[CANCER RESEARCH 32,77-82,January1972J

SUMMARY

Recognition of capillary endothelial cells was facilitated bya special staining procedure developed previously.Thymidine-labeling indices for endothelial cells were estimatedin a number of unstimulated mouse tissues and were found tobe in the range 0 to 2 .4%. There was no increase in meanlabeling index up to 3 weeks after 2000 rads or up to 2 weeksafter 4000 rads irradiation of muscle, skin, or bone. Afterfracture of the femur, the labeling index of capillaryendothelial cells in regenerating callus increased to about 10%at 3 to S days after fracture and decreased to zero by Day 16.The duration of DNA synthesis in the fractured femur wasestimated in a double-labeling experiment to be 7 ±2 hr. Theresults of a repeated labeling experiment suggested a turnovertime of 80 ±25 hr on the 6th day after fracture, with a widedistribution of intermitotic times about the mean value. At 3days after injury, when the labeling index is higher, thecorresponding estimate of turnover time is about 50 hr: thisvalue is close to a previous estimate of turnover time forcapillary endothelial cells of a mammary tumor growing in thesame strain of mouse.

INTRODUCTION

The proliferation of capillary endothelial cells plays acentral role in such diverse processes as wound healing, organtransplantation, tumor growth, and the response of tissue toradiotherapy. Indeed, the rate of extension of the capillarybed may be the limiting factor for each of these processes,inasmuch as the survival of any tissue is critically dependenton a supply of nutrients through the vascular system.However, despite the extensive studies of cell proliferation inmany types of normal and malignant tissue, there have beenfew systematic investigations of the normal rate ofproliferation of capifiary endothelial cells or of theirproliferative response to stimuli. In this study, we have used aspecial staining technique (29, 31) to obtain maximumcontrast between capillaries and other tissue elements, and we

‘Thework was supportedinpartby USPHS ResearchGrantsCA5047 and CA 6294.

2 Recipient of a Research Training Fellowship of the International

Agency for Research on Cancer, during the tenure of which this workwas done. To whom correspondence should be addressed, atDepartment of Radiology, Hospital of the University of Pennsylvania,Philadelphia, Pa. 19104.

3Present address: Department of Radiology, Kyoto PrefecturalUniversityofMedicine,Hirokouji,Kawaramachi,Kamikyo-ku,Kyoto,Japan.

Received June 3, 1971; accepted August 27, 1971.

have investigated the proliferation of endothelial cells in miceby the techniques of thymidine autoradiography. Data arepresented for endothelial cells in a variety of normal tissues, intissues damaged by irradiation, and in the callus formed afterfracture of the femur.

PROUFERATION OF ENDOTHELIAL CELLS INUNSTIMULATED TISSUE

Materials and Methods. Specific-pathogen-free C3H/He micewere used in this and all other experiments. Mice of both sexeswere used ; they were 8 to 12 weeks old and weighed 25 to 30g. The thymidine-labeling index was estimated for capillaryendothelial cells in a number of normal tissues. Five mice wereeach given i.p. injections of 50 jiCi of TdR-3H4 and werekilled 1 hr later. The following tissues were excised and placedin neutral formol-0.9% NaCl solution: liver, kidney, stomach,small intestine (jejunum), skin, muscle (both cut from theright leg), and right femur. Bones were decalcified, and 4-uparaffin sections were cut on precleaned glass slides. Sectionswere dewaxed, stained overnight in Luxol fast blue solution,and then stained with the periodic acid-Schiff reaction (31).Autoradiographs were prepared by dipping in Ilford K5emulsion, and exposure times of about 1 month were used.Slides were developed in Kodak D 19 developer and fixed inKodak fixer. After being washed for I hr, the sections werestained with Harris' hematoxylin; they were left unmountedbecause we have sometimes observed grain fading in mountedsections.

Endothelial cells usually were heavily labeled (>20grains/cell), and background was low ( < 2 grains/cell).Endothelial cells were recognized and included in cell counts ifthey were within the walls of blood vessels which were only 1cell thick (i.e., endothelium of arterioles and venules wasexcluded) and if the nuclei were long, thin, and completelysurrounded by red-stained periodic acid-Schiff-positivecytoplasm. The presence of blue-stained erythrocytes withinthe lumen assisted with recognition of capifiaries; however, weincluded empty capillaries in the counts if we were certain oftheir identity. Color photomicrographs of sections stained bythe above method have been published elsewhere (29).

Results. Estimates of labeling index for each tissue arepresented in Table 1; the low values imply a slow rate ofproliferation. Endothelial mitoses were rarely recognized inthe sections.

In many murine tissues, the duration of the S phase (Ta)has a relatively constant value in the range 7 to 9 hr (5, 24). If

4The abbreviation used is: TdR-3H, tritiated thymidine.

JANUARY 1972 77

The Proliferation of Capillary Endothelial Cells'

Ian F. Tannock2andShigejiroHayashi3

Section of Experimental Radiotherapy, University of Texas, M. D. Anderson Hospital and Tumor Institute, Houston, Texas 77025

on May 18, 2021. © 1972 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TissueNo.

ofcells

countedNo.

oflabeled

cellsLabelingindex

(%)Liver50020.4Kidney50020.4Stomach

(serosa andmucosa)29400.0Smallintestine (serosa andmucosa)50020.4Skin

(from theleg)56220.4Muscle(from theleg)30000.0Bone

(Haversian canals and other550132.4capillariesin the femur)

were anesthetized with sodium pentobarbitol, and radiationwas administered with a cesium irradiator, with the use of 2opposed sources at a dose rate of 1050 rads/min.Thymidine-labeling indices for capillary endothelial cells in theskin, muscle, and femur were evaluated from groups of 3 or 4mice that were killed at intervals up to 3 weeks afterirradiation.

The techniques of histology and autoradiography were asabove. In this and subsequent experiments, slides wererandomized to prevent subjective bias when scoringautoradiographs.

Results. Estimates of labeling index for capillaryendothelial cells at intervals after irradiation are shown inTable 2. No labeled cells were found on the 1st day afterirradiation, and DNA synthesis might be inhibited for a shortperiod. However, despite the use of high doses of irradiation,there was no apparent damage to endothelial cells and noincrease in the rate of proliferation within the interval studied.

PROUFERATION OF ENDOTHELIAL CELLS AFTERMECHANICAL INJURY

Materials and Methods. Proliferation of endothelial cells inregenerating callus after fracture of the femur was used as amodel system for studying the response to mechanical injury.Mice were anesthetized with sodium pentobarbitol, and theirright femurs were fractured by gentle digital pressure at themiddiaphysis (17, 32). No attempt was made to splint or fixthe fracture . Callus formed rapidly around the sites offracture. Serial radiographs were taken, and the early callusdeveloped into bony callus after about 3 weeks. Sufficientendothelial cells to determine labeling index could be found inregenerating callus from the 5th day after fracture, and thelabeling index was estimated on Days 5, 7, 10, 13, and 16 afterfracture. Counts of 100 to 500 cells. were used to determinelabeling index, except on Day 16 when fewer endothelial cellscould be recognized. The probable labeling index on Day 3

Days after2000 radsTissueLabeling― index (%)Days

after4000 radsTissueLabeling― index(%)1Muscle

SkinBone0

001Muscle

SkinBone0

002Muscle

SkinBone0

132Muscle

SkinBone1

025Muscle

SkinBone0

035Muscle

SkinBone1

1410Muscle

SkinBone0

0512Muscle

SkinBone0

1022Muscle

SkinBone0

04

Ian F. Tannock and Shigejiro Hayashi

Ts 15within this range for endothelial cells (see also below), anapproximate estimate of their turnover time (7') may beobtained with the formula (28):

L.I. = X(Ts)/(7) (A)

Here A is a factor that for slowly proliferating populations hasa value close to log@2and L.I. is the labeling index.

For a labeling index of 0.4%, Equation A gives a turnovertime of about 8 weeks; for the higher labeling index ofendothelial cells lining the Haversian canals of the femur(2.4%), the corresponding value of T is about 10 days. Thesevalues must, however, be regarded as approximate as theduration of the S phase was not measured. Also, precisedetermination of labeling indices with values as low as those inTable 1 requires the counting of very many cells, anddifficulties of recognition render this impractical.

PROLIFERATION OF ENDOTHELIAL CELLS AT SHORTINTERVALS AFTER IRRADIATION

Materials and Methods. The right legs of 40 mice wereirradiated with a single dose of either 2000 or 4000 rads. Mice

Table 1Labeling indices of endothelial cells in unstimulated tissues

Table2Labeling indices ofendothelial cells after irradiation

a Each estimate is based on counts of 100 endothelial cells.

78 CANCER RESEARCH VOL.32

on May 18, 2021. © 1972 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Totalcells

countedTotalpositive

cells%labeled

cellsFive

mice given injections on Day 3 and killed on Day 52766222.5 ±2.5―FivemicegiveninjectionsonDay 5 andkilledon Day 73466418.5 ±2.1

cells, a repeated thymidine-labeling experiment was started onthe 5th day after fracture. Thirty mice were given repeatedinjections of 25 jzCi TdR-3H at 6-hr intervals, and 3 mice werekilled 1 hr after each injection. Autoradiographs wereprepared, and the proportions of labeled cells were determinedfrom randomized sections.

Results. The relation between the labeling index of capillaryendothelial cells in the regenerating callus and time afterfracture is shown in Chart 1. The mean labeling indexdecreased from a mean value of about 9% on Day 5 to zero onDay 16. When animals were given injections of TdR-3H onDays 3 and 5 , and the proportions of labeled cells estimatedafter 48 hr, the results were as shown in Table 3.

The proportion of labeled cells for animals killed on Day 7at 48 hr after labeling is about twice the labeling index forcells injected at the same time (Day 5), with animals killedafter 1 hr. An increase by a little less than 2 is to be expectedif all labeled cells divide shortly after labeling and then becomedistributed around the cycle. The greater proportion of labeledendothelial cells in animals given injections of thymidine at 3days after fracture suggests a higher rate of proliferation atthat time; the 1-hr labeling index should then be about 11%(Chart 1).

The results of the double-labeling experiment were asfollows: number of cells counted, 542; number of cells labeledwith both isotopes (N), 46; number of cells labeled with 3Halone (n1 ), 7; number of cells labeled with ‘4C alone (n2), 8.The probable duration of DNA synthesis may then becalculated from the formulae:

Ts =(N+n1)/(n1)(N+n2)/(n2)hr

Proliferation of Capillary Endothelial Cells

Table3

Proportions of labeled cells in animals given injections of TdR-3H on Days 3 and 5Forfurtherdetails,seethetext.

“Mean±S.E.

0 8 124

days o(ter fracture

Chart 1. The relation between thymidine-labeing index of capillaryendothelial cells and time after fracture of the femur. ., values derivedfrom an experiment where there was a 48-hr interval between labelingand death of the animal.

was estimated by giving 5 animals injections of TdR-3 H andkilling them 48 hr later on Day 5 ; the proportion of labeledcells was compared with that for 5 animals given injections onDay 5 and killed on Day 7.

The duration of DNA synthesis (Ta) of endothelial cells wasestimated on the 5th day after fracture by a double-labelingmethod (1 , 3). Five mice were given injections of 10pCi' 4C-labeled thymidine; this was followed 1 hr later by aninjection of 250 j.zCi TdR-3 H and the mice were killed 1 hrafter the 2nd injection. Sections were cut and stained and werefirst autoradiographed by the usual procedure of dipping inIlford KS emulsion diluted 1: 1 in distilled water. Afterexposure for 1 week, slides were developed, fixed, and stainedwith hematoxylin. They were then immersed in an absorbingsolution [37.5% collodian (flexible) U.S.P., dissolved in asolvent of ether:ethyl alcohol (3: 1)1 and autoradiographed fora 2nd time in undiluted emulsion. After exposure for 4 weeks,slides were finally developed and fixed. Grains from tritiumj3-particles appeared only in the lower emulsion, and the highdose used ensured dense labeling of positive cells. Onlyj3-particles from ‘4C penetrated the absorbing layer to give aspray of grains in the upper emulsion centered on the labeledcell. Autoradiographic resolution was poorer for@@ C grains inthe upper emulsion than for tritium grains in the loweremulsion; however, by adjustment of the focus of themicroscope, it was usually possible to recognize cells labeledwith 3H alone, with ‘4Calone, or with both isotopes.

For estimation of the turnover time of capillary endothelial

(B)

The best estimate is obtained by combining the 2 expressionsto give:

Ts(2N+ni +n2)/(n1 +n2)=7.l hr (C)

The standard error of the mean estimate of T@ isapproximately 2 hr.5 The estimated duration of DNA

S From equation C

Thus

Differentiating

Ts _ 1 =(2N)/(n1 +n2)

log@(T@_ 1) = log@2—log@[(n,/N) + (n2/N)J

@Ts @(n,IN)+ @(n,/N)

Tsl (n,/PT)+(n2/N)

Binomial estimates for the standard errors of , IN and n2IN, ie.:

give i@Ts 2.0 hr.

@(n,IN )= (@J;@@;(N —,)/(N3'2)

JANUARY 1972 79

on May 18, 2021. © 1972 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

LabelingTissueSpeciesindex(%)ReferenceCapillaries

of retinaMouse0.01Engerman et aL(9)CapillariesofmyocardiumMouse0.14Capillariesof endocardium,Mouse0.7— 1.9Edwards and Klein(8)myocardium,

andepicardiumMouseAlveolarcells (types notMouse0.7Shimkin et at(25)distinguished)Kidney

stromal cells4-wk rat4-mo.rat0.9 0.25Phillips

and Leong(23)Mesenteric

arteriesAdult rat3.4Crane and Dutta(7)Capillaries(anatomical site notRabbit0.6―Spaet and Lejnieks(26)stated)Arteries

andveinsRabbit0.1—1.1―AortaRabbit0.003—0.08―Coronory

vesselsRabbit0bSpraragen et aL(27)AortaRabbit0.02Gaynor(13)Large

vessels in lung andkidney0.14—0.17AortaGuineapig0.4—2.7Payling Wright(20)Abdominal

aortaSwineO.S@O.6@Florentin et aL (11)

Jan F. Tannock and Shige/iro Hayashi

synthesis, 7 ±2 hr, is within the range found for most mousetissues (5 , 24).

The results of the repeated labeling experiments are shownin Chart 2 . By making some simplifying assumptions about thedistribution of phase times, it is possible to estimate turnovertime from these data. Since the method involves fitting a curveto a range of points, the estimated value is more reliable thanthat based only on the labeling index and duration of DNAsynthesis. The simplest model is to assume that the duration ofG1 , S, and G2 phases is constant; however, where distributionsof phase times have been estimated for slowly proliferatingtissues by the method of labeled mitoses, the distribution of

1 phase time in particular has usually been found to be

rather broad . A more reasonable model is to consider that the

duration of S and G2 phases is constant but that cells aretriggered randomly into DNA synthesis from a pool of cells ina resting (or G0) phase (4); this model is equivalent to a verywide distribution of G1 phase duration. A curve derived forthis model with parameter values T@ ‘@hr, T@ 80 hr is shownin Chart 2.

The theoretical curve of Chart 2 tends slightly tounderestimate the percentage of labeled cells at short intervalsafter the first thymidine injection and to overestimate thepercentage of labeled cells at later intervals; this effect isgreater for a curve derived by assuming constant phasedurations. Small differences of this type between experimentalpoints and derived curves are to be expected because noallowance was made for the decreasing rate of cellproliferation over the 48-hr period of the experiment (Chart1). The estimated value of turnover time is an average value,probably representative for the 6th day after fracture. Theshape of the derived curve (Chart 2) is quite sensitive to thevalue of T, once the value of T@@ known; however, it is quiteinsensitive to proportional variations in T and T@. Thus, sincethe mean value of T8 and its standard error were estimated bydouble labeling to be 7 ±2 hr, the corresponding mean valueof T and its standard error is about 80 ±25 hr. At 3 days afterfracture, when the labeling index is about 11%, the turnovertime is probably about 50 hr.

DISCUSSION

Table4

80

60

40

20

T t t t

0 6 12 18 24 30 36 42 48 The present estimates of thymidine-labeling index for

hours after first thymidrte injecton capillary endothelial cells in unstimulated mouse tissue were

Chart 2. The proportions of labeled capifiary endothelial cells in mostly in the range 0 to 0.4%, and values of turnover time areregenerating callus following repeated injections of TdR-3H. The pfl)bably about 2 months or more. Our mice, weighing 25 toexperiment was begun 5 days after fracture of the femur. Each point 30 g, grew slowly, but mainly by accumulation of fat.represents a separate animal and was derived from counts of 50 to 200 However, estimates of the distance between tattoo marks oncells. The curve was derived for a model derived in the text. the legs have demonstrated slight growth of bones in mice of

Published estimates of thymidine-labeling index for endothelial cells in unstimulated tissues

a Labeling index estimated after infusion of TdR-3 H for 7 hr.b One labeled cell per 200 high-power fields.c Labeling index determined after in vitro incubation with TdR-3 H

80 CANCER RESEARCH VOL.32

on May 18, 2021. © 1972 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Proliferation of Capillary Endothelial Cells

this size (S. Hayashi, unpublished information); this growthmay be reflected by the higher labeling index (and presumably,therefore, shorter turnover time) observed for endothelial cellsin the femur. In Table 4 we have listed published estimates ofthymidine-labeling index for endothelial cells in a variety ofspecies; most of the values obtained in mature animals are lessthan 1%, confirming that unstimulated endothelium generallyhas a slow turnover.

It has been proposed by many authors that capillary damagemight be responsible for some of the dose-limiting late effectsobserved in normal tissues after radiotherapy. Changes incapifiary function, such as penneability or blood flow, areoften observed shortly after irradiation (6, 15 , 18), but thelysis of capillary endothelial cells is seldom seen. However,Phillips (2 1, 22) has reported that rats irradiated with 1600rads or more, and protected from acute death by hypoxia,died at 3 to 6 months from acute radiation pneumonitis, withcomplete loss of endothelium from the alveolar capillaries.Strong correlations between visible capifiary damage and lateradiation necrosis of brain (19) or late myocardial fibrosis (10)have also been observed.

From the present data and from values of labeling indexlisted in Table 4, it appears that the turnover time ofunstimulated capifiary endothelium is typically a few months.In some tissues an increased rate of proliferation has beenobserved at short intervals after irradiation (2), but we foundno significant changes in labeling index up to 2 to 3 weeks.Other authors (10, 16) have also found no tendency forincreased proliferation of endothelial cells at least up to 40days after irradiation. The interval between irradiation andobservation of acute capillary damage might therefore reflectthe time for a critical number of damaged endothelial cells toenter mitosis and then become pycnotic. This effect isprobably cumulative, with other endothelial cells beingstimulated to divide in an attempt to replace those that havedied at mitosis. Thus, Fajardo and Stewart (10) have observedsome increase in thymidine-labeling index at 40 to 70 daysafter 2000 rads irradiation to rabbit hearts, and this increasedproliferative activity is followed by the appearance ofmyocardial fibrosis. Cumulative pycnosis of capillaryendothelial cells could lead to complete breakdown of thecapillary circulation, and nutritional failure might then lead tothe delayed radiation necrosis of parenchymal cells that hasbeen observed in many tissues.

The maximum labeling index of capillary endothelial cells inresponse to the mechanical injury of fracture of the femur(about 11%) was the same as the labeling index observedpreviously in a transplanted mouse mammary tumor growingin the same C3H strain of mouse (30); the corresponding valueof turnover time was about 50 hr. In the tumor,well-nourished carcinoma cells had a much greater rate ofproliferation and slowing of tumor growth resulted fromnutritional failure as the separation between neighboringcapillaries increased (30). In the regenerating callus, theperiostiocytes have likewise been reported to have a higherrate of proliferation than endothelial cells (32). Thus theproliferation of capifiary endothelial cells may be rate limitingfor both tumor growth and healing of fracture.

The relation between labeling index and time after fracture(Chart 1) suggests that the proliferation of endothelial cells

might be dependent on the release of a stimulatory factor (orrepression of an inhibitor) which then decays as healing takesplace. Continuous proliferation of endothelium in the tumorat a rate equal to the maximal response to injury would beconsistent with the continuous release of an angiogenesisfactor by tumor cells. There is now convincing evidence for astimulatory angiogenesis factor released by tumor cells. Inexperiments where tumor pieces were implanted in Milliporefilters, vascularization in surrounding normal tissue wasobserved (12, 14); implantation of normal tissue within suchfilters gave negative results (12). The stimulatory factor wasrecently isolated from several tumors and was found to be asoluble molecule the important component of which was anRNA-protein complex (12). However, if the same factor isresponsible for angiogenesis in response to injury, it must alsobe present in normal tissue, or its production must bestimulated by the injury. The lack of an acute proliferativeresponse to damage by radiation suggests that the release of anangiogenesis factor might follow the expression rather than theinduction of vascular damage. With fracture of the femur theexpression of damage is immediate and acute, but afterirradiation it appears to be delayed and chronic.

ACKNOWLEDGMENTS

We thank Dr. Herman Suit for his advice and encouragement.

REFERENCES

1. Baserga, R., and Lisco, E. Duration of DNA Synthesis in EhrlichAscites Cells as Estimated by Double-labeling with C'@ andH3-Thymidine and Autoradiography. J. Nati. Cancer Inst., 31:1559—1571,1963.

2. Brown, J. M., and Berry, R. J. Effects of X-irradiation on CellProliferation in Normal Epitheium and in Tumours of the HamsterCheek Pouch. In: Effects of Radiation on CellularProliferation andDifferentiation, IAEA Publication, pp. 475 —491. Brookhaven:Brookhaven National Laboratory Publication, 1968.

3. Burns, F. J. The Response of Irradiated Epidermal Cells to aProliferative Stimulus. Cell Tissue Kinet., in press.

4. Burns, F. J., and Tannock, I. F. On the Existence of a G0-Phase inthe Cell Cycle. Cell Tissue Kinet., 3: 321—334,1970.

5. Cameron, I. L., and Greulich, R. C. Evidence for an EssentiallyConstant Duration of DNA Synthesis in Renewing Epithelia of theAdult Mouse. J. Cell Biol., 18: 31—40,1963.

6. Casarett, G. W., and Eddy, H. A. Effects of Radiation onVasculature of Intestine. In: Time and Dose Relationships inRadiation Biology as Applied to Radiotherapy, pp. 181—185.Brookhaven, N. Y.: Brookhaven National Laboratory Publication,1969.

7. Crane, W. A. J., and Dutta, L. P. The Influence of Age on theUptake of@ â€S̃-Sulphate and@ H-Thymidine by the MesentericArteries of Rats with Regenerating Adrenal Glands. J. Pathol.Bacteriol., 88: 291—301,1964.

8. Edwards, J. L., and Klein, R. E. Cell Renewal in Adult MouseTissues. Am. J. PathoL, 38: 437—453,1961.

9. Engerman, R. L., Pfaffenbach, D., and Davis, M. D. Cell TurnoverofCapillaries. Lab. Invest., 17: 738—743,1967.

10. Fajardo, L. F., and Stewart, J. R. Capillary Injury PrecedingRadiation-induced Myocardial Fibrosis. Radiology, 101: 429—433,1971.

11. Florentin, R. A., Nam, S. C., Lee, K. T., and Thomas, W. A.Increased 3H-Thymidine Incorporation into Endothelial Cells of

JANUARY 1972 81

on May 18, 2021. © 1972 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Ian F. Tannock and Shigejiro Hayashi

Swine Fed Cholesterol for 3 Days. Exptl. MoL PathoL, 10:250—255,1969.

12. Folkman, J., Merler, E., Abernathy, C., and Williams, G. Isolationof a Ttmior Factor Responsible for Angiogenesis. J. Exptl. Med.,133: 275—288,1971.

13. Gaynor, E. Increased Mitotic Activity in Rabbit Endothelium afterEndotoxin. An Autoradiographic Study. Lab. Invest., 24:318—320,1971.

14. Greenblatt, M., and Shubik, P. Tumor Angiogenesis: TransfilterDiffusion Studies in the Hamster by the Transparent ChamberTechnique. J. NatI. Cancer Inst.,41: 111—124,1968.

15. Harris, J. W., and Noonan, T. R. Early Vascular PermeabilityChanges in Whole-Body X-irradiated Rats. Radiation Res., 34:357—365,1968.

16. Hassler, 0. Incorporation of Tritiated Thymidine into Mouse Brainafter a Single Dose of X-rays. An Autoradiographic Study. J.Neuropathol. Exptl. Neurol., 25: 97—105,1966.

17. Hayashi, S., and Suit, H. D. Effect of Fractionation of RadiationDose on Callus Formation at Site of Fracture. Radiology, 101:181—186,1971.

18. Lindop, P. J., Jones, A., and Bakowska, A. The Effect of 14 MeVElectrons on the Blood Vessels of the Mouse Earlobe. In: Time andDose Relationships in Radiation Biology as Applied toRadiotherapy, pp. 174—180. Brookhaven, N. Y.: BrookhavenNational Laboratory Publication, 1969.

19.McDonald,L.W.,andHayes,T.L.The RoleofCapillariesinthePathogenesis of Delayed Radionecrosis of the Brain. Am. J.Pathol., 50: 745—764,1967.

20. Payling Wright, H. Endotheial Mitosis around Aortic Branches inNormal Guinea Pigs. Nature, 220: 78—79,1969.

21.Phillips,T. L. An UltrastructuralStudyof theDevelopmentofRadiation Injury in the Lung. Radiology, 87: 49—54,1966.

22. Phillips, T. L. In: Time and Dose Relationship in Radiation Biologyas Applied to Radiation Therapy, pp. 194—199. Brookhaven,

N. Y.: Brookhaven National Laboratory Publication, 1969.23. Phillips, T. L., and Leong, G. F. Kidney Cell Proliferation after

Unilateral Nephrectomy as Related to Age. Cancer Res., 27:286—292,1967.

24. Pilgrim, C. H., and Maurer, W. Autoradiographische UntersuchungUber die Konstanz der D.N.S.-verdopplung-Dauer bei Zellarten vonMaus und Ratte durch Doppelmarkierung mit@ H- undI 4C-Thymidin. Exptl. Cell Res., 37: 183—199, 1965.

25. Shimkin, M. B., Sasaki, T., McDonough, M., Baserga, R., Thatcher,D., and Wieder, R. Relation of Thymidine Index to PulmonaryTumor Response in Mice Receiving Urethan and OtherCarcinogens. Cancer Res., 29: 994—998,1969.

26. Spaet, T. H., and Lejnieks, I. Mitotic Activity of Rabbit BloodVessels. Proc. Soc. Exptl. Biol. Med., 125: 1197—1201,1967.

27. Spraragen, S. C., Bond, V. P., and Dahl, L. K. Role of Hyperplasiain Vascular Lesions of Cholesterol-fed Rabbits Studied withThymidine-H3 Autoradiography. Circulation Res., 11: 329—336,1962.

28. Steel, G. G. Cell Loss from Experimental Tumours. Cell TissueKinet., 1: 193—207,1968.

29. Tannock, I. F. Effects of p02 on Cell Proliferation Kinetics. In:Time and Dose Relationships in Radiation Biology as Applied toRadiation Therapy, pp. 215—224.Brookhaven,N. Y.: BrookhavenNational Laboratory Publication, 1969.

30. Tannock, 1. F. Population Kinetics of Carcinoma Cells, CapillaryEndothelial Cells, and Fibroblasts in a Transplanted MouseMammary Tumor. Cancer Res., 30: 2470—2476, 1970.

3 1. Tannock, I. F., and Steel, G. G. Quantitative Techniques for Studyof the Anatomy and Function of Small Blood Vessels in Tumors. J.Natl. Cancer Inst., 42: 771—782, 1969.

32. Tonna, E. A., and Cronkite, E. P. Autoradiographic Studies of CellProliferation in the Periosteum of Intact and Fractured Femora ofMice Utilizing DNA Labeling with H3-Thymidine. Proc. Soc. Exptl.Biol. Med., 107: 719—721,1961.

82 CANCER RESEARCH VOL.32

on May 18, 2021. © 1972 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1972;32:77-82. Cancer Res   Ian F. Tannock and Shigejiro Hayashi  The Proliferation of Capillary Endothelial Cells

  Updated version

  http://cancerres.aacrjournals.org/content/32/1/77

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/32/1/77To request permission to re-use all or part of this article, use this link

on May 18, 2021. © 1972 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from