INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE MAINTENANCE and regulation of the umbilical-placental circulation are crucial for fetal growth and well-being. In fetal sheep this vascular bed is sensitive to several vasoconstrictors, especially ANG II (26, 27, 40, 48). This responsiveness also has been observed in human umbilical arteries studied in vitro (8, 46, 47). In intact fetal sheep the umbilical circulation is more sensitive to ANG II than the systemic vasculature (26, 27, 40, 48). However, the mechanisms responsible for these differences in vascular reactivity are unclear. One explanation is that the contractile protein composition and therefore the contractile capacity of vascular smooth muscle (VSM) in umbilical and systemic arteries differ.

In the adult, systemic VSM modulates arterial pressure and mediates responses to numerous stimuli. These differentiated VSM cells express a unique assortment of contractile and structural proteins, reflecting a mature "contractile" phenotype (37, 38). In these cells contraction results from the interaction of actin in the thin filaments and myosin in the thick filaments. Myosin molecules are composed of six subunits: two heavy chains, two regulatory, and two essential light chains. The carboxy-terminal ends of the myosin heavy chains (MHC) form a coiled-coil tail that polymerizes to form thick filaments. The amino-terminal end is folded in a globular head containing binding sites for actin and Mg²⁺-ATP, as well as for each of the light chains and is the site of chemomechanical transduction associated with Mg²⁺-ATPase activity. Smooth muscle contraction is triggered by a rise in cytosolic Ca²⁺ leading to phosphorylation of the myosin regulatory light chains, which permits actin to activate myosin Mg²⁺-ATPase (23, 45). In mature VSM these components of the contractile mechanism are present and functional.

In several species, including fetal sheep, a model widely used to study cardiovascular development, the expression of contractile proteins in VSM is developmentally regulated in a tissue-specific manner (4, 11, 18, 21, 42). In fetal sheep actin and MHC contents in systemic arteries are quite low throughout the majority of gestation, increasing just before term and during the immediate postnatal period (4, 11). There also are changes in expression of the MHC isoforms at least four of which are expressed in vertebrate smooth muscle (28, 31, 34, 35). The smooth muscle MHC isoforms SM1 (204 kDa) and SM2 (200 kDa) are alternatively spliced products of a single gene and are exclusively expressed in cells of smooth muscle lineage (33). They differ in their carboxy-terminal tail sequence and exhibit different patterns of expression during VSM development, with SM2 expression increasing in mature or contractile tissues (1, 4, 11, 18, 31, 32, 38). MHC-B, also 200 kDa, is derived from a separate gene, and its expression is associated with the "synthetic" phenotype of smooth muscle cells (31). It is found in nonmuscle cells (44), smooth muscle cells in culture (39, 44), and developing fetal VSM (4, 11, 20, 28, 31). Its expression declines during development (4, 11, 31). MHC-A, also a nonmuscle isoform, is 196 kDa and is derived from a third gene (19, 28). It is primarily expressed in platelets (19) but also in some fetal tissues, where it is developmentally regulated (18, 49), and in cultured smooth muscle cells (28, 39). The relationship between the patterns of VSM protein expression and function during development is unclear; however, it was recently observed that stress generation may be attenuated when MHC-B is present and SM2 content is low (4). This pattern of protein expression may also differ among arteries in the developing fetus, but this has not been fully explored.

To the best of our knowledge, contractile phenotype and VSM contractile function in umbilical and systemic arteries have not been compared. Therefore, the purpose of the present study was to determine whether 1) differences in VSM protein expression exist in umbilical and systemic arteries from near-term fetal sheep, 2) differences in protein expression are associated with different capacities to generate stress, and 3) umbilical artery VSM resembles adult VSM and thus is highly specialized in the fetal vasculature.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Tissue preparation. Near-term fetal (n = 12, 130-145 days of gestation, term ~145 days) and maternal (n = 12) sheep were killed by rapid intravenous injection of pentobarbital sodium (50 mg/kg) to the mother via the external jugular vein. Segments of the abdominal aorta and femoral artery were quickly obtained from fetal and adult animals, as well as segments of external and intra-abdominal umbilical arteries from fetuses, and placed into iced physiological buffered solution containing (in mM) 137.0 NaCl, 2.7 KCl, 10 Na₂HPO₄, 1.76 KH₂PO₄, and 0.1% diethyl pyrocarbonate, pH 7.4. We obtained segments of both external (from the mid-third of the umbilical cord) and intra-abdominal umbilical artery because they represent a continuum of this vessel, and we (15) recently observed differences in ANG II receptor expression in these two portions of the umbilical circulation. Arteries from pregnant ewes were used because we (2, 3) previously reported that neither the protein contents in nor the stresses generated by systemic blood vessels are altered during ovine pregnancy. Endothelium and adventitia were removed from the arteries with a soft cotton swab and sharp dissection, respectively. Strips of tissue were cut, blotted dry to remove excess water and capillary blood, frozen in liquid nitrogen, and stored at 80°C until studied. Additional segments were obtained to measure stresses (see below). These studies were approved by the Institutional Review Board for Animal Research.

Protein analysis and content. SDS homogenates were prepared from 10- to 20-mg samples of frozen tissue as previously reported (3, 11). Briefly, homogenates were divided into two aliquots. One was subjected to centrifugation at 10,000 g for 2 min, and the supernatant was removed to determine the soluble or cellular protein in each sample. The other sample was not centrifuged and was used to determine the total homogenate protein. Aliquots of both samples were analyzed for protein content by bicinchoninic acid (BCA) reagent (Pierce, Rockford, IL). Aliquots of the supernatant containing bromphenol blue and 2-mercaptoethanol were subjected to SDS-PAGE, using 3-20% and 4% polyacrylamide gels to determine the contents of total actin and MHC and the relative amounts of MHC isoforms, respectively. For each tissue mini-gels were loaded with 20-40 µg of soluble protein and subjected to electrophoresis at 200 V until the dye front reached the bottom of the gel. Gels were stained with Coomassie Brilliant Blue overnight and appropriately destained to remove background staining. Stained gels were scanned, each lane in duplicate, with a laser densitometer (LKB Instruments, Stockholm, Sweden) to estimate the relative amounts of actin, MHC, and MHC isoforms. Differences between measurements were <5%, and values for each band were averaged. The fraction of stained protein accounted for by actin and MHC was converted to micrograms using the total protein quantified by BCA reagent in each sample. Values are expressed as microgram per milligram wet weight.

SDS-PAGE and immunoblotting. Using the supernatant extracts of fetal and maternal aorta and femoral artery, as well as external and internal umbilical artery, ~200 ng of total MHC were loaded and separated on 4% polyacrylamide gels. Proteins were electrophoretically transferred to nitrocellulose paper at 80 mA overnight in the presence of methanol (20%) and SDS (0.1%). Blots were then blocked for 1 h in a buffer that contained powdered milk (0.3% wt/vol), incubated for 4 h with blocking buffer containing specific antiserum against either SM2 or MHC-B (1:4,000 and 1:20,000, respectively), and then incubated with goat anti-rabbit IgG conjugated with horseradish peroxidase (1:15,000). Antibodies to MHC isoforms were generated against short peptides derived from sequences unique to the carboxy-terminal tail of each isoform as described by Chern et al. (11). Regions containing MHC isoforms were visualized by enhanced chemiluminescence (ECL, Amersham International, Little Chalfont, UK). The blots were kept in the ECL-developing solution for 1 min, exposed on film for 5 s, and then developed.

Histological evaluation. Immediately after dissection, segments of external umbilical artery and fetal and adult femoral artery were fixed in 10% neutralized Formalin, paraffin-embedded, and stained with hematoxylin-eosin or elastin van Gieson for imaging by light microscopy.

Stress measurements. Segments of arteries were placed in chilled physiological saline solution (PSS) containing (in mM) 120.5 NaCl, 4.8 KCl, 1.2 MgSO₄, 1.2 NaH₂PO₄, 20.4 NaHCO₃, 1.6 CaCl₂, 10 dextrose, and 1 pyruvate. The adventitia was removed, and the arteries were opened along the long axis of the vessel. The endothelium was removed by gently swabbing it with a cotton swab; the efficiency of this method has been confirmed using histochemistry. Strips of equal width were cut parallel to the cellular long axis (i.e., parallel to the vessel circumference) with a double-bladed cutting tool as previously described (2). Strip widths were fixed at 1.3 mm.

Myosin light chain phosphorylation. Strips of umbilical artery and fetal and adult femoral artery were mounted for measurement of isometric force and quick frozen with tongs precooled in liquid nitrogen either at rest or at the time of maximum contraction, using 10³ M phenylephrine for femoral arteries and 10⁵ M 5-hydroxytryptamine, respectively. The frozen muscle was weighed, placed in frozen slurry of trichloroacetic acid (10%, wt/vol) in acetone that contained dithiothreitol (DTT, 10 mM), and allowed to thaw to denature cellular proteins. Thereafter the muscle was placed in 60 vol of trichloroacetic acid (10%, wt/vol) and DTT (10 mM) and homogenized. The precipitated protein was washed with diethyl ether and then suspended in urea-glycerol buffer for electrophoretic analysis of light chain phosphorylation by immunoblotting (3, 25). Samples were subjected to electrophoresis for 1 h at 400 V in a gel containing 10% polyacrylamide and 40% (vol/vol) glycerol. Protein was transferred to nitrocellulose paper, and nonphosphorylated and phosphorylated forms of the light chain were localized by Western blotting with antibodies against bovine tracheal myosin light chain (1:10,000) and peroxidase-conjugated goat anti-rabbit IgG (1:15,000) with ECL. Relative amounts of nonphosphorylated and phosphorylated light chain were quantified by laser densitometry.

Umbilical artery responses. Human umbilical arteries have been extensively studied in vitro and shown to respond differently to several constrictors (8, 46, 47). Similar studies of ovine umbilical arteries, however, are rare (17). We therefore characterized ovine umbilical artery responses to several agonists. Strips of external and internal umbilical artery were cut and hung in muscle baths as described earlier. The strips were stretched to the L_o as determined by length-force calculations, using three stretches of 1.5 g for 20 min each. The strips were then stimulated twice with 65 mM KCl. All strips were relaxed with Ca²⁺-free PSS after each contraction. The strips of external and internal umbilical artery were subsequently stimulated with 10⁶ M phenylephrine, 10⁴ M histamine, 10⁵ M 5-hydroxytryptamine (serotonin), 10⁵ M carbachol, 10⁵ M bradykinin, 10⁴ M PGF₂, 10⁵ M PGE₂, 10⁸ M ANG II, and 10⁹ M endothelin. These concentrations have been shown to cause a maximal contractions in human umbilical arteries (47).

Statistics. All data are reported as means ± SE. Data were analyzed with Student's t-test and one-way ANOVA with Newman-Keuls correction for multiple comparisons. Significance was taken as P  0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Total and soluble protein, actin and MHC contents, and histology. Total protein contents did not differ significantly among the fetal and adult arteries studied (Table 1). Although soluble protein contents in fetal femoral artery and aorta were similar, values were ~40% less than that measured in either the external or internal umbilical artery (P < 0.0001, ANOVA) but did not differ from that seen in adult femoral artery (Table 1). Furthermore, the soluble/total protein ratio was greatest in the external umbilical artery compared with all other vessels, averaging 0.92, whereas the internal umbilical artery was similar to the adult and fetal aorta, demonstrating an intermediate position between fetal systemic vessels and the external umbilical artery.

View larger version (154K): [in this window] [in a new window]   Fig. 1.   Representative examples of fetal femoral artery (A and B), adult femoral artery (C and D), and external umbilical artery (E and F) stained with hematoxylin-eosin (A, C, and E) to assess smooth muscle content and van Gieson stain to assess elastin content (B, D, and F). Femoral artery photomicrographs are at ×20 to permit a comparison of size, cell distribution, and architecture, whereas umbilical histology is shown at ×10 because of vessel size. Arrowheads, vascular endothelium; arrows, elastin fibers; m, media; a, vascular adventitia.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Total actin (A) and total myosin heavy chain (MHC, B) contents in arteries obtained from near-term fetal and adult sheep. Different letters denote significant differences between tissues (P  0.006 by ANOVA), whereas groups not significantly different from each other are noted by same letters. See Table 1 for the number of animals within each tissue group. Values are means ± SE. Ext, external; Int, internal.

MHC isoforms. The relative and absolute contents of the 204- and 200-kDa MHC isoforms present in samples of fetal and adult VSM were determined from densitometric scans of the proteins separated on 4% polyacrylamide gels. Only two MHC bands were observed for each artery studied, confirming our prior observations for ovine fetal aorta that the 196-kDa protein could not be identified (11). The 204-kDa protein accounted for ~50% of the total MHC in the fetal femoral artery and aorta. In contrast, it accounted for >62% (P = 0.0001, ANOVA) of MHC in the external umbilical artery and both adult arteries; the internal umbilical artery was intermediate at 57%. Therefore, the relative amount of the 200-kDa protein in fetal systemic VSM (45-50%) exceeded that in the external umbilical artery and adult VSM (27-36%, P < 0.0001, ANOVA); again, the internal umbilical artery was intermediate with ~45% of MHC protein the 200-kDa isoform. When we determined the absolute contents of the 204- and 200-kDa proteins, which take into account differences in total MHC content, the values were similar in the fetal systemic arteries, averaging ~1.5 µg/mg wet weight (Fig. 3). In contrast, the content of the 204-kDa MHC isoform exceeded that of the 200-kDa protein in both umbilical and adult arteries, a value that was more than twofold higher than the 200-kDa protein in both external umbilical artery (P = 0.001) and adult femoral artery (P = 0.0009, Fig. 3). Compared with all systemic arteries, the contents of both the 204- and 200-kDa proteins were greater in umbilical arteries, reflecting the greater total MHC content (Fig. 2B).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Contents of 204- and 200-kDa MHC species in arteries obtained from near-term fetal (fet) and adult (mat) sheep. Values are means ± SE. Numbers of animals studied are noted in Table 1.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Representative immunoblots of 200-kDa MHC isoforms in fetal and adult arteries. Standards (Std) for SM2 and MHC-B were adult ovine myometrium and purified bovine brain myosin (8), respectively. Approximately 200 ng of total MHC were loaded in each lane.

Contraction responses. Active stresses were determined for each artery with vascular strips obtained from four additional near-term animals and performed in duplicate. Although contraction responses could be elicited by vascular strips obtained from both umbilical arteries and either adult vessel with 65 mM KCl or 10⁶ M phenylephrine, responses by strips of fetal aorta and femoral artery were greatly attenuated to both KCl and -stimulation as well as several other agonists, including histamine, serotonin, PGF₂, ATP, ANG II, and endothelin (data not shown). As summarized in Fig. 5, stress generated by fetal systemic arteries in response to KCl, which bypasses potential maturational differences in receptor densities or coupling, were significantly less than those observed with the external umbilical artery (P < 0.0001, ANOVA), which, in contrast, generated active stresses similar to those measured for both adult arteries. It is notable that responses by the internal umbilical artery were again intermediate between the external umbilical artery and fetal systemic vessels.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Summary of active stresses obtained with strips of fetal and adult arteries using 65 mM KCl at optimal length. Tissues were obtained from 4 animals, each studied in duplicate. Different letters denote significant differences between groups (P  0.0001, ANOVA), whereas same letter signifies that groups are not significantly different. Values are means ± SE.

Myosin light chain phosphorylation. The primary regulation of force in smooth muscle occurs via Ca²⁺-dependent phosphorylation of the myosin regulatory light chain (23, 45). We therefore determined in studies of tissues from two additional animals whether the decreased stress generation in fetal systemic arteries was associated with attenuated levels of myosin light chain phosphorylation. At the time of maximum stresses (~1 min) obtained by external umbilical arteries (10⁵ M serotonin) or adult femoral arteries (10³ M phenylephrine), there was evidence of 45-57% phosphorylation of the myosin regulatory light chain (Fig. 6, A and B). In contrast, after application of 10³ M phenylephrine to fetal femoral arteries, there was no phosphorylation (Fig. 6C) or stress observed. Interestingly, to detect regulatory light chain in homogenates of fetal femoral arteries, five times the sample volume was required to be loaded compared with maternal femoral arteries.

View larger version (57K): [in this window] [in a new window]   Fig. 6.   Representative immunoblots of regulatory light chain phosphorylation in adult femoral artery (A), external umbilical artery (B), and fetal femoral artery (C). Tissue homogenates (with standard being bovine tracheal) were subjected to electrophoresis (see METHODS), which separates the nonphosphorylated (upper band) from the phosphorylated (lower band) light chain. Lanes 2 and 3 show absence of phosphorylation when vascular strips are at basal tension in presence of EGTA and absence of Ca2+ in baths. Lanes 4 and 5 show 45-57% phosphorylation at time of maximum contraction by adult femoral arteries with 103 M phenylephrine (PE) and by external umbilical arteries with 105 M 5-hydroxytryptamine (5-HT). Fetal femoral arteries required 5-fold greater sample volume to detect regulatory light chain, and phosphorylation was not observed with 103 M PE.

Umbilical artery responses. To determine the contractile responses by ovine umbilical arteries to several agonists, studies were performed comparing responses by external and internal umbilical arteries from near-term fetal sheep using doses reported by White (47). The data have been normalized as a percent of the response to 65 mM KCl, which represents 100%. The external umbilical artery (Fig. 7A) constricted in the presence of each agonist examined with the exception of bradykinin. However, only 5-hydroxytryptamine elicited a response that significantly exceeded that seen with KCl, generating fourfold greater stresses than KCl (P < 0.0001, ANOVA). The internal umbilical artery demonstrated a similar pattern of responses (Fig. 7B) but, unlike the external umbilical artery, it was unresponsive to ANG II.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Contraction responses to multiple agonists by strips of external (A) and intra-abdominal (B) umbilical arteries obtained from near-term fetal sheep. Responses are presented as percent of response observed with 65 mM KCl. Hist, histamine; BK, bradykinin; ET, endothelin; Cch, carbachol. Doses are given in METHODS. Arteries from 4 fetal sheep were studied, each in duplicate. Values are means ± SE.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In fetal sheep vascular responses to infused vasoconstrictors may be less than those seen in adult animals (5, 48). This could reflect differences in the clearance of these agents (40), the absence or immaturity of specific receptors necessary for the transduction of these messages (15, 22), or the relative immaturity of fetal VSM compared with the adult (4, 7, 11). In addition, significant differences exist between umbilical and systemic responses to vasoconstrictors (26, 40, 48). For example, in intact fetal sheep with chronically implanted flow probes, the umbilical vascular bed is more sensitive to ANG II over a wide range of doses than either the hindlimb vasculature (27) or the systemic vasculature as a whole (48). These dissimilarities, however, are not explained entirely by differences in the receptor subtype expressed in the two vascular beds (27). Thus within the fetal compartment there is increasing evidence of important differences in vascular maturation and function (12, 26, 27, 29, 40, 48). The expression of contractile proteins in ovine fetal systemic VSM is developmentally regulated, and before birth a synthetic phenotype of VSM is predominant (4, 11). However, it is unclear which factors account for the differences in umbilical and systemic vascular responsiveness and what role the expression of VSM contractile proteins might play. To address these queries we examined the expression and contents of key contractile proteins in several systemic arteries from fetal and adult sheep as well as the umbilical artery and compared this to their capacity to generate stress in vitro.

The contents of actin and myosin in vascular and visceral smooth muscle are developmentally regulated (11, 18, 42), and in ovine aorta and bladder values are extremely low until the end of gestation and during the immediate postnatal period (4, 11). Whereas actin contents in aorta of term fetal sheep increase approximately twofold at term, resulting in values that are one-half those seen in the adult, MHC contents have reached values approaching adult levels. The pattern of change in total and soluble protein contents resembles that seen with myosin, i.e., at term gestation values approach that seen in the adult (4). In the present study, contents of actin, myosin, and protein in near-term fetal and adult systemic arteries were similar to those previously observed. However, umbilical artery actin and MHC contents were 2.5- to 3.0-fold greater than those observed in fetal systemic arteries. Furthermore, total myosin content was twofold greater than that measured in adult vessels. The greater contents of actin and MHC in both umbilical arteries were associated with a greater soluble protein content, suggesting that there is not only more contractile protein in umbilical arteries but also more smooth muscle, particularly in the external umbilical artery. This conclusion is further supported by histological comparison of the vessels. The intra-abdominal portion of the umbilical artery was intermediate in this regard and in all other aspects, suggesting that it serves as a transitional artery between the fetal systemic and umbilical vasculature, an observation consistent with recent findings regarding ANG II receptor subtype expression in fetal sheep (15).

In adult mammals, VSM contains predominantly the smooth muscle MHC isoforms, SM1 and SM2, and the SM1 species is the more abundant isoform (3, 9, 18). Furthermore, SM2 accounts for all or nearly all of the 200-kDa species in adult VSM (18, 41). Thus adult VSM consists primarily of the contractile smooth muscle phenotype. In contrast, systemic arteries from the developing ovine fetus predominantly express the 200-kDa species, and SM1 does not achieve adult levels until 3-4 mo postnatal (4, 11). Furthermore, the 200-kDa species in developing systemic VSM consists primarily of the nonmuscle isoform MHC-B until after birth (4, 11). This pattern of protein expression reflects the predominance of a synthetic smooth muscle phenotype, which has been identified in proliferating smooth muscle cells in culture and within arteriosclerotic neointimas (19, 39, 44) and is associated with the potential for VSM replication (13). These differences in MHC isoform expression between fetal and adult systemic VSM were also evident in the present study. In contrast, the umbilical arteries more closely resembled mature adult arteries, i.e., predominantly expressing the SM1 isoform and having no evidence of MHC-B expression. Although we have only examined five systemic fetal arteries to date, each has consistently expressed MHC-B, and its expression has decreased with increasing age (4, 11, and unpublished observations). Notably, the bladder is the only other smooth muscle thus far identified in fetal sheep that does not express MHC-B. Like the umbilical artery, it too has a predominance of SM1 (4, 11). Thus umbilical artery VSM development is precocious within the fetal vascular compartment and, unlike systemic VSM, demonstrates only a contractile VSM phenotype at term. Advanced differentiation of the ductus arteriosus has also been described recently (12, 29); however, MHC-B expression was not examined. Because the umbilical artery and ductus arteriosus constrict at or soon after birth to redirect cardiac output, their accelerated maturation of VSM may be unique to the fetal vasculature and related to adaptative alterations necessary after birth.

Responses by vascular strips of fetal femoral artery and aorta were significantly attenuated compared with adult arteries for all agonists examined, including KCl. This is consistent with studies comparing fetal, newborn, and adult systemic arteries from sheep, rabbits, and rats (6, 24, 36, 42). Although there were quantitative differences in the measured responses to KCl or other agonists between studies, reflecting the use of strips versus rings, arteries from immature animals were consistently less responsive. In contrast, the external umbilical artery generated stresses comparable to those seen with adult vessels, whereas the internal umbilical artery stresses were intermediate. These results clearly support prior observations that the synthetic phenotype of VSM, for which MHC-B serves as a marker, is associated with decreased functional capacity (10, 43). Failure of fetal systemic arteries to contract to several agonists may be due to deficits in excitation-contraction coupling. Our finding that supramaximal concentrations of agonist did not elicit myosin light chain phosphorylation in fetal femoral arteries suggests that these agonists did not elevate intracellular Ca²⁺ concentrations; however, a more detailed time course will be required to establish whether this is the case. In preliminary studies, we have found significant expression of Ca²⁺-calmodulin-dependent myosin light chain kinase in fetal arteries at this stage of development. Thus it remains to be determined whether the immature VSM phenotype is associated with fewer receptors, channels, or other components required for excitation-contraction coupling. Supporting this notion is our recent observation that only the AT₂ subtype of the ANG II receptor, which does not couple to contraction (14), is expressed in ovine fetal systemic arteries (15, 27). It remains unclear, however, whether differences in contractile protein composition between fetal and adult arteries also contribute to the attenuated stresses. Although total MHC contents were similar in fetal and adult systemic arteries, actin contents were less and interestingly, immunoreactive regulatory light chain contents also were substantially lower in fetal VSM. Furthermore, MHC isoform composition differed, i.e., nonmuscle MHC-B was expressed only in fetal systemic arteries. Direct activation of contractile proteins by elevating Ca²⁺ in permeabilized muscle fibers will be required to directly assess the contractile capacity of fetal VSM.

Although fetal sheep are extensively used to study cardiovascular development in vivo and have provided important insights into developmental changes, important differences may exist between species. Although the human fetus cannot be studied in detail, the umbilical artery is readily available after birth, thus its responsiveness to numerous agents has been thoroughly examined (8, 46, 47). These studies, however, have focused on understanding how the umbilical artery closes at birth. Similar studies of the ovine umbilical arteries are rare (17). We therefore determined whether the external and internal umbilical arteries responded differently to several agonists and how these responses compared with those observed with human arteries. Responses by the two segments of ovine umbilical artery were similar, and 5-hydroxytryptamine elicited the greatest contractile responses by both arteries, which is consistent with earlier reports for sheep (17) and the human (8, 46, 47). Bradykinin, a potent vasoconstrictor in the human umbilical artery (46, 47), however, had no effect on either ovine artery. The reason for this disparity is not readily apparent. ANG II also contracted the external umbilical artery but had no affect on segments of the internal artery. This can be explained by the expression of AT₁ receptors in the former and AT₂ receptors, which do not couple to contraction, in the latter (15). Comparing these responses with the human is difficult because there may be regional differences in ANG II receptor expression (30) and function (46) within the external umbilical artery, which will require further evaluation. Nonetheless, evidence exists in both species that the umbilical vasculature is quite sensitive to ANG II and several other agonists, further supporting the use of the sheep as a model for studies of cardiovascular development.

In the present studies we have compared for the first time the differences in protein expression and function in fetal systemic and umbilical artery VSM to explain previous observations of differences in contractile responses in vivo (26, 27, 40, 48). We have demonstrated that unlike systemic arteries the ovine umbilical artery predominantly expresses the contractile phenotype of VSM and there are greater contents of actin and total myosin in the umbilical artery. Moreover, we have shown that the umbilical artery is functionally more like adult arteries, responding to several agonists. Thus umbilical artery smooth muscle development precedes that of the fetal systemic vasculature. Additional studies, however, are needed to determine whether the smaller systemic resistance vessels also demonstrate a slower rate of maturation.

Perspectives