Differentiation of the microvascular endothelium during early angiogenesis and respiratory onset in the chick chorioallantoic membrane

Tissue & Cell, 1995 27 (2) 159-166 © i995 Pearson Professional Ltd.

Differentiation of the microvascular . endothehum during early anglogenesls and respiratory onset in the chick chorioallantoic membrane Victor Rizzo*, Daekyung Kim t, Walter N. Durfin t, David O. DeFouw*

Abstract, The present study served to determine the extent of microvascular endothelial differentiation during early stages of morphogenesis (days 4.5-5.5 of the 21-day incubation) in the chick chorioallantoic membrane (CAM). CAM's, which serve as the embryonic lung, were prepared for intravital injections of a graded series of FITC-dextrans and subsequent ultrastructural morphometric analyses of the microvascular units. The precapillary, capillary, and postcapillary microvascular segments presented a continuous endothelium that was substantially thicker than that of adult lung endothelia (DeFouw, 1988). Further, plasmalemmal vesicles were uniformly sparse, while endothelial vacuoles, of variable diameters, were present continuously in the proliferating microvascular units. Average widths and depths of the interendothelial clefts were uniform and suggested complete structural differentiation from the onset of CAM morphogenesis. Based on our recent estimates of CAM microvascular permeability coefficients (Rizzo et al., 1995), the observed endothelial ultrastructure was associated with microvascular selectivity comparable to that of adult pulmonary microvessels (Lanken et al., 1985). Therefore, despite incomplete ultrastructural differentiation of the early CAM microvascular endothelium, these angiogenic microvessels presented adult-like barrier properties. Further they were less permeable than (Wu et ah, 1993; Yuan et al., 1993) and ultrastructurally distinct from (Kohn et al., 1992) certain tumorigenic microvessels. Thus, angiogenesis is likely not a routinely homogeneous process, and CAM microvascular permeability characteristics may be teleologically significant.

Keywerds: Angiogenesis,chorioallantoic membrane,differentiation, endothelium,ultrastructure

Introduction The chorioallantoic membrane (CAM) of the chick embryo is formed at day 4.5 of the 21-day incubation, by fusion of the vascularized allantoic sac and the chorionic ectoderm. The CAM microcirculation is part * Department of Anatomy, Cell Biology and Injury Sciences. 1 Department of Physiology, Department of Surgery and Division of MicrocircuUatoryResearch UMDNJ - New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103-2714, USA. Received 20 September 1994 Accepted 4 November 1994 Correspondence to: Victor Rizzo, PhD, Department of Anatomy, Cell Biology and Injury Sciences, UMDNJ - New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103-2714, USA.

of the respiratory exchange surface which accommodates increasing oxygen demands of the growing embryo. Hence, the CAM has been defined as an 'in ovo lung' (Hamburger and Hamilton, 1951), and threedimensional analyses of this microcirculation have revealed an architecture similar to that of the mature pulmonary microvasculature (Fuchs and Lindenbaum, 1988). Ultrastructural features of the CAM microvascular endothelium at the onset of CAM respiratory function, however, remain unknown. In the present study, the extent of ultrastructural differentiation of the CAM microvascular endothelium during days 4.5-5.5 of development was determined by comparison to the results of a previous study from this laboratory

160 R[zzo ET AL

of adult pulmonary endothelial ultrastructure (DeFouw, 1988). In a recent study, we described an increase of CAM endothelial restriction to macromolecular extravasation between days 4.5 and 5.5 of development (Rizzo et al., 1995). Therefore, despite continuous angiogenesis up to day 10 in the CAM (Ausprunk et al., 1974; Flamme et al., 1991), microvascular endothelial permeability to macromolecules was markedly reduced between days 4.5 and 5.5 of the angiogenic process. This is in marked contrast to neovascularization associated with tumorigenesis during which substantial macromolecular extravasation is a consistent feature (Ausprunk and Folkman, 1977; Gerlowski and Jain, 1986; Wu et al., 1993; Yuan et al., 1993). The present comparisons of angiogenic CAM microvascular endothelium with recent ultrastructural analyses of tumor-related angiogenic microvessels (Kohn et al., 1992; Paku and Paweletz, 1991) served to address this distinct difference in permeability between the respective angiogenic processes. The present results served to define incomplete ultrastructural differentiation of the microvascular endothelium during onset of CAM respiratory function. Despite this lack of differentiation, the angiogenic endothelium presented a highly restrictive barrier (Rizzo et al., 1995) which was comparable to that of adult pulmonary endothelia (Lanken et al., 1985), and could be teleologically relevant to the CAM's respiratory function.

Methods White Leghorn chick embryos (Truslow Farms; Chestertown, MD, USA) were incubated in ovo for 72 h at 37°C. At that time shell-less culture preparations were established and maintained in a humidified environment at 37°C (Auerbach et al., 1974). At days 4.5, 5.0, and 5.5 (developmental stages 26, 27, and 28 respectively; Hamburger and Hamilton, 1951), the culture dishes were placed on the temperature-controlled stage of an Olympus BH2 fluorescent microscope for intravital microinjections of a graded series of FITC-dextrans. Three embryos from each developmental stage were injected with either FITC-dextran 40, 70, or 150 kDa for 15rain perfusion periods. The dextrans (Sigma Chemical Co.) were prepared as a 5% solution in avian Ringers (pH 7.2) and microinjected via the vitelline vein at a dose of 100 gg/g. Thus, the injectate volumes were less than 2% of total circulating blood volume, and minimal alteration of embryonic hemodynamics would be expected (Wagman et al., 1990). Intravital observations of microvascular flow enabled precise identification of the pre- and postcapillary microvascular segments. The centripetal method of microvascular mapping was used to define the microvascular segments (DeFouw et al., 1988). In this case, capillaries served as

the point of reference and the immediately adjacent precapillary or postcapillary microvessel was defined as a first-order vessel. The union of two first-order vessels then served to define a second-order vessel. At the conclusion of the 15-rain perfusion periods, the major allantoic arteries and vein (proximally and distally remote from the observed microvascular units, respectively) were clamped prior to topical application of cold periodate-lysine-paraformaldehyde (2%) in phosphate buffer (McLean and Nakane, 1974). Since embryonic hematocrit and plasma protein concentrations are considerably lower than those in adult animals (Baumann et al., 1983), fixative-induced crosslinking between plasma proteins and between proteins and erythrocytes were less intense in the angiogenic CAM microvascular networks. Hence, vessel clamping was required to insure adequate fixation. After 30min immersion in the same fixative, each CAM was viewed with a stereomicroscope for microdissection of the firstand second-order precapillary microvessels from their postcapillary counterparts. Capillaries were routinely collected with both the pre- and postcapillary segmental samples. The respective microvascular segments were then processed for ultrastructural detection of the dextrans (Simionescu and Palade, 1971) and routinely embedded in Epon 812. During the embedding process, proper tissue orientation served to enable sequential sectioning exclusively through the capillary networks, the firstorder microvessels, and finally the second-order preand postcapillary vessels (Rizzo et al., 1993). Thin sections (60-90 nm) of the respective microvessel cross sections were stained with uranyl acetate and lead citrate and examined with a Philips EM300 microscope operating at 60 kV. Six pre- and six post-capillary samples, which included the capillaries, were evaluated at days 4.5, 5.0, and 5.5. 60 micrographs, at x 32 000 were obtained of the precapillary and postcapillary endothelia from the respective samples. In addition, 30 micrographs of the capillary endothelium were recorded from the pre- and postcapillary samples. This yielded 90 segmental micrographs per developmental stage, and a total of 270 micrographs for the morphometric analyses. Established stereological techniques (DeFouw, 1984) were applied to the respective segmental endothelia to determine: 1. average cytoplasmic thicknesses 2. average numerical densities of plasmalemmal vesicles 3. average numerical densities of vacuoles, defined as membrane-limited structures with diameters greater than 150 nm 4. mean depths (luminal to abluminal distances) and widths of the interendothelial junctional clefts. Mean values of the morphometric parameters from the precapillary, capillary, and postcapillary segments were compared within and between each developmental stage.

ENDOTHELIAL DIFFERENTIATION AND CAM ANGIOGENESIS

A one-way analysis of variance and the Bonferroni post test served to detect statistically significant differences, which were accepted at p < 0.05.

incomplete ultrastructural differentiation was presented by the uniformly low number of plasmalemmal vesicles (Table 1). The CAM precapillary and postcapillary endothelial vesicle densities were 20- to 100-fold less than those of adult lung (DeFouw, 1988). Vesicle densities in the CAM capillary endothelium were also 65- to 100-fold less than adult values. Thus, vesicular functions would likely be rudimentary in the CAM microvessels (Fig. 4). Endothelial vacuoles, in contrast to the plasmalemmal vesicles, were more prominent in the CAM than in the adult pulmonary microvasculature (Table 2). Diameters of the vacuoles were variable (400+ 125 nm) and their appearance was intermittent along the length of the microvascular unit. A definitive role of the CAM endothelial vacuoles, some of which indicate apparent continuity with the vessel lumen (Fig. 5), remains uncertain. Of the ultrastructural features assessed, the interendothelial clefts appeared most completely differentiated in the CAM microcirculation. Average widths of the clefts were uniform within the microvascular unit (Fig. 6), and these values were comparable to those (10 25 nm) within the adult microcirculation (Bundgaard and Frokjaer-Jensen, 1982). Luminal to abluminal depths of the CAM junctional clefts, which ranged from 0.6 gm

Resullts The segmental CAM microvascular endothelium at days 4.5, 5.0, and 5.5 of development was continuous but lacked an ultrastructurally distinct basal lamina (Fig. 1). Average thicknesses of the first-order and second-order precapillary endothelia (Fig. 2) ranged from 0.41 lain to 0.46 grn, and these values were substantially greater than those of adult lung precapillaries (DeFouw, 1988). Likewise, mean endothelial thicknesses of the first- and second-order postcapillaries, which ranged from 0.36 gm to 0.46 gin, were greater than those of adult lung (Fig. 2). Capillary endothelial cytoplasmic thicknesses were slightly less than the respective precapillary and postcapillary measurements; however, CAM capillary endothelia were substantially thicker than adult pulmonary capillary endothelia (Fig. 3). Thus, the early CAM segmental endothelium presented a temporally homogeneous thickness that was uniformly greater than adult pulmonary endothelial thickness. Further evidence of

L

i

5:::: <:

....

=

161

i =ii,¢

~!': ~

.

,

: ::

:

::]

Fig. 1 The junctional cleft (arrow) within the continuous endothelium of a first-order postcapillary at day 5.5 presents ultrastructural features consistent with mature tight junctional morphology. Note the absence of a distinct basal lamina. L: lumen; x 46 200.

162

p,Izzo ET AL

1.0 ~]precap-2

0,9

[]precap-1

0,8

npostcap-1

0,7

E

U'I lYl ¢.-

-,-t I---

~]postcap-2

0,6 0.5 0,4

f/// i///

i//

/I // t/ // // t/

/Yj

VA

i/J i/J

F/J

9.4, ,¢/j

0,3 0,2

T/J/ ,, "/ // t

F/.,,'J

0,1 0,0

//I /// I'// f // //1 /// ~,f, /// I/I ,'//

[/// f/l

~-"A

zA

45

N

9Z i/,,

I/..,'..'

ii,

5,0

Adult*

5,5

Days Fig. 2 Endothelial cytoplasmic thicknesses of first- and second-order pre- and postcapillary microvessels are presented. Values represent means_+ S.E.M. * adult lung values (DeFouw, 1988)
.5

.4

v i/l i11 t.2~

t~

,3

.2

rl--

I

4.5

I

5,0

I

I

5.5

Adult*

Days Fig. 3 Endothelial cytoplasmic thicknesses of the capillaries are presented. Values represent means _+S.E.M. * Adult lung values (DeFouw, 1988)
Precapillary order-2 order-1 Postcapillary order-2 order-1 Capillary

to 1.4 Ixm, were consistent along the length of the microvascutar unit, and were also comparable to adult values (Table 3).

4.5

5.0

5.5

Adult*

2.2_+0.9 3.1-+1.6

1.1 _+0.04 3.2_+1.1

1.3_+0.4 3.3_+0.9

109+_8 110_+10

Discussion

2.3-+0.7 2.2-+1.1 2.1_+0.7

2.3_+0.7 5.1_+1.6 3.0_+0.9

2.4+_0.9 3.2_+1.3 3.1_+1.1

129_+6 110_+ 10 1984-19

Formation of the CAM respiratory surface is completed between days 4.5 and 5.0 of the normal 21-day gestation. Since the CAM microcirculation undergoes progressive expansion through day 10 to satisfy increasing oxygen demands of the developing embryo, the CAM also

~ Values represent means +_S.E.M. *Mammalian pulmonary microcirculation (DeFouw, 1988).

ENDOTHELIAL DIFFERENTIATION AND CAM ANGIOGENESIS

163

Fig. 4 This portion of the endothelium from a 5.0 day postcapillary depicts the paucity of plasmalemmal vesicles (arrow). Note the portion of an erythrocyte (E) which demonstrates glycogen accumulation within the cytoplasm. L: lumen; x 52200.

Table2 Numericaldensities (number/~tm3) of cytoplasmicvacuoles within the segmentalmicrovascularendotheliat

Precapillary order-2 order-1 Postcapillary order-2 order-1 Capillary

4.5

5.0

5.5

Adult*

1.0_+0.4 1.2_+0.4

1.1_+0.4 2.1 _+0.4

1.2_+0.4 2.2_+0.7

0.3_+0.02 0

1.1_+0.4 1.0_+0.4 1.1 _+0.4

1.3_+0.9 2.2_+0.4 1.3 _+0.7

1.0_+0.7 2.3_+0.7 1.1 _+0.4

0.3_+0.03 0 0

t Values represent means ± S.E.M. *Mammalian pulmonary microcirculation (DeFouw, 1988).

serves as a model of normal angiogenesis. The present study is aimed at two primary objectives: 1. Determine the extent of endothelial differentiation (compared to adult pulmonary endothelia) during the onset of CAM respiratory function. 2. Compare CAM microvascular ultrastructure to that of tumorigenic microvessels in relation to their different permeability characteristics. Compared to results from a previous ultrastructural morphometric analysis of adult pulmonary endothelia (DeFouw, 1988), CAM microvascular endothelia were thicker, had fewer plasmalemmal vesicles, and contained large vacuoles. Only the interendothelial clefts presented adult-like tight junctional morphology. Since similar

morphometric techniques were used in both the present and previous studies from this laboratory (DeFouw, 1988), a direct comparison of the adult and embryonic endothelial ultrastructural characteristics seems reasonable. Although the relatively thick CAM capillary endothelium observed from days 4.5-5.5 would create greater diffusion distances for gas exchange than the adult airblood barrier, the CAM provides a principle respiratory surface during these early stages of embryogenesis. Perhaps, progressive reduction in capillary endothelial thickness through day 14 of development (Rizzo and DeFouw, 1993), serves to accommodate increased oxygen demands of the later stage embryo. Extravasation of dextran 40, and to a lesser extent dextran 70, from the CAM microvessels occurred transiently at day 4.5, while approximately 12 h later, the CAM endothelial barrier restricted these tracers (Rizzo et al, 1995). Further, the barrier functions observed at day 5 were maintained from days 6 through 14 of CAM morphogenesis (Rizzo et al., 1993; Rizzo and DeFouw, 1993). In adult lungs, the microvascular endothelium was defined previously to be restrictive to extravasation of dextrans ~>40 kDa (Lanken et al., 1985). Thus, it is important to note that adult-like permselectivity was apparent in the early CAM microcirculation when angiogenesis was still proceeding. Such macromolecular

164

RIZZO

ET

AL

*'4

L

4"

Fig. 5 An endothelial vacuole (v) is presented in this segment of a 4.5 day postcapillary. The presence of dextran particles (arrowhead) within the vacuole lumen suggests vacuolar continuity with the vessel lumen (L) outside the plane of section. P :perivascular cell x 34 800.

40

[~precap-2 []precap-I [-]capillary ,p0stcap-1 []p0stcap-~

35 30 ,--. E

25

r v

.c

20

4J

15 10

lliill

t=t;w. ?????2

iiiii

i:i:i:i:ii:i:?i / / / ,v.-.';:;,., Y/~ :i:;ii:i:ii: :/// /// /A/ /// ///

y-?----: .:.1.:--._--.:.

• i r • ...=,:.:,

Y//~ i:i:i~:i:)i:i:i ;//./

4.5

lii!i

='.=::'.w

5.0

5.5

Days Fig.

6

Average widths o f junctional clefts within the C A M segmental microvascular endothelia are presented. Values represent means

restriction likely contributes to maintenance o f the C A M air-blood barrier during the angiogenic process. That developing pulmonary capillary endothelia offered minimal resistance to a variety o f macromolecules (Matlon

_. S.E.M.

and Wangensteen, 1977) further highlights the unique barrier functions o f the angiogenic C A M microvessels. Macromolecular extravasation is generally related to the transcellular (vesicular) or the paracellular (junc-

ENDOTHELIAL DIFFERENTIATION AND CAM ANGIOGENESIS

Table 3 Luminal to abluminal depths (gm) of the junctional clefts within the segmental microvascular endothelia'~

Precapillary order-2 order-1 Postcapillary order-2 order-1 Capillary

4.5

5.0

5.5

Adult*

1.1_+0.2 0.9-+0.1

1.0_+0.2 1.1 -+0.2

1.2_+0.2 1.0_+0.2

0.5_+0.1 0.4_+0.03

0.6_+0.1 1.1_+0.2 0.9_+0.2

1.1 _+0.2 1.0_+0.1 1.2-+0.1

1.2_+0.2 1.4_+0.2 0.9_+0.1

0.7-+0.1 1.0+0.1 1.1 _+0.1

tValues represent means _+S.E.M. * Frog mesenteric microcirculation (Bundgaard and Frokjaer-Jansen,

1982). tional) pathways. The transcellular pathway, however, displayed rudimentary differentiation in the CAM endothelium between days 4.5 and 5.5. Such !ack of differentiation, ,characterized by a paucity of endothelial vesicles, has also been reported in the developing lung (Schneeberger, 1983). The abrupt increase in selectivity between days 4.5 and 5.0 (Rizzo et al., 1995) also occurred while vesicle densities remained uniformly low. Thus, the ontogeny of CAM endothelial selectivity is likely dependent upon cellular domains other than the vesicles. In contrast to the transcellular pathway, junctional clefts between the adjacent CAM endothelial cells met the ullLrastructural definition of tight junctions (Bundgaard, 1984) as early as day 4.5 of development. This tig]ht ultrastructural morphology is consistent with the uniform restriction of dextran 150 from days 4.5 to 14 (Rizzo et al., 1993; Rizzo and DeFouw, 1993; Rizzo et al., 1995). Endothelial restriction of dextrans 40 and 70, on the other hand, increased from days 4.5 to 5.0 (Rizzo et al., 1995). Since tight junctional morphology remained constant from day 4.5, factors other than junctional dimensions likely contribute to selectivity of the paracellular pathway. Although dextran 40 particles had access to the junctional clefts at day 4.5, the tracer was restricted to the luminal space at days 5.0 and 5.5 (Rizzo et al., 1995). A fiber matrix, comprised of glycoconjugates within the luminal plasmalemma of adult capillary endothelia has been described as a serial element of the microvascular barrier (Curry and Michel, 1980). That the matrix serves as a chromatographic filter associated with the transendothelial pathways is consistent with more recent applications of this concept (Adamson, 1992; Katz, 1992). In addition, a family of cell-cell adhesion molecules was recently identified at sites of membrane contact within junctional clefts of adult endothelia (Heimark et al., 1990; Lampugnoni et al., 1992). Thus, differential temporal or spatial expression of biomolecular domains within the CAM fiber matrix and junctional clefts between days 4.5 and 5.0 might contribute to the ontogeny of junctional cleft selectivity. Additional studies are needed to test this concept. Vacuoles are consistent features of early developing

165

microvascular networks (Feinberg, 1991). Although structural continuity of vacuoles with the luminal space was suggested by the observation of dextran particles within the vacuoles, serial section ultramicrotomy, however, is required to establish three-dimensional morphology of the vacuoles. Despite the continuous presence of vacuoles from days 4.5-5.5, extravasation of dextrans 40 and 70 was observed at day 4.5 only and extravasation of dextran 150 was not observed (Rizzo et al., 1995). These results are not consistent with an association of dextran extravasation and the endothelial vacuoles. Intracellular fusion of vacuoles, on the other hand, has been described as a mechanism for new microvessel lumen formation during embryonic angiogenesis (Clark and Clark, 1937). Accordingly, CAM endothelial vacuoles were absent at days 10 and 14 (Rizzo and DeFouw, 1993) when CAM angiogenesis and new lumen formation would have been completed. That the vacuoles were apparently continuous with the luminal space at days 4.5-5.5 also provides support to the concept that CAM endothelial vacuoles contribute to the neovascularization process. That angiogenic microvascular endothelia are generally less restrictive to macromolecules than their adult counterparts is based on studies of pathologic and tumor-induced angiogenesis. Ultrastructural studies of tumorigenic microvessels which permitted exaggerated macromolecular extravasation demonstrated expanded junctional clefts (Ausprunk and Folkman, 1977) and interconnecting transendothelial chains of vesicles and vacuoles (Kohn et al., 1992), both of which were implemented in the diminished selectivity. However, excessive macromolecular leakage and altered paracellular and transcellular pathways were not uniformly observed during the tumorigenic process (Heuser and Miller, 1986; Paku and Paweletz, 1991). Since the present results also served to define an intact endothelial barrier in the angiogenic microvessels of the embryonic CAM, angiogenesis is likely not a routinely homogeneous process. That CAM endothelial ultrastructure differed from that of certain tumorigenic microvessels is also consistent with the concept that neovascularization is regulated, in part, by local tissue factors. Further, expansion of the CAM is rapid; thus, the respective rates of angiogenesis in various tissues might be variable. Since CAM microvascular permselectivity increased over a 12-h span from days 4.5-5.0 of development (Rizzo et al, 1995), additional assessment of the CAM endothelium between day 4.0 and 4.5 would further test this concept of angiogenic heterogeneity. In conclusion, the CAM segmental microvasculature from days 4.5-5.5 displayed a uniformly thick endothelium with a paucity of plasmalemmal vesicles and apparently tight junctional clefts. Although this ultrastructural uniformity does not account for the relative increase in macromolecular selectivity observed between days 4.5 and 5.0 (Rizzo et al., 1995), it is consistent

166

RIZZO ET AL

with the respiratory function of the CAM. If the CAM endothelium, like developing pulmonary endothelia (Marion and Wangensteen, 1977), permitted excessive macromolecular leakage, the gas diffusion barrier, which includes the relatively thick endothelium at this early stage of development, would be compromised. Thus, the restrictive endothelium of the angiogenic CAM microvessels which, despite incomplete differentiation,

is comparable to that of adult pulmonary microvessels (Lanken et al., 1985), may be of teleologic significance. ACKNOWLEDGMENTS The authors thank Lisa Sweetman-Venezio for her technical assistance. This work was supported by a grant from the USPHS (NIH grant HL-47936).

REFERENCES Adamson, R.H. 1992. An extension of the fiber matrix model of vascular permeability. Microvasc. Res. 43: 352-356. Auerbach, R., Kabai, L., Knighton, D. and Folkman, J. 1974. A simple procedure for the long-term cultivation of chicken embryos. Develop. Biol. 41: 391-394. Ausprunk, D.H., Knighton, D.R. and Folkman, J. 1974. Differentiation of vascular endothelium in the chick chorioallantoic membrane: a structural and autoradiographic study. Develop. Biol. 38:237 248. Ausprunk, D.H. and Folkman, J. 1977. Migration and proliferation of endothelial cells in preformed and new blood vessels during tumor angiogenesis. Microvasc. Res. 14:53 64. Baumann, R., Padekeu, S., Haller, E.A. and Brilmayer, T. 1983. Effects of hypoxia on oxygen affinity, hemoglobin pattern, and blood volume of early chicken embryos. Am. J. Physiol. 244: R733-R742. Bundgaard, M. and Frokjaer-Jensen, J. 1982. Functional aspects of the ultrastructure of terminal blood vessels: a quantitative study on consecutive segments of the frog mesenteric microvasculature. Microvasc. Res. 23: i-30. Bundgaard, M. 1984. The three dimensional organization of tight junctions in a capillary endothelium revealed by serial-section electron microscopy. J. Ultrastruct. Res. 88: 1-17. Clark, E.R. and Clark, E.L. 1937. Observations on living mammalian lymphatic capillaries, their relationship to blood vessels. Am. J. Anat. 23:37 88. Curry, F.E. and Michel, C.C. 1980. A fiber matrix model of capillary permeability. Microvasc. Res. 20:96 99. DeFouw, D.O. 1984. Vesicle numerical densities and cellular attenuation: comparisons between endothelium and epithelium of the alveolar septa in normal dog lungs. Anat. Rec. 209: 77-84. DeFouw, D.O. 1988. Structural heterogeneity within the pulmonary microcirculation of the normal rat. Anat. Rec. 221: 645-654. DeFouw, D.O., Rizzo, V.J., Steinfeld, R. and Feinberg, R.N. 1989. Mapping of the microcirculation in the chick chorioallantoic membrane during normal angiogenesis. Microvasc. Res. 38: 136-147. Feinberg, R.N., Shumko, J.Z., Steinfeld, R. and Sweetman, L. 1991. Endothelial heterogeneity in the chick wing bud: a morphometric study. Anat. Embryol. 184:47 53. Flamme, I., Schulze-Osthoff, K. and Jurgen Jacob, H. 1991. Mitogenic activity of chicken chorioallantoic fluid is temporally correlated to vascular growth in the chorioallantoic membrane and related to fibroblast growth factors. Development 111: 683-690. Fuchs, A. and Lindenbaum, E.S. 1988. The two- and threedimensional structure of the microcirculation of the chick chorioallantoic membrane. Anat. Rec. 131: 271-275. Gerlowski, L.E. and Jain, R.K. 1986. Microvascular permeability of normal and neoplastic tissues. Microvasc. Res. 31: 288-305. Hamburger, V. and Hamilton, H. 1951. A series of normal stages in the development of the chick embryo. J. Morphol. 88: 49-92. Heimark, R.L., Degner, M. and Schwartz, S.M. 1990. Identification of a calcium-dependent cell-cell adhesion molecule in endothelial cells. J. Cell Blot. 110:1745 1756.

Heuser, L.S. and Miller, F.N. 1986. Differential macromolecular leakage from vasculature of tumors. Cancer 57:461 464. Katz, M.A. 1992. Structural change in fiber matrix allows for enhanced permeability and reduced hydraulic conductivity. Microvasc. Res. 43: 1-6. Kohn, S., Nagy, J.A., Dvorak, H.F. and Dvorak, A.M. 1992. Pathways of macromolecular tracer transport across venules and small veins. Lab. Invest. 67: 596-607. Lampugnoni, M.G., Resnati, M., Raiteri, M., et al. 1992. A novel endothelial-specific membrane protein is a marker for cell-cell contacts. J. Cell Biol. 118: 1511-1522. Lanken, P.N., Hansen-Flaschen, J.H., Sampson, P.M., Pietra, G.G., Haselton, F.R. and Fishman, A.P. 1985. Passage of uncharged dextrans from blood to lymph in awake sheep. J. Appl. Physiol. 59: 580-591. Matlon, S.V. and Wangensteen, O.D. 1977. Pulmonary filtration and reflection coefficients in the newborn rabbit. Microvasc. Res. 14: 99-110. McLean, I.W. and Nakane, P.K. 1974. Periodate-lysineparaformaldehyde fixation: a new fixative for immunoelectron microscopy. J. Histochem. Cytochem. 22:1077 1083. Paku, S. and Paweletz, N. 1991. First steps of tumor related angiogenesis. Lab Invest. 65: 334-346. Rizzo, V., Steinfeld, R., Kyriakides, C. and DeFouw, D.O. 1993. The microvascular unit of the 6-day chick chorioallantoic membrane: a fluorescent confocal microscopic and ultrastructural morphometric analysis of endothelial permselectivity. Microvasc. Res. 46:320 332. Rizzo, V. and DeFouw, D.O. 1993. Macromolecular selectivity of chick chorioallantoic membrane microvessels during normal angiogenesis and endothelial differentiation. Tissue and Cell 25(6): 847-856. Rizzo, V., Kim, D., Duran, W.N. and DeFouw, D.O. 1995. Ontogeny of microvascular permeability to macromolecules in the chick chorioallantoic membrane during normal angiogenesis. Microvac. Res. 49: 49-63. Schneeberger, E.E. 1983. Plasmalemmal vesicles in pulmonary capillary endothelium of developing fetal lamb lungs. Microvasc. Res. 25: 40-55. Simionescu, N. and Palade, G. 1971. Dextrans and glycogens as particulate tracers for studying capillary permeability. J. Cell Biol. 64: 586-607. Wagman, AJ., Hu, N. and Clark, E.B. 1990. Effect of changes in circulating blood volume on cardiac output and arterial and ventricular blood pressure in the stage 18, 24, and 29 chick embryo. Circ. Res. 67: 187-192. Wu, N.Z., Klitzman, B., Rosner, G., Needham, D. and Dewhirst, M.W. 1993. Measurement of material extravasation in microvascular networks using fluorescence video-microscopy. Microvasc. Res. 46: 231-253. Yuan, F., Leunig, M., Berk, D.A. and Jain, R.K. 1993. Microvascular permeability of albumin, vascular surface area, and vascular volume measured in human adenocarcinoma LS174T using dorsal chamber in SCID mice. Microvasc. Res. 45: 269-289.