Differential modulation of vascular cell integrin and extracellular matrix expression in vitro by TGF-beta 1 correlates with reciprocal effects on cell migration

In vitro, BAEC and BASMC migratory phenotypes are known to be reciprocally modulated by both soluble factors and extracellular matrix proteins. In addition, integrin matrix receptors mediate endothelial and smooth muscle cell attachment and migration. To further elucidate these phenomena, we studied the effects of TGF-beta 1 on integrin expression by vascular BASMC and BAEC in tissue culture. TGF-beta 1 upregulated mRNA levels and surface pools of BASMC beta 3 integrin classes without modulating beta 1 integrin mRNA levels or expression of beta 1 integrin organization. In contrast to its effects on BASMC, TGF-beta 1 increased BAEC mRNA levels and surface expression of beta 1 and beta 3 integrins without altering their organization. Conversely, extracellular matrix components (fibronectin, laminin, and fibrinogen) organized cell surface integrins in both BASMC and BAEC without affecting the size of their cell surface pools. These data are consistent with the hypothesis that SMC and EC behavior in neointimal lesions may be modulated, in part, through a coordination of soluble factor and extracellular matrix protein regulation of integrin surface expression and organization.

Acidic fibroblast growth factor-Pseudomonas exotoxin chimeric protein elicits antiangiogenic effects on endothelial cells

It has recently been shown that chimeric toxins composed of acidic fibroblast growth factor fused to mutant forms of Pseudomonas exotoxin (aFGF-PE) are cytotoxic to a variety of tumor cell lines with FGF receptors. Although aFGF-PE might be considered as a possible chemotherapeutic toxin, limited knowledge is available concerning its effect on endothelia. This study investigates whether one of the aFGF-PE fusion proteins, aFGF-PE664GluKDEL, can function as an anti-angiogenic agent. Protein synthesis studies using rat epididymal fat pad microvascular endothelial cells (RFCs) indicated that after 24 h in culture, aFGF-PE had a significant inhibitory effect on protein synthesis at concentrations greater than 100 ng/ml. In cultures incubated with 1000 ng/ml aFGF-PE, RFC protein synthesis was inhibited as much as 83%. RFCs were also cultured in a 3-dimensional type I collagen gel and incubated with either transforming growth factor beta 1, aFGF-PE, or a combination of both. Transforming growth factor beta 1 elicits in vitro angiogenesis in these 3-dimensional cultures which consist of rapid formation of complex tubular structures. Transforming growth factor beta 1-treated RFCs incubated with aFGF-PE were unable to produce this angiogenic response, nor were they able to migrate out of the 3-dimensional culture to form a monolayer as shown by controls. Cell viability analyses showed that aFGF-PE produced a dose-dependent toxic effect which ranged from 10 to 90% cell death. Competition assays in which the chimeric toxin was preincubated with antibodies to aFGF resulted in an 89% reversal of the inhibitory effects of aFGF-PE on endothelial cells. Acidic FGF-PE with a mutation in the ADP ribosylation domain of PE was inactive in both 2-dimensional and 3-dimensional cultures. These data show that aFGF-PE has specific in vitro cytotoxic, antiangiogenic, and antimigratory effects on microvascular endothelia.

Independent modulation of enterocyte migration and proliferation by growth factors, matrix proteins, and pharmacologic agents in an in vitro model of mucosal healing

BACKGROUND

Gastrointestinal mucosa heals by restitution and proliferation. These are difficult to distinguish in vivo.

METHODS

Human Caco-2 enterocytes were cultured on matrix proteins (collagen I, laminin, fibronectin) with growth factors (epidermal growth factor [EGF] and transforming growth factor-beta 1 [TGF-beta 1]) and the tyrosine kinase and prostaglandin inhibitors genistein and indomethacin. Healing was modeled by means of monolayer expansion, proliferation by means of 3H-thymidine uptake, and restitution by means of mitomycin-blocked migration.

RESULTS

Changing matrix composition failed to alter proliferation, but collagen I stimulated migration more than laminin or fibronectin (laminin/collagen, 68% +/- 2%; p less than 0.05). EGF (30 ng/ml) increased proliferation on both collagen (225% +/- 11% of basal) and laminin (206% +/- 26%) but increased migration only over laminin (210% +/- 17%) (all, p less than 0.05). TGF-beta 1 (200 pg/ml) stimulated migration over laminin (187% +/- 18%, p less than 0.005) but inhibited migration over collagen (89% +/- 3%, p less than 0.01) and did not affect 3H-thymidine uptake. When cultured on laminin, EGF but not TGF-beta 1 altered organization of the alpha 2 integrin subunit. Genistein (100 mumol/L) inhibited basal and EGF-stimulated 3H-thymidine uptake. In addition, it prevented EGF stimulation of replication-blocked migration (81% +/- 10% vs 190% +/- 20% of basal, p less than 0.0001) without altering basal replication-blocked migration. Indomethacin (10(-5) mol/L) did not alter migration but inhibited basal and EGF-stimulated proliferation by 7% +/- 1% (each, p less than 0.005).

CONCLUSIONS

Restitution and proliferation appear independently regulated by matrix and growth factors. It may be possible to individually target specific phases of mucosal healing by means of pharmacologic agents.

Indomethacin promotes germinal matrix microvessel maturation in the newborn beagle pup

BACKGROUND AND PURPOSE

Although indomethacin has been demonstrated to prevent germinal matrix and intraventricular hemorrhage in clinical and animal studies, the mechanism of action of this agent to prevent hemorrhage remains unclear. Previous studies have demonstrated both that the microvessels in the germinal matrix of newborn beagle pups undergo basement membrane maturation during the first 4 postnatal days and that indomethacin may promote laminin deposition in tumor cell culture systems.

METHODS

We employed the newborn beagle pup model to test the hypothesis that indomethacin may stimulate laminin deposition in germinal matrix microvessels. Newborn pups were randomized to receive either 0.1 mg/kg/dose i.p. indomethacin or an equal volume of saline diluent. Pups received doses of study medication once a day for 1, 2, or 3 days and were studied on postnatal days 1, 2, 3, or 4. Pups were anesthetized and systemically perfused with buffered formalin; the brains were removed and prepared for immunohistochemical study.

RESULTS

Sections stained with Bandeiraea lectin demonstrated that there was no difference in germinal matrix vessel density among the postnatal ages studied; similarly, there were no differences in vessel density between saline- and indomethacin-treated animals at any postnatal age. Quantification of germinal matrix stained intensity by confocal microscopy demonstrated significant increases in indomethacin-treated pups for both laminin staining at postnatal days 2 (p = 0.05) and 3 (p = 0.0009) and type V collagen staining at postnatal day 2 (p = 0.011). Although staining for beta 1 integrins increased across postnatal ages, there were no differences between saline- and indomethacin-treated animals.

CONCLUSIONS

These data suggest that indomethacin may stimulate basement membrane deposition in the germinal matrix microvessels of newborn beagle pups to prevent germinal matrix and/or intraventricular hemorrhage.

Identification of a structural domain that distinguishes the actions of the type 1 and 2 isoforms of transforming growth factor beta on endothelial cells

A chimeric transforming growth factor beta (TGF-beta) molecule has been synthesized to map the amino acids responsible for the substantially greater activity of TGF-beta 1 than TGF-beta 2 on growth and migration of endothelial cells. This chimera consists of a dimer of a monomeric unit composed of amino acids 1-39 of TGF-beta 2, 40-82 of TGF-beta 1, and 83-112 of TGF-beta 2. Structural identity of the purified recombinant protein has been confirmed by immunoblotting and NH2-terminal sequencing. The biological potency of the TGF-beta 2-1-2 chimera was equal to that of TGF-beta 1 in inhibition of growth of both fetal bovine heart endothelial cells and rat epididymal fat pad microvascular endothelial cells. Similarly, the TGF-beta 2-1-2 chimera was nearly equivalent to TGF-beta 1 and at least 10-fold more active than TGF-beta 2 in inhibiting migration of bovine aortic endothelial cells. These results identify the sequence between amino acids 40-82 as an important region within TGF-beta that functions to specify a TGF-beta 1- or TGF-beta 2-like activity.

Human enterocyte (Caco-2) migration is modulated in vitro by extracellular matrix composition and epidermal growth factor

The modulation of enterocyte sheet migration was studied using Caco-2 cells, a well-differentiated human colonic cell line. Although Caco-2 cells attached and spread equivalently over collagen types I, III, IV, and V and laminin, migration over laminin was significantly slower than migration over the collagen types. Fibronectin was a poor substrate for attachment, spreading, and migration. Epidermal growth factor (EGF) stimulated migration over laminin but did not alter Caco-2 migration over collagen or fibronectin. This effect was independent of cell proliferation, which was stimulated equivalently on both laminin and collagen I. Expression and organization of cell surface receptors for matrix (integrins) were studied using antibodies specific for beta and alpha integrin subunits. Integrin surface expression was assessed by immunoprecipitation of surface 125iodinated control and EGF-treated cells. Beta 1 surface pools did not change substantially in any condition studied. Alpha 1 subunit pools were decreased after EGF treatment on collagen I but alpha 1 pools increased after EGF treatment on laminin. Surface pools of alpha 2 subunits were increased following EGF treatment whether cells were cultured on laminin or collagen I. However, traditional immunofluorescent and laser confocal imaging demonstrated substantial differences in the character of alpha 2 subunit organization between collagen and laminin in the migrating cell front. Furthermore, a functional antibody to the alpha 2 subunit inhibited EGF stimulation of migration over laminin without substantial effects on basal migration over laminin or collagen I. Thus, EGF appears to exert a matrix-specific effect on enterocyte migration by modulation of integrin expression and organization.

Modulation of vascular cell behavior by transforming growth factors beta

The vascular cell responses to the type 1, 2, and 3 isoforms of transforming growth factor-beta (TGF-beta 1, TGF-beta 2, TGF-beta 3) were studied using bovine aortic endothelial (BAECs) and smooth muscle cells (BASMC3) as well as rat epididymal fat pad microvascular endothelia (RFCs). Three distinct bioassays indicated that TGF-beta elicits results that do not differ significantly from those of the TGF-beta 1 isoform in all three cell populations. These assays are: inhibition of proliferation, cell migration, and neovascularization. By contrast the cellular responses to TGF-beta 1 and TGF-beta 3 differed from those to TGF-beta 2. Three distinct receptor assays revealed the presence of type I and type II TGF-beta 1 cell surface binding proteins on BAECs, BASMCs, and RFCs. Experimentation to decipher cell surface binding by the different isoforms revealed that iodinated TGF-beta 1 bound to the surface of all three vascular cell types can be competed off in similar fashion by either TGF-beta 1 or TGF-beta 3; however, competition with TGF-beta 2 produced unique binding profiles dependent on the cell type examined. The ratios of type I to type II TGF-beta receptors in these three vascular cell types vary from 1:1 in BAECs to 1.5:1 in RFCs to 3:1 in BASMCs and can be correlated with the differences noted in cellular responses to TGF-beta 1 and TGF-beta 2 in proliferation, migration, and in vitro angiogenic assays.(ABSTRACT TRUNCATED AT 250 WORDS)

Autocrine angiotensin system regulation of bovine aortic endothelial cell migration and plasminogen activator involves modulation of proto-oncogene pp60c-src expression

Rapid endothelial cell migration and inhibition of thrombosis are critical for the resolution of denudation injuries to the vessel wall. Inhibition of the endothelial cell autocrine angiotensin system, with either the angiotensin-converting enzyme inhibitor lisinopril or the angiotensin II receptor antagonist sar1, ile8-angiotensin II, leads to increased endothelial cell migration and urokinase-like plasminogen activator (u-PA) activity (Bell, L., and J. A. Madri. 1990. Am. J. Pathol. 137:7-12). Inhibition of the autocrine angiotensin system with the converting-enzyme inhibitor or the receptor antagonist also leads to increased expression of the proto-oncogene c-src: pp60c-src mRNA increased 7-11-fold, c-src protein 3-fold, and c-src kinase activity 2-3-fold. Endothelial cell expression of c-src was constitutively elevated after stable infection with a retroviral vector containing the c-src coding sequence. Constitutively increased c-src kinase activity reconstituted the increases in migration and u-PA observed with angiotensin system interruption. Antisera to bovine u-PA blocked the increase in migration associated with increased c-src expression. These data suggest that increases in endothelial cell migration and plasminogen activator after angiotensin system inhibition are at least partially pp60c-src mediated. Elevated c-src expression with angiotensin system inhibition may act to enhance intimal wound closure and to reduce luminal thrombogenicity in vivo.

Beagle pup germinal matrix maturation studies

Intraventricular hemorrhage, or hemorrhage into the germinal matrix tissues of the developing brain, remains a common problem of preterm infants. The "risk period" for this insult is the first 3-4 postnatal days. We hypothesized that this risk period for hemorrhage is related to rapid perinatal maturation of the germinal matrix vasculature and employed the newborn beagle pup model for the study of this maturation. Newborn beagle pups (n = 30) were anesthetized and systemically perfused with buffered formalin; the brains were removed and prepared for immunohistochemical study. Sections stained with Bandeiraea lectin demonstrated that there was no difference in germinal matrix vessel density between postnatal days 1 and 4. Germinal matrix sections were also stained for antibodies to alpha-smooth muscle actin, collagen IV, collagen V, desmin, factor VIII-related antigen, fibronectin, glial fibrillary acidic protein, laminin, transferrin, and vimentin. Vasculature staining by alpha-smooth muscle actin was not noted until postnatal day 10, and differential staining was detected for antibodies to laminin and collagen V. Quantification of staining intensity by confocal microscopy demonstrated a significant increase in both extracellular matrix components at postnatal day 4 compared with day 1 (p less than 0.05 for both). These basement membrane proteins may add sufficient structural integrity to germinal matrix vessels to prevent capillary rupture and thus intraventricular hemorrhage.

Effects of soluble factors and extracellular matrix components on vascular cell behavior in vitro and in vivo: models of de-endothelialization and repair

Vessel walls are comprised of several different cell populations residing in and on complex extracellular matrices. Each of the vascular cell types has diverse and sometimes unique functions and morphologies, and each has roles in repair processes following injury. Large vessel endothelial cells are known to respond to denudation injury by sheet migration and proliferation. This is in contrast to the migration through soft tissues with tube formation and subsequent lumen formation exhibited by microvascular endothelial cells in response to injury. Vascular smooth muscle cells of larger vessels respond to injury by migration from the arterial media into the intima, proliferation, and matrix biosynthesis, ultimately causing intimal thickening. Both these cell types exhibit "dysfunctional" phenotypes during their responses to injury. Microvascular cell responses to injury, while extremely variable, are less well documented. Specifically, responses to injury by microvascular endothelial vascular cells appear to be modulated, in part, by the composition and organization of the surrounding matrix as well as by the various soluble factors and cytokines found at sites of injury, suggesting that the extracellular matrix and soluble factors modulate each other's effects on local vascular cell populations following injury.

Alternative splicing of endothelial cell fibronectin mRNA in the IIICS region. Functional significance

Transforming growth factor-beta 1 (TGF-beta 1) is thought to play a role in modulating vascular cell function in vivo. In vitro, it decreases endothelial cell proliferation and migration. We postulated that these biologic activities could be mediated through TGF-beta 1 modulation of specific gene expression. Therefore we differentially screened a human umbilical vein endothelial cell cDNA library with cDNAs prepared from both untreated and TGF-beta 1-treated bovine aortic endothelial cells. Using this technique, we isolated many TGF-beta 1-induced cDNA clones. Sequence analysis of these cDNAs showed that many of them corresponded to alternatively spliced fibronectin mRNAs. These fibronectin clones all contained the extradomain I (ED I) but three different forms of the type III connecting segment (IIICS). These different fibronectin cDNAs were expressed in bacteria and the recombinant proteins used to study the effects of IIICS alternative splicing on cell attachment, spreading, and migration in bovine aortic endothelial and smooth muscle cells and B16F10 melanoma cells. The results of these experiments show that attachment and spreading of bovine aortic endothelial and smooth muscle cells depend primarily on the presence of the Arg-Gly-Asp-Ser (RGDS) sequence in the recombinant fibronectin proteins. However attachment and spreading of bovine aortic endothelial cells are modulated by alternative splicing in the IIICS region. Specifically splicing of the IIICS region decreases spreading and increases migration rates of the endothelial cells. On the contrary, using a cell line (B16F10 melanoma cells) that is known not to require the RGDS sequence for adhesion confirmed previous findings that B16F10 melanoma cells do not require the presence of the RGDS sequence for attachment and spreading. Indeed B16F10 cells were able to attach and spread on two recombinant proteins that did not contain the RGDS sequence. However attachment and spreading of B16F10 were dramatically inhibited when a 75-base pair DNA fragment was removed from the 5' end of the IIICS region. These results suggest that various regions of the fibronectin molecule may be able to interact with different cell populations to promote cell attachment and spreading, and that alternative splicing may modulate this process.

Influence of the angiotensin system on endothelial and smooth muscle cell migration

The blood vessel wall's response to injury is an important determinant of luminal size and vessel function. The physiologic migration of endothelial cells from the edges of a wound and the pathophysiologic migration of medial smooth muscle cells into the intima are two important components of the vessel wall's response to injury. The influence of the angiotensin system on endothelial and smooth muscle cell migration have not been examined. In the present study, the influence of angiotensin system components on bovine aortic endothelial cell (BAEC) and bovine aortic smooth muscle cell (BASMC) migration after release of cultured cell monolayers from contact inhibition was determined. The angiotensin-converting enzyme (ACE) inhibitor lisinopril increased BAEC migration 41% +/- 3% (P less than 0.001), as did the specific angiotensin II antagonist sar1, ile8-angiotensin II (SAR) (41% +/- 3% (P less than 0.001). Exogenous angiotensin I and angiotensin II did not affect BAEC migration. Exogenous angiotensin II abolished the effect of lisinopril on BAEC migration. Lisinopril increased cell-associated u-plasminogen activator (u-PA) 23% +/- 3% (P less than 0.001) in migrating BAEC and angiotensin II abolished this increase. SAR increased u-PA 33% +/- 0% (P less than 0.001). In contrast, these agents had the opposite effect on smooth muscle cells. Angiotensin II increased smooth muscle cell migration 40% +/- 3% (P less than 0.001), and this effect was abolished by SAR. Angiotensin II also increased cell-associated u-PA 83% +/- 7% (P less than 0.001) in migrating BASMC. The increase in BAEC migration with inhibition of endothelial cell angiotensin II stimulation, either with lisinopril or SAR, also was associated with an increase in cell-associated u-PA. These results indicate that lisinopril interrupts an autocrine pathway in endothelial cells, in which endothelial cell-derived angiotensin I is converted to angiotensin II by ACE, and imply that angiotensin-converting enzyme inhibitors in vivo would act to reduce vessel wall injury by directly increasing the rate of endothelial cell wound closure; by increasing the antithrombotic tendency of the endothelium via enhanced u-PA; and indirectly, by decreasing production of angiotensin II and thereby the rate of smooth muscle cell migration into the intima.

Spatiotemporal segregation of endothelial cell integrin and nonintegrin extracellular matrix-binding proteins during adhesion events

Bovine aortic endothelial cell (BAEC) attachments to laminin, fibronectin, and fibrinogen are inhibited by soluble arginine-glycine-aspartate (RGD)-containing peptides, and YGRGDSP activity is responsive to titration of either soluble peptide or matrix protein. To assess the presence of RGD-dependent receptors, immunoprecipitation and immunoblotting studies were conducted and demonstrated integrin beta 1, beta 3, and associated alpha subunits as well as a beta 1 precursor. Immunofluorescence of BAECs plated on laminin, fibronectin, and fibrinogen reveals different matrix-binding specificities of each of these integrin subclasses. By 1 h after plating, organization of beta 1 integrin into fibrillar streaks is influenced by laminin and fibronectin, whereas beta 3 integrin punctate organization is influenced by fibrinogen and the integrin spatial distribution changes with time in culture. In contrast, the nonintegrin laminin-binding protein LB69 only organizes after cell-substrate contact is well established several hours after plating. Migration of BAECs is also mediated by both integrin and nonintegrin matrix-binding proteins. Specifically, BAEC migration on laminin is remarkably sensitive to RGD peptide inhibition, and, in its presence, beta 1 integrin organization dissipates and reorganizes into perinuclear vesicles. However, RGD peptides do not alter LB69 linear organization during migration. Similarly, agents that block LB69--e.g., antibodies to LB69 as well as YIGSR-NH2 peptide--do not inhibit attachment of nonmotile BAECs to laminin. However, both anti-LB69 and YIGSR-NH2 inhibit late adhesive events such as spreading. Accordingly, we propose that integrin and nonintegrin extracellular matrix-binding protein organizations in BAECs are both temporally and spatially segregated during attachment processes. High affinity nonintegrin interaction with matrix may create necessary stable contacts for longterm attachment, while lower affinity integrins may be important for initial cell adhesion as well as for transient contacts of motile BAECs.

A 48 kDa collagen-binding phosphoprotein isolated from bovine aortic endothelial cells interacts with the collagenous domain, but not the globular domain, of collagen type IV

We have identified collagen-binding proteins in detergent extracts of metabolically labelled bovine aortic endothelial cells (BAEC) by collagen type IV-Sepharose affinity chromatography. The major collagen type IV-binding protein identified by SDS/PAGE had a molecular mass of 48 kDa, which we term the 'collagen-binding 48 kDa protein' (CB48). The pI of CB48 was 8.0-8.3 in a two-dimensional gel system, running non-equilibrium pH gel electrophoresis in the first dimension and SDS/PAGE in the second dimension. Under these conditions CB48 separated into two major (a and b) and one minor isoform (c); a was the most basic of the three isoforms. Two-dimensional chymotryptic peptide maps derived from each individual isoform were virtually identical. The charge differences between the isoforms were due in part to differential H3(32)PO4 incorporation by the protein. CB48 bound to intact collagen type IV and the collagenous region of collagen type IV, but not to the globular NC1 domain. Cell-surface labelling and indirect immunofluorescence experiments localized the bulk of CB48 intracellularly in the endoplasmic reticulum Golgi region, with a minor population of molecules on the cell surface. A specific rabbit polyclonal anti-CB48 serum did not inhibit the attachment or spreading of BAEC to collagen type IV in an 'in vitro' adhesion assay, suggesting that the cell-surface population of CB48 is not involved in BAEC adhesion. We conclude that CB48 is a collagen-binding phosphoprotein that interacts with the collagenous domain of collagen type IV and may be involved in intracellular transport of collagen molecules.

Transforming growth factor beta 1 modulates extracellular matrix organization and cell-cell junctional complex formation during in vitro angiogenesis

Transforming growth factor-beta 1 (TGF-beta 1) is angiogenic in vivo. In two-dimensional (2-D) culture systems microvascular endothelial cell proliferation is inhibited up to 80% by TGF-beta 1; however, in three-dimensional (3-D) collagen gels TGF-beta 1 is found to have no effect on proliferation while eliciting the formation of calcium and magnesium dependent tube-like structures mimicking angiogenesis. DNA analyses performed on 3-D cell cultures reveal no significant difference in the amount of DNA or cell number in control versus TGF-beta 1 treated cultures. In 2-D cultures TGF-beta 1 is known to increase cellular fibronectin accumulation; however, in 3-D cultures no difference is seen between control and TGF-beta 1 treated cells as established by ELISA testing for type IV collagen, fibronectin, and laminin. In 3-D cultures there is increased synthesis and secretion of type V collagen in both control and TGF-beta 1 treated cultures over 2-D cultures. Even though an equal amount of type V collagen is seen in both 3-D conditions, there is a reorganization of the protein with concentration along an organizing basal lamina in TGF-beta 1 treated cultures. EM morphological analyses on 3-D cultures illustrate quiescent, control cells lacking cell contacts. In contrast, TGF-beta 1 treated cells show increased pseudopod formation, cell-cell contact, and organized basal lamina-like material closely apposed to the "abluminal" plasma membranes. TGF-beta 1 treated cells also appear to form junctional complexes between adjoining cells. Immunofluorescence using specific antibodies to the tight junction protein ZO-1 results in staining at apparent cell-cell junctions in the 3-D cultures. Northern blots of freshly isolated microvascular endothelium, 2-D and 3-D cultures, using cDNA and cRNA probes specific for the ZO-1 tight junction protein, reveal the presence of the 7.8 kb mRNA. Western blots of rat epididymal fat pad endothelial cells (RFC) monolayer lysates probed with anti-ZO-1 label a 220 kd band which co-migrates with the bonafide ZO-1 protein. These data confirm and support the hypothesis that TGF-beta 1 is angiogenic in vitro, eliciting microvascular endothelial cells to form tube-like structures with apparent tight junctions and abluminal basal lamina deposition in three-dimensional cultures.

Transforming growth factor beta 1 modulates extracellular matrix organization and cell-cell junctional complex formation during in vitro angiogenesis

Transforming growth factor-beta 1 (TGF-beta 1) is angiogenic in vivo. In two-dimensional (2-D) culture systems microvascular endothelial cell proliferation is inhibited up to 80% by TGF-beta 1; however, in three-dimensional (3-D) collagen gels TGF-beta 1 is found to have no effect on proliferation while eliciting the formation of calcium and magnesium dependent tube-like structures mimicking angiogenesis. DNA analyses performed on 3-D cell cultures reveal no significant difference in the amount of DNA or cell number in control versus TGF-beta 1 treated cultures. In 2-D cultures TGF-beta 1 is known to increase cellular fibronectin accumulation; however, in 3-D cultures no difference is seen between control and TGF-beta 1 treated cells as established by ELISA testing for type IV collagen, fibronectin, and laminin. In 3-D cultures there is increased synthesis and secretion of type V collagen in both control and TGF-beta 1 treated cultures over 2-D cultures. Even though an equal amount of type V collagen is seen in both 3-D conditions, there is a reorganization of the protein with concentration along an organizing basal lamina in TGF-beta 1 treated cultures. EM morphological analyses on 3-D cultures illustrate quiescent, control cells lacking cell contacts. In contrast, TGF-beta 1 treated cells show increased pseudopod formation, cell-cell contact, and organized basal lamina-like material closely apposed to the "abluminal" plasma membranes. TGF-beta 1 treated cells also appear to form junctional complexes between adjoining cells. Immunofluorescence using specific antibodies to the tight junction protein ZO-1 results in staining at apparent cell-cell junctions in the 3-D cultures. Northern blots of freshly isolated microvascular endothelium, 2-D and 3-D cultures, using cDNA and cRNA probes specific for the ZO-1 tight junction protein, reveal the presence of the 7.8 kb mRNA. Western blots of rat epididymal fat pad endothelial cells (RFC) monolayer lysates probed with anti-ZO-1 label a 220 kd band which co-migrates with the bonafide ZO-1 protein. These data confirm and support the hypothesis that TGF-beta 1 is angiogenic in vitro, eliciting microvascular endothelial cells to form tube-like structures with apparent tight junctions and abluminal basal lamina deposition in three-dimensional cultures.

Effect of platelet factors on migration of cultured bovine aortic endothelial and smooth muscle cells

Endothelial cell (EC) injury and the response of EC and smooth muscle cells (SMCs) to injury contribute to the pathophysiology in patients with vascular disease and atherosclerosis. Since platelets have been suggested to play an important role in modulating vascular injury, the present study was undertaken to examine the influence and mechanism of action of individual platelet factors on bovine aortic EC and SMC migration using an in vitro wound assay system. Serotonin decreased EC proliferation and reduced EC migration 21 +/- 1% (p less than 0.005), which was attenuated by imipramine. Transforming growth factor-beta reduced EC proliferation and decreased EC migration 52 +/- 3% (p less than 0.005). Norepinephrine increased EC proliferation but decreased EC migration 26 +/- 2% (p less than 0.005), which was abolished by phenoxybenzamine. Histamine increased EC proliferation but reduced EC migration 29 +/- 2% (p less than 0.005), which was attenuated by diphenhydramine. Platelet-derived growth factor decreased EC proliferation and decreased EC migration 40 +/- 2% (p less than 0.005). In contrast, serotonin increased SMC proliferation and increased SMC migration 31 +/- 2% (p less than 0.005), which was abolished by ketanserin. Transforming growth factor-beta increased SMC migration 35 +/- 5% (p less than 0.005). Norepinephrine increased SMC proliferation and increased SMC migration 43 +/- 4% (p less than 0.005), which was abolished by propranolol. Histamine increased SMC proliferation and increased SMC migration 38 +/- 3% (p less than 0.005), which was abolished by cimetidine. Platelet-derived growth factor increased SMC proliferation and increased SMC migration 40 +/- 3% (p less than 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)

Endothelial cell behavior after denudation injury is modulated by transforming growth factor-beta1 and fibronectin

Endothelial denudation injury after endarterectomy, autologous and synthetic grafting and balloon angioplasty leads to exposure of thrombogenic vessel wall material and may elicit an atherogenic response in the media of the affected vessels in which complete reendothelialization may not occur. While the role(s) of extracellular matrix composition and organization in this process are only incompletely understood, it is widely accepted that endothelial cells respond to matrix components in specific, complex fashions. In this report we demonstrate that large vessel endothelial cell migration is affected by the surrounding matrix and the soluble factor, transforming growth factor-beta1, which may mediate its effects, in part, by modulating endothelial cell matrix synthesis. Specifically, large vessel endothelial cell migration is decreased on a fibronectin substratum and in the presence of transforming growth factor-beta1, which increases fibronectin mRNA and protein accumulation in culture. Inhibition of sheet migration is also elicited by the addition of soluble fibronectin to the cultures. These in vitro findings are consistent with our in vivo findings of increased staining of fibronectin luminally and in the intima in the chronically deendothelialized region of a balloon catheter denuded carotid artery. Thus, reendothelialization after iatrogenic and natural injury appears to be a complex process which can be modulated by the underlying matrix and soluble factors, which may themselves modulate the matrix synthesis of local vascular cells.

Modulation of actin mRNAs in cultured vascular cells by matrix components and TGF-beta 1

Alpha-smooth muscle actin is currently considered a marker of smooth muscle cell differentiation. However, during various physiologic and pathologic conditions, it can be expressed, sometimes only transiently, in a variety of other cell types, such as cardiac and skeletal muscle cells, as well as in nonmuscle cells. In this report, the expression of actin mRNAs in cultured rat capillary endothelial cells (RFCs) and aortic smooth muscle cells (SMCs) has been studied by Northern hybridization in two-dimensional cultures seeded on individual extracellular matrix proteins and in three-dimensional type I collagen gels. In two-dimensional cultures, in addition to cytoplasmic actin mRNAs which are normally found in endothelial cell populations, RFCs expressed alpha-smooth muscle (SM) actin mRNA at low levels. alpha-SM actin mRNA expression is dramatically enhanced by TGF-beta 1. In addition, double immunofluorescence staining with anti-vWF and anti-alpha-SM-1 (a monoclonal antibody to alpha-SM actin) shows that RFCs co-express the two proteins. In three dimensional cultures, RFCs still expressed vWF, but lost staining for alpha-SM actin, whereas alpha-SM actin mRNA became barely detectable. In contrast to two-dimensional cultures, the addition of TGF-beta 1 to the culture media did not enhance alpha-SM actin mRNA in three-dimensional cultures, whereas it induced rapid capillary tube formation. Actin mRNA expression was modulated in SMCs by extracellular matrix components and TGF-beta 1 with a pattern very different from that of RFCs. Namely, the comparison of RFCs with other cell types such as bovine aortic endothelial cells shows that co-expression of endothelial and smooth muscle cell markers is very unique to RFCs and occurs only in particular culture conditions. This could be related to the capacity of these microvascular endothelial cells to modulate their phenotype in physiologic and pathologic conditions, particularly during angiogenesis, and could reflect different embryologic origins for endothelial cell populations.

Reconstitution of structure and cell function in human skin grafts derived from cryopreserved allogeneic dermis and autologous cultured keratinocytes

Grafts of allogeneic dermis plus autologous epidermal cell cultures were used to replace extensively burned skin. Cryopreserved split-thickness cadaveric skin was grafted onto debrided burn wound, and autologous keratinocytes were cultured from uninjured donor sites. Several weeks later, allograft epidermis was abraded and replaced with the keratinocyte cultures. The final grafts were thus composites of autologous cultured epidermis and allogeneic dermis. In a case with 28 months follow-up, reconstitution of the dermal-epidermal (BMZ.1) and microvascular (BMZ.2) basement membrane zones was studied immunohistochemically and ultrastructurally. Immediately before grafting, thawed cryopreserved skin reacted with antibodies against laminin and type IV collagen in normal patterns. Twenty-nine days after grafting, BMZ.1 reacted weakly with both antibodies, and anticollagen type IV reactivity was absent from BMZ.2. Antilaminin reactivity of BMZ.2, however, was moderately intense, consistent with recent neovascularization. On day 29, the allograft epidermis was replaced with autologous keratinocyte cultures. Twenty-five days later (54 d after allografting), staining of both BMZs was intense with both antibodies. Ultrastructurally, at day 76 (47 d after culture placement) BMZ.1 revealed only small hemidesmosomes, few incipient anchoring fibrils, and a discontinuous lamina densa. BMZ.2, however, was fully reconstituted. By 124 d, both BMZs appeared normal. Observations in the dermis at 76 d included the presence of lymphocytes, organellar debris, and hyperactive collagen fibrillogenesis, all indicative of dermal remodelling. The microvasculature was well differentiated, but no elastic fibers or nerves were found. In the epidermis, melanocytes and evidence of melanosome transfer were seen at 5, 47, and 95 d after grafting of keratinocyte cultures. We conclude that the composite procedure reconstitutes skin with excellent textural and histologic qualities.

Extracellular matrix specificity for the differentiation of capillary endothelial cells

Capillary endothelial cells in a fenestrated vasculature contain openings in attenuated areas of the cytoplasm which are often covered with one or two diaphragms. However, due to endothelial cell dedifferentiation upon culturing, such indicative membrane structure are often not expressed. The expression of a more differentiated phenotype can be regulated by the extracellular matrix to which the cells attach. In addition, during angiogenesis endothelial cell migration enables their interaction with other cell types and matrices which may directly affect endothelial cell growth and function. We report the ability to modify the expression of diaphragmed fenestrations and transcapillary channels in capillary endothelial cells by specific extracellular matrices. When the number of membrane openings is measured, growth of the cells on Madin--Darby canine kidney cell matrix results in the greatest number while other matrices or isolated matrix components are only partially active. Production of such a biologically active matrix not only allows an avenue to identify fenestrae-associated proteins but also provides an easy culture method to more closely mimic the in vivo phenotype of endothelial cells.

Matrix-driven cell size change modulates aortic endothelial cell proliferation and sheet migration

Current hypotheses suggest that the extracellular matrix modulates cellular function in physiologic homeostasis and during injury and repair and that selected cellular functions are cell and nuclear size dependent. For elucidation of the role of extracellular matrix-driven cell size changes in the modulation of endothelial cell proliferation and sheet migration an in vitro model system was used that allows for the culture of bovine aortic endothelial cells (BAEC) on various purified extracellular matrix components. BAEC exhibited distinct patterns in rates of attachment, spreading, migration, and proliferation when cultured on the selected extracellular matrix components laminin, types I and III collagen, type IV collagen, and fibronectin. In addition, there was a correlation between cell and nuclear size (on the various matrices tested) and rates of cell attachment and spreading. In contrast, an inverse correlation was noted between cell and nuclear size (on the various matrices tested) and proliferation and sheet migration. These data demonstrate that endothelial cells respond to matrix components in specific but complex fashions, mediated, in part, by changes in cell and nuclear size.

Distribution of a 69-kD laminin-binding protein in aortic and microvascular endothelial cells: modulation during cell attachment, spreading, and migration

Affinity chromatography and immunolocalization techniques were used to investigate the mechanism(s) by which endothelial cells interact with the basement membrane component laminin. Bovine aortic endothelial cells (BAEC) membranes were solubilized and incubated with a laminin-Sepharose affinity column. SDS-PAGE analysis of the eluted proteins identified a 69-kD band as the major binding protein, along with minor components migrating at 125, 110, 92, 85, 75, 55, and 30 kD. Polyclonal antibodies directed against a peptide sequence of the 69-kD laminin-binding protein isolated from human tumor cells identified this protein in BAEC lysates. In frozen sections, these polyclonal antibodies and monoclonal antibodies raised against human tumor 69-kD stained the endothelium of bovine aorta and the medial smooth muscle cells, but not surrounding connective tissue or elastin fibers. When nonpermeabilized BAEC were stained in an in vitro migration assay, there appeared to be apical patches of 69 kD staining in stationary cells. However, when released from contact inhibition, 69 kD was localized to ruffling membranes on cells at the migrating front. Permeabilized BAEC stained for 69 kD diffusely, with a granular perinuclear distribution and in linear arrays throughout the cell. During migration a redistribution from diffuse to predominanately linear arrays that co-distributed with actin microfilaments was noted in double-label experiments. The 69-kD laminin-binding protein colocalized with actin filaments in permeabilized cultured microvascular endothelial cells in a continuous staining pattern at 6 h postplating which redistributed to punctate patches along the length of the filaments at confluence (96 h). In addition, 69 kD co-distribution with laminin could also be demonstrated in cultured subconfluent cells actively synthesizing matrix. Endothelial cells express a 69-kD laminin-binding protein that is membrane associated and appears to colocalize with actin microfilaments. The topological distribution of 69 kD and its cytoskeletal associations can be modulated by the cell during cell migration and growth suggesting that 69 kD may be a candidate for a membrane protein involved in signal transduction from extracellular matrix to cell via cytoskeletal connections.

Phenotypic modulation of endothelial cells by transforming growth factor-beta depends upon the composition and organization of the extracellular matrix

Transforming growth factor beta (TGF-beta) is angiogenic in vivo. In vitro, endothelial cell proliferation is inhibited by TGF-beta. We have correlated this inhibitory effect with an increase in cellular fibronectin synthesis and deposition in a two-dimensional culture system using specific matrix coatings. The inhibitory effect was mimicked by addition of soluble fibronectin to cultures. In contrast, TGF-beta was found to elicit the formation of tube-like structures (mimicking angiogenesis) when microvascular endothelial cells were grown in three-dimensional collagen gels. In this culture system TGF-beta elicited rapid extensive formation of complex, branching, tube-like structures, while cell proliferation was not inhibited. These data confirm and support the hypothesis that TGF-beta is angiogenic and may exert some of its effects through modulation of matrix synthesis and are consistent with the hypothesis that the organization of the extracellular environment influences cellular responses to this "panregulin."

Inverse Ir gene control of the antibody and T cell proliferative responses to human basement membrane collagen

The immune response to pepsin-soluble human basement membrane-derived type IV collagen in mice has been characterized. Both T cell proliferative and antibody responses have been shown to be under major histocompatibility complex (MHC)-linked Ir gene control in inbred and MHC congenic mice. However, unlike previous examples studied, this response shows a separation of these two types of immunologic responsiveness. Only mice having I-As give potent in vitro T cell proliferative responses to type IV collagen whereas all mice except those having I-As give high antibody responses to this antigen. In (I-As X I-Anon-s) F1 mice, the T cell proliferative response was dominant, whereas antibody responses were markedly reduced compared with the responder parent. Given the recent demonstration that class II MHC-restricted, L3T4+ T cells can be divided into two sets, one of which helps for antibody responses and the other of which produces interleukin 2 and can also suppress such responses, it seems likely that these data can be accounted for on the basis of differential activation by this antigen of these two cell sets in mice of different MHC genotypes.

Inverse Ir gene control of the antibody and T cell proliferative responses to human basement membrane collagen

The immune response to pepsin-soluble human basement membrane-derived type IV collagen in mice has been characterized. Both T cell proliferative and antibody responses have been shown to be under major histocompatibility complex (MHC)-linked Ir gene control in inbred and MHC congenic mice. However, unlike previous examples studied, this response shows a separation of these two types of immunologic responsiveness. Only mice having I-As give potent in vitro T cell proliferative responses to type IV collagen whereas all mice except those having I-As give high antibody responses to this antigen. In (I-As X I-Anon-s) F1 mice, the T cell proliferative response was dominant, whereas antibody responses were markedly reduced compared with the responder parent. Given the recent demonstration that class II MHC-restricted, L3T4+ T cells can be divided into two sets, one of which helps for antibody responses and the other of which produces interleukin 2 and can also suppress such responses, it seems likely that these data can be accounted for on the basis of differential activation by this antigen of these two cell sets in mice of different MHC genotypes.

The basement membrane underlying the vascular endothelium is not thrombogenic: in vivo and in vitro studies with rabbit and human tissue

Studies examining the interaction of platelets with exposed subendothelium in vivo have reported conflicting results. To examine possible explanations for the apparently discrepant findings, we measured the platelet reactivity of subendothelium prepared by a number of methods both in vivo and in vitro. In addition, we examined the possibility that 13-hydroxyoctadecadinoic acid (13-HODE), an endothelial cell-derived chemorepellant, modulates the reactivity of the subendothelium to platelets. In vivo, the subendothelium of segments of rabbit carotid arteries was exposed by removing the endothelial cells by air perfusion or by balloon catheter stripping. Platelet accumulation onto the de-endothelialized segments was assessed by 3H-radioactivity uptake, using 3H-adenine-labelled platelets, and by scanning electron microscopy. In vitro, 3H-adenine-labelled platelet adhesion was measured onto plain plastic discs and onto plastic discs coated with the following purified basement membrane components: collagens type I, III, IV, V, laminin, or fibronectin. In addition, 3H-adenine-labelled platelet adhesion was measured onto plastic discs covered with human endothelial cells or onto the basement membrane underlying the endothelial cells. In vivo, there was marked 3H-platelet accumulation onto the balloon catheter carotid arteries one hour after injury. In contrast, there was no platelet accumulation onto the subendothelium of carotid arteries de-endothelialized by air perfusion. These differences were confirmed by scanning electron microscopy. Transmission electron microscopic examination demonstrated that the extracellular matrix was intact following the air perfusion injury whereas the majority of it was removed by the balloon catheter injury.(ABSTRACT TRUNCATED AT 250 WORDS)

Aortic endothelial cell proteoheparan sulfate. II. Modulation by extracellular matrix

The effects of extracellular matrix components on proteoheparan sulfate biosynthesis was studied for bovine aortic endothelial cells in tissue culture. When the cells were maintained on a variety of different purified components of the extracellular matrix, the cells expressed the same three species of proteoheparan sulfates as the cells cultured on tissue culture plastic (HS I, HS II, and HS III). However, the amounts of the three species recovered from the tissue culture medium were found to be dependent on the substrate on which the cells are grown as well as on other factors. In comparison with plastic, much less HS I was found in the medium of cells maintained on substrates containing diverse matrix molecules, whereas the amounts of HS II and HS III essentially remained the same. In contrast, when bovine aortic organ cultures were analyzed under pulsatile flow, marked differences in the profile of proteoheparan sulfate biosynthesis were observed: HS I was found exclusively associated with the plasma membrane of the endothelial cells; HS II was localized only to the subendothelial matrix; and HS III represented the only proteoheparan sulfate species in the medium. This distribution is consistent with polarized secretion and deposition into the subcellular matrix of HS III and retention of HS I in the plasma membrane in the organ culture situation, a biosynthetic phenotype which can only be approximated at best by maintaining the endothelial cells on a substrate other than plastic. When aortic media (devoid of endothelial cells) was placed in organ culture, no HS III could be detected, which suggested that the vascular endothelial cell is the major cell type responsible for its synthesis in organ culture. Thus, the extracellular matrix, depending upon its composition and organization, may play an important role in stabilizing cell polarity and thereby contribute to maintenance of the differentiated phenotype appropriate for the endothelial cell.

An immunohistochemical and serum ELISA study of type I and III procollagen aminopropeptides in primary biliary cirrhosis

By means of ELISA methodology, the aminopropeptides of Type I and Type III procollagen were measured in the serum of a group of patients with primary biliary cirrhosis. The corresponding liver biopsies were graded blindly for degrees of fibrosis and inflammation. When available, paraffin-embedded liver specimens underwent immunoperoxidase staining for mature Type I and III collagen as well as the aminopropeptides of Type I and III procollagen. Regardless of the degree of fibrosis or inflammation, serum levels of the aminopropeptide of Type I remained within normal limits. In contrast, serum levels of the aminopropeptide of Type III procollagen were elevated uniformly. Immunohistochemistry demonstrated that the aminopropeptide of Type III procollagen persists extracellularly. This finding may explain the previously reported relationship between levels of inflammation and serum levels of the Type III aminopropeptide.

Aortic endothelial cell proteoheparan sulfate. I. Isolation and characterization of plasmamembrane-associated and extracellular species

Proteoheparan sulfate biosynthesis was studied in cultured bovine aortic endothelial cells by means of pulse and pulse-chase experiments and subcellular fractionations. Three proteoheparan sulfate species were found in the medium. The major species, which the authors have called HS I, appeared in the medium only after an initial lag period and was also found associated with the plasma membrane. The other two (HS II and HS III) appeared in small amounts in the medium at early time points. At later times these were not readily observed because the large amounts of HS I present in the medium. The major medium species, HS I, appeared to be composed of approximately four heparan sulfate chains of approximately 35,000 daltons and a core protein of approximately 55,000 daltons apparent molecular weight. HS I appeared to be homogeneous by gel filtration on Sepharose CL 2B and 6B and elution from DEAE Sephacel, electrophoresis on Nu-Sieve agarose, and CsCl density centrifugation. After digestion with heparinase the core protein appeared to be homogeneous by S-200 Sephacel chromatography. HS I was also found associated with plasma membrane fractions of the cultured bovine aortic endothelial cells, and antisera raised against it stained epithelial and endothelial cells in patterns consistent with a cell surface localization. Of the other two species found in the medium, one (HS II) also appeared to be a component of the cell layer. This species appeared to contain approximately four heparan sulfate chains of approximately 20,000 daltons apparent molecular weight. Antisera raised against a similar molecule produced by HR 9 cell cultures stained basement membranes intensely, supporting the subcellular matrix localization of this molecule. The third species (HS III) was detected in culture medium only and apparently contained two heparan sulfate chains of approximately 20,000 daltons apparent molecular weight. These results support the concept of multiple endothelial cell proteoheparan sulfate species which exhibit differences in structure and localization and possibly diverse specialized functions.

Retinol and extracellular collagen matrices modulate hepatic Ito cell collagen phenotype and cellular retinol binding protein levels

The hepatic vitamin A-storing Ito cell has been implicated as a causative cell in hepatic fibrogenesis. Using a modification of a recent method (Friedman, S. L., Roll, F. J., Boyles, J., and Bissell, D. M. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 8681-8685), rat Ito cells were isolated and passaged in vitro on collagen-coated plastic dishes through cell generation 40-50. The collagen synthetic phenotype for Ito cells grown on various extracellular matrices was demonstrated by immunofluorescence and quantitated by competition enzyme-linked immunosorbent assays. When grown on a type I collagen matrix, Ito cells produced type IV greater than type III greater than type I collagen. When grown on a type IV collagen matrix, the cells produced relatively equal amounts of types I and III collagen. The absolute amounts of type I collagen produced were greater when cells were grown on type IV versus type I matrix. When 10(-5) M retinol was added to cell cultures, there was a uniform increase in type III collagen regardless of matrix type but a decrease in type I collagen when cells were grown on a type IV matrix and a large increase in type I collagen when cells were grown on a type I collagen matrix. The levels of cellular retinol binding protein, a key cytosolic retinol transport protein, were quantitated by high performance liquid chromatography and compared for cells grown on type I versus type IV collagen matrices. It was found that cells on a type I matrix contain 4.96 +/- 2.8 times more cellular retinol binding protein than do cells grown on a type IV matrix. In conclusion, Ito cell collagen synthesis may be altered by underlying extracellular matrix and exogenous retinol. This in vitro culture system should allow the study of regulatory factors and possible therapeutic anti-fibrogenic mediators.

The microvascular extracellular matrix. Developmental changes during angiogenesis in the aortic ring-plasma clot model

The composition of the extracellular matrix of developing microvessels in plasma clot cultures of rat aorta was studied with light and electron immunohistochemical techniques using affinity-purified antibodies against fibronectin, laminin, and collagen Types I, III, IV, and V. The extracellular matrix of solid endothelial sprouts in young cultures consisted of a delicate fibrillary network of fibronectin and Type V collagen and of patchy amorphous deposits of laminin and Type IV collagen. Rare fibrils of collagen Types I and III were also observed. Fibronectin stained intensely and appeared to be the predominant component of the provisional subendothelial matrix during vascular sprouting. As the cultures aged, laminin and Type IV collagen accumulated in the subendothelial space, forming a continuous feltwork around the newly formed microvessels. Patent microvessels were also surrounded by discontinuous deposits of fibronectin and by increased amounts of collagen Types I, III, and V. Ultrastructural studies revealed positive immunostaining for fibronectin, laminin, and collagen Types IV and V in the endoplasmic reticulum and in putative secretory vesicles, indicating active synthesis and secretion of these molecules by the endothelial cells. These observations indicate that the microvascular extracellular matrix undergoes significant dynamic changes during capillary development. The different composition and structural organization of the extracellular matrix at various stages of angiogenesis may have important effects on endothelial behavior and capillary morphogenesis.

Endothelial growth factors and extracellular matrix regulate DNA synthesis through modulation of cell and nuclear expansion

Studies were carried out to analyze the mechanism by which extracellular matrix (ECM) molecules and soluble growth factors interplay to control capillary endothelial cell growth. Bovine adrenal capillary endothelial cells attached to purified matrix components but spread poorly and exhibited low levels of DNA synthesis in the absence of exogenous growth factors or serum. Addition of cationic, heparin-binding growth factor purified from either human hepatoma cells or normal bovine pituitary (fibroblast growth factor) induced extensive cell spreading and up to eight fold increases in DNA synthetic rates relative to levels observed in cells on similar substrata in the absence of mitogen. However, the extent of this response differed depending upon the type of ECM molecule used for cell attachment (fold increase on type III collagen greater than gelatin greater than type IV collagen greater than fibronectin greater than type V collagen much greater than laminin). Computerized morphometry demonstrated that endothelial cell DNA synthetic rates increased in an exponential fashion in direct relation to linear increases in cell and nuclear size (projected areas). Similarly sized cells always displayed the same level of DNA synthesis independent of the type of ECM molecule used for cell attachment or the presence of saturating amounts of growth factor. In all cases, DNA metabolism appeared to be coupled to physical expansion of the cell and nucleus rather than to a specific cell morphology (e.g. polygonal versus bipolar). These findings suggest that ECM may act locally as a "solid state" regulator of angiogenesis through its ability to selectively support or prohibit cell and nuclear extension in response to stimulation by soluble mitogens.

Type I and type III procollagen peptides during hepatic fibrogenesis. An immunohistochemical and ELISA serum study in the CCl4 rat model

Serum assays of procollagen peptides have been suggested as useful clinical indicators of hepatic fibrogenesis. To evaluate these assays in an experimental situation, the carbon tetrachloride model of hepatic fibrosis was produced in the rat. With affinity-purified antibodies, ELISA assays were developed and used for detecting nanogram quantities of the aminopropeptides of Types I and III procollagen (pro-I and pro-III) in rat serum. Serum SGPT levels were also determined. Routine histologic staining, as well as immunofluorescent staining for localization of Types I, III, and IV collagen and the aminopropeptides of Types I and III procollagen were performed on corresponding livers. It was found that serum pro-III levels were elevated after 45 and 70 days of treatment (28 +/- 15.2 ng/ml and 21 +/- 3.4 ng/ml, respectively) but returned to normal levels after 90 days of treatment (less than 3 ng/ml). Serum pro-I levels were elevated after 45 days of treatment (1028 +/- 504 ng/ml) but were normal at 70 and 90 days of treatment. Serum SGPT values were elevated at 45 and 70 days of treatment but were normal at 90 days of treatment. The decline in serum pro-III levels appeared to parallel the noted decline in SGPT values. Linear regression analysis revealed a good correlation between pro-III levels and histologic inflammation (r = 0.886) but a poor correlation between pro-III levels and histologic fibrosis (r = 0.116). Immunohistochemistry revealed that the pro-III antigen persists extracellularly for at least 45 days. Therefore, serum assays of pro-III might reflect extracellular collagen degradation as well as active collagen secretion. This would limit the clinical utility of the pro-III assay as an unequivocal marker of active hepatic collagen deposition.

Endothelial cell proliferation during angiogenesis. In vitro modulation by basement membrane components

Modulation of the behavior of microvascular endothelial cells during angiogenesis has been observed to correlate with changes in the extracellular matrix. These reports prompted a comparison of the growth of microvascular endothelial cells on monolayers of various matrix components in vitro. Over a 5 day period, the proliferation of these cells was significantly greater on laminin than on either plasma fibronectin, the interstitial collagen types I and III, or on the basement membrane collagen type IV. Proliferation of the microvascular endothelial cells was compared with that of bovine aortic endothelial cells and bovine aortic smooth muscle cells on the same matrices. All three cell types grew significantly more rapidly on laminin than on fibronectin. The aortic endothelial cells differed from their microvascular counterparts in that the growth of these large vessel endothelial cells on the collagenous matrices (types I and III, or type IV) was not significantly different from that observed for laminin, but was greater than the relatively slow growth seen on plasma fibronectin. Further comparison of the growth of the microvascular endothelial cells on the two basement membrane components, laminin and type IV collagen, demonstrated that the growth of these cells on laminin can be modulated by the presence of type IV collagen. This was true either if the two matrices were combined as a mixed layer, or if the laminin was specifically bound to a layer of type IV collagen, more closely simulating the distribution of these molecules in a basement membrane. Examination by immunoperoxidase of in vivo model of neovascularization in the murine cornea revealed a temporally staggered appearance of basement membrane components. The appearance of laminin was found to occur throughout the newly formed vessels, as well as in individual cells at the migrating, proliferating tips. In contrast, the appearance of type IV collagen correlated with lumen formation and was not detected at the vessel tips. The results of this study suggest that the temporally ordered synthesis of specific matrix components plays a significant role in orchestrating the growth and differentiation of endothelial cells during the highly integrated set of responses known as angiogenesis.

A possible mechanism for inhibition of angiogenesis by angiostatic steroids: induction of capillary basement membrane dissolution

A new class of angiostatic steroids acts independently of previously identified steroid functions to inhibit angiogenesis when administered with heparin. Development of angiostatic steroids as therapeutic modulators of blood vessel growth would be greatly facilitated if their mode of action were thoroughly understood. However, the mechanism by which these steroids produce capillary regression is not known. The distributions of fibronectin and laminin were studied in growing and regressing capillaries by immunofluorescence microscopy to determine whether capillary basement membrane (BM) alterations could be involved in the mechanism of antiangiogenesis. In normal 8-day-old chick chorioallantoic membrane, fibronectin and laminin appeared in continuous linear patterns within BM surrounding growing capillaries. In contrast, chorioallantoic membranes treated with combinations of angiostatic steroid and heparin exhibited capillary BM fragmentation and eventually complete loss of fibronectin and laminin from regions of capillary involution. Capillary BM breakdown correlated with capillary retraction, endothelial cell rounding, and associated capillary regression. BM surrounding large vessels, neighboring epithelium, and nongrowing capillaries were not affected. Capillary BM dissolution is the first biochemical action identified for this new class of antiangiogenic steroids.

Mechanisms of cytoskeletal regulation: modulation of aortic endothelial cell protein band 4.1 by the extracellular matrix

The bovine aortic endothelial cell (BAEC) cytoskeleton is a complex structure modulated by many stimuli including release from contact inhibition and various components of the extracellular matrix (ECM). Transduction of information from the ECM to the cell nucleus proceeds via several complex pathways including the cytoskeleton. We have demonstrated the presence of an immunoreactive isoform of the human erythrocyte cytoskeletal protein band 4.1 (4.1) in BAEC. BAEC 4.1 is similar in molecular weight to the erythroid protein by immunoblot analyses and produces a similar pattern of cysteine specific cleavage products consistent with a cluster of cysteine residues previously described in the erythroid molecule. We have also examined the effects of defined ECM proteins on the distributions of cultured BAEC 4.1 and actin filaments (AF) at confluency and following release from contact inhibition. The distribution of 4.1 in BAEC on a plasma fibronectin substrate is complex, having partial codistribution with cytoplasmic AF and a unique perinuclear staining. In contrast, on a collagen type I/III substrate, 4.1 is localized, in part, to peripheral areas of cell-cell contact distinct from the dense peripheral band staining of AF. During migration on this substrate, 4.1 had a filamentous distribution having partial codistribution with AF. Indirect immunofluorescence staining of cross-sections of bovine calf aortae revealed a cortical staining pattern in the aortic endothelial cells with staining noted on the luminal and basolateral aspects of the cells. These data suggest that, in endothelial cells, protein 4.1 is a cortical membrane protein which may function to link actin filaments to other skeletal proteins such as spectrin. These findings also suggest an active role for protein 4.1 in cytoskeletal reorganization events which can occur in response to external stimuli, such as the extracellular matrix or contact with other cells.

Collagen types I, III, and V in human embryonic and fetal skin

The dermis of human skin develops embryonically from lateral plate mesoderm and is established in an adult-like pattern by the end of the first trimester of gestation. In this study the structure, biochemistry, and immunocytochemistry of collagenous matrix in embryonic and fetal dermis during the period of 5 to 26 weeks of gestation was investigated. The dermis at five weeks contains fine, individual collagen fibrils draped over the surfaces of mesenchymal cells. With increasing age, collagen matrix increases in abundance in the extracellular space. The size of fibril diameters increases, and greater numbers of fibrils associate into fiber bundles. By 15 weeks, papillary and reticular regions are recognized. Larger-diameter fibrils, larger fibers, denser accumulations of collagen, and fewer cells distinguish the deeper reticular region from the finer, more cellular papillary region located beneath the epidermis. The distribution of collagen types I, III, and V were studied at the light microscope level by immunoperoxidase staining and at the ultrastructural level by transmission (TEM) and scanning electron microscopy (SEM) with immunogold labeling. By immunoperoxidase, types I and III were found to be evenly distributed, regardless of fetal age, throughout the dermal and subdermal connective tissue with an intensification of staining at the dermal-epidermal junction (DEJ). Staining for types III and V collagen was concentrated around blood vessels. Type V collagen was also localized in basal and periderm cells of the epidermis. By immuno-SEM, types I and III were found associated with collagen fibrils, and type V was localized to dermal cell surfaces and to a more limited extent with fibrils. The results of biochemical analyses for relative amounts of types I, III, and V collagen in fetal skin extracts were consistent with immunoperoxidase data. Type I collagen was 70-75%, type III collagen was 18-21%, and type V was 6-8% of the total of these collagens at all gestational ages tested, compared to 85-90% type I, 8-11% type III, and 2-4% type V in adult skin. The enrichment of both types III and V collagen in fetal skin may reflect in part the proportion of vessel- and nerve-associated collagen versus dermal fibrillar collagen. The accumulation of dermal fibrillar collagen with increasing age would enhance the estimated proportion of type I collagen, even though the ratios of type III to I in dermal collagen fibrils may be similar at all ages.

Repopulation of a human alveolar matrix by adult rat type II pneumocytes in vitro. A novel system for type II pneumocyte culture

This paper describes the preparation of lung acellular alveolar matrix fragments and culture of rat type II pneumocytes directly on the alveolar epithelial basement membrane, thereby permitting study of the effect of lung basement membrane on the morphology and function of type II cells. Collagen types I, III, IV and V, laminin and fibronectin were located by immunofluorescence in the lung matrix with the same patterns as those described for the normal human lung. Transmission electron microscopy (TEM) of the fragments revealed intact epithelial and endothelial basement membranes. The matrix maintained the normal three-dimensional alveolar architecture. Glycosaminoglycans were still present by Alcian Blue staining. Isolated adult rat type II pneumocytes cultured on 150 micron thick fragments of acellular human alveolar extracellular matrix undergo gradual cytoplasmic flattening, with loss of lamellar bodies, mitochondria, and surface microvilli. These changes are similar to the in vivo differentiation of type II pneumocytes into type I pneumocytes. The type II pneumocyte behaviour on the lung epithelial basement membrane contrasted sharply with that of the same cell type cultured on a human amnionic basement membrane. On the latter surface the cells retained their cuboidal shape, lamellar bodies and surface microvilli for up to 8 days. These observations suggest that the basement membranes from different organ systems exert differing influences on the morphology and function of type II pneumocytes and that the alveolar and amnionic basement membranes may have differing three-dimensional organizations. The technique of direct culture of type II cells on the lung basement membrane provides a useful tool for studying the modulating effect of the basement membrane on alveolar epithelial cells.

Demonstration of cutaneous doxorubicin extravasation by rhodamine-filtered fluorescence microscopy

Accidental extravasation of the widely used antineoplastic anthracycline doxorubicin (adriamycin) is capable of causing serious and on-going skin and soft tissue necrosis leading to loss of limb function. The only effective cure has been the complete surgical excision of drug-contaminated tissue. Surgical excisions preparatory to skin grafting have often underestimated the extent of drug infiltration into tissue. Using an experimental technique utilizing the drug's natural property of red fluorescence, we were able to effect complete surgical excision by rhodamine fluorescence microscopy in frozen sections of resected skin.

Basement membrane as a spatial organizer of polarized epithelia. Exogenous basement membrane reorients pancreatic epithelial tumor cells in vitro

The authors have previously described a transplantable pancreatic acinar carcinoma that comprises cells which reorganize and display normal epithelial orientation only when in contact with basement membrane (BM) in vivo. In the present study, they investigated whether exogenous acellular BM or collagenous stroma (prepared from human amnion) was sufficient to reorient pancreatic acinar tumor cells in vitro. Mechanically dispersed tumor cells could not spontaneously attach to standard culture substrata; yet they adhered to exogenous intact BM or to dishes coated with purified laminin or Type IV collagen. Cell contact with amniotic BM resulted in tumor cell shape changes, assembly of intracellular actin into fibrous bundles, and restoration of normal epithelial cell-cell interactions. Computerized morphometry confirmed that tumor cells exhibited a normal polarized distribution of lipid droplets, nuclei, Golgi complexes, and zymogen granules (from base to apex) within 6 hours of culture on BM. Adhering zonules and microvilli were observed only along apical tumor cell surfaces, although full junctional complexes and distinct membrane domains did not reform. Similar attachment, cytoskeletal alterations, and reorientation occurred in the absence of protein synthesis (25 micrograms/ml cycloheximide). In contrast, tumor cells that were maintained on amniotic stroma remained round, displayed circumferential rings of actin, and appeared randomly oriented. Thus, BM may normally serve to integrate and maintain individual cells within a polarized epithelium.

Neoplastic disorganization of pancreatic epithelial cell-cell relations. Role of basement membrane

The authors have analyzed the structural relations of a nonmetastatic rat pancreatic acinar carcinoma and contrasted them with those of normal exocrine pancreas in order to better define the role of basement membrane (BM) in early stages of neoplastic disorganization. These studies showed that normal acinar cells rested on continuous BM (containing laminin, heparan sulfate proteoglycan, and Type IV and V collagens) and displayed a polarized distribution of intracellular organelles, cytoskeletal assemblies (concentration of actin within terminal web), and distinct membrane domains (apical leucine aminopeptidase). In contrast, the parenchyma of the pancreatic acinar carcinoma was free of all BM components except for a discontinuous array of laminin. In these regions, acinar tumor cells appeared randomly oriented, displayed actin in uniform cortical distributions, and lost membrane polarity. However, when tumor cells contacted mesenchymally derived connective tissue along tumor capsule and vascular adventitia, they accumulated intact BM and reoriented in a manner reminiscent of normal pancreas. Tumor cell reorganization was observed in the absence of formation of full junctional complexes or normally polarized membrane domains, although leucine aminopeptidase appeared to be excluded from regions of tumor cell surfaces that were in direct contact with BM. The loss of normal epithelial cell-cell arrangements that is the hallmark of early stages of tumor formation could therefore result from failure to match increases in cell number with commensurate BM extension.

Immunolocalization of type IV collagen and laminin in nonbasement membrane structures of murine corneal stroma. A light and electron microscopic study

The extracellular matrix of adult vertebrate corneal stroma is composed primarily of the interstitial collagen type I and smaller amounts of types III and V collagen. These collagens are organized into overlapping lamellae of striated filaments. In addition to these lamellar structures, the corneal stroma also contains 100- to 250-nm bundles of nonstriated 8- to 11-nm microfibrils. By immunofluorescent localization and electron microscopic immunolocalization, these microfibril bundles in the mouse are associated with type III collagen, type IV collagen, and laminin. By immunologic and histochemical criteria, these bundles do not contain either type I collagen, type V collagen, elastin, or oxytalan microfibrils. The cellular source, composition, and possible functions of these microfibril bundles are discussed.

Congenital polycystic tumor of the atrioventricular node. Autopsy study with immunohistochemical findings suggesting endodermal derivation

Atrioventricular node congenital polycystic tumor, a lesion that causes heart block and sudden unexpected death, has been postulated to arise from the lymphatics, foregut endoderm, cardiac mesoderm, and mesothelium since its initial description in 1911. No theory concerning the histogenesis of this tumor has been completely satisfactory, either from an embryologic or a clinical standpoint. Biologic studies including cell culture systems from the lesions as well as optimal ultrastructural examination have been compromised due to postmortem autolysis. We studied four examples of this entity by standard histochemical stains for mucopolysaccharides, and by modified immunohistochemical labeling for carcinoembryonic antigen, factor VIII, and keratin. Results suggest an origin from foregut endoderm. The lesion has not been reported to be outside the region of the atrioventricular junction, but was demonstrated in the thymus gland in one of our cases.