Structural determinants of antiproliferative activity of heparin on pulmonary artery smooth muscle cells.

In addition to its anticoagulant properties, heparin (HP), a complex polysaccharide covalently linked to a protein core, inhibits proliferation of several cell types including pulmonary artery smooth muscle cells (PASMCs). Commercial lots of HP exhibit varying degrees of antiproliferative activity on PASMCs that may due to structural differences in the lots. Fractionation of a potent antiproliferative HP preparation into high and low molecular weight components does not alter the antiproliferative effect on PASMCs, suggesting that the size of HP is not the major determinant of this biological activity. The protein core of HP obtained by cleaving the carbohydrate-protein linkage has no growth inhibition on PASMCs, demonstrating that the antiproliferative activity resides in the glycosaminoglycan component. Basic sugar residues of glucosamine can be replaced with another basic sugar, i.e., galactosamine, without affecting growth inhibition of PASMCs. N-sulfonate groups on these sugar residues of HP are not essential for growth inhibition. However, O-sulfonate groups on both sugar residues are essential for the antiproliferative activity on PASMCs. In whole HP, in contrast to an earlier finding based on a synthetic pentasaccharide of HP, 3-O-sulfonation is not critical for the antiproliferative activity against PASMCs. The amounts and distribution of sulfonate groups on both sugar residues of the glycosaminoglycan chain are the major determinant of antiproliferative activity.

Effect of fully sulfated glycosaminoglycans on pulmonary artery smooth muscle cell proliferation.

Fully sulfated heparin and other glycosaminoglycans, namely heparan, chondroitin, and dermatan sulfates, and hyaluronan have been prepared by using sulfur trioxide under mild chemical conditions. All these derivatives were assayed for antiproliferative activity on cultured bovine pulmonary artery smooth muscle cells (BPASMCs). No appreciable difference was found between heparin and fully sulfated heparin. Chondroitin and dermatan sulfates actually stimulated BPASMCs growth but full sulfonation made them strongly antiproliferative. Native hyaluronan was not antiproliferative but became strongly so after sulfonation. Neither acharan sulfate nor N-sulfoacharan sulfate had any antiproliferative activity. This suggests that O-sulfonation of the polysaccharide is critical for antiproliferative activity, whereas N-sulfonation of glucosamine residues is not.

Influence of molecular weight, protein core and charge of native heparin fractions on pulmonary artery smooth muscle cell proliferation.

Heparin macromolecules have been shown to inhibit cultured pulmonary artery smooth muscle cell proliferation in vitro and prevent hypoxic vascular remodeling in vivo. In an attempt to understand the structural determinants of heparin's antiproliferative properties, we have fractionated an antiproliferative preparation of commercial heparin into low and high molecular weight fractions. Then the high molecular weight heparin fraction was further fractionated on a DEAE-cellulose column by charge density eluting with 0 - 1 M NaCl linear gradient. The heparin protein peptides were both removed and isolated. These heparin fractions were assayed for antiproliferative effects on cultured bovine pulmonary artery smooth muscle cells. No appreciable differences were found among high and low molecular weight heparin fractions The core peptides showed no antiproliferative activity. However, higher charge density fraction was less antiproliferative.

Antiproliferative role of 3-O-sulfate glucosamine in heparin on cultured pulmonary artery smooth muscle cells.

Heparin macromolecules inhibit vascular remodeling associated with hypoxic pulmonary hypertension. Heparin's antiproliferative effect on smooth muscle cells, based on studies of synthetic pentasaccharide fragments, has been attributed to 3-O-sulfate on the internal glucosamine. To determine the role of 3-O-sulfation in smooth muscle cell growth, we treated three heparins of varying potency with heparitinases I and II, which degrade heparin fragments containing 3-O-sulfate on the glucosamine residue to delta-tetrasaccharides only. Our most antiproliferative heparin gave the least amount of delta-tetrasaccharides. This heparin was then fractionated according to degree of sulfation using ETOH precipitation. Again we found no antiproliferative difference between the highly sulfated fractions and those with a lesser degree of sulfation. These studies suggest that 3-O-sulfate of glucosamine residue is not critical in whole heparins for antiproliferative activity.

Inseparable iduronic acid-containing proteoglycan PG(IdoA) preparations of human skin and post-burn scar tissues: evidence for elevated levels of PG(IdoA)-I in hypertrophic scar by N-terminal sequencing.

Hypertrophic scarring is characterized by disordered collagen fibrils. In order to determine whether this is, in part, a result of changes in the population of proteoglycans that are thought to be involved in regulation of collagen fibril formation, we have compared PGs from post-burn normal and hypertrophic scar tissue, as well as from human dermis and epidermis. Efforts to separate the two major iduronic acid-containing proteoglycans, decorin [PG(IdoA)-II] and biglycan [PG(IdoA)-I], for quantitation were not successful. The different N-terminal sequences of these two iduronic acid-containing proteoglycans PG(IdoA-I and -II were utilized to estimate the relative amounts in the above PG(IdoA) preparations. Normal scar, dermis and epidermis were all found to contain primarily decorin with low (< 10%) levels of biglycan relative to decorin. In contrast, iduronic acid-containing proteoglycans from hypertrophic scar were found to be approximately 30% biglycan [PG(IdoA)-I]. This may be a proximal cause of altered collagen fibrils, or may result in alterations in the sequestration of growth factors, which then results in changes in collagen that effect the appearance of the scar. 1966 Elsevier Science Ltd.

Differential effect of three commercial heparins on Na+/H+ exchange and growth of PASMC.

Heparin preparations vary in chemical content and in antiproliferative activity for pulmonary artery smooth muscle cells (PASMC). Intracellular alkalinization via stimulation of the Na+/H+ antiporter appears to be a permissive event for proliferation of PASMC. We wondered whether the variable effect of heparin preparations on PASMC growth might be due to different degrees of inhibition of the Na+/H+ antiporter and whether variations in chemical formulation might correlate with the inhibition. Fluorescent microscopy of bovine PASMC was done using a dye with which fluorescence varies directly with intracellular pH (pHi). Bovine PASMC were preincubated with three heparin preparations previously shown to vary in antiproliferative activity, at 1.0 microgram/ml for 24 h. Platelet-derived growth factor (PDGF; 60 ng/ml) on PASMC without heparin resulted in a rise in pHi of 0.27 +/- 0.02 pH units. The rise in pH units in heparin-treated PASMC was 0.34 +/- 0.03 with Choay, 0.21 +/- 0.02 with Elkins-Sinn, and 0.07 +/- 0.02 with Upjohn (+/-SE; all P < 0.05; n = 5). Upjohn heparin incubation for as little as 15 min still impeded the rise in pH induced by PDGF. Heparin did not block the Na+/H+ exchanger directly, as it still restored pHi in response to an acid load. Compared with PASMC proliferation induced by 60 ng/ml PDGF, 1 microgram/ml of Choay, Elkins-Sinn, and Upjohn heparin produced -4 +/- 7.4, 1.4 +/- 4.8, and 48 +/- 2.2% inhibition of PDGF control, respectively (P < 0.05 for Upjohn compared with PDGF and Choay). The heparins varied in protein content and amino acid composition. However, amino acid and glucosamine composition, total sulfation, and extent of 3-O-sulfation did not predict their activity. Thus inhibition of PDGF activation of the Na+/H+ antiporter by a given heparin preparation correlated well with its ability to inhibit PASMC proliferation.

The in vivo effect of hyaluronan associated protein-collagen complex on wound repair.

Fetal skin wounds heal without scarring, however the underlying mechanisms remain unknown. Immunohistochemical staining and biochemical studies indicate the deposition of a collagen repair matrix that is highly organized. We have previously described a unique hyaluronan associated protein-collagen complex (HA-PC) profile present during the period of scarless healing in the sheep fetus. In this study, we examined the biologic activity of this HA-PC in an in vivo model of adult rat wound repair. A total of 84 incisional and 84 excisional wounds were examined by histology, TGF-beta immunocytochemistry and computer planimetry (excisional wounds only). Planimetry of the excisional wounds demonstrated the mean wound area remaining at day 1 was 88.7% for the control and 63.6% for the treated (p<0.01). At day 2, mean wound area was 81.5% for the control and 63.6% for the treated (p<0.01). At day 4, mean wound area was 56.6% for the control and 41.9% for the treated (p<0.01). At day 7, mean wound area was 26.9% for the control and 16.8% for the treated (p<0.01). At day 14, mean wound area was 7.9% for the control and 3.4% for the treated (p<0.05). Collagen organization was judged to be greater in the treated compared to control wounds, with a mean organization score of 2.3 vs. 1.9 (p=0.0596; Wilcoxon Signed Rank Sum Test). There were significantly more neutrophils at the wound margin of the treated compared to control wounds, 4.0 vs. 2.7 (p=0.038; Paired Two Tailed Student's t-Test). There was no difference in the number of microphages at the wound margin of the treated compared to control wounds, 6.15 vs. 6.0 (p>0.05). TGFbeta1 and beta2 staining was decreased whereas TGFbeta3 staining was increased in the HA-PC treated wounds. These results suggest that compared to control wounds HA-PC treated wounds heal more quickly, with more organized collagen, more neutrophils at the wound margin and increased TGFbeta3 expression. Furthermore, these data suggest that the manipulation of scarring in adult wounds is possible by the addition of proteins present in fetal skin.

Iduronic acid-rich proteoglycans (PGIdoA) and human post-burn scar maturation: isolation and characterization.

Proteoglycans (PGs) were extracted from human hypertrophic and normal scar tissues from two different stages of maturation after burn injury, under dissociative conditions (4 M guanidinium chloride containing proteinase inhibitors). The extracts were fractionated by ion-exchange chromatography, followed by ethanol precipitation, to give PG-containing iduronic acid (PGIdoA). The size of the PGIdoA decreased with the maturation of scars. Glycosaminoglycan (GAG) chains from PGIdoA were released by alkaline borohydride treatment, and their M(r) values were evaluated by polyacrylamide gel electrophoresis. The M(r) values for PGIdoA protein cores of the hypertrophic scars (5+ years and 2-5 years) and normal scar (5+ years and 2-5 years) were 22.6, 25, 19 and 21 kDa, respectively. The iduronic acid content of PGIdoA from both types of scar increased in their maturation phase. The M(r) values of PGIdoA decreased with maturation. PGIdoA carried the sulfate group mainly attached at C-4 of the 2-amino-2-deoxy-D-galactose residue. The NH2-terminal amino acid sequences of all the PGIdoA were similar to those of normal human skin or bone PG II (decorin) (i.e., Asp-Glu-Ala-B-Gly-Ile-Gly-Pro-Glu-Val-Pro-Asp-Asp-Arg).

Comparison of the effects of interleukin-1 beta on proteoglycan synthesis by human skin and post-burn normal scar explant cultures.

The effect of interleukin (IL)-1 beta on proteoglycan (PG) synthesis and secretion, into culture medium by normal human skin and post-burn human normal scar using tissue explants in culture, was investigated. Following exposure of different tissues to labeling with Na2[35SO4] in the presence and absence of IL-1 beta, the extractable [35SO4]PG (isolated from 0.15 M NaCl and 4 M Gdm. Cl extracts), non-extractable [35SO4]PG (isolated after papain treatment of residual tissue), and [35SO4]PG secreted into culture medium were analyzed for contents and distribution. The contents of [35SO4]PG as measured by [35SO4] incorporation indicate differences in [35SO4]PG production of extractable and non-extractable PGs and also in the PGs released into the culture medium. Examination of the sizes of [35SO4]PGs on Sepharose CL-6 beta columns with and without treatment of IL-1 beta shows that the size of non-extractable [35SO4]PG decreases after IL-1 beta treatment. Cellulose acetate plate electrophoresis of these [35SO4]PG fractions shows that the distribution of PGs alters after treatment with IL-1 beta. These results indicate that burn wound healing abnormalities (scarring) is related to a change in the level of PGs, and may be modified by IL-1 beta treatment.

Serial quantitation of hyaluronan and sulfated glycosaminoglycans in fetal sheep skin.

Fetal wounds are abundant in hyaluronic acid (HA), but little is known as to the total HA content of fetal tissues as a function of gestational age. Previous studies demonstrated scarless healing prior to approximately 130 days gestation, after which disorganized collagen deposition became prevalent. Glycosaminoglycans (GAGs) have been shown to play an important role in wound healing, cytodifferentiation, and morphogenesis. In this study, we examined the HA and GAG content of fetal sheep skin of increasing gestational age. We found that the total amount of HA and GAGs declined from a high at 80 days gestation (528 +/- 9 micrograms/gm) to a low at 130 days (174 +/- 11). Analysis of the various GAG species revealed that HA comprised the largest fraction (75-96%). The sulfated GAGs, Heparan Sulfate (HS) and Dermatan Sulfate (DS), were not present in the extracellular matrix until 120 days gestation. Both the trough of HA content and the appearance of sulfated GAGs in the extracellular fraction correspond to the appearance of scarring in fetal sheep wound repair.

Isolation and partial characterization of hyaluronan-protein-collagen complex (HA-PC) from fetal sheep skin of different gestational ages.

Hyaluronic acid (HA) has a positive effect on cell migration, differentiation and wound healing. Earlier work from our laboratory has shown the presence of biologically active proteins associated with HA. The protein associated with HA of fetal sheep skin varies in molecular weight depending on its gestational age. Specifically, the protein profile changes at 125 days of gestation, from a 60 KDa protein to a smaller protein of about 21 KDa. This time period coincides with the time that scarring becomes apparent in fetal sheep skin wounds. In this study, we have quantified changes in the proteins associated with HA with increasing gestational age, obtained amino acid profiles of these proteins with increasing gestational age, and proposed the existence of an HA-associated protein-collagen complex (HA-PC) which may serve as a scaffold for wound healing. Our results indicate that HA-PC content decreases from 42% of the dry weight at 75 days of gestation to 22% at 125 days of gestation. Protein content, in contrast, increases to 40% of the dry weight at 140 days of gestation. At the same time, collagen content increases from < 1% of the dry weight at 75 days to > 10% at 140 days. The increase in collagen content may account for the increase in total protein seen at 140 days. The expression of varying HA-PC's at different gestational ages may influence the kinetics of collagen fibrillogenesis and thus account for the previously noted late gestational change from "scarless" wound healing to "adult-like" wound healing in fetal sheep.

Hyaluronic acid of human skin and post-burn scar: heterogeneity in primary structure and molecular weight.

Hyaluronic acid (HA) was isolated from the dermis and epidermis of normal human skin and from normal and hypertrophic scar tissue, and the molecular properties of this polysaccharide were studied by circular dichroism (CD) and high performance liquid chromatography. The molecular weights of HA of normal skin and post-burn scar tissue range from 62,000 to 180,000. Hexosamine analysis showed no galactosamine contamination and 0.37 to 2.2 w/w% of protein in the HA sample. Uronic acid analysis suggests a heterogeneous distribution of glucuronic and iduronic acids. The CD profiles of these samples are similar, indicating no significant conformational variations among them. These data suggest that the variation in the molecular properties of HA between skin and scar tissue may be due to diversity of embryonic origin between epidermis HA and dermis HA, and to the diversity of the wound-healing process between normal scar HA and hypertrophic scar HA.

Proteoglycans in human burn hypertrophic scar from a patient with Ehlers-Danlos syndrome.

Proteoglycans (PGs) from human burn hypertrophic scar of a patient with Ehlers-Danlos syndrome were extracted with 4M guanidinium chloride and purified by DEAE-cellulose chromatography. Differential ethanol precipitation of the PG fraction obtained after ion-exchange chromatography yielded two low mol.-wt. PGs, on rich in glucuronic acid (PGGLCA; Mr 66 kDa) and the other rich in iduronic acid (PGIDOA; Mr 48 kDa). In PGGLCA, 84% of the glycosaminoglycan chains are composed of GlcA----GalNAc(SO4) units, whereas in PGIDOA, the chains contain 95% IdoA----GalNAc(SO4) disaccharide units. Upon treatment with testicular hyaluronidase, the PGs gave different-sized oligosaccharides. Chondroitinase ABC digestion of PGGLCA or PGIDOA gave a single protein core (Mr approximately 20 kDa). The presence of glucosamine and sialic acid in PGGLCA and PGIDOA suggests that both contain N-linked oligosaccharides.

Proteoglycan synthesis in human skin and burn scar explant cultures.

The synthesis of proteoglycans (PG) by normal human skin, and normal and hypertrophic scars were compared using tissue explants in culture. Newly synthesized PG were labelled with [35S]Na2SO4. Significant differences were found in the proportion of [35S]-radio-labelled incorporation of PG in the tissue and accumulation of [35S]PG in culture medium in the different tissues. The rate of PG biosynthesis in all three tissue types occurred in two phases. There was an initial phase of PG synthesis occurring at 0-3 h and a later phase that occurred at 3-18 h [35S]-labelled PG were isolated and characterized by Sepharose CL-6B chromatography and cellulose acetate electrophoresis. The results showed that the hypertrophic scar tissue and its culture medium contained higher proportions of dermatan sulphate (DS), chondroitin sulphate (CS) and DS' PG than the normal skin fractions. These results suggest that abnormal scarring is related to a change in the level of PG synthesis during the burn injury repair process.

Hyaluronan and wound healing: a new perspective.

Hyaluronan has long been associated with the remodelling extracellular matrix. Such remodelling occurs in development, growth and wound healing. This role has been thought to be related to the physical structure and chemical composition of the pure glycosaminoglycan chain. We question this proposition and present evidence which suggests that proteins associated with hyaluronan may be more critical determinants of tissue remodelling.

Purification and characterization of iduronic acid-rich and glucuronic acid-rich proteoglycans implicated in human post-burn keloid scar.

Small proteoglycans (PGs), extracted from human keloid scar tissue with 4M guanidinium chloride and fractionated by DEAE-cellulose chromatography, were separated by ethanol precipitation into one L-iduronic acid-rich and one D-glucuronic acid-rich fraction. The size of the L-iduronic acid-rich PG was 102 kDa with a 27 kDa glycosaminoglycan chain, that of the D-glucuronic acid-rich PG was 90 kDa with a 26 kDa glycosaminoglycan chain, and the protein core of both PGs was 14.5 kDa. The two PGs carried sulfate groups mostly attached at C-4 of the 2-amino-2-deoxy-D-galactose units. The N-terminal amino acid sequence of both was similar to human bone PGII (decorin), normal and hypertrophic scar, and human dermal tissue PG.