- Email: [email protected]
Epidermal growth factor promotes chick embryonic angiogenesis
Cell Biology
EPIDERMAL
International
Reports,
Vol. 13, No. II, November
957
1989
GROWTH
FACTOR PROMOTES CHICK EMBRYONIC ANGIOGENESIS Stewart’, J. Nelson2 and D.J. Wilson3
Rosemary Biochemistry2 and Dental Surgery3, Departments of Anatomy’, The Queen’s University of Belfast, Northern Ireland. Abstract:
The responseof the early extraembryonicvasculatureto epidermalgrowth factor (EGF) was studied. Millipore filter discs containing long-lug of EGF were placed onto the advancing edgeof the area vasculosa of 3-day chick embryos, and the effect examined macroscopically and histologically 24 hours after disc application. The capillary density at the site of application increased significantly, and the effect was seen to be dose-dependent:a similar but more marked responsewas observedin the vesselcross-sectionalareaper unit length. This change in vascularity was accompaniedby tortuous folding of the mesodermand endodenn (which normally lie parallel to the ectoderm in a trilaminar arrangement) into the yolk substance:this may indicateprecocious development of all three layers (ectoderm, mesodermand ectoderm)of the membrane,and the proliferative effects of EGF may not be confined to the vascular endotbelium.
Introduction: Numerous polypeptide growth factors have been identified as promotors of
in that they have been shown to affect one or several of the cellular activities essential for new capillary development (Folkman and Klagsbrun, 1987). These activities have been divided into several discrete steps: changes in vessel angiogenesis
permeability
with associated decrease in inter-endothelial
cell junctions,
proteolytic
enzyme dissolution of the vessel basement membrane, migration of endothelial cells
towards the angiogenic stimulus forming a capillary sprout, and formation of a patent lumen (Furcht, 1986). The growth factor EGF has been shown to stimulate both migration and proliferation of endothelial cells in vitro (McAuslan et al, 1985) although the effects of EGF in vivo have not been well documented. EGF has been shown to promote angiogenesis in the rabbit cornea (BenEzra, 1978; Gospodarowicz et al, 1979) and in the hamster cheek pouch (Schreiber et al, 1986). However, both these animal models have draw-backs, particularly in the quantification of their results and in the avoidance of an inflammatory reaction and subsequent immunological neovascularization (Folkman, 1985). The capillary network of the chick vitelline
membrane (VIM)
appears viable as an alternative
in
vivo model for the study of angiogenesis factors as it is simple and relatively undifferentiated. In addition, the chick immune system does not develop until day 7, and therefore any vascular response would not be complicated by an inflammatory reaction. 0309-I 651/89/l 10957-9/$03.00/O
0 1989 Academic
Press Ltd.
958
Cell Biology
International
Reports,
Vol. 13, No. I 1, November
1989
Materials and Methods: Fertile chick eggs which had been incubated for 2-3 days were set up as shell-less cultures (modified from Dunn and Boone, 1976; fig. la,b) and incubated further at 37’C and 1.5% CO2. Millipore discs (diameter 4.5 mm, pore size 0.45 l.rrn) containing 1 p.g and 10 ng EGF (from mouse submaxillary gland, Sigma Chemicals) were placed on the outgrowing edge of the area vasculosa (sinus terminalis) of stage 18-20 embryos (Hamburger and Hamilton, 1951), with discs containing sterile distilled water as “sham” controls. Discs were positioned on the sinus terminalis between terminal branches of the right vitelline (omphalomesenteric) vessels (fig. lc).
India Ink Injection: Twenty-four hours after their application, the discs were removed and the whole vasculature (embryonic and extraembryonic) was infused with sonicated India ink following the method of Wilson (1983). The whole area vasculosa was excised and lifted using a glass coverslip. After thorough washing in distilled water, it was flattened between two coverslips and fixed in 5% TCA for 12 hours, before dehydration through a graded series of alcohols and clearing with methyl salicylate.
Wax Histology and Image Analysis: Six cultures were treated in each group (lpg EGF, 10 ng EGF, sterile distilled water) with a further control group of six similar cultures left untreated. The cultured embryos were returned to the incubator and the macroscopic and histological effects examined 24 hours after application of the discs. The area of membrane subjacent to each disc was fixed in culture with formal saline. Following excision, fixation was continued for 3 hours and specimens were dehydrated through a graded series of alcohols, cleared with xylene and embedded in paraffin wax. Sections of thickness 7pm were cut and stained with Harris haematoxylin. To calculate the vessel cross-sectional area, tracings were made directly onto a bit-pad using a Leitz Diaplan microscope with a camera lucida attachment. These were analysed using a B.B.C. master computer with DIGIT image analysis software.
Resin Histology: To give a clearer view of histological detail, the effects of a range of doses of EGF were examined by means of semi-thin resin sections (thickness 0.75 l.trn). The procedure used was as before, but with doses of 1 pg, 100 ng, 10 ng and 1 ng EGF applied. Specimens were fixed in cold 3% glutaraldehyde, osmicated, dehydrated and embedded in Araldite resin. Sections were stained with Toluidine blue.
Cell Biology
International
Reports,
Vol. 73, No. 7 1, November
1989
959
Results: When the discs were removed, the effects of EGF on the underlying membrane could be seen with the naked eye as a distinct reddened area. In all cases, India ink injection revealed a marked increase in capillary density at the site of application of EGF. In these areas, the sinus terminalis was no longer evident as a single wide-bored vessel, but appeared as an irregular capillary plexus (fig. 2a, b). In some membranes, the outgrowth of the area vasculosa seemed to have been inhibited (fig. 2b), although this was not a constant finding and did not appear to be dose-related as a similar proportion (approximately 50%) of membranes showed this inhibition at each of the doses examined (Q.tg and 10 ng). Similar areas of inhibited outgrowth were observed in membranes treated with discs containing sterile distilled water. EGF was seen to have a pronounced effect on the histological appearance of the treated area vasculosa. The ectoderm, mesoderm and endoderm normally lie around the yolk surface in a relatively parallel, trilaminar arrangement (fig. 3a). In membranes treated with a 1 ug dose of EGF, the pattern was changed dramatically, the mesoderm and endoderm forming tortuous folds projecting deep into the yolk substance (fig. 3d). These folds, which appeared villus-like in section, consisted of a core of mesoderm, with a greatly increased capillary density, surrounded by a layer of endoderm, apparently of normal thickness. Despite this atypical appearance, the cellular detail of both mesodermal and endodermal layers appeared unaffected, in that the mesoderm consisted of loose vascular connective tissue and the endoderm of large polygonal cells packed with yolk droplets. Folding of the two basal layers into the yolk was also observed, although to a lesser extent, in membranes treated with 10 ng of EGF (fig. 3~). The ectoderm, in general, appeared unaffected by the application of EGF, although at the higher dose small focal areas of ectodermal hyperplasia were observed in some membranes (fig. 3d). Application of discs containing sterile distilled water caused no histological change in the mesodermal and endodermal layers, although in two specimens small areas of ectodermal hyperplasia, similar to those observed with the 1 l.tg dose of EGF, were seen (fig. 3b). In the remaining water-treated membranes, the ectoderm appeared normal, being composed of a thin layer of flattened cells. From the wax histological sections, it was apparent that specimens treated with EGF at both doses showed a considerable increase in vascular density. This
960
Ceil Biology
International
Reports,
Vol. 13, No. 17, November
1989
3d Fig. 1. Shell-less culture preparations (a) from above; (b) lateral view; (c) showing the position of the Millipore disc on the sinus terminalis between terminal branches of the right vitelline vessels. Fig. 2. India ink injection of regions of the area vasculosa treated with (a) 1 l.tg; and (b) 10 ng of EGF. In each case, there is a marked increase in capillary density at the site of application. Fig. 3. Wax histology of (a) an untreated region of area vasculosa, and those treated with (b) sterile distilled water; (c) 10 ng of EGF, and (d) 1 l.tg of EGF. The EGF-treated membranes show capillary proliferation accompanied by folding of the mesodermal and endodermal layers into the yolk. (All x100)
Cell Biology
International
Reports,
Vol. 73, No. 11, November
1989
961
effect was quantified by counting the number of capillaries, and measuring their cross-sectional area in a fixed length of membrane. The mean capillary density at the site of application increased in a dose-dependent manner, with an approximate rise of 20% at the 10 ng dose, and nearly a three-fold rise at the 1 p,g dose in comparison with untreated controls (fig. 4a). A similar but more marked response was observed in the vessel cross-sectional area per unit length: at the 10 ng dose, this area was nearly three times that of the corresponding control area and at the 1 pg dose, a nearly seven-fold increase was recorded (fig. 4b). As the rise in cross-sectional area of the capillaries was more pronounced than that in their number, their mean calibre must also increase. However, proliferation of small vessels was also apparent. Application of discs containing sterile distilled water caused no increase in either the number or cross-sectional area of vessels in unit length of membrane. As the parametric assumptions of normality of distribution and homogeneity of variance were not satisfied, the median values for the capillary density and crosssectional area were calculated for each specimen (Table I) and analysed using the non-parametric Kruskal-Wallis analysis of variance and a series of Mann-Whitney U tests. No significant difference in the values of either parameter were found between untreated controls and “sham” controls treated with sterile distilled water (p>O.O5), and these groups were pooled as controls in further tests. The capillary density in EGF-treated membranes at both doses was significantly greater than in the pooled control group (~~0.01 in each case), and there was a significant difference in the vessel density between these two test groups (~~0.05). Similarly, significant differences in the vessel cross-sectional area in a standard length of membrane were found for each EGF-treated group when compared with controls (~~0.01 for the 1 lrg dose; p
962
Cell Biology
4a
Reports,
Vol. 13, No. 71, November
1989
35.. 30.
E -f : 5 I s” 5s 5
International
.
25. 20. .
15.
: 10.
.
5.
1
O+ EGF lug
EGF long
I H&I Control
COntIOl
4b 16000
.
14000
ml m2000. ~10000. z 2 6000. 2
. 1
% 6000. I 4000-
. . . :
.
2000.
I i
!
H,O Conlml
COntrOl
03
I EGF lug
EGF long
Fig. 4. Scattergramsshowing the median vessel number (a) and median vessel cross-sectionalarea (b) per unit length of areavasculosafor embryos treatedwith 1 ttg and 10 ng of EGF, and for watertreatedand untreatedcontrols.
Table I. Median values of cross-sectional area (and vessel number) per unit length for each embryo studied. EGF lue
EGF 10 ne
Control (no disc)
Water Control
7802.5 (9.5)
3708.5 (6.0)
1323.0 (6.5)
1139.0 (6.0)
15328.0 (16.5)
2560.0 (6.0)
1617.5 (6.0)
1336.0 (5.5)
3584.0 (13.0)
4900.0 (7.0)
1267.0 (5.5)
2160.0 (5.0)
11518.0 (30.5)
2984.5 (6.5)
1694.5 (7.0)
1077.5 (4.0)
8066.0 (14.0) ---
2914.5 (9.5)
1416.5 (5.0)
1707.0 (6.0)
5506.5 (7.0)
897.0 (5.5)
2471.5 (7.5)
Cell Biology
International
Reports,
Vol. 13, No. 11, November
1989
963
form at lower doses, with the vessels “stacking” on top of one another (fig. 5d). The ectoderm showed little histological change, with no evidence of hyperplasia in any of the specimens studied. In the endoderm, however, the staining of the material in the yolk inclusions appeared to be less dense in treated membranes in contrast to corresponding controls. This decrease in staining intensity seemed to be dose-dependent, being noticeably less marked in membranes treated with 1 ng of EGF than in those subjected to higher doses,
Fig. 5. Resin histology of a length of untreated area vasculosa (a); and of regions treated with (b) 1 ng; (c) 10 ng; (d) 100 ng; and (e) 1 pg of EGF. (All x100)
964
Cell Biology
International
Reports,
Vol. 13, No. 11, November
1989
Discussion: The results of this study reveal that EGF has a clear promoting effect upon embryonic angiogenesis in that it causes a marked increase in both the density and cross sectional area of the capillaries per unit length of the area vasculosa. These results are in accordance with the findings of Schreiber et al (1986) that EGF stimulated angiogenesis, in the absence of an inflammatory response, in the hamster cheek pouch. EGF at a dose of 10 l.t.gshowed a small, but significant, increase in limbal capillary density in the rabbit cornea and was potent in stimulating DNA synthesis and proliferation
of cornea1 keratocytes in vitro (BenEzra, 1978).
Subsequently, Gospodarowicz et al (1979) observed a similar response to EGF at doses between 1 l.tg and 60 pg in the rabbit cornea. In addition, topical administration of EGF to corneas denuded of their epithelium resulted in not only a rapid re-epithelialization of the denuded area, but also included proliferation of the blood vessels in the limbal region (Savage and Cohen, 1972). The angiogenic activity of EGF can be explained by its ability to induce capillary endothelial cell migration and proliferation (McAuslan et al, 1985). These authors suggest that the induction results from the formation of a receptor-EGF complex which triggers off a series of intracellular events. However, it is uncertain whether the vessels are specifically affected. In the chick area vasculosa, the unusual projection of the mesoderm and endoderm into the yolk sac in response to EGF may simply be a measure to accomodate for the vast increase in vascular density in the mesodermal layer. However, it is important to note that simular tortuous folds are seen at later developmental stages and in older, more medial regions of the area vasculosa (Romanoff, 1960). The elevated capillary density, then, may be related to precocious development of all three layers of the membrane (ectoderm, mesoderm and endoderm), and the proliferative effects of EGF may not be confined to the vascular endothelium. This possibility is supported by the decreased staining intensity of the endodermal yolk inclusions in membranes treated with EGF, which could, in turn, be the result of the enzymatic degradation of their contents in the endoderm itself, and removal of the products by the mesodermal vessels. The ability of the endodermal cells to secrete the necessary proteolytic enzymes evolves naturally with the maturation of the membrane (Romanoff 1960). Significantly, EGF has already been shown to enhance development in mamalian systems, causing accelerated eruption of incisor teeth (Topham et al, 1987) and precocious eyelid opening in newborn mice (Smith et al, 1985).
Cell Biology
International
Reports,
Vol. 13, No. 1 I, November
1989
965
Bibliography: BenEzra, D. (1978). Neovasculogenic ability of prostaglandins, growth factors and synthetic chemoattractants. American Journal of Ophthalmology 86: 455-461. Dunn, B.E. & Boone, M.A. (1976). Growth of the chick embryo in vitro. Poultry Science 55: 1067-1071. Folkman, J. (1985). Tumour angiogenesis. Advances in Cancer Research 43: 175203. Folkman, J. & Klagsbum, M (1987). A family of angiogenic peptides. Nature 329: 671-672. Furcht, T.L. (1986). Critical factors controlling angiogenesis: cell matrix, and growth factors. Laboratory Investigation 55: 505-509. Gospodarowicz, D., Bialecki, H. & Thakral, K.K. (1979). The angiogenic activity of fibroblast and epidermal growth factors. Experimental Eye Research 28: 5015 14. Hamburger, V. & Hamilton, H.L. (1951). A series of normal stages in the development of the chick embryo. Journal of Morphology 88: 49-92. McAuslan, B.R., Bender, V., Reilly, W. & Moss, B.A. (1985). New functions of epidermal growth factor: stimulation of capillary endothelial cell migration and matrix dependent proliferation. Cell Biology International Reports 9: 175-l 82. Romanoff, A.L. (1960). “The Avian Embryo”, Macmillan, New York. Savage, C.R. & Cohen, S. (1972). Epidermal growth factor and a new derivative: rapid isolation procedures and biological and chemical characterization. Journal of Biological Chemistry 247: 7609-7611. Schreiber, A.B., Winkler, M.E. & Derynck, R. (1986). Transforming growth factor : a more potent angiogenic mediator than epidermal growth factor. Science 232: 1250- 1253. Smith, J.C., Spom, M.B., Roberts, A.B., Derynck, R. & Winkler, M.E. (1985). Human transforming growth factor-alpha causes precocious eyelid opening in newborn mice. Nature 315: 515-516. Topham, R.T., Chiego, D.J., Gattone, V.H., Hinton, D.A. & Klein, R.M. (1987). The effect of epidermal growth factor on neonatal incisor differentiation in the mouse. Developmental Biology 124: 532-543. Wilson, D.J. (1983). The origin of the endothelium in the developing marginal vein of the chick wing-bud. Differentiation 30: 183-187.
EPIDERMAL
International
Reports,
Vol. 13, No. II, November
957
1989
GROWTH
FACTOR PROMOTES CHICK EMBRYONIC ANGIOGENESIS Stewart’, J. Nelson2 and D.J. Wilson3
Rosemary Biochemistry2 and Dental Surgery3, Departments of Anatomy’, The Queen’s University of Belfast, Northern Ireland. Abstract:
The responseof the early extraembryonicvasculatureto epidermalgrowth factor (EGF) was studied. Millipore filter discs containing long-lug of EGF were placed onto the advancing edgeof the area vasculosa of 3-day chick embryos, and the effect examined macroscopically and histologically 24 hours after disc application. The capillary density at the site of application increased significantly, and the effect was seen to be dose-dependent:a similar but more marked responsewas observedin the vesselcross-sectionalareaper unit length. This change in vascularity was accompaniedby tortuous folding of the mesodermand endodenn (which normally lie parallel to the ectoderm in a trilaminar arrangement) into the yolk substance:this may indicateprecocious development of all three layers (ectoderm, mesodermand ectoderm)of the membrane,and the proliferative effects of EGF may not be confined to the vascular endotbelium.
Introduction: Numerous polypeptide growth factors have been identified as promotors of
in that they have been shown to affect one or several of the cellular activities essential for new capillary development (Folkman and Klagsbrun, 1987). These activities have been divided into several discrete steps: changes in vessel angiogenesis
permeability
with associated decrease in inter-endothelial
cell junctions,
proteolytic
enzyme dissolution of the vessel basement membrane, migration of endothelial cells
towards the angiogenic stimulus forming a capillary sprout, and formation of a patent lumen (Furcht, 1986). The growth factor EGF has been shown to stimulate both migration and proliferation of endothelial cells in vitro (McAuslan et al, 1985) although the effects of EGF in vivo have not been well documented. EGF has been shown to promote angiogenesis in the rabbit cornea (BenEzra, 1978; Gospodarowicz et al, 1979) and in the hamster cheek pouch (Schreiber et al, 1986). However, both these animal models have draw-backs, particularly in the quantification of their results and in the avoidance of an inflammatory reaction and subsequent immunological neovascularization (Folkman, 1985). The capillary network of the chick vitelline
membrane (VIM)
appears viable as an alternative
in
vivo model for the study of angiogenesis factors as it is simple and relatively undifferentiated. In addition, the chick immune system does not develop until day 7, and therefore any vascular response would not be complicated by an inflammatory reaction. 0309-I 651/89/l 10957-9/$03.00/O
0 1989 Academic
Press Ltd.
958
Cell Biology
International
Reports,
Vol. 13, No. I 1, November
1989
Materials and Methods: Fertile chick eggs which had been incubated for 2-3 days were set up as shell-less cultures (modified from Dunn and Boone, 1976; fig. la,b) and incubated further at 37’C and 1.5% CO2. Millipore discs (diameter 4.5 mm, pore size 0.45 l.rrn) containing 1 p.g and 10 ng EGF (from mouse submaxillary gland, Sigma Chemicals) were placed on the outgrowing edge of the area vasculosa (sinus terminalis) of stage 18-20 embryos (Hamburger and Hamilton, 1951), with discs containing sterile distilled water as “sham” controls. Discs were positioned on the sinus terminalis between terminal branches of the right vitelline (omphalomesenteric) vessels (fig. lc).
India Ink Injection: Twenty-four hours after their application, the discs were removed and the whole vasculature (embryonic and extraembryonic) was infused with sonicated India ink following the method of Wilson (1983). The whole area vasculosa was excised and lifted using a glass coverslip. After thorough washing in distilled water, it was flattened between two coverslips and fixed in 5% TCA for 12 hours, before dehydration through a graded series of alcohols and clearing with methyl salicylate.
Wax Histology and Image Analysis: Six cultures were treated in each group (lpg EGF, 10 ng EGF, sterile distilled water) with a further control group of six similar cultures left untreated. The cultured embryos were returned to the incubator and the macroscopic and histological effects examined 24 hours after application of the discs. The area of membrane subjacent to each disc was fixed in culture with formal saline. Following excision, fixation was continued for 3 hours and specimens were dehydrated through a graded series of alcohols, cleared with xylene and embedded in paraffin wax. Sections of thickness 7pm were cut and stained with Harris haematoxylin. To calculate the vessel cross-sectional area, tracings were made directly onto a bit-pad using a Leitz Diaplan microscope with a camera lucida attachment. These were analysed using a B.B.C. master computer with DIGIT image analysis software.
Resin Histology: To give a clearer view of histological detail, the effects of a range of doses of EGF were examined by means of semi-thin resin sections (thickness 0.75 l.trn). The procedure used was as before, but with doses of 1 pg, 100 ng, 10 ng and 1 ng EGF applied. Specimens were fixed in cold 3% glutaraldehyde, osmicated, dehydrated and embedded in Araldite resin. Sections were stained with Toluidine blue.
Cell Biology
International
Reports,
Vol. 73, No. 7 1, November
1989
959
Results: When the discs were removed, the effects of EGF on the underlying membrane could be seen with the naked eye as a distinct reddened area. In all cases, India ink injection revealed a marked increase in capillary density at the site of application of EGF. In these areas, the sinus terminalis was no longer evident as a single wide-bored vessel, but appeared as an irregular capillary plexus (fig. 2a, b). In some membranes, the outgrowth of the area vasculosa seemed to have been inhibited (fig. 2b), although this was not a constant finding and did not appear to be dose-related as a similar proportion (approximately 50%) of membranes showed this inhibition at each of the doses examined (Q.tg and 10 ng). Similar areas of inhibited outgrowth were observed in membranes treated with discs containing sterile distilled water. EGF was seen to have a pronounced effect on the histological appearance of the treated area vasculosa. The ectoderm, mesoderm and endoderm normally lie around the yolk surface in a relatively parallel, trilaminar arrangement (fig. 3a). In membranes treated with a 1 ug dose of EGF, the pattern was changed dramatically, the mesoderm and endoderm forming tortuous folds projecting deep into the yolk substance (fig. 3d). These folds, which appeared villus-like in section, consisted of a core of mesoderm, with a greatly increased capillary density, surrounded by a layer of endoderm, apparently of normal thickness. Despite this atypical appearance, the cellular detail of both mesodermal and endodermal layers appeared unaffected, in that the mesoderm consisted of loose vascular connective tissue and the endoderm of large polygonal cells packed with yolk droplets. Folding of the two basal layers into the yolk was also observed, although to a lesser extent, in membranes treated with 10 ng of EGF (fig. 3~). The ectoderm, in general, appeared unaffected by the application of EGF, although at the higher dose small focal areas of ectodermal hyperplasia were observed in some membranes (fig. 3d). Application of discs containing sterile distilled water caused no histological change in the mesodermal and endodermal layers, although in two specimens small areas of ectodermal hyperplasia, similar to those observed with the 1 l.tg dose of EGF, were seen (fig. 3b). In the remaining water-treated membranes, the ectoderm appeared normal, being composed of a thin layer of flattened cells. From the wax histological sections, it was apparent that specimens treated with EGF at both doses showed a considerable increase in vascular density. This
960
Ceil Biology
International
Reports,
Vol. 13, No. 17, November
1989
3d Fig. 1. Shell-less culture preparations (a) from above; (b) lateral view; (c) showing the position of the Millipore disc on the sinus terminalis between terminal branches of the right vitelline vessels. Fig. 2. India ink injection of regions of the area vasculosa treated with (a) 1 l.tg; and (b) 10 ng of EGF. In each case, there is a marked increase in capillary density at the site of application. Fig. 3. Wax histology of (a) an untreated region of area vasculosa, and those treated with (b) sterile distilled water; (c) 10 ng of EGF, and (d) 1 l.tg of EGF. The EGF-treated membranes show capillary proliferation accompanied by folding of the mesodermal and endodermal layers into the yolk. (All x100)
Cell Biology
International
Reports,
Vol. 73, No. 11, November
1989
961
effect was quantified by counting the number of capillaries, and measuring their cross-sectional area in a fixed length of membrane. The mean capillary density at the site of application increased in a dose-dependent manner, with an approximate rise of 20% at the 10 ng dose, and nearly a three-fold rise at the 1 p,g dose in comparison with untreated controls (fig. 4a). A similar but more marked response was observed in the vessel cross-sectional area per unit length: at the 10 ng dose, this area was nearly three times that of the corresponding control area and at the 1 pg dose, a nearly seven-fold increase was recorded (fig. 4b). As the rise in cross-sectional area of the capillaries was more pronounced than that in their number, their mean calibre must also increase. However, proliferation of small vessels was also apparent. Application of discs containing sterile distilled water caused no increase in either the number or cross-sectional area of vessels in unit length of membrane. As the parametric assumptions of normality of distribution and homogeneity of variance were not satisfied, the median values for the capillary density and crosssectional area were calculated for each specimen (Table I) and analysed using the non-parametric Kruskal-Wallis analysis of variance and a series of Mann-Whitney U tests. No significant difference in the values of either parameter were found between untreated controls and “sham” controls treated with sterile distilled water (p>O.O5), and these groups were pooled as controls in further tests. The capillary density in EGF-treated membranes at both doses was significantly greater than in the pooled control group (~~0.01 in each case), and there was a significant difference in the vessel density between these two test groups (~~0.05). Similarly, significant differences in the vessel cross-sectional area in a standard length of membrane were found for each EGF-treated group when compared with controls (~~0.01 for the 1 lrg dose; p
962
Cell Biology
4a
Reports,
Vol. 13, No. 71, November
1989
35.. 30.
E -f : 5 I s” 5s 5
International
.
25. 20. .
15.
: 10.
.
5.
1
O+ EGF lug
EGF long
I H&I Control
COntIOl
4b 16000
.
14000
ml m2000. ~10000. z 2 6000. 2
. 1
% 6000. I 4000-
. . . :
.
2000.
I i
!
H,O Conlml
COntrOl
03
I EGF lug
EGF long
Fig. 4. Scattergramsshowing the median vessel number (a) and median vessel cross-sectionalarea (b) per unit length of areavasculosafor embryos treatedwith 1 ttg and 10 ng of EGF, and for watertreatedand untreatedcontrols.
Table I. Median values of cross-sectional area (and vessel number) per unit length for each embryo studied. EGF lue
EGF 10 ne
Control (no disc)
Water Control
7802.5 (9.5)
3708.5 (6.0)
1323.0 (6.5)
1139.0 (6.0)
15328.0 (16.5)
2560.0 (6.0)
1617.5 (6.0)
1336.0 (5.5)
3584.0 (13.0)
4900.0 (7.0)
1267.0 (5.5)
2160.0 (5.0)
11518.0 (30.5)
2984.5 (6.5)
1694.5 (7.0)
1077.5 (4.0)
8066.0 (14.0) ---
2914.5 (9.5)
1416.5 (5.0)
1707.0 (6.0)
5506.5 (7.0)
897.0 (5.5)
2471.5 (7.5)
Cell Biology
International
Reports,
Vol. 13, No. 11, November
1989
963
form at lower doses, with the vessels “stacking” on top of one another (fig. 5d). The ectoderm showed little histological change, with no evidence of hyperplasia in any of the specimens studied. In the endoderm, however, the staining of the material in the yolk inclusions appeared to be less dense in treated membranes in contrast to corresponding controls. This decrease in staining intensity seemed to be dose-dependent, being noticeably less marked in membranes treated with 1 ng of EGF than in those subjected to higher doses,
Fig. 5. Resin histology of a length of untreated area vasculosa (a); and of regions treated with (b) 1 ng; (c) 10 ng; (d) 100 ng; and (e) 1 pg of EGF. (All x100)
964
Cell Biology
International
Reports,
Vol. 13, No. 11, November
1989
Discussion: The results of this study reveal that EGF has a clear promoting effect upon embryonic angiogenesis in that it causes a marked increase in both the density and cross sectional area of the capillaries per unit length of the area vasculosa. These results are in accordance with the findings of Schreiber et al (1986) that EGF stimulated angiogenesis, in the absence of an inflammatory response, in the hamster cheek pouch. EGF at a dose of 10 l.t.gshowed a small, but significant, increase in limbal capillary density in the rabbit cornea and was potent in stimulating DNA synthesis and proliferation
of cornea1 keratocytes in vitro (BenEzra, 1978).
Subsequently, Gospodarowicz et al (1979) observed a similar response to EGF at doses between 1 l.tg and 60 pg in the rabbit cornea. In addition, topical administration of EGF to corneas denuded of their epithelium resulted in not only a rapid re-epithelialization of the denuded area, but also included proliferation of the blood vessels in the limbal region (Savage and Cohen, 1972). The angiogenic activity of EGF can be explained by its ability to induce capillary endothelial cell migration and proliferation (McAuslan et al, 1985). These authors suggest that the induction results from the formation of a receptor-EGF complex which triggers off a series of intracellular events. However, it is uncertain whether the vessels are specifically affected. In the chick area vasculosa, the unusual projection of the mesoderm and endoderm into the yolk sac in response to EGF may simply be a measure to accomodate for the vast increase in vascular density in the mesodermal layer. However, it is important to note that simular tortuous folds are seen at later developmental stages and in older, more medial regions of the area vasculosa (Romanoff, 1960). The elevated capillary density, then, may be related to precocious development of all three layers of the membrane (ectoderm, mesoderm and endoderm), and the proliferative effects of EGF may not be confined to the vascular endothelium. This possibility is supported by the decreased staining intensity of the endodermal yolk inclusions in membranes treated with EGF, which could, in turn, be the result of the enzymatic degradation of their contents in the endoderm itself, and removal of the products by the mesodermal vessels. The ability of the endodermal cells to secrete the necessary proteolytic enzymes evolves naturally with the maturation of the membrane (Romanoff 1960). Significantly, EGF has already been shown to enhance development in mamalian systems, causing accelerated eruption of incisor teeth (Topham et al, 1987) and precocious eyelid opening in newborn mice (Smith et al, 1985).
Cell Biology
International
Reports,
Vol. 13, No. 1 I, November
1989
965
Bibliography: BenEzra, D. (1978). Neovasculogenic ability of prostaglandins, growth factors and synthetic chemoattractants. American Journal of Ophthalmology 86: 455-461. Dunn, B.E. & Boone, M.A. (1976). Growth of the chick embryo in vitro. Poultry Science 55: 1067-1071. Folkman, J. (1985). Tumour angiogenesis. Advances in Cancer Research 43: 175203. Folkman, J. & Klagsbum, M (1987). A family of angiogenic peptides. Nature 329: 671-672. Furcht, T.L. (1986). Critical factors controlling angiogenesis: cell matrix, and growth factors. Laboratory Investigation 55: 505-509. Gospodarowicz, D., Bialecki, H. & Thakral, K.K. (1979). The angiogenic activity of fibroblast and epidermal growth factors. Experimental Eye Research 28: 5015 14. Hamburger, V. & Hamilton, H.L. (1951). A series of normal stages in the development of the chick embryo. Journal of Morphology 88: 49-92. McAuslan, B.R., Bender, V., Reilly, W. & Moss, B.A. (1985). New functions of epidermal growth factor: stimulation of capillary endothelial cell migration and matrix dependent proliferation. Cell Biology International Reports 9: 175-l 82. Romanoff, A.L. (1960). “The Avian Embryo”, Macmillan, New York. Savage, C.R. & Cohen, S. (1972). Epidermal growth factor and a new derivative: rapid isolation procedures and biological and chemical characterization. Journal of Biological Chemistry 247: 7609-7611. Schreiber, A.B., Winkler, M.E. & Derynck, R. (1986). Transforming growth factor : a more potent angiogenic mediator than epidermal growth factor. Science 232: 1250- 1253. Smith, J.C., Spom, M.B., Roberts, A.B., Derynck, R. & Winkler, M.E. (1985). Human transforming growth factor-alpha causes precocious eyelid opening in newborn mice. Nature 315: 515-516. Topham, R.T., Chiego, D.J., Gattone, V.H., Hinton, D.A. & Klein, R.M. (1987). The effect of epidermal growth factor on neonatal incisor differentiation in the mouse. Developmental Biology 124: 532-543. Wilson, D.J. (1983). The origin of the endothelium in the developing marginal vein of the chick wing-bud. Differentiation 30: 183-187.