Already upon the initial isolation of natural human CXCL8 it was clear that several different NH2-terminal isoforms of CXCL8 are produced by leukocytes or fibroblasts under inflammatory conditions. Posttranslational modifications can alter the biological activity, interaction with other molecules, clearance of chemokines and hence the in vivo inflammatory properties. The most abundant CXCL8 isoforms CXCL8 and CXCL8, produced by endothelial cells, fibroblasts and monocytes respectively, have been studied extensively. These studies revealed that the removal of five NH2-terminal amino acids potentiates the neutrophil chemotactic activity of CXCL8 both in vitro and in vivo. Additional limited NH2-terminal truncations to CXCL8 and CXCL8 gradually increase the chemotactic potency in vitro, suggesting that CXCL8 is not the most active form of CXCL8. A crucial role in receptor binding and activation has been ascribed to the ELR motif in front of the CXC sequence using both chemically synthesized truncated and mutated analogs. Thus far, truncation of the NH2-terminus of CXCL8 has been shown to result in potentiation as long as proteases do not cleave in and beyond the ELR motif. In this study, the elongated natural variant of CXCL8, i.e. CXCL8 and the truncated forms CXCL8 and CXCL8 were characterized. The elongated form is presumed to arise from alternative cleavage of the signal peptide of the 99 amino acid precursor. It has been detected in medium of peripheral blood mononuclear cells conditioned with lipopolysaccharide, concanavalin A or a combination of polyriboinosinic polyribocytidylic acid and interferon-c, where it constitutes about 8 to 10% of the total amount of CXCL8 produced. In addition, the supernatant of cultured IL-1a- or TNF-a-stimulated dermal fibroblasts contains significant amounts of CXCL8. Albeit less abundant, the truncated forms CXCL8 and CXCL8 have also been isolated from the conditioned medium of PBMCs. These forms may result from aminopeptidase mediated cleavage of secreted CXCL8. The results presented here show that the NH2-terminally different isoforms described do not differ a lot in their effects on binding to and signaling through CXCR1 or CXCR2. The receptor binding and calcium signaling potency of CXCL8, CXCL8, CXCL8 and CXCL8 is similar in both CXCR1- or CXCR2-transfected cells. However, in calcium signaling and in vitro chemotaxis assays on freshly isolated blood neutrophils, expressing both CXCR1 and CXCR2, stronger responses to stimulation with CXCL8 than towards stimulation with CXCL8 were observed. Moreover, small alterations in the NH2-terminal region do influence the GAG binding affinity of CXCL8. CXCL8 and CXCL8 displayed a three-fold higher affinity for Axitinib 319460-85-0 heparin compared to CXCL8. Although not as explicit as the truncated isoforms, CXCL8 also bound heparin with higher affinity than CXCL8.