IL-5-stimulated eosinophils adherent to periostin undergo stereotypic morphological changes and ADAM8-dependent migration

BACKGROUND

IL-5 causes suspended eosinophils to polarize with filamentous (F)-actin and granules at one pole and the nucleus in a specialized uropod, the "nucleopod," which is capped with P-selectin glycoprotein ligand-1 (PSGL-1). IL-5 enhances eosinophil adhesion and migration on periostin, an extracellular matrix protein upregulated in asthma by type 2 immunity mediators.

OBJECTIVE

Determine how the polarized morphology evolves to foster migration of IL-5-stimulated eosinophils on a surface coated with periostin.

METHODS

Blood eosinophils adhering to adsorbed periostin were imaged at different time points by fluorescent microscopy, and migration of eosinophils on periostin was assayed.

RESULTS

After 10 minutes in the presence of IL-5, adherent eosinophils were polarized with PSGL-1 at the nucleopod tip and F-actin distributed diffusely at the opposite end. After 30-60 minutes, the nucleopod had dissipated such that PSGL-1 was localized in a crescent or ring away from the cell periphery, and F-actin was found in podosome-like structures. The periostin layer, detected with monoclonal antibody Stiny-1, shown here to recognize the FAS1 4 module, was cleared in wide areas around adherent eosinophils. Clearance was attenuated by metalloproteinase inhibitors or antibodies to disintegrin metalloproteinase 8 (ADAM8), a major eosinophil metalloproteinase previously implicated in asthma pathogenesis. ADAM8 was not found in podosome-like structures, which are associated with proteolytic activity in other cell types. Instead, immunoblotting demonstrated proteoforms of ADAM8 that lack the cytoplasmic tail in the supernatant. Anti-ADAM8 inhibited migration of IL-5-stimulated eosinophils on periostin.

CONCLUSIONS AND CLINICAL RELEVANCE

Migrating IL-5-activated eosinophils on periostin exhibit loss of nucleopodal features and appearance of prominent podosomes along with clearance of the Stiny-1 periostin epitope. Migration and epitope clearance are both attenuated by inhibitors of ADAM8. We propose, therefore, that eosinophils remodel and migrate on periostin-rich extracellular matrix in the asthmatic airway in an ADAM8-dependent manner, making ADAM8 a possible therapeutic target.

A mix of S and ΔS variants of STAT3 enable survival of activated B-cell-like diffuse large B-cell lymphoma cells in culture

Activated B-cell-like diffuse large B-cell lymphoma (ABC DLBCL) is characterized by increased expression and activator of signal transducer and activator of transcription 3 (STAT3). ABC DLBCL cells require STAT3 for growth in culture. In ABC DLBCL cells, eosinophils and perhaps all cells, four variant STAT3 mRNAs (Sα, ΔSα, Sβ and ΔSβ) are present as a result of two alternative splicing events, one that results in the inclusion of a 55-residue C-terminal transactivation domain (α) or a truncated C-terminal domain with 7 unique residues (β) and a second that includes (S) or excludes (ΔS) the codon for Ser-701 in the linker between the SH2 and C-terminal domains. A substantial literature indicates that both α and β variants are required for optimal STAT3 function, but nothing is known about functions of ΔS variants. We used a knockdown/re-expression strategy to explore whether survival of ABC DLBCL cells requires that the four variants be in an appropriate ratio. No single variant rescued survival as well as STAT3Sα-C, Sα with activating mutations (A661C and N663C) in the SH2 domain. Better rescue was achieved when all four variants were re-expressed or Sα and ΔSα or Sβ and ΔSβ were re-expressed in pairs. Rescue correlated with expression of STAT3-sensitive genes NFKBIA and NFKBIZ. We consider a variety of explanations why a mix of S and ΔS variants of STAT3 should enable survival of ABC DLBCL cells.

Disulfide connectivity of recombinant C-terminal region of human thrombospondin 2

The thrombospondin (TSP) family of extracellular glycoproteins consists of five members in vertebrates, TSP1 to -4 and TSP5/cartilage oligomeric matrix protein, and a single member in Drosophila. TSPs are modular multimeric proteins. The C-terminal end of a monomer consists of 3-6 EGF-like modules; seven tandem 23-, 36-, or 38-residue aspartate-rich, Ca(2+)-binding repeats; and an approximately 230-residue C-terminal sequence. The Ca(2+)-binding repeats and C-terminal sequence are spaced almost exactly the same in different TSPs and share many blocks of identical residues. We studied the C-terminal portion of human TSP2 from the third EGF-like module through the end of the protein (E3CaG2). E3CaG2, CaG2 lacking the EGF module, and Ca2 composed of only the Ca(2+)-binding repeats were expressed using recombinant baculoviruses and purified from conditioned media of insect cells. As previously described for intact TSP1, E3CaG2 bound Ca(2+) in a cooperative manner as assessed by equilibrium dialysis, and its circular dichroism spectrum was sensitive to the presence of Ca(2+). Mass spectrometry of the recombinant proteins digested with endoproteinase Asp-N revealed that disulfide pairing of the 18 cysteines in the Ca(2+)-binding repeats and C-terminal sequence is sequential, i.e. a 1-2, 3-4, 5-6, etc., pattern.

Physical characterization of the procollagen module of human thrombospondin 1 expressed in insect cells

Thrombospondin 1 (TSP1) is a homotrimeric glycoprotein composed of 150-kDa subunits connected by disulfide bridges. The procollagen module of thrombospondin 1 has been implicated in antiangiogenic activity. Procollagen modules are found in a number of extracellular proteins and are identifiable by 10 cysteines with characteristic spacing. We expressed and studied the procollagen module (C) of human TSP1, both by itself and in the context of the adjoining oligomerization sequence (o) and N-terminal module (N). The coding sequences were introduced into baculoviruses along with an N-terminal signal sequence and C-terminal polyhistidine tag. Proteins were purified from conditioned medium of infected insect cells by nickel-chelate chromatography. NoC is a disulfide bonded trimer and cleaves readily at a site of preferential proteolysis to yield monomeric N and trimeric oC. These are known properties of full-length TSP1. Mass spectroscopy indicated that C is N-glycosylated, and all 10 cysteine residues of C are in disulfides. By equilibrium ultracentrifugation, C is a monomer in physiological salt solution. Circular dichroism, intrinsic fluorescence, and differential scanning calorimetry experiments suggest that the stability of C is determined by the disulfides. The two tryptophans of C are in a polar, exposed environment as assessed by iodide fluorescence quenching and solvent perturbation. The oC far UV circular dichroism spectrum could be modeled as the sum of C and a coiled-coil oligomerization domain. The results indicate that the recombinant C folds autonomously into its native structure, and trimerization of the modules in TSP1 does not perturb their structures.

Hereditary variation in platelet integrin alpha 2 beta 1 density is associated with two silent polymorphisms in the alpha 2 gene coding sequence

The integrin alpha 2 beta 1 is a receptor for collagen that plays a fundamental role in the adhesion of blood platelets to the extracellular matrix. We previously reported that platelet alpha 2 beta 1 levels among randomly selected individuals can vary up to 10-fold and that this correlates with differences in adhesiveness to type-I or type-III collagens. We have now found two linked, allelic polymorphisms within the coding sequence of the alpha 2 gene that correlate with receptor density, TTT/TTC at codon Phe224 and ACA/ACG at codon Thr246. By Southern blot hybridization of specific antisense DNA probes to segments of genomic DNA that encompass each coding region, we have determined the gene frequencies of each allele in a random donor population (n = 65) to be 0.585 (TTC...ACG) and 0.415 (TTT...ACA). There is a statistically significant correlation between the alleles TTT...ACA (codons 224...246) and high receptor density (n = 30; P < .002), whereas the complimentary alleles TTC...ACG are associated with low receptor density. Heterozygous individuals express intermediate levels of this receptor, and familial studies confirm that these allelic polymorphisms are inherited characteristics. These findings prove that the level of platelet alpha 2 beta 1 is an inherited trait. The molecular basis for receptor density remains to be determined, but our findings establish that these silent alleles within the coding sequence of the alpha 2 gene are linked to the genetic basis for variation in receptor density.

Molecular determinants of arg-gly-asp ligand specificity for beta3 integrins

The Arg-Tyr-Asp (RYD) and Arg-Gly-Asp (RGD) sequences within the third complementarity-determining region of the heavy chain (H3) of murine recombinant Fab molecules OPG2 and AP7, respectively, are responsible for their specific binding to the platelet integrin alphaIIbbeta3. In this study, we evaluated the influence of divalent cation composition and single amino acid substitutions at key positions within H3 on the selectivity of these Fab molecules for integrin alphaIIbbeta3 versus the vitronectin receptor alphaVbeta3. The parent Fab molecule OPG2 (H3 sequence, HPFYRYDGGN) binds selectively to alphaIIbbeta3 and not at all to any other RGD-cognitive integrin, particularly alphaVbeta3, under any divalent cation conditions. The binding of the AP7 Fab molecule (HPFYRGDGGN) to alphaIIbbeta3 is not affected by the relative composition of calcium, magnesium or manganese. However, AP7 binding to alphaVbeta3, either expressed by M21 cells or as the purified integrin, is supported by manganese and inhibited by calcium. If the flanking asparagine 108 residue within the AP7 H3 loop is replaced by alanine (HPFYRGDGGA), the resulting Fab molecule AP7.4 binds selectively to alphaVbeta3 in a cation-dependent manner, but does not bind at all to alphaIIbbeta3 under any conditions. AP7.4 binding to alphaVbeta3 is supported by manganese, completely inhibited by calcium, and largely unaffected by magnesium. This behavior mimics that of the adhesive protein, osteopontin, another ligand that binds preferentially to alphaVbeta3. Despite these differences in specificity for alphaIIbbeta3 and alphaVbeta3, AP7 and AP7.4 remain selective for the beta3 integrins and do not bind to cell lines that express the RGD-cognitive integrins alphaVbeta5 or alpha5beta1. These results confirm that subtle changes in the amino acid composition immediately flanking the RGD or RYD motifs can have a profound effect on beta3 integrin specificity, most likely because they influence the juxtaposition of the arginine and aspartate side chains within the extended RGD loop sequence.

Molecular requirements for assembly and function of a minimized human integrin alphaIIbbeta3

Integrin subunit compatibility within and between species plays a major role in heterodimer assembly and ligand specificity. As an example, human alphaIIb pairs only with human beta3 and does not assemble a heterodimer with beta3 from other species. We use interspecies subunit chimeras to identify molecular requirements for subunit compatibility and show that species-restricted heterodimer assembly depends on a unique hexapeptide VGSDNH in an extended loop of the hypothetical human beta3 MIDAS domain. This allows us to express alphaIIb(1-233) and beta3(111-318) as a soluble, mini-integrin that retains RGD-dependent ligand recognition. Thus, in the case of one integrin, alphaIIbbeta3, the molecular requirements for integrin subunit compatibility and ligand recognition are intimately related.

A molecular basis for affinity modulation of Fab ligand binding to integrin alphaIIb beta3

The Arg-Gly-Asp (RGD) sequence within the third complementarity-determining region (CDR3) of the heavy chain (H3) is responsible for the binding of the recombinant murine Fab molecules, AP7 and PAC1.1, to the platelet integrin alphaIIbbeta3. AP7 binding is minimally influenced by the conformational state of this receptor, whereas PAC1.1 binds preferentially to the activated state of the receptor induced by platelet agonists. To study the molecular basis for this functional difference, we replaced the AP7 H3 loop (HPFYRGDGGN) with all or segments of the analogous sequence from PAC1.1 (RSPSYYRGDGAGP). AP7 Fd (VH domain + Cgamma1 domain) segments containing these H3 loop sequences were expressed as active Fab molecules by coinfection of Spodoptera frugiperda cell lines with recombinant baculoviruses containing Fd and AP7 kappa chain cDNA. Replacement of the entire AP7 H3 loop with that from PAC1.1 generated the mutant AP7.3 Fab molecule, which bound selectively to either activated, gel-filtered platelets or to purified alphaIIbbeta3 in a manner identical to that of PAC1.1. Identical results were obtained when solely the sequences flanking the amino side of RGD within the respective H3 loops were exchanged. AP7.3 and PAC1.1 exhibited saturable but submaximal binding to activated gel-filtered platelets. Relative to AP7, the number of AP7.3 or PAC1. 1 Fab molecules bound per platelet was 17% in the presence of 1 m Ca2+ + 1 mM Mg2+ or 40% in the presence of 10 microM Mn2+. The ratio of Fab molecules bound after versus before activation (mean =/- S.D.; n = 3) was: for AP7.3, 9.8 =/- 0.6; for PAC1.1, 8.8 +/- 0.3; and for AP7, 1.4 =/- 0.2. In addition, AP7 bound to the stably expressed integrin mutant alphaIIbbeta3(S123A), whereas AP7.3 and PAC1 did not. Because AP7.3 behaves in every respect like PAC1.1, we conclude that the ability of RGD-based ligands to distinguish activated from resting conformations of the integrin alphaIIbbeta3 can be regulated by limited amino acid sequences immediately adjacent to the RGD tripeptide. Furthermore, those Fab molecules that exhibit increased selectivity for the activated conformation of alphaIIbbeta3 bind to a subpopulation of this integrin on platelets that is modulated by divalent cations.

The exchange of Arg-Gly-Asp (RGD) and Arg-Tyr-Asp (RYD) binding sequences in a recombinant murine Fab fragment specific for the integrin alpha IIb beta 3 does not alter integrin recognition

The murine monoclonal antibody OPG2 is an excellent paradigm of natural RGD ligands and binds specifically to alpha IIb beta 3 integrin. A reactive Arg103-Tyr104-Asp105 (RYD) tripeptide is located in an extended loop, the third complementarity-determining region of the heavy chain (H3). When compared to other RGD ligands, the RYD tripeptide of OPG2 is unique, in that the side chains are fixed in a stable orientation that we have defined by x-ray crystallography. In this study, we express OPG2 H chain segments (Fd) and kappa chains as components of active, Fab heterodimers by coinfection of Spodoptera frugiperda cell lines with recombinant baculoviruses containing cDNA specific for each protein. Recombinant AP7 Fd segments are generated from the parent OPG2 Fd segments by replacement of Tyr104 with Gly, while recombinant AP7E Fd segments are produced from AP7 Fd segments, by exchange of Asp105 with Glu. Neither the free Fd segments nor the free kappa chains of OPG2 or AP7 can bind to alpha IIb beta 3. The AP7 Fab fragment, like the parent OPG2 Fab, binds strongly to purified alpha IIb beta 3 but weakly, if at all, to purified alpha V beta 3. The affinity of OPG2 and AP7 Fab fragments for gel-filtered platelets, whether nonstimulated or activated by 0.2 microM phorbol 12-myristate 13-acetate, is identical. As with other natural RGD ligands, the binding of recombinant OPG2 Fab or AP7 Fab fragments to purified alpha IIb beta 3 or to gel-filtered platelets is completely inhibited by the peptide RGDW or by addition of EDTA, AP7E Fab fragments do not bind at all to either purified alpha IIb beta 3 or platelets. Our results demonstrate, for the first time within a natural protein ligand, that the tripeptides RGD and RYD exhibit equivalent binding capacity and specificity for the integrin alpha IIb beta 3.

Activation of calpain I and hydrolysis of calpain substrates (actin-binding protein, glycoprotein Ib, and talin) are not a function of thrombin-induced platelet aggregation

Calcium-activated neutral proteinase (calpain) has been shown to cleave proteins involved in the maintenance of cell structure. In human platelets, substrates of calpain include glycoprotein Ib (GPIb), actin-binding protein (ABP), and talin. GPIb-ABP complexes can be isolated in detergent extracts and are thought to represent membrane-cytoskeleton attachment sites. It has been hypothesized that the hydrolysis of GPIb-ABP by calpain is regulated by the extent of binding of this proteinase to the plasma membrane-cytoskeleton interface with platelet activation. Recently, another calpain substrate (talin) has been shown to redistribute from the cytoplasm to the plasma membrane-cytoskeleton interface as the result of thrombin stimulation. To investigate the intracellular distribution of calpain I, we employed the monoclonal antibody B27D8, specific for the heavy chain (catalytic subunit) of calpain I. Indirect immunofluorescent staining of resting human platelets revealed undetectable surface antigen. Permeabilization with Triton X-100, however, revealed a diffuse intracellular antigen consistent with a cytosolic distribution. To determine whether this antigen distribution reflected the proenzyme or the activated form of calpain I and to assess the degree of hydrolysis of ABP, GPIb, and talin, we employed B27D8 and murine monoclonal antibodies against ABP (1B3 and 3D1), GPIb (LJIb10), and rabbit polyclonal antibodies against talin (A2 and B11) in a quantitative immunotransblot assay. Examination of resting platelets revealed that calpain I existed as the 85-kd proenzyme form and that ABP, GPIb, and talin existed in their native intact forms. When platelets were aggregated with thrombin, autoproteolysis of calpain I occurred within the 30 seconds required to completely solubilize platelet aggregates in sodium dodecyl sulfate-containing buffer and not as a direct result of thrombin-induced activation.(ABSTRACT TRUNCATED AT 250 WORDS)

Nucleotide sequence of the human autoantibody 2E7 specific for the platelet integrin IIb heavy chain

The platelet glycoprotein (GP) IIb-IIIa complex figures prominently as an immunogen in autoimmune (idiopathic) thrombocytopenic purpura (ITP). 2E7 is a human monoclonal IgM autoantibody, derived from splenocytes of a patient with ITP, that recognizes a specific octapeptide amino acid sequence, Phe-Asp-Gly-Tyr-Trp-Gly-Tyr-Ser, on the heavy chain of GPIIb. This represents the first precise identification of an epitope on GPIIb-IIIa recognized by a human antibody. In this study, we have isolated total mRNA from 2E7, synthesized the corresponding cDNA using reverse transcriptase, and amplified the immunoglobulin mu and kappa chain cDNA by the Taq 1 polymerase chain reaction (PCR) using specific primers. The 2E7 mu chain variable region is encoded by a VH3 gene segment that is 98% homologous to the germline gene VH1.9III, a D-gene that is not homologous to any of the germline D-genes reported to date, and a JH6 gene segment that is essentially germline. The heavy-chain sequence, save for the unique D-gene, is similar to that of a number of human autoantibodies. The 2E7 kappa variable region is encoded by a Vk1 gene segment linked to a Jk1 gene segment. The Vk1 sequence of 2E7, with the exception of one nucleotide, is identical to that of autoantibody HF2-1/17, a prototype of SLE-associated anti-DNA autoantibodies bearing the 16/6 idiotype. The single base substitution results in a relatively conservative exchange of Asp for Glu at position 70 of the protein sequence. Despite this near identity in sequence, 2E7 does not bind to either single-stranded or double-stranded DNA. From these results, we conclude that specificity of 2E7 is likely to reside in either or both the D-JH region (CDR3) of the mu chain and the Jk region (CDR3) of the kappa chain. In addition to the identification of a novel D-gene, we also provide evidence that the 2E7 VHIII gene is probably a prototype of a VHIII subfamily, that the germline Vk1 gene shared by 2E7 and autoantibodies of the 16/6 idiotype probably represents a separate Vk family, and that this Vk gene cannot itself attribute specificity for DNA.