|
|
|
 |
||
|
Introduction to the Clinical Molecular Profiling Core
Figure 1. High-resolution array-based comparative
genomic hybridization (array-CGH) of a breast cancer. Chromosome 8 is illustrated
with gains on the right and losses on the left. Note the complex pattern with
precisely defined boundaries of copy number change. This type of data, which can
be generated rapidly from small, clinically practical samples, might be useful
for tumor classification and gene discovery.
ncologists caring for patients are acutely aware that each patient is an individual and that each tumor has its characteristic biological properties. These differences may not be important if there are only a few options for therapy, but now, in the age of targeted therapies with numerous new agents appearing in the clinic and many more on the horizon, matching the right treatment to the right patient appears increasingly important. Identifying the biological differences between tumors and defining the mechanisms by which theses features affect clinical outcomes are key components of contemporary clinical research. Biological data linked to clinical trials can enhance the value of those studies as CCR investigators work toward the rational implementation of targeted therapy. The Clinical Molecular Profiling Core has been created to facilitate biological data collection on every tumor entered into a CCR trial. Fortunately, technologies for the molecular profiling of cancer have advanced substantially. It is now possible to obtain a tremendous amount of information from clinical specimens. The goal of the Clinical Molecular Profiling Core will be to provide every CCR clinical investigator with access to a suite of technologies for the characterization of biospecimens collected in the course of clinical trials. Because genetic and epigenetic changes are fundamental to cancer progression, the Core will focus primarily on genomic technologies: gene expression profiling, comparative genomic hybridization, high-density single nucleotide polymorphism (SNP) analysis, DNA sequencing, and related assays. Access to the Core will spare clinical investigators the need to develop technical expertise in these areas. Efforts are being made to offer a range of technologies to accommodate the realities of specimen collection in a variety of clinical situations. Specimen tracking, handling, and assays will follow standard procedures to maximize data reliability and maintain compliance with NCI specimen-collection guidelines. Core scientific staff will be available to consult with clinical investigators about assay selection, study design, and specimen requirements. Core staff will also support data analysis and interpretation. This is an exciting period in the history of cancer research, in large part because there is a sense that advances in cancer biology are leading to real progress in cancer therapy. We trust that the Clinical Molecular Profiling Core will create opportunities for CCR clinical investigators to bring these technologies to bear on their efforts to develop more effective therapies for their patients.
Targeting Cancer Cells by Exploiting Karyotypic Complexity and Chromosomal Instability
We have completed a detailed analysis of the chromosomal aberrations present in the drug-discovery panel of 60 human cancer cell lines (the NCI-60), used by the NCI Developmental Therapeutics Program (DTP) to screen compounds for anticancer activity (Roschke AV et al. Cancer Res 63: 8634–47, 2003). Measures of karyotypic complexity include the number of clonal chromosomal rearrangements present in a cell line (structural complexity), the number of chromosome deviations from the ploidy level (numerical complexity), and modal chromosome number. Measures of cell-to-cell chromosomal variability, which reflect the degree of ongoing instability, include numerical and structural heterogeneity. The NCI-60 cancer cell lines show wide variation in these parameters. Karyotypes of these cell lines have been made publicly available on two Web sites (http://www.ncbi.nlm.nih.gov/sky/skyweb.cgi and http://home.ncifcrf.gov/CCR/60SKY/new/demo1.asp). We then looked for relationships between markers of the chromosomal state and drug resistance or sensitivity. As a first snapshot, we used a 1,429-drug subset of the more than 100,000 compounds tested against the cell lines in a short-term cytotoxicity assay. This subset was selected because each agent had been tested at least four times on all or most of the NCI-60. It includes most of the drugs currently used clinically for cancer treatment, along with many candidates that have reached clinical trials. A correlation analysis was performed comparing sensitivity data (expressed as the negative logarithm of GI50 [growth inhibition of 50%]) and each of the karyotypic parameters. A positive correlation between sensitivity to a given compound and an increased level of a given karyotypic parameter means that cell lines with higher values for that specific parameter would be more sensitive to the growth inhibitory action of that agent. The positive correlations between drug sensitivity and karyotypic complexity and heterogeneity found in this analysis (122 statistically significant positive correlations, P < 0.05) allowed us to identify agents that are more active against karyotypically complex and chromosomally unstable cancer cells. Grouping of selected agents based on their functional classification or chemical structure yielded seven distinct groups of chemical compounds. Relationships between karyotypic parameters and sensitivity of cancer cells to identified classes of agents are diagrammed in Figure 1. Figure 1. Stratification of compound groups based on activity associated with a particular karyotypic parameter. NC, numerical complexity; NH, numerical heterogeneity; SC, structural complexity; SH, structural heterogeneity. To explore the possibility that an agent targeted a particular cell lineage that just happened to be more karyotypically complex or that other cellular “states,” such as mismatch repair status or p53 gene status, might be the critical factors acting as determinants of sensitivity or resistance to these compounds, we reanalyzed the data for selected compounds from each group. We did this sequentially, leaving out one and then another of each of the nine lineages in the panel, or leaving out the six mismatch repair-defective cell lines or the 18 p53 wild-type cell lines present in the panel. The essential features of the correlations that we described and the groups of compounds that we identified were not changed by these additional tests. In collaboration with David G. Covell, PhD, Anders Wallqvist, PhD, and Ruili Huang, PhD (Screening Technologies Branch, NCI), we performed a much larger-scale correlation analysis of karyotypic parameters using data obtained from approximately 30,000 chemical compounds tested on the NCI-60 cell lines, and identified additional classes of chemical agents associated with karyotypic parameters (Wallqvist A et al. Mol Cancer Ther 4: 1559–68, 2005). As an aid to this analysis, we employed computational tools based on methods of self-organizing maps (SOMs) used by the Covell lab to cluster the NCI’s database of GI50 measurements of these 30,000 compounds across the panel of NCI-60 cancer cell lines (http://spheroid.ncifcrf.gov). When we made projections on these maps of the compounds that have been identified as positively and significantly correlated with karyotypic parameters, they mainly hit a relatively unexplored region in the SOM, where standard anticancer drugs are not, for the most part, present, and where mechanisms of action of chemical compounds are among the least elucidated. These findings suggest that these “lead” compounds identified as active against karyotypically complex and/or chromosomally unstable cancer cells may, indeed, represent new classes and mechanisms of action for potential anticancer agents. The karyotypic parameters associated with the activities of these compounds may well be markers for underlying genes or pathways that are the true targets of these agents. It is equally important, however, to recognize that certain agents may be active against the “state” of complexity or instability itself rather than against any specific gene product or pathway. It is plausible that the assessment of the chromosomal state of a cancer cell population could serve as a future guide for the selection of drugs active against aggressive and intractable cancers.
FGF8 Takes Center Stage During Kidney Development
Vertebrates are segmented, as demonstrated by somite formation (Figure 1, part A)—blocks of mesoderm lining the anterior-posterior embryonic axis and giving rise to muscle, dermis, and vertebrae. Manipulation of the chick embryo suggested that FGF8 regulates somitogenesis by keeping the presomitic mesoderm unsegmented until the appropriate cue induces the next somite in the embryo’s tail end. However, this idea could not be tested genetically for lack of an appropriate Cre mouse that would inactivate Fgf8 expression in forming somites and yet allow normal gastrulation. In our recent publication cited above, we characterize such a mouse line, called T-Cre, because Cre is controlled by regulatory elements of the T (or Brachyury) gene and hence is expressed prior to somitogenesis in the early mesoderm as it forms during gastrulation. T-Cre–mediated Fgf8 inactivation yielded embryos that gastrulated normally and generated presomitic mesoderm devoid of Fgf8 gene product (Figure 1, part A). These embryos gave us a surprising result: FGF8 is not required for somitogenesis because the somites and their derivatives were normal (Figure 1, part A). Our current unpublished work addresses this conundrum by demonstrating that the role of FGF8 in this process is partially redundant with a subset of five other FGF genes coexpressed in this region. Although somitogenesis was unaffected in these mutants, neonates died because they lacked functional kidneys (Figure 1, part B). A central event during kidney development is a reciprocal induction between two lineages: the ureteric bud (UB) and metanephric mesenchyme (MM). As a result, the MM condenses and converts to an epithelium that undergoes a series of morphogenetic changes to form the structures of the nephron. In turn, the UB branches outward toward the periphery of the growing kidney where this mutual induction event repeats as the kidney grows. The end result is a functioning kidney consisting of a large number of nephrons connected by the UB-derived collecting ducts. We determined that Fgf8 was expressed in the condensing mesenchyme and that mutants suffered aberrant apoptosis in the MM of the kidney cortex, preventing new nephron formation (Figure 1, part C). Besides this role as a survival factor for this progenitor population, we also found that FGF8 regulates the expression of specific genes crucial for normal kidney development. Microarray analysis of microdissected kidneys at 12.5 days gestation, when mutant and control kidneys cannot be distinguished grossly, revealed a number of genes misregulated in the mutant tissue. Follow-up work led us to focus on two of these genes that proved to be pivotal to understanding the Fgf8 kidney phenotype. One of these genes encodes the secreted signaling molecule WNT4 and the other the transcription factor LIM1. MM-specific loss of either gene causes an arrest in kidney development that resembles the T-Cre–mediated inactivation of Fgf8. By determining the expression pattern of each gene in mouse mutants lacking one of the other two genes, the epistatic relationship was determined. This information, along with in vitro explant experiments indicated that both FGF8 and WNT sources are required in parallel for normal development, resulting in the model shown in Figure 1, part D. Figure 1. A) Somites (arrows) of mutant embryos at 9.5 days of gestation display no Fgf8 expression. B) Kidneys of mutant neonates are hypoplastic and nonfunctional. C) A ring of aberrant cell death (red) occurs in mutant kidneys at 14.5 days of gestation, where nephrogenesis would normally be taking place. D) Data from mutant analysis and in vitro explant experiments were used to generate a model wherein FGF8 induces Wnt4 gene expression and then both FGF8 and WNT4 are required for Lim1 gene expression and nephrogenesis. ad, adrenal gland; bl, bladder; ki, kidney; ov, ovary; ur, ureter; ut, uterus. It is intriguing that the FGF/WNT nexus we uncovered occurs in such processes as brain and limb development as well as during oncogenesis. Therefore, the task before us is to determine how the molecular interactions of these signaling pathways regulate normal development and how they cause disease when they go awry.
Regulation of MHC Class I Expression by a T Lymphocyte–Specific Enhanceosome
MHC class I molecules bind peptide fragments derived from degraded intracellular proteins and present them on the cell surface. Complementary receptors on T lymphocytes recognize foreign, non-self, peptide cargo (such as from degraded viral or transformed proteins) bound to MHC class I molecules. Whereas the immune system is tolerant to “self” peptides, recognition of “non-self” results in a peptide-specific immune response to eliminate cells bearing the foreign peptide signature. Thus, expression of MHC class I molecules is essential for immune surveillance and destruction of virally infected and transformed cells. Regulation of MHC class I expression is mediated through a series of promoter proximal elements. Over the past several years, work from a number of laboratories has demonstrated that constitutively high levels of MHC class I expression in B lymphocytes is attributable to assembly of a B lymphocyte–specific enhanceosome complex in the proximal promoter region. The B-lymphocyte enhanceosome represents the coordinated recruitment of multiple, widely expressed DNA binding transcription factors anchored by a coactivator, the class II transactivator (CIITA), which is constitutively expressed only in B lymphocytes and other antigen-presenting cells. In most other tissues, CIITA can be induced by gamma-interferon (IFNγ) to dynamically upregulate MHC class I expression. Although T lymphocytes express high levels of MHC class I, they do not normally express CIITA. Prior to the described studies, the molecular basis for high constitutive MHC class I expression in T lymphocytes was not known. However, it was known that T lymphocyte–specific genes, such as those encoding the T-cell receptor, were regulated by a T lymphocyte–specific enhanceosome (TCE). The TCE consists of the widely expressed coactivator Aly, T cell–specific transcription factors RUNX1 and LEF1, and the RUNX1 cofactor CBFβ. Aly, like CIITA and other coactivators, does not bind proximal promoter DNA sequences directly, but is recruited by DNA-bound LEF-1 and RUNX1/CBFβ transcription factors to form the TCE, which potently increases expression of cognate genes. In a recent study, we examined the ability of T-lymphocyte enhanceosome components, RUNX1/CBFβ, LEF1, and Aly to mediate T lymphocyte–specific MHC class I expression. The first evidence of a role for the TCE in regulating MHC class I gene expression came from our finding that disrupting RUNX1 in the Jurkat T-lymphocyte cell line resulted in a 50% reduction in endogenous MHC class I expression. In complementary experiments, we reconstituted the T-lymphocyte enhanceosome in HeLa epithelial cells and determined its effects on MHC class I expression. Partial reconstitution of the TCE with RUNX1/CBFβ and LEF1 enhanced the activity of an MHC class I promoter reporter construct 3 to 5 fold; neither RUNX1/CBFβ nor LEF1 alone had an effect. However, complete reconstitution of the TCE with RUNX1/CBFβ, LEF1, and Aly resulted in profound increases, both in exogenous MHC class I promoter activity and transcription of endogenous class I genes. Furthermore, chromatin immunoprecipitation from splenic lymphocytes demonstrated the association of the T-cell enhanceosome component RUNX1 with MHC class I proximal promoter sequences, providing evidence that the TCE directly regulates MHC class I gene expression in vivo. The above findings establish that the TCE—minimally composed of RUNX1/CBFβ, LEF1, and Aly—functions as part of the “intrinsic” pathway regulating MHC class I gene expression. This enabled us to examine the mechanism by which regulatory signals of the tissue-specific “intrinsic” pathways are integrated with dynamic “extrinsic” signals (such as those induced by hormones and cytokines) on a common promoter region. To this end, we examined the effect of the IFNγ-inducible CIITA “extrinsic” pathway on “intrinsic” pathway T lymphocyte–specific class I expression in HeLa epithelial cells. Interestingly, assembly of both “extrinsic” (CIITA) and “intrinsic” T-lymphocyte (RUNX1/CBFβ, LEF1, and Aly) enhanceosomes synergistically activated MHC class I transcription, suggesting that the “intrinsic” and “extrinsic” pathways are distinct and complementary, and target distinct promoter elements. These findings provide both a molecular basis for the constitutively high levels of MHC class I in T lymphocytes and their further synergistic induction by IFNγ.
WNT3A and Embryonic Development: Telling Left from Right
Powerful “organizer” signals emanating from a transient embryonic structure called the node play an important role in specifying the LR axis, thus defining a spatial, three-dimensional framework for embryonic development. Although the bone morphogenetic protein (BMP) family member, Nodal, is known to function in the node as the left determinant, little is known about the organizer signals that control its activation. Members of the WNT family of secreted signaling molecules are known to play important roles in development and, when misregulated, in diseases such as cancer. WNT3A is a particularly good candidate for an organizer molecule since it is expressed prior to LR axis formation and has powerful axis-inducing activity when overexpressed in frog and chick embryos. We hypothesized that WNT3A functions as an organizer during mouse embryogenesis. Analysis of embryos carrying targeted null alleles of Wnt3a demonstrates that it is indeed necessary for organ asymmetry. Heart looping is randomized, with approximately half of the mutants displaying normal rightward looping, and half displaying hearts that looped to the left (a condition known as situs inversus) or that remained in the midline (situs ambiguus) (Figure 1, parts A through C). Similar laterality defects were also observed in the lungs, liver, and gut, indicating that WNT3A plays an early global role in LR determination. To determine when, and where, Wnt3a is required for LR determination, we examined mutant embryos for the expression of several asymmetrically expressed genes known to regulate LR specification. In situ hybridization analyses indicated that asymmetric gene expression in the lateral plate mesoderm (a tissue that ultimately directs organ asymmetry) was absent (Figure 1, parts D and E). Furthermore, asymmetric gene expression in the mutant node was aberrant, even though general markers of the node and other axial tissues were expressed normally. In particular, Nodal was expressed in an unusually small, symmetrical, posterior domain in the mutant node. These results suggest that WNT3A specifically regulates a genetic program that controls Nodal expression in the node. Does WNT3A regulate Nodal gene expression in the node directly? WNTs activate target gene expression by stabilizing β-catenin, a multifunctional protein that possesses a transactivation domain. To determine which embryonic tissues respond directly to WNT signals, we examined the expression of a WNT/β-catenin–responsive lacZ reporter transgene in vivo. Examination of this reporter in wild-type and Wnt3a–/– mutant embryos illustrated that it is normally expressed in the node and adjacent mesoderm and that this expression is dependent on Wnt3a. These studies suggest that WNT3A signals to these tissues directly, and that it does so through β-catenin. Interestingly, recent reports indicate that Nodal expression in the node is controlled by the Notch signaling pathway (Raya A et al. Genes Dev 17: 1213–8, 2003; Krebs LT et al. Genes Dev 17: 1207–12, 2003), while the Notch ligand Delta-like1 (Dll1) is a target of the WNT/β-catenin pathway (Galceran J et al. Genes Dev 18: 2718–23, 2004; Hofmann M et al. Genes Dev 18: 2712–17, 2004). This suggests that WNT3A modulates Nodal expression indirectly, through the activation of Dll1. Indeed, we found that Dll1 expression is perturbed in the Wnt3a mutants, in a manner sufficient to explain the aberrant Nodal expression. Furthermore, we showed through genetic interaction studies that Dll1 and Wnt3a function in the same genetic pathway to regulate heart asymmetry. Since Dll1 is expressed only in the mesoderm immediately surrounding the node and not in the node itself, it appears that WNT3A directly activates Dll1 in the mesoderm, which then activates Notch signaling and Nodal expression in the adjacent node (Figure 1, part F). Figure 1. WNT signaling regulates left-right (LR) determination. (A through C) Scanning electron microscopy micrographs of ventral views of developing hearts on embryonic day 9.5 (E9.5). (A) The wild-type outflow tract (ot) loops to the right. Wnt3a–/– hearts display normal looping (B), inverted looping (situs inversus) (C), or loops that remain in the midline (situs ambiguus, not shown). Ventral-posterior views of E8.2 wild-type (D) and mutant (E) embryos processed for two-color whole-mount in situ hybridization showing Nodal (orange) and Lefty1 (purple) expression in the node (n), and overlapping Nodal (orange) and Lefty2 expression (purple) in the left lateral plate mesoderm (lpm). Note the absence of staining in the Wnt3a–/– lpm and the aberrant expression in the node. (F) WNT3A regulates LR determination and segmentation via the Delta/Notch pathway. The diagram depicts a perspective similar to that shown in (D) and (E). Wnt3a is expressed in the primitive streak and posterior node (red stippling) where it directly activates Dll1 expression in the mesoderm surrounding the node (blue arrows). In turn, DLL1 directly activates Nodal (red N) expression at the node/mesoderm boundary to set up the LR axis, and regulates segmentation (dashed arrows) in the presomitic mesoderm (psm). How LR asymmetric gene expression (green gradient) is established in the node is poorly understood, but involves node cilia. L, left; R, right; A, anterior; P, posterior; lv, left ventricle; s, somite; L1, Lefty1; L2, Lefty2. Wnt3a is also required for the formation of the repeated structures of the developing trunk known as segments or somites. These structures give rise to the skeletal muscles and vertebrae of the trunk. The number of somites, and therefore the number of vertebrae, can vary enormously among vertebrates—ranging from six in some adult frogs to several hundred in some shark species. Our demonstration that Wnt3a simultaneously controls LR axis determination and somite number through the Notch pathway reveals that Wnt3a links the body plan with fundamental morphogenetic events that are critical for determining the final shape of an organism. Given the high degree of conservation of WNT genes among metazoans (Kusserow A et al. Nature 433: 156–60, 2005), we suggest that WNTs such as WNT3A played an important role in the evolution of complex body plans.
Scientific Advisory CommitteeIf you have scientific news of interest to the CCR research community, please contact one of the scientific advisors (below) responsible for your areas of research.
CCR Frontiers in Science—StaffCenter for Cancer Research Robert H. Wiltrout, PhD, Director Deputy Directors Douglas R. Lowy, MD Editorial Staff Sue Fox, BA/BSW, Senior Editor * Palladian Partners, Inc. FOR INTERNAL USE ONLY |
||
 |
ost cancers have an abnormal chromosomal content, called aneuploidy, characterized by changes in chromosomal structure and number. Chromosomal aberrations tend to be more numerous in malignant tumors than in benign ones, and karyotypic complexity is associated with poorer prognoses and aggressive clinical and distinctive histopathologic features. Therefore, quantitative or qualitative changes in the karyotypic state of malignancy could represent determinants for anticancer therapies and might ultimately allow targeting of the most aggressive and incurable cancers.
he family of human and mouse fibroblast growth factors (FGFs) is large, numbering 22. Originally named for their effect on cultured cells, they regulate a wide variety of cellular and morphogenetic processes. Arguably, FGF8 is the busiest family member. It was isolated from mammary tumor cells and has since been implicated in the oncogenesis of sex hormone–related cancers of the breast and prostate. Most of our knowledge of how FGF8 controls morphogenesis comes from studying its various roles during mouse development. FGF8 is required for normal gastrulation, the embryonic stage when the three germ layers—mesoderm, ectoderm, and endoderm—are formed. Thanks to techniques of conditional mutagenesis, we also know that FGF8 is required for left/right asymmetry and regulates the development of different brain regions, the eyes, heart, limbs, and face. Equally diverse are the cellular processes that FGF8 regulates; depending on the embryonic stage, it controls cell growth, apoptosis, migration, and gene expression.
ccurate tissue-specific gene expression is, in general, the result of diverse transcriptional reflexes in response to both pre-set “intrinsic” and dynamic “extrinsic” stimuli. Many genes, including members of the major histocompatibility complex (MHC) class I family, are regulated by complex and overlapping developmental, tissue-specific, and inducible stimuli. Although MHC class I genes are ubiquitously expressed, distinct pre-set “intrinsic” stimuli lead to dramatically different tissue-specific levels of expression. For example, the lowest MHC class I levels are observed in neural tissues and germ-line cells, whereas the highest levels occur in lymphoid tissues, especially among T and B lymphocytes. Further, numerous “extrinsic” stimuli, in particular hormones and cytokines, are capable of dynamically modulating pre-set tissue-specific MHC class I expression patterns. The molecular mechanisms driving tissue-specific and dynamically modulated modes of MHC class I transcription are only partially understood and our ongoing efforts to understand these processes help us to understand how complexly regulated genes achieve proper levels of expression. 
