Intraneoplastic diversity

A quantitative view on multivalent nanomedicine targeting

Although the concept of selective delivery has been postulated over 100 years ago, no targeted nanomedicine has been clinically approved so far. Nanoparticles modified with targeting ligands to promote the selective delivery of therapeutics towards a specific cell population have been extensively reported. However, the rational design of selective particles is still challenging. One of the main reasons for this is the lack of quantitative theoretical and experimental understanding of the interactions involved in cell targeting. In this review, we discuss new theoretical models and experimental methods that provide a quantitative view of targeting. We show the new advancements in multivalency theory enabling the rational design of super-selective nanoparticles. Furthermore, we present the innovative approaches to obtain key targeting parameters at the single-cell and single molecule level and their role in the design of targeting nanoparticles. We believe that the combination of new theoretical multivalent design and experimental methods to quantify receptors and ligands aids in the rational design and clinical translation of targeted nanomedicines.

Angiogenic heterogeneity in head and neck squamous cell carcinoma: biological and therapeutic implications

The literature contains numerous references describing heterogeneity for tumor phenotypes including cell proliferation, invasiveness, metastatic potential, and response to therapies. However, data regarding angiogenic heterogeneity are limited. In this study, we investigated the degree of intertumoral angiogenic heterogeneity present in head and neck squamous cell carcinomas (HNSCC). In addition, we investigated the biological relevance that this heterogeneity may have in the context of cytokine directed antiangiogenic therapy. Keratinocytes were harvested from HNSCC specimens using laser capture microdissection (LCM). Gene expression profiling of the RNA extracted from these specimens demonstrated variability in the expression of angiogenesis-related genes. Hierarchical clustering and principal component analyses (PCA) demonstrated the presence of unique patient clusters, suggesting that there may be two potentially distinct pathways by which HNSCC induce angiogenesis. Immunohistochemistry for VEGF, IL-8/CXCL8, HGF, and FGF-2, cytokines that play functional roles in HNSCC angiogenesis was performed on the original patient samples as well as a larger panel of normal, dysplastic and HNSCC specimens to validate the heterogeneous expression observed in the gene expression profiling studies. Finally, the therapeutic response of HNSCC tumor xenografts to anti-VEGF therapy was found to be dependent on the amount of VEGF produced by the tumor cells. These findings support the hypothesis of intertumoral angiogenic heterogeneity. They imply that there are differences with regard to the specific molecular mechanisms by which individual tumors within the same histological type induce angiogenesis. Moreover, they demonstrate the need for a more in-depth understanding of the variability of the angiogenic phenotype within a given type of neoplasm when designing cytokine targeted antiangiogenic therapies. Finally, they suggest that studies in conjunction with the ongoing clinical trials that explore the correlation between target expression and clinical outcome are warranted.

Dynamic imaging of emerging resistance during cancer therapy

One of the greatest challenges in developing therapeutic regimens is the inability to rapidly and objectively assess tumor response due to treatment. Moreover, tumor response to therapeutic intervention in many cases is transient, and progressive alterations within the tumor may mask the effectiveness of an initially successful therapy. The ability to detect these changes as they occur would allow timely initiation of alternative approaches, maximizing therapeutic outcome. We investigated the ability of diffusion magnetic resonance imaging (MRI) to provide a sensitive measure of tumor response throughout the course of treatment, possibly identifying changes in sensitivity to the therapy. Orthotopic 9L gliomas were subjected to two separate therapeutic regimens, with one group receiving a single 5-day cycle (1omega) of low-dose 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) and a second group receiving two cycles at the same dose, bisected with 2 days of rest (2omega). Apparent diffusion coefficient maps were acquired before and throughout treatment to observe changes in water mobility, and these observations were correlated to standard measures of therapeutic response and outcome. Our results showed that diffusion MRI was indeed able to detect the emergence of a drug-resistant tumor subpopulation subsequent to an initially successful cycle of BCNU therapy, leading to minimal gains from a second cycle. These diffusion MRI findings were highly correlated with tumor growth delay, animal survival, and ex vivo growth inhibition assays showing emerging resistance in excised tumors. Overall, this study highlights the ability of diffusion MRI to provide sensitive dynamic assessment of therapy-induced response, allowing early opportunities for optimization of therapeutic protocols.

Phase II study of 4-ipomeanol, a naturally occurring alkylating furan, in patients with advanced hepatocellular carcinoma

BACKGROUND/PURPOSE

4-Ipomeanol (IPO; NSC 394438), a naturally occurring furan isolated from common sweet potatoes (Ipomoea batatas) infected with the fungus Fusarium solani was the first agent to be developed by the National Cancer Institute based on a biochemical-biological rationale as an anticancer agent targeted specifically against lung cancer. Prior to clinical development, IPO was shown to induce pulmonary toxicity in the lungs of several mammalian species because the agent is metabolized to a highly reactive furan epoxide by specific cytochrome P450 monooxygenases found in pulmonary Clara cells and type II pneumocytes, which share biochemical features with bronchogenic carcinoma. However, instead of inducing the anticipated lung toxicity in patients with lung cancer in disease-directed phase I studies, hepatotoxicity was the principal toxic effect of IPO in humans. Based on the presumption that IPO may be preferentially activated by cytochrome P450 monooxygenases in liver cells and biochemically-related hepatic malignancies, a phase II study was conducted to determine the activity and evaluate the toxicity of IPO in patients with advanced hepatocellular carcinoma.

PATIENTS AND METHODS

Nineteen patients with advanced measurable hepatocellular carcinoma were enrolled on the phase II trial. All patients had an Eastern Cooperative Oncology Group performance status of at least two, no evidence of pulmonary dysfunction, and had either no prior treatment or minimal prior therapy. Patients were treated with IPO at a dose of either 1032 mg/m2, which was the maximum tolerated and recommended phase II dose previously derived for patients with normal hepatic function (15 patients) or 826 mg/m2 if they had serum bilirubin concentrations in the range of 2.0 to 3.0 mg/dL (four patients). Treatment was repeated every three weeks. Objective tumor response, the primary endpoint of the study, was assessed after every two courses of treatment, and both pulmonary function and lung density were rigorously monitored using successive pulmonary function testing and computerized tomography.

RESULTS

All nineteen patients were evaluable for both response and toxicity. No major objective responses were observed. One patient had a minor, brief reduction in lung metastases. Although marker lesions and overall disease remained stable for at least 12 and 24 months in three and two patients, respectively, the median time to progression was three months and the median survival was five months for all patients. The principal toxicity was reversible elevations in hepatic transaminases, which occasionally resulted in dose reduction. No clinically-significant pulmonary toxicity was noted.

CONCLUSION

IPO at a dose of either 826 or 1032 mg/m2 administered every three weeks did not demonstrate a relevant degree of clinical activity against advanced hepatocellular carcinoma. Further evaluations of TO is not recommended for this disease.

Heterogeneous vascular dependence of tumor cell populations

Cells within a tumor are highly heterogeneous with respect to a wide range of genotypic and phenotypic characteristics. The latter include such properties as growth, survival, invasion, and metastasis. We asked whether the degree to which individual tumor cells rely on a tumor's vasculature might also be heterogeneous. By adapting an intravital Hoechst 33342 staining technique, we labeled and isolated tumor cells based on their relative proximity to perfused vessels. Because tumor regions distal to the vasculature are likely hypoxic, we examined cells deficient for hypoxia-inducible factor-1alpha (HIF-1alpha), a transcription factor that has been shown to mediate hypoxia-induced responses, including apoptosis. Despite reduced vascularization in HIF-1alpha-/- embryonic stem cell-derived tumors, their growth in vivo was found to be accelerated relative to HIF-1alpha+/+ tumor counterparts. We hypothesized that this paradoxical observation is because of decreased apoptotic rate, resulting in diminished vascular dependence of HIF-1alpha-/- cells. Analysis of heterogeneous tumors established from mixtures of HIF-1alpha+/+ with HIF-1alpha-/- cells revealed that the proportion of cells expressing wild-type HIF-1alpha was increased in perivascular areas and decreased in distal tumor regions. Thus, cells expressing HIF-1alpha were found to be highly dependent on proximity to blood vessels for their growth and survival in vivo, whereas cells that had lost HIF-1alpha expression were much less so. Heterogeneity in angiogenesis dependence was also observed among cell subpopulations isolated from human melanoma xenografts. This potential for selection of less vascular-dependent tumor cell variants throughout the course of disease progression may have important implications for the long-term efficacy of anti-angiogenic therapy.

Three cell lines showing androgen-dependent, -independent, and -suppressed phenotypes, established from a single tumor of androgen-dependent Shionogi carcinoma 115

We investigated the heterogeneity of cells in terms of androgen responsiveness within a single tumor mass of Shionogi carcinoma SC-115 showing androgen-dependent growth. After cloning of the tumor by the limiting dilution method in the presence of androgen, we isolated 40 clones at random. Twenty-two clones required androgen for growth (androgen-dependent phenotype), 16 did not (androgen-independent phenotype), and the remaining two clones showed growth inhibition when androgen was added (androgen-suppressed phenotype). In addition, 22 androgen-dependent clones showed heterogeneity in growth factor sensitivity in the absence of androgen. All clones were sensitive to both acidic and basic fibroblast growth factor (FGF), 7 of 22 clones were sensitive to epidermal growth factor (EGF) and transforming growth factor (TGF)-alpha, and 2 of 22 clones were sensitive to TGF-beta. This preexisting heterogeneity may be partly responsible for the growth of androgen-dependent tumor under hormone-deprived circumstances. Three typical clones, SC2G, SC1G, and SC4A, were selected from androgen-dependent, -independent, and -suppressed phenotypic groups, respectively. These clones, as well as original solid tumors, were found to produce heparin-binding growth factors of heterogeneous elution positions. The molecular nature of these growth factors is not yet known. Neither anti-basic FGF antibody nor anti-EGF antibody inhibited the cell growth when added in cell culture, suggesting the factors were distinct from basic-FGF and EGF.

Proliferative activity of primary breast cancer and of synchronous lymph node metastases evaluated by [3H]-thymidine labelling index

The [3H]-thymidine labelling index ([3H]TdR LI) has been used to evaluate and comparatively analyse the proliferative activity of different tumour lesions from the same patient. The analysis was performed on the primary tumour and its synchronous lymph node metastasis from 210 patients operated on for breast cancer. A direct relation was observed between the proliferative activity of the two different lesions (Spearman correlation coefficient = 0.46, P less than 0.0001), but there was considerable scatter amongst the data. The [3H]TdR LI of primary and of metastatic lesions belonged to the same proliferation classes in only 47% of the cases. Higher or lower [3H]TdR LI values, categorized on the basis of the tertiles of the frequency distribution, occurred in the node metastasis than in the primary tumour in an almost similar percentage of the remaining cases. Menopause, receptor status and pathological features did not affect interlesion kinetic patterns. The prognostic role of the proliferative activity of the two different lesions was investigated on 107 patients with stage II tumours homogeneously treated with surgery and systemic adjuvant therapy. Relapse-free survival at 3 years was significantly affected by the proliferative activity of the primary tumour but not by that of the lymph node metastasis.

Analysis of clonal evolution in a tumor consisting of pSV2neo-transfected mouse fibrosarcoma clones

The process of clonal evolution was analyzed in a line of methylcholanthrene-induced mouse fibrosarcomas. The tumor cells were transfected with pSV2neo gene and 22 clones were randomly isolated. Genetically tagged clones were mixed and inoculated into syngeneic mice. Southern blot analysis revealed that one of the clones, no. 11, dominated both in tumors in situ and in lung metastatic nodules. No. 11 clone and other clones were similar in growth rates in vitro and in vivo, in spontaneous and experimental metastatic abilities, in immunogenicity, and in the capacity of intercellular communication in vitro. Although no. 11 clone overgrew other clones in vivo, this was not the case when clones were mixed and maintained in vitro. We conclude that clonal interactions in vivo may be responsible for the dominance of no. 11 clone in the tumor. It is likely that the preferential metastasis of no. 11 clone to the lung may be a simple reflection of the proliferative advantage of the dominant clone in the tumor in situ.

Different effects of cyclophosphamide in vivo and phosphamide mustard in vitro on two cell clones of chemically induced mammary carcinoma of the rat

Two cell clones of a methylnitrosourea-induced rat mammary carcinoma, a hyperdiploid (44 chromosomes, clone A) and a hypertetraploid clone (88 chromosomes, clone B) were cultured and transplanted subcutaneously into three groups of eight rats. Group 1 was treated with 62.6 mg cyclophosphamide/kg, group 2 with 41.8 mg/kg once weekly for 3 weeks. The volume of tumors derived from clone B cells was diminished by the administration of the agent, whereas clone A cell tumors did not respond. Incubation of cells of both clones with phosphamide mustard in vitro showed that cells of clone B are much more sensitive to the activated cyclophosphamide, especially after incubation in low concentrations of 40 microM and 20 microM, than those of clone A. It is concluded that the initial success of cyclophosphamide therapy on chemically induced tumors is due to the different sensitivities of the tumor cell populations.

Analysis of tumour cell composition in tumours composed of paired mixtures of mammary tumour cell lines

In order to quantitate the effects of tumour subpopulation interactions, we have devised a method to determine the subpopulation composition of tumours by using paired tumour cell lines able to grow in different selective media. Line 4T07 forms colonies in thioguanine but not in HAT and line 168 forms colonies in HAT but not in thioguanine. An independent technique of determining tumour cell content was used to validate this method: line 168 and 4T07 cells are distinguishable by flow cytometry after staining with propidium iodide for DNA content. Mixtures of cell suspensions prepared from each unmixed tumour, as well as from tumours arising from mixtures of these lines, were analysed by both the colony formation assay and by the DNA content assay. The colony formation assay yielded values in good agreement with the DNA content assay, but was considerably more sensitive in that it was able to quantitate minority subpopulations that constituted less than 10% of the tumour. Both methods revealed that in tumours arising from mixtures, the tumour cells were almost entirely line 4T07, even when the inoculum had contained a high proportion of 168 cells. Since line 168 cells are very tumorigenic per se, these results suggest that line 4T07 cells are capable of interfering with 168 proliferation in mixed tumours, either directly or through a host-mediated mechanism.

Cell cycle responses of heterogeneous human colon adenocarcinoma subpopulations to X-irradiation

The cell cycle responses of two exponentially growing subpopulations of cells (clones A and D), originally obtained from a human colon adenocarcinoma to X-irradiation, were studied using centrifugal elutriation. Cell suspensions were separated by changing counter-current flow rate while keeping the rotor speed constant (1600 rpm) and the composition of eluted fractions was determined using flow cytometry. The X-ray sensitivity of unseparated clone D cells was somewhat greater than that of clone A cells (e.g. 10% greater at the 10% level of survival). This difference appeared to be due to a greater value of the alpha parameter (one-hit cell killing), using the linear-quadratic equation in which the relative survival S/S0 = exp - (alpha D + beta D2) with dose (D) in Gy. This finding was confirmed in the cell cycle studies where the alpha parameter was always greater for the clone D cells than for the clone A cells. The beta parameter was essentially the same for both cell lines through the cell cycle.

Characterization of radiation sensitivity of human squamous carcinoma A431 cells

Studies have been performed to investigate the radiosensitivity of human squamous carcinoma cells. A431 cells were grown in vitro as exponential and fed-plateau monolayer cultures or as multicellular spheroids. Radiobiological studies of various cultures showed that fed-plateau phase cells were more sensitive (D0 = 1.3 Gy) than exponentially growing cells (D0 = 1.5 Gy). After a single dose of 12 Gy or two doses of 6 Gy irradiation, A431 cultures exhibited a large capacity for potentially lethal damage (PLD) repair (PLD repair factor = 17), but a relatively small sublethal damage (SLD) repair. In order to measure the radiation sensitivity of proliferating (P) and quiescent (Q) cells, enriched populations of P- and Q-cells were isolated from A431 spheroids. Flow cytometric analysis with acridine orange (AO) staining demonstrated that there was a shift of the RNA histograms in fed-plateau and spheroid cultures towards lower values, suggesting the presence of a subpopulation of Q-cells. Centrifugal elutriation was used to isolate the Q-cells from dissociated spheroid cells. Coulter cell volume distributions and flow cytometric analysis showed that Q-cells had a small cell volume (approximately 1380 microns3), low RNA content and a G1-like DNA content. Continuous labelling experiments with tritiated thymidine confirmed the non-proliferating nature of the Q-cells. Irradiation of the Q-cells after isolation from spheroids with between 0 to 10 Gy showed that they were more radiosensitive (decreased D0) than the P-cells isolated from these spheroids. The latter were, however, similar in radiosensitivity to exponential G1 cells.

Immunohistochemical localization of adenosine deaminase complexing protein in intestinal mucosa and in colorectal adenocarcinoma as a marker for tumour cell heterogeneity

Adenosine deaminase complexing protein (ADCP), a dimeric glycoprotein, has been reported to be decreased or deficient in transformed or cancer-derived cell lines, indicating its potential significance as an indicator of malignant transformation. A similar deficiency was reported in total homogenates of tumours of colon, kidney, lung and liver. In previous biochemical studies we failed to confirm the consistent reduction in ADCP concentration in cancer tissues. A possible explanation for our findings was thought to be intercellular heterogeneity in ADCP expression in individual tumour cells. To study ADCP expression in individual cells, we developed an immunohistochemical method which was applied to tissue sections. Paraformaldehyde--lysine--periodate (PLP) solution was found to be a suitable fixative. Fixed tissue samples were paraffin-embedded, sectioned and stained for ADCP, using an indirect peroxidase-labelled antibody procedure. The protein was localized in normal colonic mucosa, mainly in the brush border region of the luminal epithelium and in cytoplasmic granules. Intense ADCP immunoreactivity was found also in the basal part of some cells. In cancer cells, three staining patterns were observed: membranous, diffuse cytoplasmic and granular cytoplasmic. The adenocarcinomas exhibited significant intratumour and intertumour heterogeneity in their staining types. Further studies on ADCP expression in colorectal cancer in relation to clinical and histopathological characteristics are warranted in order to fully evaluate the potential significance of ADCP as a cancer associated antigen.

Progression to steroid autonomy in S115 mouse mammary tumor cells: role of DNA methylation

Although monoclonal in origin, mammary tumors acquire a marked heterogeneity of cell phenotypes, including a mixture of steroid hormone-sensitive cells and insensitive cells. We describe here long-term studies on the effects of androgen withdrawal on cloned androgen-responsive S115 mouse mammary tumor cells as a model system to investigate mechanisms by which tumor cells lose their steroid sensitivity. In the prolonged absence of androgen, the cells lost hormone-sensitive parameters reproducibly, including loss of proliferative response, saturation density response, cell morphology response, and mouse mammary tumor virus long terminal repeat (MMTV-LTR)-related RNA. These experiments have demonstrated that when deprived of hormone in the long term, a clone of responsive cells gives rise reproducibly to a population of unresponsive cells in an ordered series of phenotypic changes. At the time when the cells lost all androgen response in terms of cell biology and MMTV-LTR-RNA, increased methylation of MMTV-LTR sequences in the DNA was detected. Thereafter recovery of androgen sensitivity has not been achieved in any of these parameters. The possible role of de novo DNA methylation in the progression to androgen autonomy of S115 cells is discussed.

Early origin and pervasiveness of cellular heterogeneity in some malignant transformations

Nontransformed BALB/3T3 cells were passaged weekly in monolayer culture on plastic dishes and aliquots were regularly assayed for colony production when suspended in agar. During several months of passaging, a single large colony arose once in the agar assay, and its constituent cells were isolated to form a subline of transformed cells. These cells had the fusiform and rounded morphologies characteristic of transformed cells and had a colony-forming efficiency in agar (CFEag) of approximately 10%. Five of the agar colonies were isolated at random and the cell populations of these primary subclones were further analyzed. Four of the five subclonal populations differed from each other in appearance slightly, but consistently, while the fifth had a markedly different colonial morphology. They also differed from one another in CFEag as well as in average diameter of the agar colonies. These general differences among the five subclones remained, although the CFEag and colony sizes changed recognizably in repeated weekly transfers. All of the subclones produced sarcomas in nude mice but did so at different rates. A secondary generation of five subclones was derived from each of three of the original subclonal populations. All the secondary subclones had the same morphology as the primary subclones from which they were derived. The averages of the CFEag and colony sizes of the secondary subclones from each of the subclones differed as a group from the other groups, but they also differed, to a lesser extent, among themselves. Despite the differences noted among the primary subclones in morphology, growth in agar, and tumor production in mice, they were virtually identical in their rapid growth rate on a plastic surface and all shared a high rate of glucose consumption. A second transformed clone arose among the continuously passaged nontransformed cells 5 months after the first one had appeared. Subclones of this clone did not differ recognizably from one another in morphology or in CFEag. The results indicate that recognizable heterogeneity can arise in some tumors during the earliest stages of their development and involve a high proportion of their constituent cells, while it may not become evident in other tumors until much later.

Solute concentration effects on the expression of cellular heterogeneity of anchorage-independent growth among spontaneously transformed BALB/c3T3 cells

Clones were derived in culture from a tumor initiated by spontaneously transformed 3T3 cells and tested for their colony-forming efficiency in agar (CFEag). Incubation of petri dish cultures was done in subsaturation humidity to minimize mold contamination. There was great variation in CFEag between clones but also, under certain conditions, within clones. The most prominent condition that generated phenotypic diversity in CFEag was partial evaporation of the medium, which may occur during the protracted development of a mass population from a single cell. Evaporation was disproportionately great in 35-mm dishes and peripheral wells of multiwell plates. If the supraphysiological solute concentration resulting from evaporation was greater than 133% of normal, there was progressive suppression of cell growth in the succeeding transfer in agar or on plastic, even if isotonic medium was substituted 1 d before transfer. The effect of supraphysiological concentrations of all the solutes of the medium could be reproduced by simply increasing the NaCl concentration. Damaged cells were restored to their full growth potential after 3 d in isotonic medium. When nontransformed cells were chronically exposed to increased salt, irreversible increases in 2-deoxyglucose uptake were produced. With continued exposure of these cells to high salt, they became morphologically transformed, produced colonies in agar with high efficiency, and formed sarcomas when inoculated into nude mice.

Distribution of anionic surface sites on human melanocytes and human melanoma cells in culture

With cationized ferritin (CF) as an ultrastructural marker for anionic cell surface sites, cultured guinea pig melanocytes display a uniquely homogeneous labelling pattern and a striking absence of redistribution of marker material. In the present study, we applied the same technique to normal human melanocytes and melanoma cells. Unfixed primary human mixed epidermal cell cultures displayed CF labelling patterns identical to those in guinea pig epidermal cells: on keratinocytes, CF was found in distinct aggregations which, upon prolonged incubation, clustered and were shed. Melanocytes, in contrast, bound CF to the cell surface as a uniform diffuse monolayer. There were no signs of cluster formation or shedding. Melanoma cell cultures were derived from 10 primary (2 lentigo maligna, 1 lentigo maligna melanoma, 4 superficial spreading melanomas, 2 nodular portions of superficial spreading melanomas, 1 nodular melanoma) and from 7 metastatic melanomas (4 cutaneous, 3 lymph node metastases). The CF labelling patterns encountered were heterogeneous. Three out of 10 primary tumors and 5 out of 7 metastases showed alterations of the normal melanocyte labelling pattern: regions of typical CF distribution were irregularly interrupted by stretches of membrane free of marker. In some areas, CF occurred in small globular aggregates. There was considerable heterogeneity of CF labelling patterns in different clones of a given culture. Altered CF binding patterns in melanoma cells appear to be associated with high metastasizing protential of the cell clones and may thus represent an unfavourable prognostic sign.

Mitogenic hormones and tumor promoters greatly increase the incidence of colony-forming cells bearing amplified dihydrofolate reductase genes

Previous work has shown that the presence of a phorbol ester tumor promoter, phorbol 12-myristate 13-acetate (PMA), during a single-step selection for methotrexate (MTX)-resistant mouse 3T6 cells results in an up to 100-fold increase in the incidence of MTX-resistant, colony-forming cells. MTX resistance of most of these cells is due to amplification of the gene for dihydrofolate reductase (DHFR), the target enzyme for MTX. We show here that other active, noncytotoxic phorbol ester tumor promoters, such as phorbol 12, 13-didecanoate and 20-phorbol 12,13-butyrate, at their optimal concentrations (approximately equal to 0.1 microM) are approximately equal to PMA in increasing the incidence of MTX-resistant 3T6 colonies. Mezerein, a potent second-stage tumor promoter, but a weak complete promoter, increases the incidence of MTX resistance up to 350-fold, the strongest effect for any of the agents so far tested. PMA analogs that are inactive as tumor promoters, such as phorbol or phorbol 12,13,20-triacetate, have no effect on the incidence of MTX-resistant 3T6 colonies. Anthralin, a nonphorbol tumor promoter, is approximately equal to 40% as active as PMA in the MTX selection assay. Remarkably, the hormones insulin, arginine vasopressin, and epidermal growth factor, all of which are mitogenic for 3T6 cells, also exert a strong PMA-like effect on the incidence of MTX-resistant 3T6 colonies under conditions of MTX selection. The effect of insulin at its optimal concentration (approximately equal to 1 microgram/ml) is approximately equal to 70% that of PMA. Although the effect of PMA on the incidence of MTX-resistant 3T6 colonies does not significantly depend on the initial density of seeded cells or volume of the medium added, the analogous effect of insulin is strongly influenced by these parameters. Mevalonic acid, arachidonic acid, thymidine, caffeine, and nicotine, all of which are known to influence patterns of DNA synthesis in mammalian cells, were tested at their highest noncytotoxic concentrations and failed to produce any significant effect on the incidence of MTX-resistant 3T6 colonies. We discuss possible mechanisms of hormone- and tumor promoter-facilitated gene amplification in mammalian cells, relationship of mitogenic hormones to tumor promoters, and also implications of our findings for the problem of drug resistance in cancer chemotherapy.