antibodies significantly reduce neutrophil chemotactic activity of sputum from patients with COPD ¹⁶². CXCR2 inhibition with a small-molecule antagonist reduces neutrophilic inflammation in lungs of mice exposed to acute cigarette smoke, suggesting that CXCL8 plays an important role in lung inflammation that contributes to the development of COPD ¹⁶³.

Cellular crosstalk between alveolar macrophage-secreted CXCL8 and CXCR2-expressing neutrophils contributes to COPD. Therefore, blocking the CXCL8-CXCR1/2 pathway could be beneficial in treating COPD.

Asthma

Asthma is characterized by episodes of reversible airflow obstruction, bronchial constriction, and lung inflammation induced by allergens ¹⁶⁴. Neutrophils and eosinophils are increased in the lung epithelium and sputum during severe asthma exacerbations accompanied by increased expression of ELR+ chemokines and its receptors ^165-171. CXCR1/2 is also expressed on airway smooth muscle and mediates cell contraction and migration to enhance airway responsiveness and remodeling (bronchoconstriction) that is observed in asthma ^{172, 173}. Lastly, CXCR2 deficient mice exhibit reduced bronchial hyper-responsiveness and neutrophil recruitment induced by ozone challenge compared to wild-type mice ¹⁷⁴. CXCL8-CXCR2 dependent neutrophil recruitment is important for the development of asthma and the blockade of this signaling pathway may provide a new approach to treating asthma.

Cystic fibrosis

Cystic fibrosis (CF) is an autosomal recessive genetic disorder caused by genetic mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene that leads to abnormal transport of chloride and sodium ions across the epithelium of tissue organs, such as the lungs and pancreas ^{175, 176}. The most affected organs of CF are the lungs, which exhibit increased mucus buildup. Subsequent cycles of lung infection and neutrophilic inflammation lead to bronchiectasis and respiratory failure ¹⁷⁷. As with other lung diseases discussed thus far, the airways of CF patients also have elevated levels of CXCL8 and other chemotactic cytokines (IL-1, IL-6, TNF) that coordinate the infiltration of neutrophils ^178-183. Antibodies against CXCL8 significantly inhibit the chemotactic activity of sputum from CF patients ¹⁸⁴. Though CXCR1 and CXCR2 expression in airway smooth muscles from CF and non-CF patients are about the same, CXCL8 induces greater contractions in airway smooth muscle of CF patients, which might be due to increased myosin light chains that contribute to bronchoconstriction observed in CF ¹⁸⁵. CXCR1, but not CXCR2, promotes bacterial killing; however, this function is lost in the airways of CF patients ¹⁰². CXCR1 is cleaved by airway proteases and the fragments of CXCR1 stimulate bronchial epithelial cells to secrete CXCL8 via Toll-like receptor 2 (TLR2) ¹⁰². CF patients carrying a particular CXCR1-2_HA are associated with decreased CXCR1 combined with increased CXCR2 mRNA and protein expression, and impaired antibacterial function. These haplotypes are also associated with decreased lung function in CF patients ¹⁸⁶.

Inflammatory bowel diseases

PMN recruitment guides the CXCL8-CXCR1/2 axis that plays a significant role in the pathogenesis of IBD ^187-189. Neutrophils are rapidly recruited at the site of infection within the intestine in response to CXCL8. This is caused by the increase of migratory capacity of PMNs leading to the overexpression of CXCR1/2 ¹⁹⁰. Bacterial infection triggers epithelial cells to release CXCL8, which mediates migration of PMNs from blood circulation ¹⁹⁰. Patients with ulcerative colitis have elevated levels of CXCL8 mRNA expression in colonic mucosa compared to healthy volunteers ¹⁹¹. CXCR2 antagonist SB225002 attenuated severity of the disease in DSS-induced experimental colitis in paired immunoglobulin-like type 2 receptor alpha (PILRα)-deficient mice by negatively regulating neutrophil function ¹⁹². These studies support the relevance of the CXCL8-CXCR1/2 axis in the pathogenesis of IBD; therefore, disruption of this signaling may be an attractive therapeutic strategy for the treatment of IBD.

Neuro-inflammatory diseases

Neuro-inflammatory diseases, such as Neuro-Sweet disease, are characterized by neutrophil infiltration due to the abnormal chemotaxis of neutrophils mediated by the CXCL8-CXCR2 axis. Patients with this disease exhibit elevated levels of CXCL8 in comparison to healthy subjects ¹⁹³.

CXCR2, expressed in neutrophils and oligodendrocyte progenitor cells, is reported to promote demyelination leading to multiple sclerosis. Inhibition of CXCR2 by a CNS penetrating antagonist or conditional knockout of Cxcr2 improves remyelination in a mouse model ^{122, 194, 195}. Cxcr2^-/- mice were resistant to cuprizone-induced demylination as compared to Cxcr2^+/+ mice ¹²⁰. A CNS penetrating CXCR2 antagonist (compound 22) exhibited significant remyelination at 100 mg/kg twice daily dose for 9 consecutive days in cuprizone-induced demyelination mouse model ¹⁹¹. In a separate study, it was reported that tamoxifen-treated Cxcr2-conditional knock-out mice showed modestly, but significantly accelerated remyelination (26.2% PLP-IR area fraction) after 2 weeks recovery from 6 weeks cuprizone feeding compared to tamoxifen-treated control mice (24.1% PLP-IR area fraction) ¹⁹².

Suppressor of cytokine signaling 3 (SOCS3) deficient neutrophils activate Stat3 and produce high levels of CXCL2. CXCL2 plays a critical role in the development of atypical experimental autoimmune encephalomyelitis (EAE), a model for multiple sclerosis in mice. The CXCR2 inhibition attenuated atypical EAE by blocking the neutrophil from infiltrating the cerebellum and brainstem ¹⁹⁶.

CXCR1/2 signaling is also implicated in various neurodegenerative diseases, such as Alzheimer's disease (AD) and stroke ^197-201. Increased expression of CXCL8 in serum and brain of HIV-1 patients is associated with neurocognitive disorders ¹⁹⁹. Expression of CXCR2 and CXCL8 increases in microglia and inhibition of CXCR2 reduced microgliosis and conferred neuroprotection in Aβ1-42 induced rat model of AD ²⁰². CXCR2 gene polymorphism is also associated with the risk of ischemic stroke in patients with essential hypertension ¹⁹⁸. Since CXCR2 and its ligand CXCL8 are upregulated in the AD tissues and CXCR2-dependent inflammatory responses play a critical role in progression of AD, inhibition of CXCR2 would be an effective neuroprotective strategy.

Vascular diseases

Myocardial and cerebral ischemia and infarction produce reactive oxygen species as well as elevated levels of pro-inflammatory cytokines and chemokines, triggering recruitment of circulating leukocytes and neutrophils into the ischemic tissues. As a consequence, the activated neutrophils can cause ischemic injury to the tissue. Targeting CXCL8-CXCR1/2 axis has proved to be an effective approach to treat ischemic injury ^203-205. Treatment with anti-CXCR2 antibody significantly blocked neutrophil infiltration into the infarcted area and reduced the size of infarction after long and delayed treatment in a mouse model of chronic myocardial infarction ²⁰⁶. Expression levels of CXCL5, macrophage migration-inhibitory factor (MIF) and CXCR2 were shown to be elevated in human atherosclerotic coronary artery. It was proposed that the CXCL5-CXCR2 and/or MIF-CXCR2 interactions could increase the risk of coronary artery disease ^{207, 208}. Contrary to the CXCR2 action on neutrophil, high levels of expression of CXCR2 on endothelial progenitor cells (EPCs) regulate their homing to regressing plaques and promoted plaque resolution ²⁰⁹. All of these previously described observations suggest the importance of CXCR2 in the development of coronary artery disease and disruption of CXCL5-CXCR2 and/or MIF-CXCR2 interactions by pharmacological antagonism of CXCR2 could be a therapeutic approach to this disease.

CXCR2 also plays a role in the development of hypertension. Infiltration of pro-inflammatory CXCR2⁺ cells causes vascular dysfunction and hypertension. Knocking out of CXCR2 gene or treatment with CXCR2 antagonist SB265610 attenuates angiotensin II-induced hypertension, vascular remodeling of aortic macrophage infiltration in mice ²¹⁰. Therapeutic intervention by pharmacological antagonist of CXCR2 could be an attractive strategy to treat hypertension where infiltration of CXCR2+ cells plays important role in disease development.

Arthritis

CXCR2 signaling plays an important role in maintaining cartilage homeostasis. ELR+ CXC chemokines attract CXCR2 expressing inflammatory cells such as neutrophils in inflammatory arthritis ²¹¹. CXCR1/2 is responsible for pathogenesis of rheumatoid arthritis (RA) as an inducing factor for neutrophil adhesion to the synovial microvascular endothelium and thus, promotes neutrophil migration into the joints ²¹². CXCR2 also mediates neutrophil recruitment in Brucelas-induced arthritis ²¹³.

While activation of CXCR1/2 signaling is critical for the pathogenesis of most types of arthritis, including RA, this is not the case with osteoarthritis. ELR+ chemokines are expressed in healthy cartilage and maintain viability and differentiation of chondrocytes. Disruption in the CXCR1/2 signaling induces phenotypic instability of chondrocyte, a leading cause of osteoarthritis ²¹¹. While high CXCR1/2 signaling contributes to the development of chronic arthritis, low CXCR1/2 signaling may serve as protective factor against the development of chronic osteoarthritis. Thus, maintaining optimum level of CXCR1/2 signaling is important for articular cartilage.

Psoriasis

Psoriasis is an autoimmune inflammatory skin lesion in which high epidermal cells build up. CXCR2 guided neutrophil accumulation in the skin is one of the histological features of psoriasis. Chemokines, CXCL1, CXCL2, and leukotriene B4 are upregulated in psoriatic skin and attract CXCR2 expressing neutrophils into the psoriatic lesions. CXCL1/2-CXCR2 axis initiates neutrophil infiltration into the psoriatic skin, which is further driven by LTB₄-BLT1 axis ²¹⁴. Interleukin-17A (IL-17A) induces CXCL8 production in keratinocytes. Targeting IL-17A by an antibody secukinumab reduces CXCL8 expression and induces neutrophil clearance from psoriatic skin ²¹⁵. CXCR2 antagonist, SB225002 also reduces neutrophil infiltration and partly alleviated psoriatic skin inflammation in a mouse model of psoriatic skin ²¹⁴. Although, ABX-IL-8 (Abgenix, Fremont, CA), a fully humanized antibody against CXCL8, failed to show efficacy in phase II clinical trials, ABCream (Anogen, Missassauga, ON, Canada), a formulation of CXCL8-blocking monoclonal antibody is reported to be effective in psoriatic treatment and has been approved in China ^{214, 216}. CXCR2 mediated neutrophil recruitment also contributes to other skin lesions, such as livedoid vasculopathy ²¹⁷ and atopic dermatitis ²¹⁸.

These observations corroborate the role of CXCR2 and its ligands in the development of psoriasis and other skin lesions, and the disruption of CXCR2 signaling by CXCL8-blocking antibodies or CXCR2 antagonists could be effective in the treatment of these skin lesions.

Other inflammatory disorders

CXCR2 is also important for other inflammatory conditions. CXCR2-mediated neutrophil recruitment is essential for clearing microbial spores from infected lungs. Cxcr2^-/- mice develop severe hypoxia and inflammation during pulmonary aspergillosis compared to the wild-type mice ²¹⁹.

CXCR2 signaling pathway is also reported to have a role in nociception. Overexpression of GRK6 or treatment with CXCR2 siRNA suppresses CXCR2 expression in dorsal root ganglion and significantly reduces pain in animal model of neuropathic pain ^{220, 221}.

Homing of mesenchymal stem cells (MSCs) to the site of injury is important for wound healing and this is regulated by the human natural killer (NK) cells. Interaction between CXCR2 expressed by MSCs and NAP-2 secreted by NK cells is one of the driving forces of tissue repair or regeneration at the site of injury ²²².

CXCL1 and CXCL8 are significantly upregulated in various inflammatory liver diseases, including alcoholic hepatitis and are responsible for formation of Mallory-Denk Bodies in the liver. Inhibition of CXCR2 signaling blocks neutrophil infiltration and reduces liver injury ^{223, 224}. CXCR2 signaling is also involved in the development of pancreatitis and the inhibition of CXCR2 suppresses pancreatic inflammation ²²⁵. CXCR2 signaling is also associated with Sjogren's syndrome ²²⁶, HIV-associated nephropathy ²²⁷ and type 1 diabetes ²²⁸.

Maintaining CXCR2 homeostasis is very important for microbial infections, wound healing as well as various inflammatory diseases, where CXCR2 signaling is either beneficial or detrimental. Further studies are needed to better elucidate the role of CXCR2 signaling and its inhibition in these diseases.

The CXCL8-CXCR1/2 Axis in Cancer

Chemokines for CXCR1/2 including CXCL1, 5, 7, and 8 are secreted and expressed by various cancer cell types and stimulate cancer cell proliferation and migration in an autocrine fashion ^229-236. Oncomine analysis reveals that the CXCL8 expression is significantly higher in colon, head and neck, pancreatic and esophageal cancers as compared to healthy tissues (Figure 7) ¹³⁴. Chemokine expression also correlates with tumor grade and metastatic potential in human tumors. For example, CXCL8 serum levels are increased in patients with prostate cancer as compared to healthy volunteers, and it correlates with the stages of metastasis ²³⁷. In pancreatic cancer, CXCL5 expression is associated with shorter patient survival (25.5 months shorter than patients expressing low CXCL5) and correlates with clinical stage ²³⁸. CXCL8 is also overexpressed in bladder cancer and the overexpression of CXCL8 is associated with late stage disease. Overall survival rate of bladder cancer patients is significantly reduced with high expression of CXCL8 ²³⁹. Vascular invasive cancer phenotype is associated with overexpression of CXCL1-3 and CXCL8, and inhibition of CXCR2 signaling reduces tumor invasion ²⁴⁰. Tumor-associated macrophage-derived CXCL8 induces suppression of estrogen receptor-α via HOXB13 and promotes tumor cell invasion and metastasis in endometrial cancer ²⁴¹. We compared the gene expression level of CXCL8 and its receptors in the NCI-60 cell lines and in 1036 cell lines from the Cancer Cell Line Encyclopedia (CCLE) (Figure 8). While CXCL8 is overexpressed in various cancer cell lines, its receptors, CXCR1 and CXCR2, are only overexpressed in very few cancer cell lines. Normal tissue distributions of CXCL8 and CXCR1/2 are plotted in Figure 9. In accordance with their role in the inflammatory response, tissues from the immune system express high levels of these genes. We investigated the functional importance of the CXCL8-CXCR1/2 axis using the dataset from the Project Achilles database and observed that CXCL8, CXCR1 or CXCR2 knockdown negatively impacted cell survival and proliferation (Figure 10) ²⁴². The list of top cell lines showing greatest sensitivity to the corresponding shRNA treatments are shown in supplementary Tables S1-S5. A summary of the roles and involvement of the CXCL8-CXCR1/2 axis in major cancers is shown in Table 3.

Figure 7.

CXCL8, CXCR1/2 expression in cancer vs normal tissues from Oncomine analysis. Cancer vs normal tissue expression of CXCL8, CXCR1 and CXCR2 (A) where threshold p-value = 0.0001, fold change = 2 and gene rank = top 10%. CXCL8 expression (Log2 median) in patient samples of Kaiser Colon (B), Hu Esophagus (C), Ye Head-Neck (D) and Badea Pancreas (E). AC = Adenocarcinoma.

Figure 8.

Gene expression data of CXCL8, CXCR1/2 in the panel of cancer cell lines. Gene expression of CXCL8, CXCR1/2 in the NCI-60 cell lines (A) and 1036 cell lines from the Cancer Cell Line Encyclopedia (CCLE) (B).

Figure 9.

CXCL8 and CXCR1/2 gene expression distribution in normal tissues. Gene expression data of CXCL8 (A), CXCR1 (B) and CXCR2 (C) in various human tissues (NextBio, http://www.nextbio.com). Data are plotted as mean ± standard deviation. Dotted line represents the average value of all tissues.

Figure 10.

shRNA activity of CXCL8 (A), CXCR1 (B) and CXCR2 (C) in a panel of cancer cell lines. Data obtained from Project Achilles Data Portal of Broad Institute (https://www.broadinstitute.org/achilles ). shRNA score denotes log2 based decrease in CXCL8, CXCR1 or CXCR2 shRNA compared to pooled shRNA in cancer cell lines after several rounds of proliferation post-shRNA infection. A negative shRNA score indicates decreased cancer cell proliferation post-shRNA transfection.

Cancer cells are also stimulated by other sources of chemokines, mainly derived from tumor-associated macrophages. The hypoxic and stressed tumor microenvironments stimulate macrophages to secrete CXCL chemokines, a process mediated by NF-κB ^{24, 243}. CXCR1/2 ligands stimulate CXCR1/2 expressing endothelial cells and promote tumor angiogenesis. In order for a tumor to progress beyond 2-3 mm³, it must acquire the capacity to induce angiogenesis ²⁴⁴. The tumor vasculature delivers essential nutrients and oxygen to the tumor cells that facilitate the uncontrolled growth and invasion of tumor cells. For example, upon CXCL8 stimulation, endothelial cells begin the angiogenic process by secreting matrix metalloproteinases (MMPs) to break down the extracellular matrix (ECM) and start to proliferate and initiate the formation of capillaries ²⁴⁵. The involvement of CXCR2 in tumor progression and angiogenesis is further demonstrated by several in vivo cancer models that showed depletion of chemokines and/or the receptor significantly reduced tumor growth associated with decreased microvessel density ¹⁰⁴. For example, the CXCR2 knockout mice implanted with Lewis lung carcinoma (LLC) exhibit reduced tumor growth, vascular density, and spontaneous metastases in orthotopic tumor models compared to the wild-type ¹⁰⁴. Similar results were observed in prostate and pancreatic cancers, in which CXCR2 knockout mice had smaller tumors and reduced tumor angiogenesis ^{129, 246}.

Small-molecule inhibitors of CXCR2 have also shown promising anticancer effects in preclinical studies. Treatment with SCH527123 reduced tumor growth and angiogenesis as well as improved sensitivity to oxaliplatin treatment in colon cancer mice xenografts ²⁴⁷. However, SCH527123 inhibited colon cancer liver metastases but had no effects on tumor growth in mice xenografts ²⁴⁸. Both CXCR2 knockout studies using shRNA and small-molecules against CXCR2 did not significantly affect cancer cell proliferation in in vitro studies, further supporting that CXCR2 and its ligands play an essential role in the tumor microenvironment.

CXCL8 is also up-regulated in response to various anticancer agents and may contribute to chemoresistance. For example, NF-κB-mediated CXCL8 synthesis and secretion was elevated in prostate cancer cells treated with oxaliplatin or 5-FU ^{249, 250} and in pancreatic cells treated with gemcitabine ²⁵¹. Additionally, the inhibition of CXCR2 with small-molecule antagonists or shRNA knockdown significantly enhanced sensitivity to oxaliplatin, 5-FU, ansamycin, TRAIL, doxorubicin, paclitaxel, and lapatinib in prostate, colon, and breast cancer cells ^{247, 249, 250, 252-256}. Taken together, the CXCL8-CXCR1/2 axis facilitates tumor progression by stimulating tumor cells and it is a critical component of the tumor microenvironment (Figure 11).

Figure 11.

The multiple roles of CXCL chemokines and CXCR1/2 during tumor development. CXCR1/2 and CXCL promote tumor growth through several mechanisms. Secretion of CXCL8 by tumor cells and tumor-associated macrophages stimulates cancer cell proliferation, survival, and chemoresistance. CXCL8 secretion also mediates neutrophil recruitment to the tumor site and stimulates neutrophils to secrete growth factors (GF) and matrix metalloproteinase (MMPs) to facilitate cancer cell migration, invasion, and metastases. CXCL8 stimulates CXCR2-expressing endothelial cells that form blood vessels within the tumor and stimulate tumor angiogenesis.

Interestingly, the volatile anesthetics used during surgery are reported to enhance cell proliferation and metastasis, leading to relapse of tumor. CXCR2 signaling is one of the pathways activated by such anesthetics. Inhibition of CXCR2 by siRNA abrogates migratory activities of anesthetics on ovarian cancer cells ^{257, 258}. In separate studies it was shown that the Kaposi's sarcoma-associated herpesvirus (KSHV) miRNA, miR-K3, promotes angiogenesis, cell migration and invasion through GRK2/CXCR2/AKT signaling ^{259, 260}.

Above observations suggest that the CXCL8-CXCR1/2 signaling pathway plays significant roles in the tumor progression and metastasis. CXCR1/2 as well as CXCL8 are overexpressed in various tumors and are correlated with tumor stages and grades. Knocking down CXCR2 and CXCL8 as well as disruptions of the CXCL8-CXCR1/2 signaling by small-molecules and antibodies are effective in blocking tumor growth and metastasis as well as in improving sensitivity towards other chemotherapy.

CXCR2: linking COPD to lung cancer

The manifestations of lung cancer and COPD are diametrically opposed. Lung cancer is characterized by uncontrolled cell proliferation, whereas COPD is characterized by inflammation-mediated destruction of the extracellular matrix and cell death ^{261, 262}. Hence, treatment for these two diseases has been separately developed, targeting different cellular pathways. However, several studies suggest that inflammation may be an underlying mechanism that contributes to the development of both diseases. Mouse models with an activating K-ras mutation suggest COPD-like airway inflammation promotes the progression of lung cancer development in mice. K-ras mutations are found in 30% of all lung adenocarcinomas from smokers ²⁶³. Exposure of mice with an activating K-ras mutation to aerosolized NTHi lysate (Haemophilus influenza, commonly found in lower respiratory tract of COPD patients) resulted in neutrophil/macrophage/CD8 T cell-associated COPD-like airway inflammation. These mice exhibited a 3.2-fold increase in lung surface tumor number (156±9 versus 45±7). The study concludes that COPD-like airway inflammation promotes lung carcinogenesis in a background of an activated K-ras allele in airway secretory cells ²⁶⁴. Observational studies have also found that smokers with COPD have a 1.3 to 4.9 fold increased risk of lung cancer compared to smokers without COPD ^265-267, which suggest that COPD and lung cancer are linked. Studies among ex-smokers with COPD showed that concurrent, regular use of inhaled corticosteroids reduced risk of lung cancer by 50%, suggesting that reduced inflammation in COPD patients offers a protective effect against cancer ²⁶⁸.

Houghton et al. proposed a model that explores the common origins of lung cancer and COPD. Upon cigarette smoke or pathogen exposure in the lungs, inflammatory cells are recruited and activated, releasing serine, MMPs and reactive oxygen species (ROS). Emphysema, a major type of COPD, occurs when extracellular matrix destruction and cell death exceed reparative capacity resulting in airspace enlargement. To compensate for the loss of damaged alveolar cells, bronchioalveolar stem cells (BASCs) proliferate. However, the over-compensation of BASC proliferation predisposes these cells to become malignant ²⁶⁹. According to this model, both diseases arise from inflammatory recruitment of macrophages and neutrophils to the lungs. Hence, therapeutics that will reduce the inflammatory response upon pathogen exposure will diminish episodes of emphysema and prevent the development of cancerous cells. Since CXCR1/2 and their ligands play an important role in inflammatory response and have been implicated in both lung cancer and COPD, it may be a potential common molecular pathway that links these two diseases together.

CXCL8-CXCR1/2 axis in immunogenic cell deaths

Chemotherapy induces immunogenic cell death (ICD) of tumor cells by releasing damage-associated molecular patterns (DAMP). ICD is a multistep process, which include a) the secretion of “find me” signals, such as fractalkine, nucleotides, and ATP, by dying cells to chemo-attract phagocytes or dendritic cells (DCs), b) the expression of “eat me” signals (calreticulin and phosphatidylserine) that enable phagocytes or DCs to attack the cancer cells, and, c) the release of danger signals, such as high mobility group box 1 protein (HMGB1) aiding dying tumor cells to loose tolerance ³⁴¹. As a result, anticancer immune responses are developed, leading to the cancer cell death. siRNA screening of several GPCRs and their ligands revealed that the depletion of CXCR1 and its ligand CXCL8 inhibit immunogenic translocation of calreticulin to the cell surface in the methotrexate treated cancer cells. In fact, methotrexate induced production of CXCL8 by human cancer cells in vitro. Transcriptome analyses in in vivo tumor-bearing mice models revealed that ICD is preceded by the transcriptional activation of Cxcl2, the mouse homolog of CXCL8 ³⁴². HMGB1 also stimulates production of pro-inflammatory cytokines, such as CXCL8 from neutrophils and macrophages and thus, accelerates ICD ³⁴³. Although CXCL8-CXCR1/2 axis promotes tumor progression, invasion and metastasis in both spontaneous and inflammation-driven tumor models, it may also block the growth of early neoplastic lesions by inducing cell senescence and promotes the recruitment of innate immune effectors which mediates immunogenic cell death. Therefore, therapeutic application of CXCR1/2 agonist or antagonist should be explored with caution depending on cancer type, stage and certain drug combinations.

Immune checkpoint inhibition and CXCL8-CXCR1/2 axis

Suppression of host's immune response plays an important role in cancer progression. There are several immune checkpoints which facilitate immune escape of cancer cells by suppressing host's immune response. Expansion of myeloid-derived suppressor cells (MDSCs) is used as one of these checkpoints by various cancers, such as pancreatic and prostate to evade immune system. CXCR2 signaling is responsible for trafficking of these MDSCs to the tumor bed ^{325, 344-348}. Cancer stem cells produce high levels of macrophage migration inhibitory factor (MIF), which increases the production of the immune suppressive enzyme arginase-1 in MDSCs in a CXCR2 dependent manner. Inhibition of MIF receptor, CXCR2 by antibody decreased production of arginase-1 and thus, blocked immune evasion of tumor cells ³⁴⁹. Inhibition of programmed death-1 (PD-1) by anti-PD-1 antibody prevented tumor growth during early phase tumor progression, but delayed treatment with anti-PD-1 therapy showed little benefits ³⁴⁶. MDSCs cause resistance to anti-PD1 therapy by inhibiting T-cell infiltration and activation ^{325, 346}. Treatment with CXCR2-blocking antibody or CXCR2 antagonist AZ13381758 along with anti-PD-1 antibody exhibited greater benefit on tumor growth inhibition as well as survival in mice model compared to either agent alone ³⁴⁶. Therefore, CXCR2 signaling is important for regulating tumor immune checkpoints and disruptions of this signaling pathway by antagonist or anti-CXCR2 antibody in combination with chemotherapy could block immune evasion by cancer cells.

CXCL8-CXCR1/2 axis and cancer stem cells

Cancer stem cells (CSCs) can initiate and maintain the growth and progression of cancer due to their self-renewal and differentiation properties. The importance of CSCs has been established in many cancers including breast, liver, colon, and pancreatic cancers ³²¹. Recent studies suggest that the CXCL8-CXCR1/2 axis may play an important role in the tumor progression and metastasis by regulating CSC proliferation and self-renewal ^{292, 293, 321, 350-354}. CXCR1 positively correlates with the CSC markers CD44 and CD133 in pancreatic cancer and the CXCL8-CXCR1 axis induces stem-cell like mammosphere formation and increase in CSC population in pancreatic cancer cells in vitro ³²¹. CXCR1- and CXCL8-specific blocking antibodies reverse these effects on pancreatic CSCs ³²¹. CXCL8 promotes epithelial-mesenchymal transition (EMT) of human breast cancer cells in an autocrine/paracrine manner. EMT induces metastatic and stemness characteristics as well as intrinsic resistance of tumor cells ^{344, 355}.

Suppression of CXCR2 in human pluripotent stem cells (hPSCs) by siRNA led to inhibition of the maintenance of stemness characteristics and proliferation, and causes a significant decrease in the expression of pluripotency markers OCT-4, Nanog, and Rex-1 and an increase in the expression of germ layer markers NESTIN and GATA3 ³⁵⁰. Inhibition of CXCR2 hindered hPSC self-renewal and resulted in a gradual increase in cell differentiation ³⁵⁰. The CXCL8-CXCR1/2 signaling regulates maintenance and cellular growth of glioblastoma stem cells. Glioblastoma endothelial cells cultured in 3D scaffold-based system enhanced CXCL8 levels and upregulated CXCR1/2 leading to increased migration, growth and stem cell characteristics of CSCs ³⁵⁶. CXCR2 plays a critical role in the pathogenesis of glioblastoma as silencing of CXCR2 in CSCs abrogated tumor-promoting effects of CSCs in vivo ³⁵⁶. Recombinant CXCL8-induced activation of CXCR1/2 in breast cancer cells resulted in an increased pool of CSCs and cell self-renewal. CXCL8-CXCR1/2 cascades trans-activated HER2 signaling mediated by SRC, PI3K, and MEK in metastatic breast CSCs ³⁵⁷. Mesenchymal stem cells (MSCs) in the bone promoted mammary cancer cell migration in vitro via the CXCR2 receptor. CXCR2, CXCL1 and CXCL5 antibodies and a small molecule inhibitor of CXCR2, SB265610 significantly abrogated the migratory effect of the PyMT cells to MSC conditioned media ²⁹⁰. CXCR2 is also highly expressed in stem cell populations of acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) and plays critical role in tumor progression ³⁵⁸.

As CSCs are thought to be responsible for tumor initiation, progression and metastasis, the CXCL8-CXCR1/2 axis may be an important therapeutic target for cancer. Combination of CXCL8-CXCR1/2 interaction inhibitor and other targeted therapy, such as HER2 inhibitor may be a novel approach to treat cancer. CXCR2 not only regulates CSCs, but also the survival and self-renewal of human pluripotent stem cells (hPSCs) as well as hematopoietic stem cells (HSCs). Cxcr2 knockout studies in mice as well as CXCR2 knockdown studies in cultured-cells demonstrated reduced pluripotency and self-renewal capacity of hPSCs and HSCs ^359-361.

CXCR2 interacts with PSD-95/DlgA/ZO-1 (PDZ) scaffold protein NHERF1 to regulate endothelial progenitor cell (EPC) homing and angiogenesis. Disruption of this interaction leads to decrease in in vivo angiogenic activities of EPC in mice ³⁶². Therefore, targeting CXCR2 could be a plausible therapeutic strategy for angiogenesis-related diseases.

CXCL8-CXCR1/2 axis and the tumor microenvironment

The tumor microenvironment plays an important role in the progression and metastasis of tumors. The tumor microenvironment consists of a variety of non-malignant stromal cells, such as adipocytes, fibroblasts, and migratory hematopoietic cells. Along with these components of the tumor microenvironment, tumor-associated macrophages (TAMs), which are migratory hematopoietic cells, have a pivotal role in tumor progression and metastasis ³⁶³. Many cancer cells secrete high levels of CXCR1/2 ligands without expressing the receptors, suggesting that CXCR1/2 ligands are also involved in the tumor microenvironment in a paracrine fashion (199, 203). Indeed, CXCR1/2 ligands played a significant role in neutrophil tumor infiltration that facilitated cancer cell proliferation, invasion, and chemoresistance via increased levels of angiogenic and growth factors produced by these tumor-associated neutrophils ^{125, 364-366}. CXCL8 is one of the most abundant cytokines in TAMs and mediates TAM-dependent tumor invasion and metastasis of papillary thyroid carcinoma (PTC) in vivo ³⁶⁷. Obesity-associated adipose stromal cells (ASCs), which aggravate certain cancers, such as prostate cancer, are recruited into tumor microenvironment by CXCL1/8-CXCR1/2 axis ³⁶⁸.

CXCR2 signaling in the microenvironment regulates breast cancer growth and metastasis. There is a decrease in overall tumor cell proliferation, angiogenesis, apoptosis and metastasis in Cxcr2^-/- mice compared to wild-type mice ³⁶⁹. Optical coherence tomography and laser-induced fluorescence imaging assays demonstrated that CXCR2 expression in the tumor microenvironment positively correlates with tumor burden in colon cancer ³⁷⁰. CXCL8 and the CXCR2 in the tumor microenvironment stimulated cell growth, progression and metastasis of colon cancer ²⁸³. The LKR-13 lung adenocarcinoma cells derived from Kras^LA1 mice are resistant to anti-CXCR2 antibody in vitro. However, the same cells established as syngeneic tumors in wild type mice are sensitive to the antibody, supporting the role of CXCR2 in the tumor microenvironment ²⁷⁴. CXCL1 elicits paracrine functions though CXCR2 receptor expressed on stromal cells in tumor microenvironment of an osteogenic prostate tumor ³⁷¹.

Pancreatic ductal adenocarcinoma (PDAC) cells mixed with pancreatic fibroblasts exhibited faster tumor growth than PDAC cells alone in mice. Treatment with a CXCR2 inhibitor or knock down of CXCR2 in the stromal cells delayed tumor growth in a mixed cell xenograft model ^{372, 373}. These observations further support the notion that interaction between CXCL8 in cancer cells and stromal CXCR2 is important for tumor progression, invasion and metastasis, and validate the CXCL8-CXCR2 interaction as an important therapeutic target for cancer.

Targeting CXCR1/2 signaling axis for theranostics

CXCR2 is an excellent target for in-vivo molecular imaging to determine therapeutic outcome. Increased expressions of CXCR2 and its ligands have been observed in cancer in cases of aberrant angiogenesis and higher grades and stages of malignancy. Tc-99m labeled ELR-containing 6-mer peptide (ELR-ECG) can potentially be used as a marker to target CXCR2 ³⁷⁴. CXCL5, a ligand for the receptor CXCR2, was reported as novel independent serum prognostic marker in patients with colorectal cancer (CRC). Preoperative serum CXCL5 is associated with CRC patients compared to healthy volunteers (p=0.013), liver metastasis (p=0.0040), and poor overall survival in univariate analysis (p=0.0002) and in multivariate analysis (p=0.038) ³⁷⁵. CXCL1-CXCR2 axis was predicted as a prognostic marker for hepatocellular carcinoma (HCC). While, CXCR2⁺ peri-tumoral stromal cell density was associated with reduced disease free survival (p = 0.015) and overall survival (p = 0.002) of HCC patients, CXCL1 was positively correlated with density of CD15+ neutrophils in tumor ³³⁴. Cxbladder, a non-invasive urine-based laboratory test to detect bladder cancer, measures urine CXCR2 level along with four cancer biomarkers to rule out false positives as CXCR2 is highly expressed in neutrophils associated with non-malignant inflammatory bladder conditions ³⁷⁶. CXCL8 along with CXCL10 and CXCL13 were reported as potential biomarkers for diagnosis of neurosyphilis patients, especially in asymptomatic neurosyphilis patients. Level of CXCL8 in CSF is more than 3 fold higher in neurosyphilis patients than in non-neurosyphilis patients ³⁷⁷.

These studies suggest that CXCR2 axis can be exploited as a theranostics tool. Similar studies with CXCR4 are well documented ^378-380.

CXCR1/2 Inhibition

Given the therapeutic potential of CXCR1/2 inhibition and the druggability of these receptors, several pharmaceutical companies have developed potent inhibitors during the past two decades. Several classes of small-molecule compounds as well as peptide-based inhibitors selectively inhibit CXCR1/2 receptors with IC₅₀ values in the sub-nanomolar range (Table 4). In general, many of the early small-molecule inhibitors were designed to be selective for CXCR2 over CXCR1 (with the exception of repertaxin). CXCR2-selective inhibitors are sought after for several reasons. First, since CXCR2 binds to all ELR+ chemokines, CXCR2 inhibition might provide a wider therapeutic application, especially in pathologies that may predominantly involve CXCR2-selective ligands. Second, as CXCR1 and CXCR2 play a critical role in the immune system, complete inhibition of both receptors might compromise immune response. Third, it was previously thought that most preclinical disease models (mouse and rat) used to assess the efficacy of these compounds only expressed the CXCR2 homologue and not CXCR1.

Table 4.

Select examples of CXCR1/2 inhibitors and their biological properties

CXCR1/2 Inhibitors	Activities of CXCR1/2 inhibitors
(R)-Ketoprofen	R)-Ketoprofen inhibited CXCL8-mediated PMN migration (IC50 = 34 nM) and interacts with the TM2 and TM7 region of CXCR1 [398].
Repertaxin	Repertaxin inhibited human PMN migration induced by CXCL8 (IC50 = 1 nM) and CXCL1 (IC50 = 400 nM) [392]. Repertaxin inhibited CXCL8-induced migration in CXCR1-transfected (IC50 = 1 nM) and CXCR2-transfected (IC50 = 100 nM) cells [392]. Repertaxin reduced lung neutrophil recruitment and vascular permeability by 50% in LPS-induced acute lung injury model at 15mg/kg [399]. Repertaxin reduced acute inflammation and autonomic dysreflexia in a model of spinal cord injury in rats [400]. Repertaxin inhibited CXCL8-induced PMN adhesion to fibrinogen, CD11b upregulation, and neutrophil activation (granule release) [394]. Reparixin in combination with paclitaxel inhibited brain tumour metastasis due to CSC [293]. Repertaxin inhibited CXCL8-induced T-cell and NK cell migration [394]. Repertaxin reduced granulocyte graft infiltration and serum creatinine post syngeneic transplantation in Lewis rats [401]. Repertaxin inhibited neutrophil recruitment into reperfused livers and reduced myeloperoxidase content in a rat model of liver post-ischaemia reperfusion injury [392]. Repertaxin reduced levels of hypertension-related mediators associated with reduced blood pressure in spontaneous hypertensive rat models [402]. Repertaxin reduced oligodendrocyte apoptosis and neutrophil migration to site of injury post traumatic spinal cord injury in rats [403]. Repertaxin inhibited CXCL8- or CINC-1-induced migration and calcium flux in human or rat neutrophils [395]. Repertaxin reduced neutrophil influx and vascular permeability in a model of mild and severe ischemia/reperfusion (I/R) injury in rats [395].
DF2156A	DF2156A is a noncompetitive allosteric inhibitor that is predicted to be stabilized by a direct ionic bond interaction with Lys99 on CXCR1 and Asp293 on CXCR2 [396]. DF2156A inhibited chemotaxis in CXCR1 and CXCR2 over-expressing transfectants and leukocytes [396]. DF2156A inhibited mice sponge-induced angiogenesis by reducing leukocyte influx and vessel formation [396]. DF2156A inhibited CXCL8-mediated HUVEC proliferation, migration and tube formation [396]. DF2156A decreased PMN influx in a rat model of ischemia and reperfusion injury [396]. DF2156A is a dual inhibitor of CXCR1 and CXCR2. It inhibited human and rat neutrophil migration in response to CXCL1 and CXCL8 [404]. DF2156A has improved in vivo half-life compared to repertaxin [404]. DF2156A decreased cerebral artery PMN infiltrate and improved neurological function in cerebral ischemia/reperfusion rat model [404].
DF2162	DF2162 prevented chemotaxis of rat and human neutrophils induced by chemokines acting on CXCR1/2. It is orally bioavailable and inhibited neutrophil influx and production of inflammatory factors in an arthritis rat model [405]. DF2162 reduced neutrophil accumulation in airway, but increased neutrophils in lung parenchyma in a bleomycin-induced pulmonary fibrosis mouse model [406].
Analogue 1 Analogue 2 Analogue 3	Analogues 1-3 inhibited CXCL1- and CXCL8-mediated human PMN migration with IC50 < 10 nM [393].
DF2755A	Allosteric inhibitor, DF2755A, selectively inhibited CXCL8-induced chemotaxis without affecting ligand binding to neutrophils and also reduced inflammatory and post-operative pain in several mouse models. It inhibited both CXCR1 (IC50 = 4.2 nM) and CXCR2 (IC50 = 2.1 nM) [407].
SCH527123	SCH527123 inhibited CXCL8 binding to CXCR1 (IC50=36 nM) and CXCR2 (IC50=2.6 nM) and inhibited neutrophil chemotaxis to CXCL8 (IC50=16 nM) and CXCL1 (IC50<1 nM) [408]. Optimization to improve potency of SCH527123 was performed [409-412]. SCH527123 bound to CXCR2 receptors of mice (Kd=0.2 nM), rat (Kd=0.02 nM), and monkey (Kd=0.08 nM) [388]. SCH527123 exhibited less affinity for monkey CXCR1 (5-fold decreased) and was >100 fold less potent in CXCR1-mediated chemotaxis [388]. SCH527123 blocked LPS induced pulmonary neutrophils (ED50=1.2-1.8 mg/kg) in mice and rat [388]. SCH527123 bound to CXCR1 (Kd=3.9 nM) and CXCR2 (Kd=0.049 nM) in CXCR1/2-overexpressing cell lines and inhibited CXCL1 and CXCL8-mediated neutrophil chemotaxis and myeloperoxidase release [389]. SCH527123 reduced sputum neutrophils in patients with severe asthma but had no effect on FEV1, sputum myeloperoxidase, CXCL8 or elastase [390]. SCH527123 reduced ozone induced sputum neutrophil in healthy volunteers. Treatment was safe and well-tolerated (4 day, 50 mg once daily dose) [243]. SCH527123 reduced LPS-induced sputum neutrophil influx (79% inhibition) compared to healthy volunteers [413]. SCH527123 and SCH479833 inhibited colon cancer metastasis, reduced tumor neovascularization, and increased tumor cell apoptosis in mice model [248].
SCH479833	SCH527123/SCH479833 inhibited melanoma cell proliferation, chemotaxis, and invasivon in vitro and reduced tumor growth associated with decreased tumor cell proliferation and microvessel density in an in vivo mouse model of melanoma [317]. SCH527123 was well tolerated with neutropenia without severe myelosuppresive effects in phase 1 clinical trial [414]. SCH527123 at 50 mg significantly improved FEV1 in patients with COPD in phase 2 clinical trial although at higher doses led to discontinuation due to decrease in absolute neutrophil count in serum [415].
SCH563705	SCH563705 exhibited potent inhibitory activities against CXCL8 binding to CXCR2 (Ki = 1 nM) and CXCR1 (Ki = 3 nM) [416]. SCH563705 inhibited CXCL1 and CXCL8 induced human PMN migration (CXCR2: IC50=0.5 nM, CXCR1: IC50=37 nM) [416]. SCH563705 reduced inflammation and bone and cartilage degradation in a mouse model of anti-collagen antibody-induced arthritis [417].
N-(3-bromo-4-cyano-2-hydroxyphenyl)-N-(2-bromophenyl)urea	N-(3-bromo-4-cyano-2-hydroxyphenyl)-N-(2-bromophenyl)urea is a competitive CXCR2 inhibitor. It inhibited human PMN chemotaxis mediated by CXCL1 (IC50 = 14 nM) and CXCL8 (IC50 = 35 nM) and inhibited CXCL8-mediated neutroprenia in rabbit models [418].
SB332235	SB332235 inhibited human CXCL8 binding to rabbit CXCR2 (IC50=40.5 nM) and CXCL8-induced calcium mobilization (IC50 = 7.7 nM). It had lower affinity for CXCR1 (IC50 > 1000 nM) and was less active against CXCR1-mediated calcium mobilization (IC50 = 2200 nM). SB332235 inhibited human CXCL8-induced chemotaxis of rabbit neutrophils (IC50 = 0.75 nM) [26]. SB332235 was optimized to improve pharmacokinetics [419]. SB332235 blocked T-cell entry when rat hippocampus was injected with amyloid β [420].
SB225002	SB225002 inhibited CXCL8 binding to CXCR2 (IC50 = 22 nM) and binding to CXCR1 with IC50 > 150 fold higher than CXCR2. SB225002 inhibited CXCL1 and CXCL8-induced chemotaxis in rabbit and human neutrophils [381]. SB225002 reduced alveolar neutrophil and exudate macropage influx in mice infected with S. pneumoniae [421]. SB225002 inhibited CXCL8 binding to CXCR2 (IC50 = 9.9 nM) and CXCL1 binding to CXCR2 (IC50 = 87.9 nM) [89]. In a mouse model of hepatic ischemia and reperfusion, post treatment with SB225002 increased hepatocyte proliferation and regeneration, similar to CXCR2-/- mice [422]. Optimization and other analogues of SB225002 were performed and reported [423]. SB225002 exhibited antinociceptive effects in several mouse models of pain (spontaneous nociception) [424]. SB225002 enhanced the activities of agonists for the δ opioid receptor acting in an allosteric fashion [425]. SB225002 reduced tumor progression in a mouse model of pancreatic cancer associated with reduced angiogenesis and improved survival [373].
SB265610	SB265610 reduced superoxide accumulation and lipid peroxidation in lungs, and preserved alveolar development in hypoxic newborn rats (exposed to 60% oxygen) [426]. SB265610 inhibited BAL neutrophil influx and myeloperoxidase accumulation in the lungs in hypoxia-induced newborn rat lung injury model [427]. SB265610 inhibited rat neutrophil calcium mobilization (IC50=3.7 nM) and chemotaxis (IC50=70 nM) to CINC-1 [427]. SB265610 acts as allosteric, inverse agonist. SB265610 reduced maximal [125I]-CXCL8 binding without affecting its Kd. It also reduced agonist-induced (CXCL1 and CXCL8) CXCR2 activation and basal [35S]-GTPγS binding [428].
SB656933	SB656933 reduced LPS-induced sputum neutrophil influx (52% inhibition) in healthy volunteers [413]. SB656933 inhibited neutrophil CD11b upregulation (IC50=261 nM) and shape change (associated with chemotaxis, IC50=311 nM) in COPD patients [429]. SB656933 inhibited CXCL1-induced CD11b expression on peripheral blood neutrophils (70% inhibition) and reduced sputum neutrophils which correlated with reduced myeloperoxidase concentrations in ozone-induced airway inflammation in healthy patients. SB656933 was safe and well-tolerated at single doses as high as 1100 mg [385]. SB656933 was well-tolerated at 50 mg in cystic fibrosis patients treated for 28 days, and patients had reduced sputum neutrophils (30% reduction) and elastase (26% reduction) compared to baseline. Patients had increased blood levels of fibrinogen, C-reactive protein (CRP), and CXCL8 compared to placebo. No changes in lung function were observed (NCT00903201) [386].
SB455821 (undisclosed structure)	SB455821 inhibited MIP-2-induced neutrophil migration in in vitro (IC50~20 nM) and in in vivo mice models [430].
SB-517785-M (undisclosed structure)	SB-517785-M reduced angiotensin II-induced neutrophils and mononuclear cell recruitment, arteriolar mononuclear leukocyte adhesion, and levels of MIP-1 and RANTES [431].
Danirixin (GSK1325756)	Danirixin, a selective CXCR2 antagonist inhibited CXCL1-induced neutrophil activation (CD11b expression) in healthy adults after repeated daily doses in a dose depedent manner [432].
Compound 22	A CNS penetrating analog of SB656933, compound 22 inhibited CXCR2 in a Tango assay with IC50 less than 1 nM and showed efficacy in treatment of CNS demyelinating disorders [194].
1-Phenyl-3-(cyclopent-2-en-1-yl) urea derivative 2	1-Phenyl-3-(cyclopent-2-en-1-yl) urea derivative 2 inhibited CXCR2 in a Tango assay with pIC50 ≥ 9.0 [433].
Diazafluorenones	Diazaflurenones inhibited CXCL8 binding to isolated human neutrophils with IC50 values from 0.05 to 12 μM [434].
Benzoimidazole analog 2	Benzoimidazole analog 2 inhibited CXCL8 binding to recombinant CXCR2 receptor expressed in CHO-K1 cells with IC50 value of 0.322 μM [435].
Napthalenecarboxamide 10	Napthalenecarboxamide 10 inhibited CXCR2 in calcium fluorescence assay (FLIPR) with an IC50 value of 2.2 μM [436].
AZD8309	AZD8309 reduced leukocyte count (48% inhibition) post nasal LPS challenge in healthy volunteers. No adverse effects were detected after 3 days of dosing [437].
Thiazolopyrimidine 29	Thiazolopyrimidine 29 inhibited CXCL8 binding to CXCR2 (IC50=14 nM) and calcium mobilization (IC50=40 nM) [438].
thiazolo[4,5-d]pyrimidine-2(3H)-ones	Thiazolo[4,5-d]pyrimidine-2(3H)-ones inhibited CXCL8 binding to CXCR2 (IC50=1-60 nM). They showed improved potency and oral bioavailability over thiazolopyrimidines [439].
AZ10397767	AZ10397767 increased cytotoxicity of geldanamycin and 17-AAG (HSP90 inhibitors) in PC3 but not DU145 prostate cancer cells [253]. AZ10397767 reduced neutrophil infiltration into A549 (NSCLC) spheroids and A549 tumor xenograft models in mice. Treatment did not reduce microvascular density [275]. Optimization studies to improve potency of Thiazolopyrimidine analogs were performed and reported [440]
AZD5069	AZD5069 inhibited radio labeled CXCL8 binding to human CXCR2 with IC50 = 0.8 nM [441]. Recent clinical data showed that AZD5069 did not adversely affect neutrophil mobilization from bone marrow to peripheral circulation, and thus, did not interfere with normal function of neutrophils in the phagocytosis or the oxidative burst response to bacterial pathogens [442]. AZD5069 significantly reduced sputum neutrophil counts, but failed to improve clinical outcomes in bronchiectasis patients [443]. AZD5069 was well tolerated with no increase in infection rates in phase 2 clinical trial in patients with COPD [444].
Compound 43	Compound 43 inhibited CXCR2 in low picomolar range (pIC50 , 8.4) with lower intrinsic renal clearance and good half life (t1/2, 3.2 h) in mice [445].
3-Arylamino-2H-1,2,4-benzothiadiazin-5-ol 1,1-dioxides	3-Arylamino-2H-1,2,4-benzothiadiazin-5-ol 1,1-dioxides inhibited CXCL8 binding to CXCR2 (IC50 = 30 nM) and CXCR1 (IC50 = 3.2µM), and inhibited FLLPR CXCR2 calcium assay (IC50 ~ 300-600 nM) [446].
2-Aminopyrimidin-4(1H)-one analog 3e	2-Aminopyrimidin-4(1H)-one analog 3e, a bioisostere of urea, inhibited CXCR2 (CXCR2-β-arrestin pIC50 = 8.2) [447].
Compound 6i	Compound 6i inhibited CXCL8 binding to CXCR1 (IC50 = 6.2 µM) and CXCR2 (IC50 = 30 nM) [448].
Compound 6j	Compound 6j was a dual CXCR1 and CXCR2 antagonist and inhibited CXCL8 binding to CXCR1 and CXCR2 with similar IC50 values (~20 nM) [448].
3,4-Diamino-1,2,5-thiadiazoles	3,4-Diamino-1,2,5-thiadiazoles inhibited CXCL8 binding to CXCR2 (IC50 = 13-126 nM) and CXCR1 (IC50 = 44 nM-10µM) [449, 450]. 3,4-Diamino-1,2,5-thiadiazoles were selective CXCR2 antagonists [449, 450].
6-amino-4-oxo-1,3-diphenyl-2-thioxo-1,2,3,4-tetrahydropyrimidine-5-carbonyl derivative 17	6-amino-4-oxo-1,3-diphenyl-2-thioxo-1,2,3,4-tetrahydropyrimidine-5-carbonyl derivative 17 inhibited CXCL8-induced PMN migration (IC50 = 0.02 nM) [451].
Acylsulfamide 6	Carboxylic acid bioisostere acylsulfamide 6 inhibited CXCL8 binding to CXCR2 (IC50 = 50 nM) and CXCL8-induced calcium mobilization (IC50 = 5 nM) [452]. Acylsulfamide 6 inhibited rabbit neutrophil chemotaxis (IC50 = 700 nM) [452]. Acylsulfamide 6 inhibited hyperoxia induced neutrophil (BAL) accumulation (~50% inhibition) in newborn rats at a dose of 10 mg/kg [452].
Triazolethiol 45	Triazolethiol 45 inhibited CXCL8 binding to CXCR2 (IC50 = 28 nM) and calcium mobilization (IC50 = 48 nM), and exhibited good bioavailability [453].
Imidazolylpyrimidine 1	Imidazolylpyrimidine 1 blocked CXCL8 binding to CXCR2 with Ki = 60 nM [454]. Imidazolylpyrimidine 1 bound to CXCR2 in transmembrane helices 3, 5, and 6 [455].
Imidazolylpyrimidine 40	Imidazolylpyrimidine 40 blocked CXCL8 binding to CXCR2 with Ki = 25 nM [454]
CXCL8(3-73)K11R	Blocked CXCL8 binding to human neutrophils (IC50 = 1.8 pM) and CXCL1 binding with less potency [456].
6-chloronicotinamide N-oxide 4a	6-chloronicotinamide N-oxide 4a inhibited CXCL8-induced human neutrophil chemotaxis (IC50 = 1.3-2.3 μM) [457]. Inhibited CXCL8 binding to CXCR1 and CXCR2 with similar IC50 values (~1µM) [457]. Well-tolerated in mice and stable to rat liver microsomes [457].
CXCL8(3-72)K11R/G31P (G31P) CXCL8 peptide	G31P was effective at 10 ng/mL in vitro. G31P inhibited both CXCR1- and CXCR2-mediated neutrophil migration and calcium mobilization. G31P blocked neutrophil infiltration in guinea pig model of airway endotoxemia [458]. G31P protected against ischemia and reperfusion injury in rats [203, 204]. G31P had no agonistic or chemotactic activity and antagonized the binding of antibodies to CXCR1/2 [459]. Other analogs and variations of this peptide was also generated [460]. G31P treatment prior to E.coli and LPS challenge in guinea pigs reduced pulmonary neutrophil recruitment (85% inhibition) [461]. G31P treatment significantly inhibited human lung cancer growth and metastasis by blocking the activity of CXCR1 and CXCR2. The treatment also significantly reduced expression of VEGF and NF-κB-p65 as well as phosphorylation of ERK1/2 and AKT [462].
SX-517	SX-517 inhibited CXCL1- and CXCL8-stimulated Ca2+ flux in human polymorphonuclear cells (hPMNs) with IC50 values of 38 nM and 36 nM, respectively. It also antagonized CXCL8-stimulated [35S]GTPγS binding (IC50 = 60 nM) in HEK293 cells that stably expressed human recombinant CXCR2 [463]. SX-517 acted as a noncompetitive inhibitor of CXCR2 as it failed to compete with the binding of [125I]-CXCL8 to CXCR2 [463].
SX-576	SX-576 inhibited GRO-α-mediated intracellular calcium release in isolated human PMNs with an IC50 value of 22 nM [464]. SX-576 exhibited significant microsomal stability in rat and monkey liver microsomes with more than 90% still intact after 60 min [464].
Compound 7	Compound 7 inhibited CXCL8-mediated calcium flux in stably CXCR1 or CXCR2 transfected RBL cells with IC50 values of 7 nM and 4 nM, respectively. Compound 7 has improved aqueous solubility, oral bioavailability and plasma stability compared to its analogs [465]. Compound 7 at a dose of 1 mg/kg also significantly reduced neutrophil influx in the BAL fluid of an in vivo rat model [465].
Triazolopyrimidine 14	Triazolopyrimidine 14 inhibited the binding of GRO-α to human recombinant CXCR2 expressed in CHO membranes with an IC50 of 0.33 μM [466]. Triazolopyrimidine 14 showed reasonable lead-like properties with 52% bioavailability (po), 9 h half-life (iv) and low microsomal clearance in rat [466].
Triazolopyrimidine 20	Triazolopyrimidine 20 inhibited the binding of GRO-α and GTPγS to human recombinant CXCR2 expressed in CHO membranes with IC50 values of 0.04 μM and 0.14 μM, respectively [467].
CX4338	CX4338 selectively inhibited CXCR2-mediated recruitment of β-arrestin-2 (IC50 = 6.3 μM) and receptor internalization [468]. CX4338 also enhanced CXCR2-mediated MAPK activation [468]. CX4338 inhibited CXCL8-mediated chemotaxis in CXCR2-overexpressing cells at a concentration as low as 1 μM [468]. CX4338 significantly reduced neutrophils in bronchoalveolar lavage in a LPS-induced mouse inflammation model [468].
CX797	CX797 inhibited IL8-induced chemotaxis in CXCR2-bla U2OS Tango cells [469].
CX25	CX25, which was identified using a ligand-based pharmacophore approach, inhibited CXCR2 in a Tango assay (IC50 = 0.36 μM). It also inhibited CXCR4 with an IC50 of 0.59 µM [20].
CX4152	CX4152 inhibited CXCR2 (IC50 = 7.6 μM) and exhibited selectivity over CXCR4 (IC50=64.7 µM) in the Tango assay [20]. CX4152 induced receptor internalization in a dose- and time-dependent manner. It also down-regulated expression of total CXCR2 at 5 h treatment [20]. CX4152 did not induce rapid calcium mobilization in CXCR2-overexpressing cells (293T-CXCR2-GFP), whereas CX25 induced peak calcium flux within 1 min stimulation [20]. CX4152 inhibited CXCL8-induced chemotaxis in a concentration-dependent manner with an IC50 of 51 μM [20]. CX4152, at a dose of 50 mg/kg, significantly inhibited polymorphonuclear leukocyte migration (about 2-fold decrease) in a murine model of neutrophilic airway inflammation induced with LPS [20].
CX815	CX815 inhibited CXCR2 (IC50 = 0.4 µM) and exhibited selectivity over CXCR4 (IC50 > 50 µM) [20].

Several small-molecule antagonists have advanced to clinical trials for various inflammatory-mediated diseases including asthma, COPD, cystic fibrosis, and cancer (Table 5). In general, most of these studies show that CXCR1/2 inhibition is safe and well tolerated with few adverse events. The first series of small-molecule CXCR2 antagonists were phenol-containing diarylureas developed by GlaxoSmithKline ³⁸¹. Further optimization to increase potency and reduce clearance resulted in the addition of a sulfonamide substituent adjacent to the phenol group ³⁸². SB656933, a representative of this class of CXCR2 antagonists, has been tested in patients with cystic fibrosis and COPD ³⁸³. SB656933 is CXCR2-selective with an IC₅₀ value of 5 nM for CXCL8 inhibition (CXCR1 IC₅₀ >1µM) ³⁸⁴. Previously disclosed clinical trial results showed that SB656933 reduced ozone-induced sputum neutrophils by 74% when pretreated with a single dose at 150 mg in healthy patients ³⁸⁵. SB656933 also reduced sputum neutrophils by 30% compared to baseline in CF patients treated with 50 mg of SB656933 for 28 days ³⁸⁶. However, no change in lung function was observed, suggesting that a longer treatment duration may be required or CXCR2 inhibition alone is not sufficient to enhance lung function in CF patients. Another CXCR2 inhibitor from the same class, SB225002, was reported to improve antitumor and antiangiogenic response in preclinical models of ovarian cancer in combination with VEGFR inhibitor sorafenib ³¹¹.

Table 5.

CXCR1/2 inhibitors tested in clinical studies ⁴⁷⁰

Inhibitors	Highest Phase	Structures	Indications	Sponsors
Repertaxin	Phase 3		Metastatic breast cancer (Phase 2), pancreatic islet transplantation in type 1 diabetes (Phase 3); Kidney transplantation (Phase 2); Lung transplantation (Phase 2)	Dompé Farmaceutici S.p.A
Navarixin (SCH 527123, MK-7123)	Phase 2		COPD, Asthma, Psoriasis	Merck Sharp & Dohme Corp.
Danirixin (GSK1325756)	Phase 2		COPD	GlaxoSmithKline
AZD5069	Phase 2		Asthma (Phase 2); COPD (Phase 2); bronchiectasis (Phase 2); Head and Neck cancer (Phase 1b/2)	AstraZeneca
DF2156A	Phase 2		Bullous Pemphigoid	Dompé Farmaceutici S.p.A
AZD8309	Phase 1		Rheumatoid arthritis; COPD	AstraZeneca
SB656933	Phase 1		COPD	GlaxoSmithKline

Isosteric replacements of the urea from early CXCR2 antagonists led to phenol-containing N, N'-diarylsquaramides. In these compounds, the urea is replaced with 3,4-diamino-1,2-dioxocyclobutene (squaramide) ³⁸⁷. A compound from this class, SCH527123, was tested in a Phase II clinical trial for COPD. SCH527123 is a highly potent, non-competitive allosteric CXCR2 antagonist (CXCR2 K_d=49 pM; CXCR1 K_d=3.9 nM) ^{388, 389}. SCH527123 reduced ozone-induced sputum neutrophils in healthy patients ²⁴³ and reduced sputum neutrophils (36% reduction) in asthmatic patients treated with a daily dose of 150 mg of SCH527123 for 4 weeks. However, it showed no changes in lung function ³⁹⁰. In another clinical trial SCH527123 significantly reduced CXCL-8-induced migration of neutrophils in the peripheral blood and sputum of mild atopic asthma patients, but had no effect on the migration of bone marrow neutrophils ³⁹¹. In preclinical studies, SCH527123 exhibited anticancer effects in colon and melanoma cancer mouse models by inhibiting tumor growth and microvessel density ^{247, 317}.

Reparixin is a non-competitive CXCR1 and CXCR2 dual inhibitor designed using molecular modeling studies with CXCR1, and it is structurally different from the earlier classes of antagonists. Mutation analysis and molecular modeling showed that reparixin binds to a pocket in the transmembrane allosteric region of CXCR1 and inhibits CXCL8 induced receptor signaling in intracellular compartments without altering CXCL8 binding affinity ^{392, 393}. Reparixin potently inhibited CXCL8-induced human PMN migration (IC₅₀ = 1 nM) in in vitro studies ³⁹⁴. In reperfusion injury/ischemia rat models, reparixin successfully prevented neutrophil influx and significantly reduced organ/tissue damage ³⁹⁵. In a breast cancer preclinical model, reparixin in combination with paclitaxel inhibited tumor metastasis and reduced CSC population ²⁹³. Phase I clinical trials showed that reparixin in combination with paclitaxel is safe and well-tolerated in patients ^{396, 397}. A phase II study of the same combination has been initiated ³⁹⁷. CXCR1/2 inhibition may lead to some adverse effects such as neutropenia, susceptibility to opportunistic infection as a consequence of impaired neutrophil recruitment and decreased bacterial clearance ¹⁰⁷.

Antibody Therapy

Blockade of the CXCL8-CXCR1/2 axis using neutralizing antibodies is safe and efficacious in different diseases. A brief summary of previous studies with select antibodies is discussed below.

Therapeutic antibodies targeting CXCL8

Humanized monoclonal antibody against CXCL8 effectively reduces severity of acute lung injury in rabbits by preventing neutrophil infiltration in the lung ⁴⁷¹. Gemcitabine induces CXCL8 expression and promotes tumor neovascularization. Anti-CXCL8 antibody treatment attenuates tumor formation as well as intra-tumoral vascularity in nude mice transplanted with Mia PaCa-2 cells and treated with gemcitabine ²⁵¹.

ABX-IL8, a fully humanized antibody against CXCL8 developed by Abgenix (Fremont, CA) was tested in Phase II clinical trials for psoriasis, rheumatoid arthritis (RA), and COPD. Unfortunately, the clinical trials against psoriasis and RA were discontinued due to lack of efficacy ²¹⁶. The clinical trial against COPD yielded some positive results as it significantly improved shortness of breath compared to placebo in patients with COPD ^{216, 472, 473}. However, there has been no report on the advancement of this agent for almost a decade. ABX-IL8 showed some promising results against melanoma in preclinical studies. Tumor growth in mice bearing human melanoma cells, A375SM (high CXCL8 producer) and TXM-13 (intermediate CXCL8 producer), was significantly reduced in the ABX-IL8 treated group when compared to control IgG-treated animals ⁴⁷⁴. ABX-IL8 downregulates expression of matrix metalloproteinase-2 (MMP2), inhibits angiogenesis and increases apoptosis in melanoma cells in vivo ⁴⁷⁴.

ABCream, which is a topical formulation of CXCL8 monoclonal antibody developed by Anogen (Canada), is marketed in China for the treatment of psoriasis. A 60% improvement of psoriasis in approximately 50% of patients after six-week treatment with ABCream was observed ²¹⁶. Another CXCL8 neutralizing antibody, HuMab 10F8, antagonizes CXCL8-dependent neutrophil activation and migration and significantly improved disease conditions in patients suffering with palmoplantar pustulosis, a chronic inflammatory skin disease. It caused more than 50% reduction in the formation of fresh pustules and a dose-dependent decrease in CXCL8 levels ⁴⁷⁵. Anti-CXCL8-neutralizing antibody abolishes TAM-dependent invasion of papillary thyroid carcinoma cells in a dose dependent manner ³⁶⁷. Although several preclinical and clinical studies were performed using CXCL8 blocking antibody, the success of these studies may be limited to inflammatory skin diseases. Thus far, no significant clinical success of CXCL8 antibodies against cancer treatment has been achieved. Proper selections of cancer types and stages may be important for the success of CXCL8 antibody therapy.

Therapeutic antibodies targeting CXCR1/2

Treatment with a CXCR1 antibody significantly reduced CXCL8-induced cell proliferation in small-cell lung cancer (SCLC) cells ²⁷². CXCR2 antibody significantly prevented systemic neovascularization in a mouse lung following left pulmonary artery ligation ⁴⁷⁶. CXCR2 neutralizing antibodies attenuated premalignant alveolar lesions in a Kras^LA1 mouse model by inducing apoptosis in the vascular endothelial cells within the lesions ²⁷⁴. CXCR2 neutralizing antibodies also inhibited proliferation of LKR-13 lung adenocarcinoma cells that are derived from Kras^LA1 mice and established as syngeneic tumors in wild-type mice ²⁷⁴. Treatment with a CXCR2 neutralizing antibody significantly inhibited ELR(+)-CXC-chemokine-induced proliferation, invasion, and tube formation of human umbilical vein endothelial cells (HUVEC) in vitro and reduced tumor volume, Ki-67 proliferation index, and microvessel density in a pancreatic cancer mouse model ³¹⁹. A CXCR2 monoclonal antibody discovered using a combination of in vitro techniques specifically inhibited CXCL8 and Gro-α-induced ß-arrestin recruitment in CXCR2 overexpressing cells with IC₅₀ values of 0.3 and 0.2 nM, respectively ⁴⁷⁷. In spite of several successful preclinical studies using CXCR1/2 neutralizing antibodies in the treatment of cancer and COPD, to date none of the antibodies have been advanced to clinical trials for these diseases.

Targeting CXCL8-CXCR1/2 axis with miRNA

microRNAs (miRNAs) are small non-coding ~22nt RNAs that regulate gene expression by guiding Argonaute (AGO) proteins to RNA targets to cause mRNA degradation and/or translational repression ⁴⁷⁸. Like other signaling pathways, the CXCL8-CXCR1/2 axis also can be modulated by miRNA therapy. miR-708 inhibited mRNA expression as well as release of CXCL8 along with other chemokines and asthma related genes in human airway smooth muscle cells. Another miRNA, miR-140-3p, also inhibited CXCL8 mRNA expression, but not the CXCL8 release ⁴⁷⁹. There are several miRNAs that are aberrantly expressed in our body, indirectly regulating CXCL8 release. miR-155, which is highly expressed in CF lung, indirectly induces IL-8 production by regulating SHIP1. On the other hand, overexpression of miR-93 and miR-17 inhibited CXCL8 production and release in CF bronchial epithelial ⁴⁸⁰. Thus, modulating miRNAs by directly or indirectly regulating CXCL8 expression could be an attractive therapeutic strategy for the diseases where CXCL8 plays an important role.

miR141 has been reported to inhibit tumor growth and metastasis by targeting CXCL1-CXCR2 axis and recruitment of regulatory T-cells. Decreased expression of miR141, which resulted in increased level of CXCL1 and recruited regulatory T-cells in the tumor microenvironment to promote immune escape of cancer cells, is associated with reduced median survival time of patients with NSLC and malignant pleural effusion (MPE) ⁴⁸¹. The miR141-CXCL1-CXCR2 signaling axis in MPE may act as an important immune checkpoint for cancer cells to evade immune attacks and thus is associated with shorter survival of patients with NSCLC. Therapeutic intervention in this signaling axis may be an attractive strategy to treat the patients with NSCLC and MPE.

Inhibition of CXCL8 Expression

Inhibition of CXCL8 expression can be achieved in several ways, such as by targeting signal transduction pathways that trans-activate CXCL8 expression, or by targeting transcription factors affecting CXCL8 gene expression.

Several kinases are involved in the signal transduction pathways that regulate CXCL8. Inhibitors of these kinases indirectly block CXCL8 by down-regulating CXCL8 expression (Table 6). For example, several p38 MAPK inhibitors have been reported to down-regulate CXCL8 expression. MEK1-specific inhibitors PD98059 and U0126 blocked adenovirus serotype 7 (Ad7)-induced release of CXCL8 through inhibition of ERK pathways in a dose-dependent manner ⁴⁸². p38 MAPK inhibitor SB203580 inhibited TNFα-induced CXCL8 production ⁴⁸³. SB203580 significantly reduced IL-1β-induced CXCL8 mRNA expression and protein secretion in Caco-2 and HT29 cells. Pretreatment of Caco-2 cells with SB203580 (10 μM) down-regulated IL-1β-induced CXCL8 expression by 62% and reduced CXCL8 secretion by 80% ⁴⁸⁴. Another p38 MAPK inhibitor, SB202190, inhibited TNFα and LPS-induced CXCL8 mRNA expression in monocytes. SB202190 blocked LPS response in monocytes with an EC₅₀ of 100 nM ⁴⁸⁵. PD98059, SB203580 and SB202190 blocked adenosine A_2B receptor-mediated CXCL8 production in human mast cells (HMC-1) with IC₅₀ values of 3 μM, 0.3 μM and 0.3 μM, respectively ⁴⁸⁶. Treatment of human rhinovirus (HRV)-infected human bronchial epithelial (BEAS-2B) cells with SB203580 and SB239063 (both at 3 μM) resulted in 49% and 85% inhibition of CXCL8 secretion, respectively ⁴⁸⁷. RWJ 67657, another p38 MAPK inhibitor inhibited transcription and secretion of CXCL8 in monocyte-derived macrophages in a dose dependent manner with an IC₅₀ value of 0.3 μM in healthy controls and 1.2 μM in rheumatoid arthritis patients ⁴⁸⁸.

Table 6.

Inhibitors of CXCL8 expression

Inhibitors	Function
MEK1 inhibitors
PD98059 U0126	PD98059 and U0126 attenuated Ad7-induced induction of CXCL8 expression [482].
p38 MAPK inhibitor
SB203580 SB202190 SB239063 RWJ67657	SB203580 (10 μM) reduced IL-1β-induced CXCL8 mRNA expression by 62% and CXCL8 release by 80% [484]. SB202190 inhibited TNFα or LPS-induced CXCL8 mRNA expression in monocytes [485]. SB203580 and SB202190 inhibited A2B receptor-mediated CXCL8 secretion in HMC-1 with same IC50 value of 0.3 μM [486]. SB203580 and SB239063 at 3 μM reduced CXCL8 secretion in HRV-infected BEAS-2B cells by 49% and 85%, respectively [487]. RWJ67657 inhibited CXCL8 production with IC50s of 0.3 μM and 1.2 μM in monocyte-derived macrophages from healthy human and rheumatoid arthritis patients, respectively [488].
JNK inhibitor
SP600125	The JNK inhibitor SP600125 (10 μM) significantly downregulated CXCL8 expression in glioblastoma U87Δ cells [494].
PI3K Inhibitors
Wortmanin LY294002	LY294002 and Wortmannin at 15 μM significantly inhibited 90% and 72% of TNFα-induced CXCL8 production, respectively [489].
GDC-0941 BEZ-235 SHBM1009 (Structure not disclosed)	GDC-0941, BEZ-235 and SHBM1009 inhibited CXCL8 production induced by betacellulin in A549 cells at 1 μM, 10 μM and 10 μM, respectively [491].
EGFR Inhibitor
Erlotinib	Erlotinib significantly downregulated CXCL8 expression in cells as well as in cancer patients [491] [492].
NF-KB inhibitors
BAY11-708 (IKK inhibitor) Dexamethasone (glucocorticoid receptor agonist)	BAY11-708 inhibited IL-1β-mediated CXCL8 mRNA expression and protein secretion in a dose-dependent manner [484].
Proteasome inhibitors
Bortezomib MG132	Proteasome inhibitors bortezomib and MG132 inhibited CXCL8 expression in cancer cells. Bortezomib inhibited CXCL8 expression in bladder and prostate cancer cell lines [472]. MG132 reduced TNFα-induced CXCL8 production in human lung-carcinoma cells [495].
Others
N-Pyrazolyl-N'-alkyl/benzyl/phenylureas	N-Pyrazolyl-N'-alkyl/benzyl/phenylureas inhibited human CXCL8-induced neutrophil migration with IC50 values as low as 10 nM [497]. These compounds did not bind CXCR1/2, but inhibited phosphorylation of protein tyrosine kinases in the 50-70 kDa region [497].

PI3K inhibitors, LY294002 and Wortmannin, significantly inhibited TNFα-induced CXCL8 production in liver cancer cells ^{489, 490}. The percent inhibitory effects of LY294002 and Wortmannin at 15 μM were 90% and 72%, respectively ⁴⁸⁹. Other PI3K inhibitors, BEZ235 (1 μM), GDC0941 (10 μM) and SHBM1009 (10 μM), significantly reduced betacellulin-induced CXCL8 production in human lung cancer A549 cells by inhibiting the activation of the ERK signaling pathway ⁴⁹¹.

Treatment with EGFR inhibitor erlotinib (0.1 - 10 μM) significantly inhibited betacellulin-induced over-production of CXCL8 in a dose-dependent manner ⁴⁹¹. Erlotinib down-regulated the CXCL8 expression in cancer patients and the reduced levels of serum CXCL8 is associated with stronger EGFR inhibition and prolonged overall survival ⁴⁹².

JNK inhibitor SP600125 blocked CXCL8 expression and secretion by regulating NF-kB activation in human esophageal epithelial cells ⁴⁹³. The JNK inhibitor SP600125 (10 μM) significantly attenuated CXCL8 production in glioblastoma U87Δ cells by reducing AP-1 reporter activity ⁴⁹⁴.

Proteasome inhibitors regulate the degradation of IκB and hence inhibit NF-κB transcriptional activity. The proteasome inhibitor bortezomib inhibited proliferation and CXCL8 production in bladder and prostate cancer cell lines ⁴⁷². Another proteasome inhibitor MG132 blocked TNF α -induced NF-κB activation and CXCL8 secretion in human lung carcinoma A549 cells ⁴⁹⁵.

Several NF-κB pathway inhibitors (e.g., dexamethasone, BAY 11-7082) have been reported to inhibit CXCL8 expression ⁴⁷². BAY 11-7082, an irreversible inhibitor of IKK, inhibited IL-1β-mediated CXCL8 production in a dose-dependent manner by blocking phosphorylation of IκB and attenuating downstream NF-κB activation ⁴⁸⁴. Dexamethasone, a glucocorticoid receptor agonist, reversed IL-1β induced CXCL8 expression in human cancer cells ^{472, 496}.

Conclusions

The pathophysiological role of the CXCL8-CXCR1/2 axis in inflammatory diseases including COPD, asthma, cystic fibrosis, inflammatory bowel diseases, psoriasis, arthritis, Alzheimer's disease, and stroke has been well established. PMN recruitment by CXCL8 plays key role in host innate responses of inflammatory diseases. In response to external stimuli, lung macrophages secrete CXCL8, which coordinates PMN migration to the lungs, and causes airway inflammatory diseases. The CXCL8-CXCR1/2 axis has been implicated in progression of colorectal, breast, lung, melanoma, pancreatic, ovarian, and prostate cancers. More significantly, genes implicated in the CXCL8-CXCR1/2 signaling pathway play important role in the tumor microenvironment and cancer stem cells, where they actively participate in tumor progression and stem cell migration.

CXCR2 signaling also plays an important role in immune evasion of tumor cells by trafficking MDSCs, an immune checkpoint into the tumor microenvironment, and causes resistance to several chemotherapeutic treatments. Combination of anti CXCR2 and other chemotherapy, particularly anti-PD1 therapy against pancreatic cancer proved more beneficial than either of single agents alone ^{325, 346}. CXCL8-CXCR1/2 axis also acts as oncosuppressing signaling in immunogenic cell death by various chemotherapeutic agents.

Targeting the receptors CXCR1/2, in particular CXCR2 may be the most effective strategy within the CXCL8-CXCR1/2 axis rather than blocking expression or secretion of CXCL8 as it may be compensated by other chemokine ligand which can still activate CXCR1/2 signaling. Currently, several inhibitors and antibodies targeting CXCL8-CXCR1/2 pathway are under various stages of clinical development for inflammatory diseases. In general, these agents tend to be well tolerated in human and are especially suited for use in combination with chemotherapy in select cancers.

As the CXCL8-CXCR1/2 axis has both tumor promoting and tumor suppressing properties, great care should be taken while developing either inhibitors or stimulators of CXCL8-CXCR1/2 axis. The applications of these modulators should be based on tumor types, grades, stages, immunogenic conditions as well as co-administered drugs. Blockade of CXCL8-CXCR1/2 axis may also have some adverse effects, such as neutropenia and increased susceptibility to opportunistic infections. Choosing optimal dosing for anti-CXCR2 therapy is most vital to reduce the severity of these adverse events.

Supplementary Material

Supplementary figures and tables.