Tumor metastasis is not due to adaptation of cells to a new organ environment

DNA barcoded competitive clone-initiating cell analysis reveals novel features of metastatic growth in a cancer xenograft model

A problematic feature of many human cancers is a lack of understanding of mechanisms controlling organ-specific patterns of metastasis, despite recent progress in identifying many mutations and transcriptional programs shown to confer this potential. To address this gap, we developed a methodology that enables different aspects of the metastatic process to be comprehensively characterized at a clonal resolution. Our approach exploits the application of a computational pipeline to analyze and visualize clonal data obtained from transplant experiments in which a cellular DNA barcoding strategy is used to distinguish the separate clonal contributions of two or more competing cell populations. To illustrate the power of this methodology, we demonstrate its ability to discriminate the metastatic behavior in immunodeficient mice of a well-established human metastatic cancer cell line and its co-transplanted LRRC15 knockdown derivative. We also show how the use of machine learning to quantify clone-initiating cell (CIC) numbers and their subsequent metastatic progeny generated in different sites can reveal previously unknown relationships between different cellular genotypes and their initial sites of implantation with their subsequent respective dissemination patterns. These findings underscore the potential of such combined genomic and computational methodologies to identify new clonally-relevant drivers of site-specific patterns of metastasis.

Amigo2-upregulation in Tumour Cells Facilitates Their Attachment to Liver Endothelial Cells Resulting in Liver Metastases

Since liver metastasis is the main cause of death in cancer patients, we attempted to identify the driver gene involved. QRsP-11 fibrosarcoma cells were injected into the spleens of syngeneic mice to isolate tumour sub-populations that colonize the liver. Cells from liver metastatic nodules were established and subsequently injected intrasplenically for selection. After 12 cycles, the cell subline LV12 was obtained. Intravenous injection of LV12 cells produced more liver metastases than QRsP-11 cells, whereas the incidence of lung metastases was similar to that of QRsP-11 cells. LV12 cells adhered to liver-derived but not to lung-derived endothelial cells. DNA chip analysis showed that amphoterin-induced gene and open reading frame 2 (Amigo2) was overexpressed in LV12 cells. siRNA-mediated knockdown of Amigo2 expression in LV12 cells attenuated liver endothelial cell adhesion. Ex vivo imaging showed that suppression of Amigo2 in luciferase-expressing LV12 cells reduced attachment/metastasis to liver to the same level as that observed with QRsP-11 cells. Forced expression of Amigo2 in QRsP-11 cells increased liver endothelial cell adhesion and liver metastasis. Additionally, Amigo2 expression in human cancers was higher in liver metastatic lesions than in primary lesions. Thus, Amigo2 regulated tumour cell adhesion to liver endothelial cells and formation of liver metastases.

Combined Effect of Transfusion and Blood Groups on the Survival of Patients With Breast Cancer. A Clinical Study of 901 Patients †

Previous reports by the authors and by others have shown that trans fusion adversely affects the survival of cancer patients. To delineate fur ther the mechanism of action of blood transfusion on biological behavior of cancer, the authors studied the role of the major blood groups ABO in 901 patients who had all undergone radi cal surgery for operable primary breast cancer and among whom 294 had received at least one unit of whole blood perioperatively.

Multivariate analysis using Cox's regression models on life tables was used. It was found that the adverse effect of transfusion on survival was a blood group-related phenomenon manifesting its action in patients with blood groups A and B (P = 0.0006 and P = 0.006 respectively). The most im pressive effect of transfusion was seen in patients whose age was be tween forty-one and fifty (51.9% ten- year actuarial survival in 47 transfused patients compared with 73.9% in 70 nontransfused ones).

A hypothesis is presented suggest ing that transfusion exerts its effect on the hemostatic system, promoting clotting in group A and B patients and fibrinolysis in group O and AB patients, and thus affects metastatic cascade, with results that have been presented in this series.

Selection of metastatic breast cancer cells based on adaptability of their metabolic state

A small subpopulation of highly adaptable breast cancer cells within a vastly heterogeneous population drives cancer metastasis. Here we describe a function-based strategy for selecting rare cancer cells that are highly adaptable and drive malignancy. Although cancer cells are dependent on certain nutrients, e.g., glucose and glutamine, we hypothesized that the adaptable cancer cells that drive malignancy must possess an adaptable metabolic state and that such cells could be identified using a robust selection strategy. As expected, more than 99.99% of cells died upon glutamine withdrawal from the aggressive breast cancer cell line SUM149. The rare cells that survived and proliferated without glutamine were highly adaptable, as judged by additional robust adaptability assays involving prolonged cell culture without glucose or serum. We were successful in isolating rare metabolically plastic glutamine-independent (Gln-ind) variants from several aggressive breast cancer cell lines that we tested. The Gln-ind cells overexpressed cyclooxygenase-2, an indicator of tumor aggressiveness, and they were able to adjust their glutaminase level to suit glutamine availability. The Gln-ind cells were anchorage-independent, resistant to chemotherapeutic drugs doxorubicin and paclitaxel, and resistant to a high concentration of a COX-2 inhibitor celecoxib. The number of cells being able to adapt to non-availability of glutamine increased upon prior selection of cells for resistance to chemotherapy drugs or resistance to celecoxib, further supporting a linkage between cellular adaptability and therapeutic resistance. Gln-ind cells showed indications of oxidative stress, and they produced cadherin11 and vimentin, indicators of mesenchymal phenotype. Gln-ind cells were more tumorigenic and more metastatic in nude mice than the parental cell line as judged by incidence and time of occurrence. As we decreased the number of cancer cells in xenografts, lung metastasis and then primary tumor growth was impaired in mice injected with parental cell line, but not in mice injected with Gln-ind cells.

AACR centennial series: the biology of cancer metastasis: historical perspective

Metastasis resistant to therapy is the major cause of death from cancer. Despite almost 200 years of study, the process of tumor metastasis remains controversial. Stephen Paget initially identified the role of host-tumor interactions on the basis of a review of autopsy records. His "seed and soil" hypothesis was substantiated a century later with experimental studies, and numerous reports have confirmed these seminal observations. An improved understanding of the metastatic process and the attributes of the cells selected by this process is critical for the treatment of patients with systemic disease. In many patients, metastasis has occurred by the time of diagnosis, so metastasis prevention may not be relevant. Treating systemic disease and identifying patients with early disease should be our goal. Revitalized research in the past three decades has focused on new discoveries in the biology of metastasis. Even though our understanding of molecular events that regulate metastasis has improved, the contributions and timing of molecular lesion(s) involved in metastasis pathogenesis remain unclear. Review of the history of pioneering observations and discussion of current controversies should increase understanding of the complex and multifactorial interactions between the host and selected tumor cells that contribute to fatal metastasis and should lead to the design of successful therapy.

AACR centennial series: the biology of cancer metastasis: historical perspective

Metastasis resistant to therapy is the major cause of death from cancer. Despite almost 200 years of study, the process of tumor metastasis remains controversial. Stephen Paget initially identified the role of host-tumor interactions on the basis of a review of autopsy records. His "seed and soil" hypothesis was substantiated a century later with experimental studies, and numerous reports have confirmed these seminal observations. An improved understanding of the metastatic process and the attributes of the cells selected by this process is critical for the treatment of patients with systemic disease. In many patients, metastasis has occurred by the time of diagnosis, so metastasis prevention may not be relevant. Treating systemic disease and identifying patients with early disease should be our goal. Revitalized research in the past three decades has focused on new discoveries in the biology of metastasis. Even though our understanding of molecular events that regulate metastasis has improved, the contributions and timing of molecular lesion(s) involved in metastasis pathogenesis remain unclear. Review of the history of pioneering observations and discussion of current controversies should increase understanding of the complex and multifactorial interactions between the host and selected tumor cells that contribute to fatal metastasis and should lead to the design of successful therapy.

CD26/dipeptidyl peptidase IV regulates prostate cancer metastasis by degrading SDF-1/CXCL12

Stromal derived factor-1 (SDF-1 or CXCL12) expressed by osteoblasts and endothelial cells, and its receptors CXCR4 and CXCR7/RDC1 are key molecular determinants in prostate cancer (PCa) metastasis. What drives PCa cells into the extravascular marrow space(s) once they make contact with the blood vessel endothelium, however remains unclear. Here, we evaluated whether degradation of CXCL12 facilitates PCa cell entry into the marrow cavity by locally lowering CXCL12 levels intravascularly. To explore this possibility, co-cultured conditioned media from PCa cells and endothelial cells were evaluated for their ability to degrade biotinylated CXCL12 (bCXCL12). Co-culture of PCa cells/endothelial cells resulted in greater digestion of CXCL12 than was achieved by either cell type alone, and this activity regulated invasion in vitro. The ability to degrade CXCL12 was not however observed in PCa and osteoblasts co-cultures. Fractionation and inhibitor studies suggested that the activity was CD26/dipeptidyl peptidase IV (DPPIV) and possibly other cysteine/serine proteases. By inhibiting CD26/DPPIV, invasion and metastasis of PCa cell lines were enhanced in in vitro and in vivo metastasis assays. Together, these data suggest that the degradation of CXCL12 by CD26/DPPIV may be involved in the metastatic cascades of PCa, and suggests that inhibition of CD26/DPPIV may be a trigger of PCa metastasis.

The role of sialomucin CD164 (MGC-24v or endolyn) in prostate cancer metastasis

BACKGROUND

The chemokine stromal derived factor-1 (SDF-1 or CXCL12) and its receptor CXCR4 have been demonstrated to be crucial for the homing of stem cells and prostate cancers to the marrow. While screening prostate cancers for CXCL12-responsive adhesion molecules, we identified CD164 (MGC-24) as a potential regulator of homing. CD164 is known to function as a receptor that regulates stem cell localization to the bone marrow.

RESULTS

Using prostate cancer cell lines, it was demonstrated that CXCL12 induced both the expression of CD164 mRNA and protein. Functional studies demonstrated that blocking CD164 on prostate cancer cell lines reduced the ability of these cells to adhere to human bone marrow endothelial cells, and invade into extracellular matrices. Human tissue microarrays stained for CD164 demonstrated a positive correlation with prostate-specific antigen levels, while its expression was negatively correlated with the expression of androgen receptor.

CONCLUSION

Our findings suggest that CD164 may participate in the localization of prostate cancer cells to the marrow and is further evidence that tumor metastasis and hematopoietic stem cell trafficking may involve similar processes.

Biological properties and gene expression associated with metastatic potential of human osteosarcoma

Lung metastasis has a great influence on the prognosis of patients with osteosarcoma. We previously established two high-metastatic sublines, M112 and M132, from the HuO9 human osteosarcoma cell line by in vivo selection. In this study, we newly isolated a high-metastatic subline, H3, and three low-metastatic sublines, L6, L12 and L13, from HuO9 by the dilution plating method. Three high-metastatic sublines produced more than 200 metastatic nodules in the lung, while three low-metastatic sublines produced no or few nodules after injection of 2 x 10(6) cells into the tail vein of nude mice. There were significant differences in the motility and invasiveness between high- and low-metastatic sublines, whereas the growth rates in vitro and the tumorigenicity in vivo showed no correlation with their metastatic abilities. Early adherence to culture plates was significantly lower in two of three low-metastatic sublines, which occupied smaller surface areas on the culture plates than other sublines did. Comparison of the expression of 637 cancer-related genes by cDNA microarray revealed that seven genes were differentially expressed between high- and low-metastatic sublines. Among them, five genes (AXL, TGFA, COLL7A1, WNT5A, and MKK6) were associated with adherence, motility, and/or invasiveness. These results suggest that the differences in motility/invasiveness and adhesive abilities are key determinants of lung metastasis in osteosarcoma.

High collagenolytic activity in spontaneously highly metastatic variants derived from a human pancreatic cancer cell line (SUIT-2) in nude mice

Cell lines with high metastatic capacity to the lung were established by sequential passage of a human pancreatic cancer cell line (SUIT-2) through the lung of a nude mouse, via the lateral tail vein and from a subcutaneous inoculum. Cells of the parental SUIT-2 and sublines S2-VPx (x-cycle selection from SUIT-2 cells, by Vein-Pulmonary metastasis-culture) and S2-CPx (x-cycle selection, by Cutis-Pulmonary metastasis-culture) were injected intravenously or subcutaneously into nude mice to produce experimental or spontaneous lung metastasis. The S2-VP10 cell line produced pulmonary metastases in 100% of the nude mice, when injected intravenously. It failed, however, to produce more lung colonies than its parent cell line, when injected subcutaneously. The S2-CP8 cell line produced extensive pulmonary metastases in 100% of the nude mice, when injected either intravenously or subcutaneously. This study indicates that the nude mouse provided a good model for in vivo selection of metastatic cells from SUIT-2 cells both experimentally and spontaneously, and that the SUIT-2, S2-VPx, and S2-CPx cell lines will be valuable in the study of human cancer metastasis. We previously reported high levels of ezrin expression in the S2-VP10 and S2-CP8 cell lines. Here we show that these cell lines exhibit a greater capacity to invade or attach to various extracellular matrix components than the parent SUIT-2 cells. The S2-CP8 cell lines also exhibit greater level of type-I and type-IV collagen-degrading activity than the parent SUIT-2 cell line and the S2-VP10 cell line, which shows similar collagen-degrading activity to the parent SUIT-2 cells. In RT-PCR studies, SUIT-2, S2-CP8 and S2-VP10 cell lines constitutively expressed many matrix metalloproteinases (MMP-1, MMP-2, MMP-3, MMP7, MMP-9, MMP-10 and MMP-14). These results suggest that some parameters that enhance adhesion and invasion are important to both experimental and spontaneous metastasis and the collagen degrading enzymes are predicted to play a key-role during spontaneous metastasis.

Brain metastases: biology and the role of the brain microenvironment

Metastatic lesions constitute the most frequently occurring malignancy in the brain, and their detection portends a grim prognosis. Efforts to treat these lesions have failed partly because the biologic processes that govern their development are poorly understood. In recent years, it has become evident that metastases occur as a result of a multistep process involving a rigorous natural selection of cells in the primary tumor that bear molecular and biologic characteristics permitting brain metastasis. In addition, recent studies have uncovered the importance of the brain microenvironment and its contribution to the metastatic process. The development of targeted therapies against brain metastases demands a better understanding of these molecular processes and the factors that influence them. This review examines the interplay between tumor cells and host brain tissue in the context of our current understanding of the role of various molecules involved in the metastatic process.

A dynamical model for the growth and size distribution of multiple metastatic tumors

Metastasis is the spread of tumors culminating in the establishment of one or more secondary tumors at remote sites. In deciding the best treatment for cancer therapy, estimations of the colony size of metastatic tumors and predictions of the future spread of colonies are needed. A dynamical model for the colony size distribution of multiple metastatic tumors is presented here. The dynamics is described by equations that incorporate both the colonization by metastasis and the growth of each colony. When the colony growth is subject to the Gompertz function, the explicit solution obtained tends to an asymptotic stable distribution that shows a monotonically decreasing or U-shaped pattern according to the values of clinically significant parameters, such as the colonization coefficient and the fractal dimension of blood vessels. This predicted colony size distribution agrees well with successive data of a clinically observed size distribution of multiple metastatic tumors of liver cancer. The combined analysis of the theoretical colony size distribution and clinical data will give useful information on the diagnosis and the therapy for cancer patients.

Growth patterns of pulmonary metastases and primary tumours from five murine fibrosarcoma cell clones

The growth patterns, including the size, shape and regional preferences, of lung metastases from five murine fibrosarcoma cell clones were studied. Spontaneous metastases developed from tumours formed by subcutaneous inoculation of the cell clones. Lung colonies (experimental metastases) were established by i.v. injection of cells. The numbers of both spontaneously and experimentally formed subpleural lung metastases were counted through a stereomicroscope. The fraction of colonies that was located subpleurally was determined in histological sections of lungs. The growth kinetics of clonally derived primary tumours, and the number of spontaneous and experimental lung metastases, differed greatly between certain cell clones. The number of spontaneous lung metastases was correlated with the maximum size of primary tumours. No close correlation was observed between the size of the primary tumours and the size of experimental metastases. There were differences between the cell clones in the shape and regional preferences of their lung deposits. The subpleural colonies were generally larger than the intrapulmonary ones. Thus, both the regional distribution and the growth pattern of lung deposits differed between the clones.

Drug sensitivity and metastatic ability in B16 melanoma cells

We have previously reported that highly metastatic cell lines derived from KHT fibrosarcoma (KHT 35L1) and B16 melanoma (B16F10) are more resistant to N-phosphonacetyl-L-aspartate (PALA) and methotrexate (MTX) than the parental cell lines. This correlation between drug resistance and metastatic ability suggested the possibility that both phenotypes might have arisen in parallel as a result of a similar mechanism. In this study, we examined this possibility by reproducing the selection procedure for B16F10 cells (by serial passage of B16F1 cells as lung nodules) and testing the cells at each passage for changes in resistance to PALA and MTX. The results confirm that serial passage of B16F1 cells as lung nodules (LP) selects for cells with increasing metastatic ability (100-fold after seven passages), but these cells did not develop increased resistance to PALA and became more sensitive to MTX. For comparison B16F1 cells were also serially passaged (six passages) as leg tumors (LT). These cells became slightly more metastatic (3-fold) than B16F1 cells maintained in tissue culture, and demonstrated a small increase in sensitivity to MTX, as in the LP lines. There was also an apparent increase in resistance to PALA. In no instance was there a parallel increase in drug resistance and metastatic ability indicating that these two phenotypes do not necessarily arise in parallel in this cell line.

Different lectin-binding patterns in primary breast cancers and their metastases

The expression of glycoconjugates in primary tumors and their metastases in 18 consecutive cases of metastasized breast cancer was studied by use of lectin histochemistry. Paraffin sections were stained with a panel of seven fluorochrome-labeled lectins with defined sugar specificities. The study revealed variation in the lectin binding patterns of individual cancers. In the primary tumors the lectin reactivity was diversified, whereas in their respective metastases it was rather homogeneous. This finding indicates that there is intratumoral heterogeneity in the primary cancers, whereas the selected subclones of malignant cells with restricted glycoconjugate expression seem to give rise to metastases.

Orthotopic implantation is essential for the selection, growth and metastasis of human renal cell cancer in nude mice [corrected]

Human neoplasms are heterogeneous for a variety of biological properties that include invasion and metastasis. The presence of a small subpopulation of cells with a highly metastatic phenotype has important clinical implications for diagnosis and therapy of cancer. For this reason, it is important to develop an animal model for the selection and isolation of metastatic variants from human neoplasms and for testing the metastatic potential of human tumor cells. We have implanted human renal cell carcinoma (HRCC) cells (obtained from a surgical specimen) into different organs of nude mice and then recovered the tumors and established each in culture. The 5 established lines differed in their biological-metastatic properties and had a unique karyotype, indicating that growth at different organs selects for different subpopulations of HRCC. Moreover, the HRCC did not metastasize unless they were implanted orthotopically. These findings indicate that the appropriate nude mouse model for studying the biology and therapy of HRCC must be based on the orthotopic implantation of tumor cells.

In vivo selection of human renal cell carcinoma cells with high metastatic potential in nude mice

Studies were made to determine whether the orthotopic implantation of human renal cell carcinoma cells (HRCC) into nude mice will produce distant metastases, thus allowing for the selection of variant cells with high metastatic potential. The parental SN12C line was established in culture from a surgical specimen of HRCC. The renal subcapsule (RSC) of adult nude mice was injected with SN12C cells; the mice were killed when they became moribund. Cell lines were established from either single or multiple lung HRCC metastases. The intravenous injection of many (but not all) of the metastasis-derived lines produced significantly more experimental metastases than did the parental cells. The injection of cells into the RSC demonstrated that, in general, cells derived from spontaneous metastases were more metastatic than cells of the parental line. Hence adult nude mice can be used to select HRCC cells with high metastatic potential. These HRCC variant lines offer a good model for studying the cell properties of metastatic HRCC.

Cultured brain endothelium inhibits the cytocidal action of natural killer cells on glioma

The killing of GL26 and YAC-1 cells by natural killer cells (NKC) is reduced in the presence of a monolayer of endothelial cells. This reduction in cytotoxicity correlates with the degree of adhesion between the tumor cells and the endothelial monolayers. The cytotoxicity of NKC toward glioma was 10% when carried out on plastic, but a monolayer of endothelium derived from brain inhibited the cytotoxicity by about 90%. Endothelium from thoracic duct and lung also inhibited cytotoxicity by about 90%, endothelium from aorta inhibited by 55% and that from ovary by only 45%. Cytotoxicity of NKC toward YAC-1 (a control NK target) was 40% on plastic, but a monolayer of endothelium from thoracic duct inhibited the cytotoxicity by 75%. Endothelium from brain and lung inhibited cytotoxicity by about 60%, aorta by 50%, and ovary by 40%. Interactions between tumor cells and the host-organ microvascular endothelium appear to protect neoplastic cells from natural surveillance mechanisms and may play a role in the formation of metastatic tumor deposits.

Growth regulation of cancer metastases by their host organ

We analyzed mechanisms responsible for organ-specific metastasis by using two melanoma sublines derived from the same mouse tumor, of which one colonizes the lungs (F10) and the other colonizes the liver (L8) after intravenous injection. Both lines were obtained by selective growth in lung or liver after injection of tumor cells into a tail vein or portal vein. Contrary to common concepts, the cells of the liver-colonizing melanoma line do not accumulate preferentially in the liver after intravenous administration in vivo. However, the selective survival and proliferation of these melanoma cells in the target organ (liver) may be explained by the unexpected observation that they can be specifically stimulated to proliferate in the presence of hepatocytes, whereas the cells of the lung-colonizing line cannot. Growth promotion under coculture conditions in vitro was monitored both by thymidine incorporation into DNA and by increase in cell numbers. The proliferative stimulus is not mediated by an easily diffusible factor but rather depends upon direct contact between liver cells and those tumor cells that metastasize to that particular organ.

Organ specificity of tumor metastasis: role of preferential adhesion, invasion and growth of malignant cells at specific secondary sites

The locations of distant secondary tumors in many clinical cancers and animal tumors are nonrandom, and their distributions cannot be explained by simple anatomical or mechanical hypotheses based on the simple lodgment or trapping of tumor cell emboli in the first capillary bed encountered. Evidence from certain experimental tumor systems supports Paget's 'seed and soil' hypothesis on the nonrandom distributions of metastases, in which the unique properties of particular tumor cells ('seeds') and the different characteristics of each organ microenvironment ('soil') collectively determine the organ preference of metastasis. Experimentally, differential tumor cell adhesion to organ-derived microvessel endothelial cells and organ parenchymal cells, differential invasion of basement membranes and organ tissues, and differential responses to organ-derived growth-stimulatory and -inhibitory factors all appear to be important determinants in explaining the organ preference of metastasis. Each tumor system may achieve organ specificity because of its own unique set of multiple metastasis-associated properties and responses to host microenvironments. As neoplasms progress to more highly malignant states multisite metastases are more likely and organ-specific metastases may be masked or circumvented owing to stochastic events, tumor cell diversification, host selection processes, and increased production of tumor autocrine molecules that may modulate adhesion, invasion, growth, and other properties important in metastasis. The importance of each of these properties, however, appears to vary considerably among different metastatic tumor systems. These and other tumor cell and host properties may eventually be used to predict and explain the unique metastatic distributions of certain human malignancies.

In vivo emergence of a highly metastatic tumour cell line from a rat rhabdomyosarcoma after treatment with an alkylating agent

Rats bearing a transplanted nickel-induced rhabdomyosarcoma (RMS 9-4/0), treated with chlorozotocin (CZT), an alkylating agent, showed an amplified metastatic invasion of the lung (median of 165 lung tumour nodules, compared to 3 for untreated controls). A higher level of metastatic invasion (200 nodules) was reached spontaneously after the grafting of the S4T line, which was obtained by successive in vivo passages of RMS 9-4/0 cells in CZT treated rats. S4T tumour cells also invaded the liver and a considerable proportion of the lymph nodes. The NT4T line, obtained by successive in vivo passages in untreated rats, showed a lesser degree of enhancement of metastatic capacity (57 nodules). Both derived lines proved to be more aggressive than the parental, proliferated more rapidly, and were resistant to CZT toxicity. Only the non-treated lineage became more resistant to NK lysis. The S4T line lost its myogenic differentiation and was best described as a fibrohistiosarcoma, whereas NT4T did not. Chromosome analysis demonstrated a reduced range of chromosome number per cell in both lines. We conclude that both S4T and NT4T tumours became more metastatic than RMS 9-4/0 as the result of tumour progression through in vivo passages, and that in addition S4T acquired a spontaneously higher metastatic potential, similar to that which occurred in rats grafted with RMS 9-4/0 or NT4T tumours and treated by CZT. This suggests an inheritable mutation in the S4T line.

The influence of the microenvironment of liver-specific tumor cell colonization in a murine tumor model

Malignant tumors often show an organ-specific metastatic spread. Some cells of the primary apparently bear an affinity for growing in the microenvironment of certain organs. After i.v. injection of myofibrosarcoma cells from the primary ER 15-P into the tail vein of male C57/Bl6J mice, metastases developed in various organs. A tumor cell line (ER 15-Me3) isolated from liver metastases of the primary was found to colonize preferentially to the liver. To find out whether the liver specificity of the tumor cell line ER 15-Me3 depended on the hepatic microenvironment, tumor cells from this line were transplanted i.m. into the thighs of mice once (ER 15-Me3 i.m.1) or 5 times (ER 15-Me3 i.m.5), and then injected into the tail vein of mice. A part of the 5-times passaged tumor cell line was also injected into the mesenteric vein (ER 15-Me3 i.m.5-Me1) prior to reinjection into the tail vein.

RESULTS

After i.v. administration of tumor cells from the first i.m. passage of the tumor cell line (ER 15-Me3 i.m.1) into the tail vein, the liver-specific metastatic behavior of tumor cells remained stable. Following the i.v. injection of tumor cells from the 5th i.m. transplant generation of the tumor cell line (ER 15-Me3 i.m.5) into the tail vein, organ distribution was similar to that of the primary. After only 1 mesenteric vein passage of the 5-times i.m. transplanted line ER 15-Me3 i.m.5-Me1 followed by i.v. injection into the tail vein, did tumor cells regain their liver-specific colonizing potential. Thus, the liver-specific tumor cell line seems to contain a small number of other tumor cell populations from the unselected primary. In the muscle, these tumor cells have a growth advantage over the liver-specific cells, while the latter will grow better in the liver. This indicates that the microenvironment may be one important factor influencing the organ-specific metastatic pattern of tumor cells.

Potassium inhibition of transforming protein P85gag-mos and reversal of the transformed phenotype in 6m2 cells

K+ at high concentrations (52-72 mM hypertonic KCl) has been reported to induce reverse transformation in the 6m2 cell, which is a clone of normal rat kidney cells (NRK) infected with a temperature-sensitive transformation virus. When exposed to high K+, 6m2 cells grown at the permissive temperature (33 degrees C) exhibit normal morphology and reduced soft agar growth, characteristics of cells grown at nonpermissive temperature (39 degrees C). In the current study, flattening of cells and rearrangement of surface microvilli were demonstrated by scanning electron microscopy to occur within 6 hr of exposure to high K+, similar to the effect of temperature shift to 39 degrees C. Exposure to K+ resulted in a 90% inhibition of P85gag-mos-associated serine kinase activity within 5 min, with a subsequent reduction of up to 75% of the synthesis of this protein. These alterations in the putative transforming protein were similar to those induced by temperature shift and were considered to be the basis for retrotransformation. The cell microtubular system and F-actin cables were affected more slowly by K+ than by a temperature shift to 39 degrees C. The former did not achieve the fine reticulum network seen in NRK cells until 72 hr later, but the latter remained aberrant. The effect on the enzyme might be mediated by alteration in phosphorylation, but the mechanism by which kinase inactivation induces retrotransformation is not yet known.

Mechanism of liver-specific metastatic tumor spread in a murine tumor model

Malignant tumors frequently show an organ-specific metastatic spread, the causes of which are still largely unknown. Using an experimental tumor model, a methylcholanthrene-induced pleomorphic myofibrosarcoma ER 15-P of the C57Bl6J mouse, we wanted to find out whether this phenomenon is due to an adaptation or to a selection of tumor cells. After i.v. injection of tumor cells from the primary ER 15-P into the tail vein of male mice, metastases were regularly found in the lungs, mediastinal lymph nodes, and brain, as well as in the liver and kidneys, and occasionally in the adrenals. The following experimental procedures were used to isolate a tumor cell line with a possible liver preference: (1) Tumor cells from the primary ER 15-P were injected into a mesenteric vein of male mice. Tumor cells from the resulting liver colonies were again injected into the portal system of one group of mice. In a second group, part of the same cell suspension was injected into the tail vein. This procedure was performed four times. (2) Tumor cells from the primary ER 15-P were applied into the tail vein of male mice. Tumor cells from the resulting liver metastases were reinjected directly into the tail vein. This experiment was repeated three times. (3) Tumor cells from the primary ER 15-P were injected into the tail vein of male mice. Tumor cells from liver metastases were then injected, first, into the portal system of one group of male mice, and thereafter into the tail vein of another group of animals. This experiment was repeated twice. The following results were obtained: (1) By a repeated adaptation of tumor cells from the primary ER 15-P to liver tissue, no tumor cell line could be isolated that would show a preferential metastatic spread to this organ after tail-vein injection. (2) Repeated i.v. passages of tumor cells from liver metastases into the tail vein led to the selection of a tumor cell line with a tendency to liver metastasis. (3) Tumor cells selected from liver metastases induced via tail-vein injection showed, after a prolonged stay in the liver and a successive i.v. passage into the tail vein, a marked specificity for this organ. These results indicate that the liver-specific spread of tumor cells in our model is based on the selection of a tumor cell line from the primary ER 15-P influenced by the hepatic microenvironment.

Homing receptors and metastasis

As discussed in the preceding sections, there are several indications that the lymphocyte homing receptors involved in the normal process of lymphocyte recirculation are also relevant to the behavior of metastatic cells. Cell fusion experiments indicate that previously nonmetastatic cells can acquire metastatic capacity from fusion with normal lymphocytes. Murine T lymphomas that bear high levels of functional homing receptors can metastasize to peripheral lymphoid organs, whereas those lymphomas lacking homing receptors cannot. Virtually all lymph node metastases of lymphomas contain a high proportion of MEL-14hi cells, even if the primary tumor has been selected to be relatively deficient in these cells. Further investigations of the biology of lymphocyte homing receptors will reveal whether or not there are additional lymphocyte homing receptors and will clarify the role of lymphocyte homing receptors in metastasis. Antibodies against three lymphocyte homing receptors could therefore be useful for diagnosis and treatment of metastatic disease.

Growth in an organ microenvironment as a selective process in metastasis

The purpose of these studies was to determine whether the ability of tumor cells to grow in an organ parenchyma selects for cells with enhanced potential to metastasize to this organ. B16-F1 melanoma cells (with low metastatic potential) were cultured in vitro on fragments of mouse lung or kidney suspended in medium supplemented with only 1 per cent fetal bovine serum. Seven days later, the organs were enzymatically dissociated and tumor cells recovered and expanded in monolayer culture. Tumor cells were then harvested and seeded onto fresh organ fragments. This sequence was repeated six times. The cells designated as B16 Lung-6 and B16 Kidney-6 were then injected intravenously into C57BL/6 mice and the number of experimental metastases counted after 21 days. B16 Lung-6 and, to a lesser degree, B16 Kidney-6 cells produced significantly more lung tumor colonies than B16-F1 cells. Some factor(s) in the organ environment did appear, therefore, to select out cells with greater metastatic potential from the low metastatic B16-F1. Forty-five clones of the B16-F1 melanoma, isolated by limiting dilution, were screened for their ability to grow on explants of mouse lung in a low-serum medium. Four clones exhibiting least growth and four clones exhibiting most growth, as assessed by examining histological sections of the lung explants, were injected intravenously into syngeneic mice. The eight clonal populations differed in experimental metastatic potential, but this behavior did not correlate with the ability of cells to grow in lung explants in vitro. The data suggest that selecting for cells with enhanced ability to grow in an organ is a necessary but not sufficient condition for isolating cells with high metastatic potential.

Factors influencing the growth and metastatic behavior of tumors

The factors influencing the growth and metastatic behavior of experimental animal tumors are examined. 10(4) TA3Ha cells were injected intraperitoneally, intravenously, subcutaneously into the flank, and subcutaneously into the tail tissues of syngeneic strain A mice. The tumor takes from these injections were 50/50 (100%), 1/10 (10%), 10/10 (100%), and 7/12 (60%), respectively. The frequency of lung metastasis from these sites was 0, 100, 50, and 100%, respectively. At the time of host death, the flank and tail tumors were, respectively, 2.2 +/- 0.5 cm (geometric mean diameter) and 1.2 +/- 0.3 cm in the TA3Ha and 0.9 +/- 0.1 cm and 0.4 +/- 0.1 cm in the L1210 systems. TA3Ha tumors metastasized regularly to the lymph nodes but the L1210 tumors seldom metastasized to the lymph nodes. Tail implants of TA3Ha tumors behaved similarly in the athymic nude mice and strain A mice. TA3Ha cells inoculated into the Millipore chambers and maintained in mice for greater than 150 days were viable and able to form tumors. The results demonstrate that the anatomic location of the tumor affects the growth and metastatic behavior of the tumors, and that the tumors of different histologic origin metastasize differently even when grown in corresponding locations.

Organ specificity of metastatic tumor colonization is related to organ-selective growth properties of malignant cells

In cancers such as malignant melanoma, tumor spread or metastasis occurs preferentially to certain organ sites. Mouse B16 melanoma cell sublines that have been selected sequentially in vivo for enhanced blood-borne colonization of lung, brain, or ovary were tested for their survival and growth stimulation in vitro by soluble factors released from suspensions of mouse lung, brain, liver, ovary or kidney tissues. In general, the growth rate of lung-colonizing B16 cells was stimulated by high concentrations of lung-tissue-derived factors significantly more than by factors from the other tissues, whereas the growth rate of ovary-colonizing B16 cells was stimulated by ovary or lung-tissue-derived factors significantly more than by factors from the other tissues. In contrast, the growth of brain-colonizing B16 cells was not stimulated by factors released from brain tissue. When it occurred, stimulation of B16 cell growth by factors released from mouse organ tissues was dose-dependent. Liver tissue factors, and at high concentrations kidney tissue factors, inhibited cell growth of the B16 sublines, an inhibition correlating with the low potential of B16 cells to colonize liver and kidney in vivo. In addition to preferential target-organ cell adhesion found previously with B16 sublines (Nicolson et al., 1985a), the present results suggest that metastasis to specific organ sites is also dependent on the survival and growth of B16 cells affected by soluble organ-derived factors.

Rationale and methods for the use of nude mice to study the biology and therapy of human cancer metastasis

Human neoplasms are biologically heterogeneous. The extensive cellular diversity found in malignant neoplasms is generated by the rapid emergence of clonal subpopulations of tumor cells with different properties that include invasion, metastasis and responsiveness to treatment. Studies in rodent systems have indicated that cancer metastases can be clonal in their origin and that different metastases can originate from different progenitor cells from the primary tumor. This metastatic heterogeneity of tumor cells has many ramifications for studies of tumor biology, in general, and studies of therapy, in particular. The heterogeneous nature of metastatic human neoplasms can now be studied under defined conditions in healthy athymic nude mice. The neoplasms must be free of mouse pathogens and the mice must be kept in specific-pathogen-free conditions. Careful consideration must be given to the intimate tumor-host relationship for each tumor system studied, because the metastatic potential of human neoplasms can vary with the site of implantation into nude mice. Several methods for studying the biology of human neoplasms in the nude mouse are described as well as techniques to assure the success of these studies. The data show that the healthy young nude mouse can be a useful in vivo model for ascertaining the metastatic potential of human neoplasms, for selecting and maintaining cell variants of high metastatic potential from heterogeneous human tumors, and for studying therapeutic agents directed against metastatic cells proliferating in visceral organs.

Tumor cell metastasis

Local tissue invasion and the formation of metastatic lesions are characteristic properties of many malignant tumors. The formation of metastases is a complex process involving the passage of tumor cells from the site of the primary bulk tumor through successive connective tissue barriers, ultimately resulting in the growth of secondary tumor cell colonies in distinct target organ locations. At many stages in the metastatic process, tumor cells interact with multiple components of the extracellular matrix. Recently, the importance of basement membrane as a barrier to invasive cells has been recognized. In the course of the transition from in situ to invasive carcinoma, normal or dysplastic epithelial cells residing on a basement membrane are replaced by neoplastic cells which subsequently invade the basement membrane and enter the underlying stroma. Once in the stroma, tumor cells can then penetrate the walls of blood vessels or the lymphatic system and enter into the circulation. Circulating tumor cells next arrest in the lumina of small vessels, invade the vessel wall, and leave the circulation. These cells are now directly exposed to the extracellular matrix of a target organ where they may grow to form secondary tumors. Throughout the metastatic process tumor cells are thus in contact with, and are potentially responsive to, various components of the extracellular matrix. This review provides a survey of the recent advances in our understanding of the interactions of metastatic tumor cells with the extracellular matrix. Specifically, the role of basement membrane as a barrier to metastatic tumor cells is examined.

Atherosclerosis. I. A leiomyoproliferative disease of the arteries resulting from breakdown of the endothelial barrier to potent blood growth factors. II. Perspectives in atheroprophylaxis

For a historical survey of the pathogenesis of atherosclerosis the reader is referred to references 1 and 2. Comprehension of the vessel-blood interface homeostasis hinges upon an understanding of the pathophysiology of angio-lymphoid relationships. Even in the smooth contact of the intact hydrophobic intimal lining with the marginal flow of the circulatory stream, small amounts of thrombin and small aggregates of aging platelets float by under physiologic conditions. Since endothelial cells of the vascular intima contain receptors for thrombin, filling these receptors with thrombin becomes a stimulus for the production of prostacyclin (PGI2) by the endothelial cells; PGI2 in turn inhibits adherence of the small platelet aggregates by ADP; homeostasis is maintained. The size of physiologic thrombin-producing platelet microaggregates is controlled by physiologic levels of antithrombin III.

An investigation into the association between liver damage and metastatic uveal melanoma

We examined the effect of liver damage induced by carbon tetrachloride on the incidence of hepatic metastasis in the Greene hamster melanoma. Nine of 17 animals with liver damage had melanoma growth in the liver after the intraportal vein injection of tumor cells, whereas tumor growth occurred in four of 27 control animals (P less than .01). These results suggest that, in this model, liver damage predisposes to metastatic involvement by melanoma.

Relationship of macrophage content, immunogenicity, and metastatic potential of a murine osteosarcoma of recent origin

The purpose of these studies was to determine whether the high macrophage content (greater than 50 per cent) of the 90Sr-induced osteogenic sarcoma J (Os-J) of recent origin correlated with its immunogenicity or low metastatic potential. Cloning experiments demonstrated that the Os-J tumor is heterogeneous with regard to the production of experimental pulmonary metastases. Immunization-challenge studies in syngeneic mice and comparisons of tumor growth in normal or nude mice established that the slow growing Os-J tumor is poorly immunogenic. In vitro studies demonstrated that the Os-J tumor is highly susceptible to macrophages-mediated lysis. This may explain the slow growth of the tumor in normal recipients with an intact mononuclear phagocyte system, as compared with the more rapid emergence of tumors in macrophage-suppressed mice. However, spontaneous metastases of the Os-J tumor were not observed either in normal or macrophage-suppressed mice. Although a high macrophage infiltration of neoplasms could slow tumor growth, this was not associated with the immunogenicity of the neoplasm and did not appear to limit the spontaneous metastasis of this essentially benign neoplasm.

Pattern of organ colonization of metastasizing mouse L929-MM4 cells and the stimulation of lung tumor formation by cyclophosphamide

The pattern of lung colonization of L929-MM4 metastasis-derived cells was studied in C3HA/Hab mice. The incidence of mice with lung tumors and the number of tumor foci increased with the number of tumor cells in the i.v. inoculum until a 'plateau' was reached. However, some 30 per cent of the animals continued to show no macroscopic or microscopic evidence of metastases. Twenty days after the i.v. injection of 2.5 X 10(5) L929-MM4 cells, viable tumor cells were recovered from 80-90 per cent of the animals as judged by in vivo or in vitro assay methods, even though survival of mice under these conditions was 50 per cent at the end of a 3-month period. These results are discussed in relation to possible tumor dormant states. Cyclophosphamide stimulated both the incidence and the number of lung tumor foci after i.v. injection of tumor cells, if the drug was administered before the tumor cell inoculum. On the basis of the observed time dependency of this effect, and the low immunogenicity of L929-MM4 cells in C3HA/Hab mice, the results could be explained on the basis of drug damage to normal cells.

Cell surface molecules and tumor metastasis. Regulation of metastatic phenotypic diversity

Metastatic tumor cells are characterized by quantitative alterations in cell surface and other properties that confer to these cells their abilities to invade, disseminate, implant, survive and grow at secondary sites. Metastasis is also determined by a variety of host factors that prevent, allow or even stimulate metastatic processes. The emergence of diversified cell subpopulations in malignant tumors insures that some cells will ultimately become highly metastatic, resulting in tumor progression towards characteristics which are the most favorable for survival and growth. Unknown mechanisms appear to stimulate and then to control phenotypic diversification of tumor cell subpopulations. These mechanisms may be altered by genetic (mutational) and/or epigenetic (non-mutational) modifications that individually influence cells within a malignant neoplasm.

Differences in the peripheries of Lewis lung tumor cells growing in different sites in the mouse

The peripheries of Lewis lung (3LL) tumor cells growing in different organs of the mouse were studied by cell electrophoresis and electron microscopic quantitation of colloid iron hydroxide (CIH) adsorption before and after incubation with neuraminidase. The results show that cells growing in the kidney after direct injection have significantly higher anodic mobilities than cells growing in the subcutaneous sites from which they were derived, or in intramuscular sites, liver or spleen. The proportional contributions of cell surface sialic acids were similar in all sites. Electron microscopy of cells reacted with CIH indicates that the increased surface charge density of the tumor cells growing in the kidney is due to the presence of increased densities of non-CIH-binding ionogenic groups. Before neuraminidase treatment, the surface distribution patterns of CIH were indistinguishably random for 3LL cells growing in all sites. After neuraminidase treatment, significantly more clustering of CIH particles was observed on 3LL cells with a history of growth in the kidney than in subcutaneous sites. The changes observed in the 3LL cells growing in the kidney were irreversible, and persisted on multiple back-transplantation to subcutaneous sites. Detailed analysis of the results shows the changes to be due to the preferential selection of a subpopulation consisting of approximately 10% of the original (subcutaneous) tumor-cell population. This evidence for an irreversible site-induced selection of a pre-existing sub-population of 3LL cells contrasts with the reversible (modulation) site-induced adaptation previously observed by us in Walker-256 cancer cells, and therefore indicates that both selective and adaptive processes can occur. Even in the case of the 3LL cells, site-specific selection is not general, since the changes were observed in tumors growing in the kidney but not in the other anatomic sites. At present we cannot comment on the relevance of these reported changes to naturally occurring metastasis.

Variant selection and blood-borne "clonogenic" tumour cells in metastasis of FSA cell clones

Our studies of the metastatic behaviour of FSA1231 and FSA1233 show that the role of selection processes of metastatic variants at the secondary sites is minimal in the spontaneous metastasis of these systems. In contrast, tumour-cell release efficiency (number of blood-borne clonogenic tumour-cells) correlates well with the difference in spontaneous lung metastasis efficiencies of the cell clones FSA1231 and FSA1233. Also, the different tumour-cell release efficiencies could explain the discrepancy between artificial and spontaneous metastasis of these cell clones.

Metastatic clones selected from an RSV-induced mouse sarcoma share a common marker chromosome

Previous work has shown that the metastatic potential of RSV-transformed fibroblasts is correlated with the ability to form colonies in 0.6% ("hard") agar. Metastatic subclones were selected by this property from the non-metastasizing fibrosarcoma B77-313 line. A marker chromosome was found at high frequency (90% of cells) in all the subclones studied. This marker was detectable in only 0.5% of the parental B77-3T3 cells, demonstrating that metastatic clone precursors pre-existed, as a small minority, in the parental line. The genotypic marker appeared to be steadily associated with the metastatic phenotype since, after prolonged in vitro propagation, the subclones retained both the marker chromosome and the high metastatic potential. Although the marker chromosome was constantly present, chromosomal numerical and structural aberrations were also found in 20% of the long-term-propagated subclone cells, supporting the suggestion that metastatic properties are associated with cytogenetic instability.

Organ specific metastasis with special reference to avian systems

Many malignant tumors demonstrate a definite propensity for metastasis to specific organs despite the fact that tumor cells with the potential for metastasis may circulate randomly throughout the body. Current concepts of organ specific metastasis (OSM) center around the generation of tumor cell variants with enhanced capacity for metastasis to specific organs. At present three hypotheses, mechanical, seed and soil, and specific tumor cell adherence (STCA), stand out as possible explanations for OSM. These possible mechanisms of OSM are by no means mutually exclusive. Recent efforts to understand OSM have included the selection of organ-specific metastasizing variants from tumor cell lines and an examination of their surface and metastatic properties. OSM-selected cell lines from many different tumor systems have been used to examine the relative contributions of the three mechanisms. While examples of each mechanism have been reported, the relative contributions of each for different tumor systems may differ substantially. Therefore, generalizations about the behavior of tumors based on studies with just a few tumor lines and systems may not be valid. There is substantial evidence that cell surface molecules are important in the process of OSM and homing of lymphocytes to specific lymph nodes. Monoclonal antibodies have been produced against putative cell surface receptors and initial biochemical characterization has begun. There is much evidence that cell surface glycoconjugates can serve as specific recognition structures on normal cells and in addition, may play important roles in OSM. The role of these carbohydrates is discussed. The chick embryo as a model system is discussed as it offers several advantages for the study of metastasis in general and OSM in particular. A variety of human and murine tumors, including some freshly isolated, have been shown to grow and metastasize in these embryos. Furthermore, cell lines which have been selected for OSM in adults show similar patterns of metastasis in chick embryos indicating that this system may be an especially attractive one for the analysis of OSM.

The selective nature of metastasis

The issue of whether metastases result from the random survival of cells released from a primary tumor or from the selective growth of specialized tumor subpopulations endowed with metastatic properties is important to our understanding of the metastatic process and to the development of therapeutic modalities against metastatic disease. We have found that the tumor cells populating spontaneous metastases are more metastatic than the cells populating the parent neoplasm, clearly indicating that metastasis is selective and not random. The selective nature of metastasis is a consistent observation, however, only when tumor cells are obtained from spontaneous metastases from mice bearing heterogenous, poorly metastatic tumors. Tumor cells from spontaneous metastases from mice bearing tumors that have been selected for metastatic potential or that are homogeneous (cloned) do not differ significantly in metastatic potential from tumor cells populating the parent tumor. Thus, under some conditions the process of metastasis can appear random. Although tumor cells from different individual metastases may be homogeneous with regard to a metastatic phenotype, they may be heterogeneous with regard to their sensitivity to chemotherapeutic agents. Thus, although metastasis selects for metastatic variants, resulting in the population of metastatic foci with tumor cells endowed with metastatic properties, it does not appear to select for phenotypes irrelevant to the process of metastasis such as sensitivity to therapeutic agents.

Immunology of metastasis. Can the immune response cope with disseminated tumor?

The application of immunologic methods to the treatment of neoplasia has been a goal of research in tumor immunology. Unfortunately, no clearly defined success for such therapy has been achieved. However, the most recent advances in tumor biology have provided for a more valid conceptual framework upon which to plan further research in this area. The more general awareness of tumor progression and heterogeneity, particularly in the context of tumor metastasis, while imposing a sense of gloom regarding all therapeutic modalities, shifted immunologic thinking toward the development of nonspecific modalities. We herein propose that this 'shift' may be premature and that immunotherapy using cytolygic T cells could still be feasible. Our views are based on newer approaches for selecting immunogenic variants of malignant tumors and a better understanding of the relationship of the immune response to metastases.

Biological diversity in metastatic neoplasms: origins and implications

Whether neoplasms are unicellular or multicellular in their origin, the process of tumor evolution and progression can rapidly generate biological diversity. Metastases result from the survival and proliferation of specialized subpopulations of cells within the parent tumor. Metastases may have a clonal origin and different metastases may develop from different progenitor cells. However, as with the primary tumor, the origin of metastases is unimportant since the process of tumor evolution and progression can generate biological diversity within and among different metastatic foci.