Microarray analysis of homocysteine-responsive genes in cardiac neural crest cells in vitro.

The amino acid homocysteine increases in the serum when there is insufficient folic acid or vitamin B(12), or with certain mutations in enzymes important in methionine metabolism. Elevated homocysteine is related to increased risk for cardiovascular and other diseases in adults and elevated maternal homocysteine increases the risk for certain congenital defects, especially those that result from abnormal development of the neural crest and neural tube. Experiments with the avian embryo model have shown that elevated homocysteine perturbs neural crest/neural tube migration in vitro and in vivo. Whereas there have been numerous studies of homocysteine-induced changes in gene expression in adult cells, there is no previous report of a homocysteine-responsive transcriptome in the embryonic neural crest. We treated neural crest cells in vitro with exogenous homocysteine in a protocol that induces significant changes in neural crest cell migration. We used microarray analysis and expression profiling to identify 65 transcripts of genes of known function that were altered by homocysteine. The largest set of effected genes (19) included those with a role in cell migration and adhesion. Other major groups were genes involved in metabolism (13); DNA/RNA interaction (11); cell proliferation/apoptosis (10); and transporter/receptor (6). Although the genes identified in this experiment were consistent with prior observations of the effect of homocysteine upon neural crest cell function, none had been identified previously as response to homocysteine in adult cells.

Genes, folate and homocysteine in embryonic development.

Population-based studies of human pregnancies show that periconceptional folate supplementation has a significant protective effect for embryos during early development, resulting in a significant reduction in developmental defects of the face, the neural tube, and the cono-truncal region of the heart. These results have been supported by experiments with animal models. An obvious quality held in common by these three anatomical regions is that the normal development of each region depends on a set of multi-potent cells that originate in the mid-dorsal region of the neural epithelium. However, the reason for the sensitive dependence of these particular cells on folic acid for normal development has not been obvious, and there is no consensus about the biological basis of the dramatic rescue with periconceptional folate supplementation. There are two principal hypotheses for the impact of folate insufficiency on development; each of these hypotheses has a micronutrient component and a genetic component. In the first hypothesis the effect of low folate is direct, limiting the availability of folic acid to cells within the embryo itself; thus compromising normal function and limiting proliferation. The second hypothetical effect is indirect; low folate disrupts methionine metabolism; homocysteine increases in the maternal serum; homocysteine induces abnormal development by inhibiting the function of N-methyl-D-aspartate (NMDA) receptors in the neural epithelium. There are three general families of genes whose level of expression may need to be considered in the context of these two related hypotheses: folate-receptor genes; genes that regulate methionine-homocysteine metabolism; NMDA-receptor genes.

Partial cloning and sequencing of chick fibrillin-1 cDNA.

The recent identification of numerous matrix genes and gene products has allowed a detailed examination of their roles in development. Two of these extracellular matrix proteins, fibrillin-1 and fibrillin-2, are components of the elastin-associated microfibrils. Given what is known about the distribution of the fibrillins in normal tissues and the abnormalities that result when mutations occur, a basic hypothesis has emerged: fibrillin-1 is primarily responsible for load bearing and providing structural integrity, whereas fibrillin-2 may be a director of elastogenesis. Nevertheless, examination of phenotypes in disorders caused by mutations in fibrillin-1 or fibrillin-2 suggests some common functions. To better understand these similar and diverse roles, it would be helpful to examine these proteins during chick development. To accomplish this goal, it is first necessary to characterize the chick homologs of the known fibrillins. In this study, the partial chick FBN1 cDNA was identified by polymerase chain reaction-aided cloning as a first step toward elucidating these goals. Sequence analysis indicated that there is striking conservation between chick and mammalian fibrillin-1 at the DNA and protein levels. Antisense and sense riboprobes were synthesized and used in in situ hybridization in stage 14 chick embryos and high levels of FBN1 transcripts were observed in the heart.

N-methyl-D-aspartate receptor agonists modulate homocysteine-induced developmental abnormalities.

We showed previously that the induction of neural crest (NC) and neural tube (NT) defects is a general property of N-methyl-D-aspartate receptor (NMDAR) antagonists. Since homocysteine induces NC and NT defects and can also act as an NMDAR antagonist, we hypothesized that the mechanism of homocysteine-induced developmental defects is mediated by competitive inhibition of the NMDAR by homocysteine. If this hypothesis is correct, homocysteine-induced defects will be reduced by NMDAR agonists. To test the hypothesis, we treated chicken embryos during the process of neural tube closure with sufficient homocysteine thiolactone to induce NC and NT defects in approximately 40% of survivors or with homocysteine thiolactone in combination with each of a selected set of NMDAR agonists in 0. 05-5000 nmol doses. Glutamate site agonists selected were L-glutamate and N-methyl-D-aspartate. Glycine site agonists were glycine, D-cycloserine, and aminocyclopropane-carboxylic acid. Glycine was the most effective overall, reducing defects significantly at two different doses (each P>0.001). These results support the hypothesis that homocysteine may affect NC and NT development by its ability to inhibit the NMDAR. One potentially important consequence of this putative mechanism is that homocysteine may interact synergistically with other NMDAR antagonists to enhance its effect on development.

ERK2 activation by homocysteine in vascular smooth muscle cells.

Homocysteine at abnormally high levels is an independent risk factor for atherosclerosis and may be a key factor in atherogenesis. Since homocysteine (Hcys) has been shown to promote cell proliferation and induction of the gene transcription factor c-fos in vascular smooth muscle cells (VSMCs), effects which can be mediated by MAP kinase, we hypothesized that homocysteine activates a MAP kinase-dependent signal transduction pathway. In this study, we find that homocysteine transiently activates MAP kinase (ERK2 isoform) in cultured VSMCs from chick embryos. Homocysteine activation of ERK2 is dose-dependent with an EC50 of approximately 500 nM and blocked by the MAP/Erk kinase (MEK) inhibitor PD98059. VSMC embryonic lineage is another determinant of homocysteine sensitivity. These findings demonstrate that homocysteine activates the MAP kinase signal transduction pathway and thus support the hypothesis that homocysteine may promote atherosclerosis by stimulation of growth promoting signal transduction pathways.

Dextromethorphan and other N-methyl-D-aspartate receptor antagonists are teratogenic in the avian embryo model.

N-Methyl-D-aspartate (NMDA) receptors are a calcium-conducting class of excitatory amino acid receptors that are involved in neuronal development and migration. Certain well known teratogens (e.g. homocysteine, ethanol, and chloroform) that induce congenital neural tube and neural crest defects also have the capacity to act as NMDA receptor antagonists. We hypothesized that teratogenicity was a general property of NMDA receptor antagonists, and that high affinity NMDA receptor antagonists would induce neural tube and neural crest defects. Chicken embryos were given 5, 50, or 500 nmol/d of selected NMDA receptor antagonists for 3 consecutive days during the process of neural tube closure, beginning 4 h after the beginning of incubation. Selected NMDA receptor antagonists represented three classes of antagonists: ion channel blockers, glycine site antagonists, and glutamate site agonists and antagonists. All classes of NMDA receptor antagonists induced embryonic death and congenital defects of the neural crest and neural tube; however, the channel blockers were the most potent teratogens. Dextromethorphan at 500 nmol/embryo/d killed more than half the embryos and induced congenital defects in about one-eighth of the survivors; dextromethorphan was also highly lethal at 50 nmol/embryo/d. Glutamate site NMDA receptor agonists (NMDA and homoquinolinic acid) displayed weak toxicity relative to their known NMDA receptor potency. Taken together, these data indicate that NMDA receptor antagonists, particularly channel blockers, are potent teratogens in the chicken embryo model. Because dextromethorphan is a widely used nonprescription antitussive, its strong teratogeneticity using this model is particularly noteworthy.

Homocysteine signal cascade: production of phospholipids, activation of protein kinase C, and the induction of c-fos and c-myb in smooth muscle cells.

Hyperhomocysteinemia has been recognized as an independent risk factor for cerebral, coronary, and peripheral atherosclerosis. To examine the contribution of homocysteine (H[cys]) in the pathogenesis of vascular diseases, we sought to determine whether the H[cys] effect on vascular smooth muscle (VSMC) proliferation is mediated by a specific receptor/transporter or is due to an interaction with growth factors or cytokines. We show that H[cys] induced c-fos and c-myb and increased DNA synthesis and cell proliferation 12-fold in neural crest-derived VSMC (N-VSMC). The H[cys] effect on N-VSMC proliferation is inhibited by Mk-801, a noncompetitive antagonist of the N-methyl-D-aspartate (NMDA) receptor, a glutamate-gated calcium ion channel receptor, and CGS 19755, a competitive antagonist of NMDA-type glutamate receptor. H[cys] stimulates the synthesis of mass amounts of sn-1,2 diacylglycerol, and activates protein kinase C translocation from the nucleus and cytoplasm to cell membranes. Furthermore, protein kinase C inhibitors block the growth effect mediated by H[cys]. These findings indicate that H[cys]-mediated responses are coupled to diacylglycerol-dependent protein kinase C activation. Our results suggest that homocysteine activates a receptor/transporter-like factor in neural crest derived smooth muscle.

Differential response of mesoderm- and neural crest-derived smooth muscle to TGF-beta1: regulation of c-myb and alpha1 (I) procollagen genes.

Previously, we demonstrated that avian vascular smooth muscle cells (VSMC) derived from embryonic abdominal and thoracic aorta grow differently in the presence of transforming growth factor beta (TGF-beta1) and platelet-derived growth factor (PDGF-BB) (Wrenn et al., In Vitro Cell. Dev. Biol. 29, 73-78, 1992). The thoracic VSMC (N-VSMC) are derived from neural crest, and therefore differentiate from ectoderm; the abdominal VSMC (M-VSMC) are derived from mesoderm. The present study was designed to identify factors that mediate the differential responses of the VSMC to TGF-beta1. We found that TGF-beta1 increased DNA synthesis by approximately sevenfold in N-VSMC. Levels of both alpha1 (I) procollagen and c-myb mRNAs were markedly induced in N-VSMC treated with TGF-beta1. Chimeric plasmids containing up to 3.5 kb of alpha1 (I) procollagen 5' flanking DNA were induced to equivalent levels as procollagen mRNA in N-VSMC. However, TGF-beta1 increased DNA synthesis by threefold in M-VSMC; there was no effect on alpha1 (I) procollagen expression, and c-myb was not expressed, as demonstrated by immunohistochemistry staining and RNA analyses. Antisense c-myb oligodeoxynucleotides blocked the TGF-beta1 induction of alpha1 (I) procollagen and the growth of N-VSMC. The increase in DNA synthesis by M- and N-VSMC was correlated with the secretion of PDGF-AA, and staurosporine and antibodies directed against PDGF-AA suppressed DNA synthesis. Our results demonstrate that TGF-beta1 activity and c-myb expression modulate the expression of alpha1 (I) collagen and cell proliferation in neural crest-derived smooth muscle. The regulation of these events by TGF-beta1 may be important during morphogenesis of blood vessels and vascular diseases.

Homocysteine induces congenital defects of the heart and neural tube: effect of folic acid.

The biological basis or mechanism whereby folate supplementation protects against heart and neural tube defect is unknown. It has been hypothesized that the amino acid homocysteine may be the teratogenic agent, since serum homocysteine increases in folate depletion; however, this hypothesis has not been tested. In this study, avian embryos were treated directly with D,L-homocysteine or with L-homocysteine thiolactone, and a dose response was established. Of embryos treated with 50 microliters of the teratogenic dose (200 mM D,L-homocysteine or 100 mM L-homocysteine thiolactone) on incubation days 0, 1, and 2 and harvested at 53 h (stage 14), 27% showed neural tube defects. To determine the effect of the teratogenic dose on the process of heart septation, embryos were treated during incubation days 2, 3, and 4; then they were harvested at day 9 following the completion of septation. Of surviving embryos, 23% showed ventricular septal defects, and 11% showed neural tube defects. A high percentage of the day 9 embryos also showed a ventral closure defect. The teratogenic dose was shown to raise serum homocysteine to over 150 nmol/ml, compared with a normal level of about 10 nmol/ml. Folate supplementation kept the rise in serum homocysteine to approximately 45 nmol/ml, and prevented the teratogenic effect. These results support the hypothesis that homocysteine per se causes dysmorphogenesis of the heart and neural tube, as well as of the ventral wall.

Embryonic lineage of vascular smooth muscle cells determines responses to collagen matrices and integrin receptor expression.

Developmental studies have demonstrated that the vascular smooth muscle cells (VSMC) present within the elastic arteries are differentiated from two definitive origins, the neural crest and the mesoderm. Cells from these distinct progenitors differ in their ability to determine long-range spatial order of the extracellular matrix, in proliferative responses, and in the expression of critical proteins. The present study utilizes collagen gel contraction assays and the analysis of integrin receptor subunit expression to evaluate cell-matrix interactions. In the presence of serum and transforming growth factor-beta 1 (TGF) or TGF-beta 1 alone, VSMC isolated from the abdominal aorta (AA-VSMC) were found to contract collagen matrices to a significantly greater extent than VSMC from the thoracic aorta (TA-VSMC). However, in TA-VSMC, beta 1 integrin and gel contraction were stimulated only in the presence of serum factors. Metabolic labeling and immunoprecipitation of integrin subunits revealed that TGF-beta 1 induced beta 1 and alpha 5 integrin subunits in AA-VSMC four-and ninefold, respectively. AA-VSMC gel contraction stimulated by serum and TGF-beta 1 alone was inhibited with anti-beta 1 integrin antibody by 70 and 100%, respectively. However, the beta 1 integrin-specific antibody inhibited serum-induced TA-VSMC gel contraction by 25%. The data suggest that vascular smooth muscle cell ontogeny is an important determinant of cell function, phenotype, and response to growth factors such as TGF-beta 1.

Expression of collagens and decorin during aortic arch artery development: implications for matrix pattern formation.

The elastic matrix of the large arteries shows a high level of spatial order. However, the mechanisms by which such order is established and maintained are largely unknown. The embryonic development of the avian heart and great vessels provides an appropriate model to investigate these mechanisms. In control embryos, an elastic matrix with a high level of spatial order develops in the nascent great vessels. But after the normal vascular smooth muscle (VSM) progenitor cells in the great vessels are experimentally replaced by other VSM progenitor cells, the elastic extracellular matrix is congenitally disordered. The present study used this model to test the hypothesis that the proteoglycan decorin was involved in the establishment and maintenance of the normal three-dimensional spatial order of the vascular elastic matrix. The temporospatial expression of decorin was analysed during development of normal vessels and in experimental vessels with surrogate VSM. The results showed the following: (1) the expression of decorin was related in time and space to the establishment of large helical collagen type III fibers that are characteristic of the normal elastic extracellular matrix; (2) in the experimental extracellular matrix there were few helical fibers of collagen type III, but those that were present remained positive for decorin; and (3) in both control and experimental vessels, decorin associated with neither fibers of collagen type I nor fibers of collagen type III in any conformation other than the large helical fibers. These data indicate a previously unrecognized relationship between decorin and the spatial order of the physiologically significant helical fibers of collagen type III.

Expression of elastin, smooth muscle alpha-actin, and c-jun as a function of the embryonic lineage of vascular smooth muscle cells.

In the avian embryo, vascular smooth muscle cells (VSMC) in the aortic arch (elastic) arteries originate in the neural crest, whereas other VSMC develop from local mesoderm. These two lineages have been shown previously to be significantly different in the timing and expression of the smooth muscle phenotype and in their respective abilities to produce an orderly elastic matrix. Two differing kinds of VSMC also have been shown in mammals. In the experimental absence of neural crest (NC) in the avian embryo, the matrix is spatially disordered. The molecular basis of the difference between the normal NC-VSMC and the surrogate mesodermal (MDM)-VSMC has not previously been investigated. In this study the expression of vascular smooth muscle alpha-actin, tropoelastin, c-fos and c-jun were examined via immunoblotting, immunohistochemistry, Northern blot, and/or transcription run-on assays. Control avian VSMC of NC origin were compared with experimental MDM-derived VSMC that populate the cardiac outflow after surgical ablation of the NC. The results show that, when they are grown under identical conditions in vitro or freshly removed from an embryonic vessel, surrogate MDM-VSMC express about 10 times more alpha-actin and tropoelastin than the normal NC-VSMC; and MDM-VSMC express up to 15 times more c-jun, whereas c-fos was not different. These results show profound heterogeneity in the regulation of VSMC-specific genes that is based in the embryonic lineage of the cells.

Transforming growth factor-beta: signal transduction via protein kinase C in cultured embryonic vascular smooth muscle cells.

Transforming growth factor-beta (TGF-beta), an ubiquitous regulatory peptide, has diverse effects on the differentiation and behavior of vascular smooth muscle cells (VSMC). However, the molecular mechanism through which TGF-alpha exerts its effects remains obscure. We investigated the phosphoinositide/protein kinase C [PKC] signaling pathway in the action of TGF-beta on cultured embryonic avian VSMC of differing lineage: a) thoracic aorta, derived from the neural crest; and b) abdominal aorta, derived from mesenchyme. The second messenger responsible for activation of PKC is sn-1,2-diacylglycerol [DAG]; TGF-beta increased the mass amounts of DAG in the membranes of neural crest-derived VSMC concurrent with translocation of PKC from the soluble to the membrane fraction, but TGF-beta had no effect on the DAG or PKC of mesenchyme-derived VSMC. TGF-beta potentiated the growth of platelet-derived growth factor (PDGF)-treated, neural crest-derived VSMC; but abolished PDGF-induced growth of mesenchymal cells. It is concluded that molecular and functional responses of VSMC to TGF-beta are heterogeneous and are functions of the embryonic lineage of the VSMC.

Coronary artery development in the chick: origin and deployment of smooth muscle cells, and the effects of neural crest ablation.

Previous studies of coronary artery ontogeny have stressed early development and therefore have dwelt mainly upon the origin of the endothelium of the nascent coronary artery stem. This study has analyzed the ontogeny of the vascular smooth muscle cells (VSMC) in the coronary arteries of the domestic chicken, by establishing the timing and deployment of smooth muscle alpha-actin (SMAA). Anti-SMAA was applied to sections of normal embryos, and to sections of experimental embryos that had undergone surgical ablation of the neural crest over somites 1-3. The results show an orderly symmetrical deployment of SMAA in control coronary arteries. SMAA was expressed significantly earlier in the coronary artery VSMC compared with those of the cardiac outflow vessels; this early expression may indicate a unique responsiveness to induction of the smooth muscle phenotype. The normal orderly development of coronary artery VSMC was dependent upon the presence of the neural crest, and therefore was disrupted in the experimental embryos whose neural crest was ablated.

Influence of vascular smooth muscle heterogeneity on angiotensin converting enzyme activity in chicken embryonic aorta and in endothelial cells in culture.

The smooth muscle of the abdominal region of the chicken aorta derives from locally recruited mesenchyme (mesenchymal smooth muscle), whereas that of the thoracic region derives from the neural crest (ectomesenchymal smooth muscle). We hypothesized that this smooth muscle heterogeneity might affect important enzymatic functions of the vessel wall. Therefore, we measured angiotensin converting enzyme (ACE) activity in homogenates of chicken thoracic and abdominal aorta at different embryonic stages (days 10, 14, and 18 of gestation). ACE activity increased in both regions over the time of gestation (p less than 0.001 in both cases); the increase was steeper and ACE activity was higher in thoracic than in abdominal segments (p less than 0.001). Km values were similar (approximately 7 microM) at all times and between the two segments, whereas changes in Vmax values closely paralleled those in enzyme activity, indicating gestation-dependent increases in the amount of enzyme. Neural crest ablation at an early developmental stage resulted in an increase of ACE activity in thoracic homogenates (p less than 0.001), predictably leaving that in abdominal homogenates unaffected. Bovine pulmonary artery endothelial cell monolayers exposed to media conditioned with cultured mesenchymal or ectomesenchymal smooth muscle cells exhibited elevated ACE activity (46% and 83%, respectively, relative to control medium, with p less than 0.01 in both cases; p less than 0.05 between the two media). Increases in endothelial cell ACE activity corresponded to proportional increases in ACE protein determined by enzyme-linked immunosorbent assay (r = 0.99) and were interpreted as indicative of enhanced enzyme synthesis subsequent to exposure of endothelial cells to smooth muscle-conditioned media.(ABSTRACT TRUNCATED AT 250 WORDS)

Spatial disorder of collagens in the great vessels, associated with congenital heart defects.

Surgical ablation of the cardiac neural crest from the chicken embryo results in persistent truncus arteriosus (PTA) and a change in the elastic laminae of the great vessels, wherein elastin and the elastin microfibril show significant spatial disorder. The purpose of this study was to test the hypothesis that the interstitial collagens would also be disordered in the elastic laminae of chicken embryos with PTA. The birefringence characteristics of interstitial collagen were examined to evaluate spatial ordering. The results showed that collagen in the elastic laminae assumed an orderly configuration of well-defined fiber bundles in the great vessel walls of control embryos, whereas vessels from embryos with PTA lacked any distinct spatial order. Collagens type I and III were localized in the vessel walls. Type III collagen was the principal collagen of the elastic laminae, but was absent from the intima of all vessels. In the elastic laminae of vessels from control embryos, collagen type III showed well-defined fiber bundles whereas embryos with PTA had diffuse collagen type III in poorly defined laminae that were not separated by discrete layers of smooth muscle cells. Collagen type I was a minor component of the elastic laminae but formed robust pericellular fiber bundles throughout the media and intima. Collagen type I fibers appeared to be coarsened and less uniform in the vessels from embryos with PTA.

Development of the musculoelastic septation complex in the avian truncus arteriosus.

It is now well established that cells from the cardiac neural crest (CNC) are essential for normal conotruncal septation. The truncal septation complex consists of the aorticopulmonary (AP) septum and the myocardial sheath of the truncus. The principal role of the CNC cells during septation appears to be their differentiation into the elastogenic smooth muscle that forms the AP septum proper. The objective of this study was to integrate serial reconstruction and specific histochemical markers in order to provide a unified analysis of the relationships between the CNC and the other components of the truncal septation complex. The development of the septation complex was compared normal embryos vs. embryos from which the CNC had been surgically ablated. Embryos from each group were harvested after incubation periods of 4-8 days (Hamburger-Hamilton stages 23-34). Histochemical procedures were performed for positive identification of the elastic matrix and smooth muscle alpha-actin; the presence of these proteins was used as the criterion for "septal cells" and to define the boundaries of the septum. The results indicate that the shape, components, boundaries, and degree of organization of the septation complex may be different from previous descriptions. Furthermore, all of the components of the truncal septation complex are dysgenic in the absence of the CNC. Of special significance in the absence of CNC. Of special significance in the absence of CNC are: 1) the failure of the myocardial sheath to retract; 2) the apparently random distribution of surrogate ectomesenchyme; and 3) the impairment of truncal elastogenesis. These results indicate that the cells of neural crest origin interact with the surrounding mesenchyme during septation and that the entire septation complex depends upon the presence of the neural crest cells for normal development.

Impaired elastic matrix development in the great arteries after ablation of the cardiac neural crest.

The cells that form the aorticopulmonary septum in the avian embryo have been shown to be similar to the cells that form the walls of the great vessels in two ways: both are derived from the cardiac neural crest and both are able to synthesize an elastogenic matrix in the early embryo. Because of these similarities, and because ablation of the cardiac neural crest causes congenital defects of the outflow tract that are related to failure of proper septation, it was hypothesized that such an ablation also would cause the walls of the great vessels to be defective. The purpose of this study was to compare the elastic matrix in the mediae of the great vessels of normal embryos with those from which the cardiac neural crest had been ablated. The results show that the elastic matrix in the great vessels of the experimental embryos was impaired 1) in the rate of downstream propagation of the initiation of elastogenesis among younger embryos, incubation days 4-8 and 2) in the spatial configuration of the elastic matrix among the older embryos, incubation days 16-20. These results may provide a biological explanation for the elastin defect that affects the pulmonary artery of many patients with cyanotic congenital heart defects.

Smooth muscle cells of neural crest origin form the aorticopulmonary septum in the avian embryo.

Previous studies have shown that the cells of the aorticopulmonary (AP) septum are similar to the smooth muscle cells of the mediae of the great vessels in their common origin from the cardiac neural crest and in their common expression of an elastic extracellular matrix. The purpose of this study was to test the cells of the AP septum for the presence of certain cytoplasmic proteins, especially smooth muscle alpha-actin (SMAA) whose presence is definitive of smooth muscle. A monoclonal antibody against SMAA was applied to normal chicken embryos at 3.5-8 days of incubation and to age-matched embryos from which the cardiac neural crest had been ablated surgically. Antibodies against the intermediate filaments desmin, cytokeratin, and vimentin also were applied. The results showed that the AP septal cells expressed SMAA during the process of septation, days 5-8; but when the cardiac neural crest was ablated and septation was defective, no cells in the conotruncal connective tissue expressed SMAA. None of the intermediate filament proteins were detected in the septum. These results indicate that the AP septal cells are smooth muscle and therefore may be hypothesized to have an active role in septation.