Mechanisms that regulate the transition of micrometastases from clinically undetectable and dormant to progressively growing are critically important but poorly understood in cancer biology. Here we examined the effect of a primary tumor on the growth of solitary tumor cells in the mouse liver, as well as on the development of tumor angiogenesis in a dorsal skin-fold chamber. s.c. placement of a CT-26 (BALB/c-derived mouse colon carcinoma) primary tumor markedly inhibited development of liver metastasis in BALB/c mice after subsequent intraportal injection of tumor cells. Dorsal skin-fold chamber experiments showed that this growth inhibition paralleled a strong antiangiogenic effect by the primary tumor. Furthermore, intravital microscopy of the liver after intraportal injection of green fluorescent protein-expressing tumor cells showed that primary tumors promoted dormancy of single tumor cells for up to 7 days. Immunohistological staining for Ki-67 confirmed that these solitary cells were indeed dormant. In contrast, in the absence of a primary tumor, GFP-expressing tumor cells quickly developed into micrometastases. Thus, primary CT-26 tumor implants nearly abrogated tumor metastasis by inhibition of angiogenesis and by promoting a state of single-cell dormancy. Knowledge of the mechanism underlying this dormancy state could result in the development of new therapeutic tools to fight cancer.

Tumor dormancy provides the conceptual framework to explain a prolonged quiescent state in which tumor cells are present, but tumor progression is not clinically apparent. Clinical examples of tumor dormancy are numerous, including cases of melanoma and breast carcinoma, where periods of clinical latency may last for a decade or more. Although a dormant population of tumor cells is clinically undetectable, they can be activated and thus pose a constant risk for tumor recurrence. Until recently, this phenomenon received surprisingly little attention. However, the demonstration in mice that a primary tumor inhibited angiogenesis in remote metastases led Holmgren et al. (1) to propose that tumor dormancy may be explained by the lack of vascularization of such micrometastases. Under conditions of angiogenic suppression, similar rates of proliferation and apoptosis would then result in viable, but nongrowing, dormant metastases. This scenario, where inhibition of angiogenesis appears to limit tumor growth by elevating the incidence of tumor cell apoptosis, was also observed in animal models of concomitant tumor resistance using Lewis lung carcinoma cells (2). A similar mechanism underlies the therapy with endogenous antiangiogenic agents such as angiostatin or endostatin (2, 3, 4, 5, 6, 7). Recent work by Luzzi et al. (8) provides an additional concept of dormancy, where single tumor cells remain within the host tissue for prolonged periods of time, neither proliferating nor undergoing apoptosis. This implies that metastatic growth could then be halted at a stage where angiogenesis is not required, thus suggesting mechanisms that underlie induction, maintenance, and loss of this type of dormancy are of great clinical interest. In a first step, we tested the effect of a s.c. CT-26 primary tumor on tumor growth and angiogenesis of a second tumor implant in a dorsal skin-fold chamber preparation. In a second approach, we followed the fate of solitary, stably GFP-expressing CT-26 tumor cells in the liver of mice that either had, or did not have, a s.c. primary tumor. On the basis of these experiments, we describe the influence of a primary tumor on both tumor angiogenesis and dormancy of solitary disseminated tumor cells.

Animals, Cell Culture, and GFP Transfection.

Male 6–8-week old BALB/c (Charles River, Sulzfeld, Germany) or SCID-beige mice (Harlan Sprague Dawley, Borchen, Germany) were used. Mice were housed under standard conditions. Animals were anesthetized with ketamine hydrochloride and xylazine (7.5 and 2.5 mg, respectively, per 100 mg body weight) prior to all surgical procedures. All experiments involving mice were approved by the Animal Care Committee from the Government of the Oberpfalz, Germany.

The well-described CT-26 cell line (9), derived from a murine BALB/c colon adenocarcinoma, was maintained in RPMI 1640 (Biochrom, Berlin, Germany) supplemented with 10% FCS, 1% penicillin/streptomycin, and 1% l-glutamine. The enhanced GFP4 expression vector (Clontech, Palo Alto, CA) was transfected into CT-26 cells using Lipofectamine (Life Technologies, Inc., Karlsruhe, Germany), according to the instructions of the manufacturer. Using flow cytometry, the brightest fluorescent cells (>95th percentile) were sorted and expanded in culture. Cells were maintained in 200 μg/ml geneticin-containing selection medium (Life Technologies, Inc.). Single-cell suspensions were obtained by treating monolayers of cells with 0.25% trypsin in calcium and magnesium-free Dulbecco’s PBS.

Experimental Metastasis Assay.

Tumor cells were adjusted to a final concentration of 5 × 107 cells/ml, and 0.02 ml were injected s.c. into the posterior mid-dorsum of mice. Tumor volumes were estimated by the formula V=π/6 × a2 × b, where a was the short axis, and b the long axis. When tumors reached a size >500 mm3 (∼10 days after implantation), mice were challenged intraportally with 3 × 105 CT-26 cells, and liver surface metastases were counted 10 days later. Control animals had no s.c. primary tumor.

Dorsal Skin-Fold Chamber Assay.

Tumor angiogenesis was specifically quantified using the transparent dorsal skin-fold model, as described in detail by Asaishi et al. (10), Carmeliet and Jain (11), and Guba et al. (12). Briefly, a 1-cm-diameter flap of skin was dissected away from opposing surfaces of the dorsal skin-fold of anesthetized BALB/c mice, leaving a fascial plane with associated vasculature. The hole was held vertically away from the body with a pair of identical titanium frames that were sutured to both sides of the flap. The underlying surgical site was then sealed with a glass window enclosed in one of the frames. After a recovery period of 2 days, the glass window was removed, and CT-26 cells (105 cells/animal) were carefully placed on the upper tissue layer, and the chamber was closed again. For intravital microscopy, mice were immobilized in a tube with a longitudinal slit, from which the protruding skin-fold chamber was locked into a fixed position that could be viewed by intravital microscopy through the frame-mounted coverslip. Quantitative vascular analysis on days 1, 4, 7, and 10 included determination of the tumor area and MVD (cm−1), which was defined as the length of all newly formed microvessels in three randomized areas within the tumors.

In Vivo Liver Microscopy.

To follow the fate of individual cancer cells in vivo, IVM was performed using a modification of a technique described previously (8, 13, 14). Briefly, GFP-labeled CT-26 cells, together with 5 × 104 red fluorescent microspheres, were inoculated into a mesenteric vein of mice that either had, or did not have, primary tumors. Analysis by IVM was performed 4 and 7 days after injection. During the IVM procedure, mice were anesthetized and placed on a heated platform. Through an abdominal midline incision, the left liver lobe was exteriorized, immobilized on a specially designed stage, and covered with a thin coverglass. Individual cells could be identified by their fluorescence. Solitary tumor cells, multicellular foci (3–10 cells), micrometastases (< or >200 μm in diameter) and trapped microspheres were counted separately.

IVM Technique.

In vivo microscopy was performed using a modified Axiotech Vario microscope (Zeiss, Oberkochen, Germany) equipped with a green (520–570-nm) filter block. Observations were made using ×2.5, ×10 long distance and ×20 water immersion working objectives. Epifluorescence for liver microscopy was generated by a 100-W mercury lamp. Observations of the window chambers were carried out using a white-light transillumination technique. Images were recorded through a video camera (PCO, Kehlheim, Germany) on S-VHS tapes for subsequent off-line analysis. Individual video frames were captured and digitalized for detailed analysis. Off-line measurements and calculation of microvascular parameters were assisted by specifically designed image software (Dr. Günther Ackermann, University of Regensburg, Regensburg, Germany).

Accounting Technique.

For quantification of cell survival and metastatic growth over time, a cell accounting technique was used as described previously by Luzzi et al. (8). Briefly, 9-μm red fluorescent microspheres were mixed with the cell suspension to produce a cell:microsphere ratio of 10:1. Red fluorescent microspheres were used as permanent markers to determine the status of injected cells. More specifically, after injection, cells and microspheres became arrested in the microcirculation of the liver in a fixed proportion. At later time points, the ratio of red fluorescent microspheres to green fluorescent solitary tumor cells, multicellular foci, or micrometastases was used to quantify these individual metastatic populations and to help determine cell survival. Because metastatic tumor growth is of clonal origin, individual metastases were assumed to represent the survival of a single injected cell.

Proliferation of Solitary Cells and Metastases.

We performed immunohistological staining of liver sections for the proliferation marker Ki-67 to determine whether solitary tumor cells or cells within metastases were undergoing proliferation. After IVM on days 4 and 7, we examined serial sections from livers of three mice in each group to evaluate the relative number of tumor cells staining for Ki-67. In the first step of this process, fluorescent tumor cells were counted and localized in frozen liver tissue by fluorescence microscopy. In a second step, immediately adjacent serial cryosections were stained for the proliferation marker Ki-67. Acetone-fixed cryosections were blocked with S3022 solution (Dako, Hamburg, Germany) for 30 min and were then incubated at 37°C for 1 h with a 1:100 dilution of rabbit polyclonal antimouse Ki-67 antibody (Dianova, Hamburg, Germany). After they were washed three times in PBS for 15 min, the sections were incubated with (1:100) alkaline phosphate-conjugated antirabbit IgG (Roche, Mannheim, Germany). After washing with three exchanges of PBS, an enzymatic reaction with the nitroblue tetrazolium chloride, 5-bromo-4-chloro-3-indolylphosphate substrate (Dako) was allowed to proceed until the desired color intensity was reached. Finally, slides were counterstained with hematoxylin reagent.

Statistical Analysis.

Data are given as the mean ± SE in quantitative experiments. More specifically, an average value for each microcirculatory and histomorphological parameter was determined in each animal, and these values were used to calculate the mean among all of the animals in each group of experiments. For analysis of differences between the groups, a one-way ANOVA, followed by unpaired Student’s t test, was performed. Results with P < 0.05 were considered significant.

Growth of Metastases Is Severely Curtailed by a Primary Tumor.

The ability of a primary tumor to inhibit growth of metastases at a second site was confirmed after intraportal injection of tumor cells in the presence or absence of a primary tumor. Ten days after the second tumor inoculum, animals were sacrificed, and liver surface metastases were counted (Fig. 1). In mice bearing a primary tumor liver, metastasis was almost completely abrogated, with only 1 animal of 10 showing a minimal tumor burden at day 10. At the same time point, all nonprimary tumor-bearing mice developed a large number of metastases. Therefore, results from the CT-26 liver metastasis model showed that substantial concomitant tumor resistance does occur in the presence of a CT-26 primary tumor.

Primary Tumors Cause a Systemic Antiangiogenic Effect.

To determine whether angiogenesis was inhibited by a primary tumor, we examined the effect of a s.c. tumor on growth and angiogenesis of second tumor implants in transparent dorsal skin-fold chambers. Results showed that the presence of a large primary tumor almost completely inhibited growth and neovascularization of second CT-26 tumor implants grown in the dorsal skin-fold chamber of BALB/c mice, whereas tumor growth was unaffected in nonprimary tumor-bearing mice (n = 6 in each group). More specifically, on day 10 the tumor area was 30 times smaller in primary tumor-bearing mice, as compared with nonprimary tumor-bearing controls (Fig. 2,A). In parallel, the MVD increased rapidly with time in mice without primary tumor. In contrast, mice with a large primary tumor developed no new vessels (Fig. 2 B). Also, we noted that the size of the primary tumor affected the potency of the inhibitory effect on the second tumor. A primary tumor volume of <500 mm3 showed less extensive inhibition compared with primary tumors >500 mm3 (data not shown).

An overview of neovascularization of a second CT-26 tumor implant in tumor-bearing and control mice is shown after 10 days in Fig. 3. After implantation, tumor cell masses were seen as shadowy areas at low magnification. At day 3, the area of the tumor implant showed a nonspecific vascular reaction, where underlying host vessels became dilated and tortuous. At this stage, the vascular reaction was not different between mice with or without a primary tumor. After 7 days, tumors in mice without a primary tumor had developed a functional neovascular network, with sprouts and early vessels covering the entire tumor area. Three days later, these tumors nearly doubled in diameter, and their neovascular network had matured, showing large tortuous vessels with abrupt changes in diameter (Fig. 3,B). In direct contrast, tumors of mice with primary tumors remained small and avascular until day 10 (Figs. 2 and 3).

A Primary Tumor Promotes Dormancy of Solitary Cells.

We performed IVM at days 4 and 7 after intraportal delivery of tumor cells to determine the effect of a primary tumor implant on early development of liver metastases. Survival and relative distribution of solitary cells, multicellular foci, and micrometastases were determined by means of a cell accounting procedure using reference microspheres; the dormancy of cells was assessed by Ki-67 immunohistochemistry. To avoid an immune response to GFP protein (15), we used immunodeficient SCID-beige mice (16) for this set of experiments.

Immediately after intraportal injection, cells and microspheres became arrested by size restriction in the periportal liver sinusoids, as described recently (13, 14, 17, 18). During the first 4 days, the remaining tumor cells successfully completed their extravasation process into the liver parenchyma, without any difference between non- or primary tumor-bearing mice (data not shown). Notably, a large number of cells were lost (∼55%) during the first 4 days, but the same loss occurred in non- and primary tumor-bearing mice (Fig. 4). From days 4–7, there is no further loss of tumor foci in primary tumor-bearing mice. However, in mice without a primary tumor, the number of tumor foci decreased between days 4 and 7. Interestingly, Fig. 5,A shows that on day 4, ∼40% of tumor cells were present as solitary cells (average cell diameter, 19 ± 2 μm) in primary tumor-bearing animals. In controls without a primary tumor, we found only 25% remaining as solitary cells; however, some of the tumor cells at this time point appear to have already replicated and formed multicellular foci, which consisted of 3–10 cells (average diameter, 48 ± 3 μm) and small micrometastases below a diameter of 200 μm. Compared to day 4, we did not observe an appreciable decrease in solitary cells in primary tumor-bearing mice by day 7 (Figs. 5,B and 6,A), suggesting that this population of cells persists and could be dormant. In contrast, the number of single cells decreased in controls without a primary tumor, and single cells already formed micrometastases with a diameter >200 μm (Fig. 6 B), suggesting that proliferating cells may be more susceptible to cell death. The hypothesis that solitary tumor cells are kept in a dormant state by a primary tumor implant was further substantiated by Ki-67 staining of corresponding liver tissue sections on days 4 and 7. On both days, we found only a negligible (∼0.2%) fraction of Ki-67-positive solitary cells in primary tumor-bearing mice. Although a higher percentage of solitary cells was positive for Ki-67 in mice without a primary tumor (∼5%), these cells were also typically not proliferating. In contrast, at this same time point ∼65% of tumor cells within variably sized micrometastases were Ki-67 positive. Together, these results suggest that the vast majority of tumor cells in primary tumor-bearing mice persist in a state of dormancy without apparent signs of proliferation or cell death until day 7. Without the control of a primary tumor, however, tumor cells rapidly replicate to form micrometastases that show a high Ki-67 proliferation index.

Concomitant resistance, the phenomenon by which tumor-bearing hosts are able to inhibit distantly placed second implants of the same tumor, has been related in some studies to the inhibition of angiogenesis in the vascular bed of metastases or other secondary tumors (1, 4, 11, 19, 20, 21). The present study provides an additional explanation for this phenomenon where a primary tumor is shown to halt the metastatic progression of solitary tumor cells, keeping them in a dormant state that is far below the critical size requiring the induction of neovascularization for nutrient supply.

Concomitant resistance in our model was confirmed in an artificial liver metastasis model in the presence or absence of a primary tumor. The inhibition of metastatic growth was not attributable to induction of an immune response driven by the primary tumor implant, because similar results were obtained in SCID-beige mice lacking functional T and B lymphocytes and natural killer cells (16). Recent work by Folkman and coworkers (2, 3, 4, 5, 7) and O’Reilly (6) demonstrated convincingly that some primary tumors inhibit the growth of their metastases by antiangiogenic mediators that are released by the primary tumor. These endogenous antiangiogenic proteins, such as angiostatin or endostatin, kept tumors below a critical size (∼1 mm in diameter), where sufficient nutrient supply occurs by diffusion. At this stage, avascular micrometastases show balanced proliferation and apoptosis until angiogenesis allows for further growth (1, 4). In the dorsal skin-fold chamber model, we confirmed the ability of a primary tumor to inhibit neovascularization of a second tumor implant. Moreover, inhibition of angiogenesis in our model was dependent on primary tumor size, where a primary tumor volume <500 mm3 was found to produce incomplete inhibition. These results are in accordance with observations made by Sckell et al. (19), where inhibition of basic fibroblast growth factor-induced angiogenesis was suppressed by implantation of PC-3 tumors in a size-dependent manner.

Notwithstanding effects on neovascularization, the present study describes a new mechanism by which a primary tumor can suppress the growth of metastases. This mechanism precedes the above described inhibition of vascularization. Using a model with stably GFP-transfected CT-26 tumor cells in combination with Ki-67 staining, we found that a primary tumor in our system inhibited not only angiogenesis but also inhibited the initiation of growth of solitary, dormant, tumor cells. This was evident from the large fraction of nondividing single tumor cells persisting from days 4 to 7 in primary tumor-bearing mice. The vast majority of these solitary cells was found to be negative for the proliferation marker Ki-67, whereas most of the tumor cells within metastatic foci were positive for Ki-67. Because Ki-67-negative dormant tumor cells were regularly found in the extravascular parenchymal tissue, dormancy of solitary tumor cells occurred after successful extravasation. This observation is in accordance with recent work by Luzzi et al. (8) and Cameron et al. (22), who proposed that a major contributor to metastatic inefficiency was the failure of extravasated cells in a target organ to initiate growth. It is notable that a limited number of tumor cells had proliferated in tumor-bearing mice but formed predominantly multicellular foci of 4–10 cells, or small micrometastases. In contrast, animals without primary tumor had, at this same time point, already reached a stage of extended avascular micrometastases. Solitary tumor cells and multicellular foci with <10 cells are well below the size where induction of angiogenesis is required (1). Therefore, these results suggest that primary tumors interfere with the initial proliferative development of metastases before angiogenesis begins to control further growth to macroscopic lesions.

The mechanism by which the primary tumor induces dormancy of solitary tumor cells is not known. It is of interest to note that at least in vitro all known endogenous inhibitors of angiogenesis inhibit proliferation of endothelial cells but show no antiproliferative effects on the tumor cell itself (5). One intriguing possibility would be that endothelial cells, which may secrete chemotactic cytokines, or cytokines with proliferative activity for tumor cells, are inhibited by these antiangiogenetic mediators. This interpretation is consistent with a recent observation by Li et al. (23), where the application of antiangiogenic soluble vascular endothelial growth factor receptors led to inhibition of tumor cell migration toward preexisting blood vessels and subsequent tumor cell death. This mechanism might also explain why repetitive antiangiogenic treatment can prevent reoccurrence of tumors after cessation of drug application (4, 24). Yet another conceivable, and even more intriguing possible explanation, is that tumor cells are put to “sleep” by a novel mechanism, independent and separate from the action of antiangiogenic factors. It may be that the primary tumor secretes a currently unknown molecule, which directly silences disseminated extravasated tumor cells. Therefore, this study raises the possibility that besides targeting angiogenesis by antiangiogenic substances, metastatic tumor growth could also be stopped at an even earlier stage than was thought previously. Our findings suggest it may be of critical importance to further explore the mechanisms controlling this population of dormant cells.

Fig. 1.

The effect of a primary tumor implant on hematogenous metastasis of CT-26 cells in mouse liver. The number of tumors observed per liver, 10 days after intraportal injection, are indicated for individual mice that either have, or do not have, an established primary tumor. Horizontal line, mean values. The difference between the two groups was significant (P < 0.001; n = 10).

Fig. 1.

The effect of a primary tumor implant on hematogenous metastasis of CT-26 cells in mouse liver. The number of tumors observed per liver, 10 days after intraportal injection, are indicated for individual mice that either have, or do not have, an established primary tumor. Horizontal line, mean values. The difference between the two groups was significant (P < 0.001; n = 10).

Close modal
Fig. 2.

Inhibition of tumor growth and angiogenesis of a second tumor implant in the presence or absence of a primary tumor. Tumor area (A) and MVD (B) were measured in the transparent chambers at different time points after implantation of CT-26 tumor cells into non- or primary tumor-bearing mice (n = 6). Data are presented as the means; bars, SE. *, P < 0.05 versus primary tumor-bearing animals.

Fig. 2.

Inhibition of tumor growth and angiogenesis of a second tumor implant in the presence or absence of a primary tumor. Tumor area (A) and MVD (B) were measured in the transparent chambers at different time points after implantation of CT-26 tumor cells into non- or primary tumor-bearing mice (n = 6). Data are presented as the means; bars, SE. *, P < 0.05 versus primary tumor-bearing animals.

Close modal
Fig. 3.

The effect of a primary tumor on neovascularization of a second CT-26 tumor implant in transparent chambers of BALB/c mice. A, the photomicrograph shows the second CT-26 tumor implant in a primary tumor-bearing mouse. The development of a vascular network is completely curtailed by the established primary tumor. The inoculation site appears as a shadowy area; only a few dilated vessels are observed within the tumor cell mass. B, at the same time point, intense neoangiogenesis occurs in second tumor implants of animals without a primary tumor. A typical mature functional tumor neovascular network can be seen.

Fig. 3.

The effect of a primary tumor on neovascularization of a second CT-26 tumor implant in transparent chambers of BALB/c mice. A, the photomicrograph shows the second CT-26 tumor implant in a primary tumor-bearing mouse. The development of a vascular network is completely curtailed by the established primary tumor. The inoculation site appears as a shadowy area; only a few dilated vessels are observed within the tumor cell mass. B, at the same time point, intense neoangiogenesis occurs in second tumor implants of animals without a primary tumor. A typical mature functional tumor neovascular network can be seen.

Close modal
Fig. 4.

Percentage of tumor foci (solitary cells, multicellular foci, and micrometastases) surviving on day 4 (n = 4) and 7 (n = 6) after intraportal CT-26 GFP cell and microsphere injection into SCID-beige mice. During the first 4 days, survival of tumor foci was significantly reduced, regardless of the presence of a primary tumor. After day 4, further cell foci were lost in nonprimary tumor-bearing mice. In primary tumor-bearing mice, the number of tumor foci remained constant from days 4 to 7. Data are presented as the means; bars, SE. *, P < 0.05 versus mice with no primary tumor.

Fig. 4.

Percentage of tumor foci (solitary cells, multicellular foci, and micrometastases) surviving on day 4 (n = 4) and 7 (n = 6) after intraportal CT-26 GFP cell and microsphere injection into SCID-beige mice. During the first 4 days, survival of tumor foci was significantly reduced, regardless of the presence of a primary tumor. After day 4, further cell foci were lost in nonprimary tumor-bearing mice. In primary tumor-bearing mice, the number of tumor foci remained constant from days 4 to 7. Data are presented as the means; bars, SE. *, P < 0.05 versus mice with no primary tumor.

Close modal
Fig. 5.

Distribution of single cells, multicellular foci, and micrometastases < and >200 μm in diameter 4 (A) and 7 (B) days after intraportal injection of GFP-expressing CT-26 tumor cells. Each column represents the mean value obtained from four to six animals; bars, SE. *, P < 0.05 versus animals with no primary tumor.

Fig. 5.

Distribution of single cells, multicellular foci, and micrometastases < and >200 μm in diameter 4 (A) and 7 (B) days after intraportal injection of GFP-expressing CT-26 tumor cells. Each column represents the mean value obtained from four to six animals; bars, SE. *, P < 0.05 versus animals with no primary tumor.

Close modal
Fig. 6.

Influence of the primary tumor on the sequential development of metastases in mouse liver, as visualized on day 7 after intraportal injection of GFP-expressing CT-26 cells. In A, in primary tumor-bearing mice, a large fraction of injected CT-26 cells remained solitary in the postextravasational stage. Only rarely, tumor cells were found in these animals to initiate further growth into multicellular foci or small avascular micrometastases. #, solitary extravasated tumor cells within the liver parenchyma. In B, in mice without a primary tumor, most of the tumor cells remaining by day 7 proliferated to micrometastases. ○, micrometastases; *, fluorescent microsphere trapped in the microcirculation.

Fig. 6.

Influence of the primary tumor on the sequential development of metastases in mouse liver, as visualized on day 7 after intraportal injection of GFP-expressing CT-26 cells. In A, in primary tumor-bearing mice, a large fraction of injected CT-26 cells remained solitary in the postextravasational stage. Only rarely, tumor cells were found in these animals to initiate further growth into multicellular foci or small avascular micrometastases. #, solitary extravasated tumor cells within the liver parenchyma. In B, in mice without a primary tumor, most of the tumor cells remaining by day 7 proliferated to micrometastases. ○, micrometastases; *, fluorescent microsphere trapped in the microcirculation.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This research was supported by Grant STE 960/1 of the Deutsche Forschungsgemeinschaft and a grant from the University of Regensburg (to M. S. and M. G.).

4

The abbreviations used are: GFP, green fluorescent protein; MVD, microvascular density; IVM, intravital microscopy; SCID, severe combined immunodeficient.

The authors thank M. Cetto and K. Pollinger for excellent technical assistance. We thank Dr. L. Kuntz-Schugarth specifically for sorting the GFP-transfected cells.

1
Holmgren L., O’Reilly M. S., Folkman J. Dormancy of micrometastases: balanced proliferation and apoptosis in the presence of angiogenesis suppression.
Nat. Med.
,
1
:
149
-153,  
1995
.
2
O’Reilly M. S., Holmgren L., Shing Y., Chen C., Rosenthal R. A., Moses M., Lane W. S., Cao Y., Sage E. H., Folkman J. Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma.
Cell
,
79
:
315
-328,  
1994
.
3
O’Reilly M. S., Holmgren L., Shing Y., Chen C., Rosenthal R. A., Cao Y., Moses M., Lane W. S., Sage E. H., Folkman J. Angiostatin: a circulating endothelial cell inhibitor that suppresses angiogenesis and tumor growth.
Cold Spring Harb. Symp. Quant. Biol.
,
59
:
471
-482,  
1994
.
4
O’Reilly M. S., Holmgren L., Chen C., Folkman J. Angiostatin induces and sustains dormancy of human primary tumors in mice.
Nat. Med.
,
2
:
689
-692,  
1996
.
5
O’Reilly M. S., Boehm T., Shing Y., Fukai N., Vasios G., Lane W. S., Flynn E., Birkhead J. R., Olsen B. R., Folkman J. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth.
Cell
,
88
:
277
-285,  
1997
.
6
O’Reilly M. S. Angiostatin: an endogenous inhibitor of angiogenesis and of tumor growth.
EXS
,
79
:
273
-294,  
1997
.
7
O’Reilly M. S., Pirie-Shepherd S., Lane W. S., Folkman J. Antiangiogenic activity of the cleaved conformation of the serpin antithrombin.
Science (Wash. DC)
,
285
:
1926
-1928,  
1999
.
8
Luzzi K. J., MacDonald I. C., Schmidt E. E., Kerkvliet N., Morris V. L., Chambers A. F., Groom A. C. Multistep nature of metastatic inefficiency: dormancy of solitary cells after successful extravasation and limited survival of early micrometastases.
Am. J. Pathol.
,
153
:
865
-873,  
1998
.
9
Drixler T. A., Rinkes I. H., Ritchie E. D., van Vroonhoven T. J., Gebbink M. F., Voest E. E. Continuous administration of angiostatin inhibits accelerated growth of colorectal liver metastases after partial hepatectomy.
Cancer Res.
,
60
:
1761
-1765,  
2000
.
10
Asaishi K., Endrich B., Gotz A., Messmer K. Quantitative analysis of microvascular structure and function in the amelanotic melanoma A-Mel-3.
Cancer Res.
,
41
:
1898
-1904,  
1981
.
11
Carmeliet P., Jain R. K. Angiogenesis in cancer and other diseases.
Nature (Lond.)
,
407
:
249
-257,  
2000
.
12
Guba M., Steinbauer M., Buchner M., Frolich D., Farkas S., Jauch K. W., Anthuber M. Differential effects of short-term ace- and AT1-receptor inhibition on postischemic injury and leukocyte adherence in vivo and in vitro.
Shock
,
13
:
190
-196,  
2000
.
13
Morris V. L., MacDonald I. C., Koop S., Schmidt E. E., Chambers A. F., Groom A. C. Early interactions of cancer cells with the microvasculature in mouse liver and muscle during hematogenous metastasis: videomicroscopic analysis.
Clin. Exp. Metastasis
,
11
:
377
-390,  
1993
.
14
Naumov G. N., Wilson S. M., MacDonald I. C., Schmidt E. E., Morris V. L., Groom A. C., Hoffman R. M., Chambers A. F. Cellular expression of green fluorescent protein, coupled with high-resolution in vivo videomicroscopy, to monitor steps in tumor metastasis.
J. Cell Sci.
,
112
:
1835
-1842,  
1999
.
15
Stripecke R., Carmen V. M., Skelton D., Satake N., Halene S., Kohn D. Immune response to green fluorescent protein: implications for gene therapy.
Gene Ther.
,
6
:
1305
-1312,  
1999
.
16
Mosier D. E., Stell K. L., Gulizia R. J., Torbett B. E., Gilmore G. L. Homozygous scid/scid-beige/beige mice have low levels of spontaneous or neonatal T cell-induced B cell generation.
J. Exp. Med.
,
177
:
191
-194,  
1993
.
17
Guba M., Bosserhoff A. K., Steinbauer M., Abels C., Anthuber M., Buettner R., Jauch K. W. Overexpression of melanoma inhibitory activity (MIA) enhances extravasation and metastasis of A-mel 3 melanoma cells in vivo.
Br. J. Cancer
,
83
:
1216
-1222,  
2000
.
18
Morris V. L., Schmidt E. E., MacDonald I. C., Groom A. C., Chambers A. F. Sequential steps in hematogenous metastasis of cancer cells studied by in vivo videomicroscopy.
Invasion Metastasis
,
17
:
281
-296,  
1997
.
19
Sckell A., Safabakhsh N., Dellian M., Jain R. K. Primary tumor size-dependent inhibition of angiogenesis at a secondary site: an intravital microscopic study in mice.
Cancer Res.
,
58
:
5866
-5869,  
1998
.
20
Volpert O. V., Lawler J., Bouck N. P. A human fibrosarcoma inhibits systemic angiogenesis and the growth of experimental metastases via thrombospondin-1.
Proc. Natl. Acad. Sci. USA
,
95
:
6343
-6348,  
1998
.
21
Wu Z., O’Reilly M. S., Folkman J., Shing Y. Suppression of tumor growth with recombinant murine angiostatin.
Biochem. Biophys. Res. Commun.
,
236
:
651
-654,  
1997
.
22
Cameron M. D., Schmidt E. E., Kerkvliet N., Nadkarni K. V., Morris V. L., Groom A. C., Chambers A. F., MacDonald I. C. Temporal progression of metastasis in lung: cell survival, dormancy, and location dependence of metastatic inefficiency.
Cancer Res.
,
60
:
2541
-2546,  
2000
.
23
Li C. Y., Shan S., Huang Q., Braun R. D., Lanzen J., Hu K., Lin P., Dewhirst M. W. Initial stages of tumor cell-induced angiogenesis: evaluation via skin window chambers in rodent models.
J. Natl. Cancer Inst.
,
92
:
143
-147,  
2000
.
24
Cao Y., O’Reilly M. S., Marshall B., Flynn E., Ji R. W., Folkman J. Expression of angiostatin cDNA in a murine fibrosarcoma suppresses primary tumor growth and produces long-term dormancy of metastases.
J. Clin. Investig.
,
101
:
1055
-1063,  
1998
.