Main

Uterine blood flow increases markedly in mammalian pregnancy. The increase is due to a fall in vascular resistance, which results in a rise in cardiac output and a redistribution of blood flow to the uterine vascular bed. In mammals with hemochorial placentae, the reduction in uterine vascular resistance is accomplished by the enlargement of the main uterine artery and radial vessels (the vessels deriving from the main uterine artery and entering the uterine wall) and the erosion of the vascular wall of the spiral arteries by the invading trophoblast(1). The stimuli and the mechanisms responsible for this dramatic vascular remodeling remain unclear.

We sought to determine whether all layers of the uterine vessels are stimulated to grow during pregnancy and whether the growth preceded or coincided with the rise in uterine blood flow. The enlargement of the main uterine artery and radial vessels during pregnancy is likely to involve both hypertrophy and hyperplasia. Hypertrophy is supported by an increase in uterine artery medial cell size, decrease in cell number per unit area, and decrease in DNA relative to RNA content(2–4). Hyperplasia is demonstrated by increased DNA synthesis in radial artery segments and in the intimal and medial layers of the uterine and radial arteries(3, 5). Unknown is whether hyperplasia occurs in the adventitia, known to be important for elaborating and amplifying growth signals in other circulations(6, 7), or if growth occurs in the uterine venous circulation. In addition, although compositional changes occur during pregnancy in nonuterine vessels(8, 9) and aortic smooth muscle cells from pregnant animals have increased responsiveness to PDGF in vitro compared with cells from nonpregnant animals(10), whether a proliferative response to pregnancy occurs outside the uterine circulation in vivo has not been tested directly.

Relatively little is known concerning the time course and distribution of the proliferative response to pregnancy. The enlargement of the uterine artery during human pregnancy is complete by mid-pregnancy(11). Consistent with this is an early report in the guinea pig in which the proliferative response of the radial vessels nearest the placenta underwent the greatest and most prolonged enlargement(12), but a more recent study in rats found that vascular proliferation in the radial arteries peaked earlier than in the more distal, uterine artery(3). Thus, unclear is whether proliferation in the various layers of the vessel wall moves progressively downstream (from the uterine to the radial arteries), upstream (from the radial to the uterine arteries), or occurs simultaneously throughout the uterine circulation. Increased flow and shear stress [reviewed in Mulvany(13)] and short-term estradiol treatment(5, 10) have been implicated as important stimuli for vascular proliferation in several systems. Given that the hormonal stimuli are present early but the greatest rise in flow occurs late in pregnancy, we considered that examining the distribution and time course of the proliferative response in the uterine circulation in relation to the rise in uterine blood flow would be informative for determining the extent to which the primary stimulus in each of the vessels examined was mechanical(e.g. increased blood flow or shear stress) or was related to the hormonal characteristics of pregnancy.

We hypothesized that pregnancy stimulated new cell growth in all layers of the major vessels in the uterine circulation (uterine artery, radial artery, uterine vein) as well as possibly in vessels outside the uterine circulation, and that this growth occurred before the greatest rise in uterine blood flow. We further hypothesized that treatment with estradiol mimicked this pregnancy-stimulated growth response. BrdU, a thymidine analog whose uptake provides an index of DNA synthesis, was implanted s.c. over four 2-wk periods throughout the guinea pig's 9-wk pregnancy. The adventitia, media, and intima were examined in each layer of vessels in the uterine (uterine artery, radial artery, uterine vein) and nonuterine (mesenteric artery and thoracic aorta) circulations. The time course of the proliferative response in these vessels was compared with the rise in uterine blood flow. To determine the effect of prolonged estradiol treatment on DNA synthesis, BrdU incorporation was also measured in ovariectomized animals treated for 2 wk with 17β-estradiol or vehicle. We considered that these studies would be informative about the mechanisms involved in vascular proliferation during pregnancy and, in turn, provide insight into conditions in which pregnancy-induced remodeling does not occur(14, 15).

METHODS

Animals and treatments. A total of 41 female guinea pigs from Sasco (Omaha, NE) were studied. Maternal body weights (total - uterine contents) were similar among nonpregnant, pregnant, vehicle- and estradiol-treated groups at the time of sacrifice. Pregnancy duration was calculated as days after conception as judged by the appearance of a vaginal plug. Litter size averaged 4 ± 1 (SEM) pups. One animal (studied at 14-28 d) had a unilateral pregnancy with one fetus on the left side. Animals were ovariectomized by Sasco approximately 1 mo before study, and absence of ovaries was confirmed at sacrifice. Serum samples were obtained from nonpregnant, pregnant, and ovariectomized animals by cardiac puncture at time of sacrifice and kept frozen at -70 °C until time of assay. 17β-Estradiol was measured by RIA (Diagnostic Products, Los Angeles, CA).

BrdU pellets (400 mg 14-d release, Innovative Research, Toledo, OH) were implanted s.c. in the infrascapular region using local anesthetic (lidocaine) in three nonpregnant, nine pregnant, and eight ovariectomized animals. BrdU is a thymidine analog which is incorporated into the DNA of cells during S phase. Pellets were implanted in guinea pigs over the following 14-d intervals (term= 63 d): d 0-14 (n = 2), d 14-28 (n = 2), d 28-42(n = 3), d 42-56 (n = 2). BrdU was well tolerated as judged by absence of abortions and normal weight gain. In ovariectomized animals, three were implanted with vehicle and five with 17β-estradiol(2.5 mg, 14-d release, Innovative Research, Toledo, OH) at the same time as the BrdU pellets were placed. At the end of the 14-d labeling period, animals were killed after being anesthetized with 0.5 mL of Rompum (Miles, Inc., Swanee Mission, KS) (xylazine) and 10 mL/kg Ketaset (Fort Dodge Laboratories, Inc., Fort Dodge, IA) (ketamine). The animals were perfused with 1% heparinized saline, and the vasculature was fixed with 10% buffered formalin under 50 mm Hg pressure. Tissues were removed, preserved overnight in 10% buffered formalin, embedded in paraffin, and processed for histology.

DNA synthesis measurements. Sections were prepared for each animal from one uterine artery, the uterine vein, three to six radial arteries, one mesenteric artery, and the thoracic aorta. For the uterine artery, sections were obtained from the ovarian, middle, and cervical segments. All other vessels were sampled at their midpoints. Radial arteries(the vessels deriving from the main uterine artery and entering the uterine wall) were evenly divided between those supplying and not supplying placentae. Uterine and radial arteries from both horns were sampled from the animal with the unilateral pregnancy, but results are presented from the pregnant side unless specified otherwise.

Sections of formalin-fixed tissue were deparafinized with xylene and ethanol and then rinsed in 0.01 M PBS. Endogenous peroxidase activity was quenched by a 30-min soak in 0.3% hydrogen peroxide in absolute methanol and rinsed in PBS. Tissue sections were hydrolyzed with protease XXIV (Sigma Chemical Co., St. Louis, MO), and DNA was denatured with 2 N hydrochloric acid. Sections were then incubated in horse serum and primary BrdU antibody. Adventitial, medial, and intimal layers were distinguished by the locations of the internal elastic lamina, the external elastic lamina, the fatty tissue surrounding the adventitia, cell shape, and cellularity. In the uterine vein, all vessel layers were counted together. A sufficient number of sections were prepared to enable 500 adventitial, 1000 medial, and 250 intimal cells to be counted from each type of artery and each animal. From 500 to 1000 uterine vein cells were counted per animal.

Vessels were stained immunohistochemically with a MAb against BrdU(16). Antigen:antibody complexes were visualized with an anti-mouse IgG antibody avidin-biotin-horseradish immunoperoxidase system (ABC Kit 32028, Pierce, Rockford, IL). The diaminobenzidine tetrahydrochloride substrate for the peroxidase resulted in BrdU-labeled nuclei appearing brown under light microscopy. Sections were lightly counterstained with hematoxylin such that unlabeled nuclei appeared blue. Labeling indices were calculated as the number of BrdU-labeled (brown) nuclei divided by the total number (brown and blue) of nuclei. A section of guinea pig small bowel with rapidly dividing crypt cells was used as a control to verify BrdU uptake.

Morphometric measurements. Morphometric measurements of uterine artery medial area were made in the middle segment of the uterine artery. Still video images were recorded, transferred to computer image files using Photoshop, version 4.0, software (Adobe Systems Inc., Mountain View, CA), and processed using Image software (version 1.35, National Institutes of Health Research Services). The external and internal perimeters of the media were defined by the external and internal elastic internal laminae, respectively, and outlined. Total medial area was calculated as (the area within the external perimeter) - (the area within the internal perimeter).

Uterine blood flow. Uterine blood flow was measured using radiolabeled microspheres in four nonpregnant and 18 pregnant animals housed under the same conditions as the animals participating in the BrdU studies. Blood flow measurements were performed as described previously(17). Briefly, in anesthetized guinea pigs (25-50 mg/kg intramuscular ketamine, 1-2.5 mg/kg intramuscular xylazine), polyvinyl catheters were advanced through the carotid artery to the left ventricle (0.58 mm inside diameter, 0.99 outside diameter) and through the jugular vein to the superior vena cava (0.28 mm inside diameter, 0.64 outside diameter). Approximately 5-7 d later, once the animals had recovered from the catheterization procedure as judged by return to presurgery weight and rate of weight gain, cardiac output was measured in awake animals by green dye dilution without blood loss as previously described(18). Once duplicate (within 10%) values were obtained, a minimum of 105 15-μm diameter (DuPont NEN, Boston, MA) microspheres labeled with one of three radionuclides (46Sc,113 Sn, 85Sr) were injected into the left ventricle. The animals were killed by pentobarbital sodium overdose. The uterus and placentae were removed and counted with an Auto-Gamma Scintillation Spectrometer. Fetuses were of appropriate size for gestational age. No reabsorption sites were observed at the time of sacrifice nor were any spontaneous abortions detected. Uterine (nonpregnant) or uterine (pregnant) flows were expressed as organ flow(mL/min) = [(cardiac output) × (organ cpm)]/(total injected cpm).

Statistics. Data in the text, tables, and figures are reported as the X ± SEM. Replication indices were averaged for each vessel from each animal. Sample sizes were considered as the number of animals in each group. One-way analysis of variance with Fisher's protected least significant difference multiple comparison test was used to compare nonpregnant and pregnant or vehicle and estradiol-treated groups (Statview, version 4.1, Abacus Concepts Inc., Berkeley, CA). Correlations were calculated using linear regression techniques. Probabilities were combined from tests of significance using the Fisher technique(18). Comparisons were considered significant when p < 0.05.

RESULTS

DNA synthesis in uterine vessels. Pregnancy increased BrdU labeling indices, a measure of DNA synthesis, in all layers of the uterine artery wall (Fig. 1). Peak DNA synthesis in each layer of the uterine artery occurred by mid-pregnancy, d 28-42 of the guinea pig 63-d gestation, at which time DNA synthesis had increased 4-fold in the adventitia, 41-fold in the media, and 14-fold in the intima. The labeling indices remained elevated in the adventitia but fell near term in the media and intima (Fig. 1). No differences were observed in the labeling indices from the ovarian, middle, and cervical segments of the uterine artery, and hence values were combined.

Figure 1
figure 1

Pregnancy increases the BrdU labeling index, a measure of DNA synthesis, in all layers of the uterine artery. The BrdU labeling index was calculated as the number of BrdU-labeled cells divided by the total number of cells. Values are X ± SEM. *p < 0.05 comparison with nonpregnant group. **p < 0.05 for comparisons with all groups.

Full size image

Pregnancy increased DNA synthesis in all layers of the radial artery (Fig. 2) and the uterine vein (Fig. 3). Peak labeling indices in the radial artery adventitia were greater than those in the media or intima. Maximal values occurred at mid-pregnancy (d 28-42) in the radial artery adventitia and uterine vein, whereas values continued to rise until near-term in the radial artery media and intima. Comparing peak pregnant with nonpregnant values, the labeling indices rose 12-fold in the adventitia, 23-fold in the media, 5-fold in the intima, and 12-fold in the uterine vein. In the unilateral pregnant animal, the uterine and radial arteries on the pregnant side had higher adventitia, media, and intima labeling indices than those from the nonpregnant side (uterine artery adventitia = 6.6 versus 4.0; media = 5.0 versus 1.2; intima = 1.5 versus 1.0; radial artery adventitia = 11.1versus 2.8; media = 11.4 versus 3.0; intima = 3.7versus 0.6, respectively).

Figure 2
figure 2

Pregnancy stimulates DNA synthesis in the radial artery adventitia, media, and intima. Note the different y axis scale for each vessel layer. Values are X ± SEM. *p< 0.05 for comparisons with nonpregnant group. **p < 0.05 for comparisons with all groups.

Full size image
Figure 3
figure 3

DNA synthesis increases in the uterine vein. BrdU labeling indices were calculated for all layers combined. Values are X ± SEM. **p < 0.05 for comparisons with all groups.

Full size image

DNA synthesis in nonuterine vessels. Labeling indices for all cell layers of the mesenteric artery and aorta were unchanged at any time during pregnancy compared with the nonpregnant condition(Table 1). Likewise, labeling indices for all pregnancy times combined did not exceed nonpregnant values in any layer in any vessel.

Table 1 BrdU labeling indices (number BrdU labeled cells/all cells,%) in the mesenteric artery and thoracic aorta from nonpregnant and pregnant guinea pigs (mean ± SEM)
Full size table

Uterine artery morphometric changes. Uterine artery medial area doubled during pregnancy, attaining maximal values at d 28-42(Table 2). The enlargement of medial area was accompanied by greater external and internal perimeters, indicating a probable enlargement of luminal diameter. The medial area tended to decline near term toward but not to nonpregnant values. Greater uterine artery medial area (y) was associated with higher labeling indices (x) among all animals(r = 0.74, y = 0.004x + 0.051, p < 0.05). In the one animal with a unilateral pregnancy, uterine artery medial area external and internal perimeters were greater on the pregnant than the nonpregnant side (medial area = 0.08 versus 0.06 mm2,Pext = 1.30 versus 1.12 mm, Pint = 0.73 versus 0.53 mm, respectively).

Table 2 Uterine artery morphometric values in nonpregnant and pregnant guinea pigs (mean ± SEM)
Full size table

Relationship to uterine blood flow. Uterine artery blood flow(y) rose modestly by mid-pregnancy (d 40) and exponentially thereafter (Fig. 4). Thus, the major rise in uterine artery blood flow occurred after peak DNA synthesis in the uterine artery, uterine vein, and radial artery adventitia and during the period of peak DNA synthesis in the radial artery media and intima. Cardiac output during pregnancy (295 ± 12 mL/min) was above nonpregnant values (275 ± 13 mL/min).

Figure 4
figure 4

Uterine blood flow (y) increases exponentially with advancing pregnancy (x, term = 63 d). Each data point is an individual awake, healthy guinea pig with a litter size of two to four pups in whom regional blood flows were determined using microspheres (see“Methods”). Open circles with SEM bars are previously published values(24).

Full size image

Relationship to estradiol. Serum estradiol levels rose progressively during pregnancy (nonpregnant = 19 ± 3, 0-14 d = 34± 15, 14-28 d = 38 ± 6, 28-42 d = 44 ± 2, 42-56 d = 56± 0 pg/mL). Estradiol levels correlated with labeling indices in the uterine artery adventitia and in all layers of the radial artery(Table 3). Estradiol treatment for 14 d raised labeling indices in the radial artery adventitia and tended (p = 0.08) to increase DNA synthesis in the media of all the vessels examined but not in any other vessel or vessel layer (Table 4).

Table 3 Correlation coefficients (r values) between serum estradiol levels and labeling indices in the uterine circulation among all nonpregnant and pregnant animals (n = 12)
Full size table
Table 4 Effects of estradiol treatment on labeling indices (% BrdU labeled cells) in vessels from three vehicle-treated and five estradiol-treated animals (mean ± SEM)
Full size table

DISCUSSION

We found that pregnancy stimulated DNA synthesis in the uterine artery, the radial artery, and the uterine vein but had little effect on vessels outside the uterine circulation. The maximal increase in DNA synthesis preceded the time of greatest uterine artery blood flow in all layers of the uterine artery, the radial artery adventitia, and the uterine vein. However, maximal proliferation occurred at the time of greatest uterine blood flow in the radial artery media and intima. A role for estradiol in the pregnancy-induced rise in DNA synthesis was supported by finding a positive relationship between serum estradiol levels and labeling indices throughout pregnancy and stimulatory effects of 17β-estradiol treatment on DNA synthesis in the radial artery adventitia and media of all vessels examined but estradiol was not able to account fully for the pregnancy-associated changes observed.

Issues for the interpretation of our study results concerned the choice of experimental animal, study design, and method for assessing cellular replication. We used BrdU to measure DNA synthesis in vivo over four 2-wk periods comprising nearly all of the guinea pig's 9-wk gestation. BrdU is a thymidine analog that is taken up by the cells during S phase. The BrdU labeling index, the percentage of cells that have taken up BrdU, has been widely used as a direct, quantitative measure of DNA synthesis(16). The guinea pig has proved a valuable model for studying the effects of pregnancy and chronic hypoxia on the maternal vasculature(17, 19). We chose it for the present study because its long gestation permits sampling a sufficient number of time points to define the time course of the proliferative response. Labeling cells for 2 wk also had the advantage of yielding an adequate number of cells in each layer of the vessel wall. Although a limited number of animals were studied at a given time point, the total number of cells counted was considerably greater than in previous studies(3). Moreover, the effects of pregnancy on increasing DNA synthesis were consistent among animals. A difficulty for techniques based on BrdU or thymidine incorporation is that the number of cells going through S phase is not equivalent to cellular replication if cells exhibit ploidy (multiple nuclei). Vascular smooth muscle and endothelial cells but not, to our knowledge, fibroblasts exhibit ploidy, but it is unlikely that the magnitude of increases observed were due to ploidy alone.

Our study results are consistent with previous observations of uterine and radial artery hyperplasia during pregnancy but extend these findings to include all layers of the vessel wall in uterine and nonuterine vessels, the time course of the proliferative response throughout pregnancy, and the relationship of the proliferative response(s) to the rise in uterine artery blood flow. Like previous studies, which documented DNA synthesis by short-term BrdU labeling and thymidine incorporation(3, 5), we found that pregnancy increased DNA synthesis in the medial and intimal layers of the uterine and radial arteries. We also found increased DNA synthesis in the adventitial layer of these vessels. The importance of hyperplasia in the enlargement of the uterine artery during pregnancy was demonstrated in the present study by the relationship between increased DNA synthesis and greater uterine artery medial area.

The time course of the proliferative response to pregnancy varied by vessel and by layer of the vessel wall. In the uterine artery media and intima, DNA synthesis peaked by mid-pregnancy and then declined near term, whereas DNA synthesis in the radial artery media and intima continued to rise until term. The mid-pregnancy peak in uterine artery DNA synthesis coincided with maximal expansion of medial area. The proliferative response in the adventitial layer of the uterine and radial arteries and in the uterine vein exhibited a somewhat different time course; maximal DNA synthesis occurred at d 28-42, but the increase was sustained at d 42-56. Comparing vessels, the increase in DNA synthesis in the uterine artery media and intima appeared essentially complete by mid-pregnancy, whereas it continued until term in the medial and intimal layers of the radial artery. These findings differ from those obtained in the rat. BrdU labeling over two 24-h periods in the rat revealed higher DNA synthesis in the radial artery media and intima at mid-pregnancy than at term and a decrease in DNA synthesis from mid- to late-pregnancy in the uterine artery media and intima(3). Our results are similar to those obtained in the guinea pig in which the greatest and most prolonged growth in the uterine and radial arteries occurred nearest the placenta(12). The differences in time course may reflect species variation, although both the rat and guinea pig have hemochorial placentae and generally similar vascular responses to pregnancy. Alternatively, because only 2% of rat pregnancy whereas 89% (four 14-day periods) of guinea pig pregnancy were sampled, the changes observed in the rat may not have been representative of those occurring throughout pregnancy. Comparing the layers of the arterial wall suggested that the adventitia may be playing an especially important role. The increase in DNA synthesis occurred earliest in the radial artery adventitia, then moved inward to the media and intima, and was retained in the uterine artery adventitia even after DNA synthesis had declined in the media and intimal layers. We thus concluded that medial and intimal growth of the uterine artery was complete by mid-pregnancy, whereas it continued until term in the medial and intima layers of the radial artery.

The relationship of the proliferative responses in the various vessels to the rise in uterine blood flow is informative about the factors responsible for increasing uterine blood flow during pregnancy and, in turn, the influences of blood flow on the proliferative response. Since maximal DNA synthesis and uterine medial area were attained before uterine blood flow had increased appreciably (Fig. 4), the increase in uterine artery DNA synthesis appears important for initiating the rise in uterine blood flow. On the other hand, the similarity in the time course of the radial artery proliferative response and the rise in uterine artery blood flow suggests that the increase in radial artery DNA synthesis was important in sustaining the late pregnancy rise in uterine artery blood flow. Increased flow and shear stress have been implicated as important stimuli for vascular proliferation in several systems(13). As judged by a comparison of their time courses, increased flow and shear stress are likely to have contributed to the increased DNA synthesis in the radial artery but less likely to have affected the uterine artery media and intima because proliferation was complete before uterine blood flow had increased appreciably. Similar findings are seen in women for whom the major increase in uterine artery diameter precedes the rise in uterine artery blood flow(11). In the animal with the unilateral pregnancy, the higher labeling indices from the pregnant side suggested that greater flow likely stimulated DNA synthesis in the uterine and radial arteries but that its effects were more marked in the radial artery. Thus, we concluded that the proliferative response and medial enlargement of the uterine artery were important in initiating the rise in uterine artery blood flow. The resultant increase in flow and shear stress likely stimulated DNA synthesis in the radial artery and helped to sustain the rise in flow near term.

We considered that hormonal influences and, in particular, estradiol-stimulated proliferation of the uterine vessels early in pregnancy. Short-term estradiol treatment increases thymidine incorporation in radial arteries in vitro and prompts histologic changes in nonuterine vessels(5, 8). We have reported that estradiol increases the growth response to PDGF in cultured uterine artery smooth muscle cells via nuclear and nonnuclear receptor mechanisms(10). In the present study, estradiol levels correlated with the labeling indices during pregnancy. Further, 2-wk estradiol treatment raised DNA synthesis in the radial artery adventitia to early pregnancy levels and tended to increase DNA synthesis in the media of all vessels examined. The adventitia is known to elaborate growth factors and extracellular matrix proteins important in pulmonary vascular remodeling(6, 7). Perhaps stimulation of the adventitia by estradiol early in pregnancy precipitates medial and intimal proliferation and progressive remodeling from the radial arteries to the main uterine artery and vein. However, estradiol treatment did not fully reproduce the effects of pregnancy; the rise in DNA synthesis was not as great, and not all layers of the vessel wall were affected. The lesser effect of estradiol treatment may have been due to the shorter period of hormonal elevation, because we saw little change in uterine or radial artery DNA synthesis during the first 14 d of pregnancy. Even though several hours of estradiol treatment stimulated vascular smooth muscle cell growth in culture(10) and thymidine incorporation in vitro(5), a longer period of hormonal elevation may be required to overcome local antiproliferative mechanisms present in vivo. Other hormonal and pregnancy-related characteristics are likely involved.

The lack of change in DNA synthesis in the vessels examined outside the uterine circulation indicates that the stimulatory effects of estradiol, blood flow, or other characteristics of pregnancy are affected by factors operating within the uterine circulation. Although it is possible that our limited sample sizes did not permit detection of subtle increases in DNA synthesis, no labeling indices for any layer in any of the nonuterine vessels were above nonpregnant levels, even when time points were combined from all pregnant animals. The lack of proliferative response was not due solely to differences in blood flow, because an increase in blood flow occurs in the aorta and mesenteric arteries during guinea pig pregnancy, and as reviewed above, the uterine artery's proliferative response preceded the rise in uterine blood flow. The localization to the uterine circulation could have been due to regional differences in growth factors or antiproliferative mechanisms. Estradiol and progesterone concentrations in uterine lymphatics are higher than in circulating levels(20). In the uterine circulation, a delicate balance likely exists between locally acting vasodilators and mitogens to allow for both uterine artery remodeling and vasodilation of pregnancy. For example, increased nitric oxide activity has been implicated in the vasodilation of pregnancy(21, 22), yet nitric oxide in relatively high concentrations is an inhibitor of smooth muscle cell growth(23). Anti-proliferative mechanisms may serve to limit the proliferative effects of pregnancy in vivo to the uterine circulation. Even though we observed that estradiol increases the growth response to PDGF in cultured aortic as well as uterine artery smooth muscle cells, no stimulatory effects in the aorta were observed in vivo(10).

In summary, we have demonstrated that pregnancy stimulates DNA synthesis in the adventitial, medial, and intimal cells of the uterine artery, radial artery, and uterine vein. The time course of the proliferative response and the effect of estradiol treatment support the likelihood that local hormonal influences in combination with increased blood flow stimulate the increase in DNA synthesis. Needed are further studies to define the precise mechanisms by which hormonal influences and/or increased flow stimulate DNA synthesis. Such studies are likely to be important for understanding the abnormalities in vascular growth seen in preeclampsia(14) and during chronic hypoxia(15).