Between epigenomic extremes

This post continues yesterday's discussion of epigenomic cancer.

There seem to be only two tumors in humans that can be conceptualized as pure epigenomic tumors: malignant rhabdoid tumors and germ cell tumors. These two tumors occupy opposite extremes of epigenomic alteration. In rhabdoid tumors, something happens that wrecks epigenomic control (bi-allelic INI1 loss), and this results in a tumor that has no specifiable developmental lineage. In germ cell tumors, the epigenome "reset" button is pressed, and the resulting germ cell tumors either have a globally "erased" epigenome, as in the case of true germinomas (e.g., seminomas in males), or they take any of the differentiation paths open to them (e.g., embryonal carcinomas, teratocarcinomas, choriocarcinomas, etc.).

Between the malignant rhabdoid tumors and the germ cell tumors are all the remaining tumors of humans (several thousand kinds). These tumors tend to have varying amounts of epigenomic and genomic alterations. We know this to be true because we see both types of changes in almost every malignant tumor. We see the evidence of genetic changes when we look at the many chromosomal anomalies found in malignant cells. We see the epigenomic changes when we see cytologic atypia characterized by light and dark areas of the nuclei (corresponding to alterations in chromosomal proteins), variations of chromatin distribution on the nuclear borders, abnormally shaped and sized nucleoli, etc.

Of course, there is an interplay between epigenome and genome in cancers. A good example is seen in acute promyelocytic leukemia.

In acute promyelocytic leukemia, a gene translocation produces the PML/RAR(alpha) fusion protein.[1] Normally, promyelocytes differentiate to become non-dividing myelocytes (neutrophils). Neutrophils are the major circulating nucelated cell and play a crucial role in inflammation and the body's defenses against infections. This fusion protein causes the promyelocyte to divide, producing more promyelocytes and fewer neutrophils. Eventually, the population of clonal promyelocytes arising from the neoplastic progenitor cell will attain a sufficiently large number to be recognized clinically as a promyelocytic leukemia.

Acute promyelocytic leukemia is one of the few cancers that can achieve clinical remission without treatment with cytotoxic agents. Remission is achieved with all-trans retinoic acid, which somehow causes the neoplastic promyelocytes to differentiate and become non-dividing mature myelocytes.[2]

The mechanism by which the neoplastic fusion protein, PML/RAR(alpha), induces a neoplastic phenotype and the mechanism whereby all-trans retinoic acid reverses the neoplastic phenotype seems to be mediated through the epigenome. It is hypothesized that PML/RAR(alpha) modifies histone deacetylase complexes resulting in the inappropriate transcriptional repression of genes would normally inhibit promyelocyte proliferation. All-trans retinoic acid is though to reverse this effect.[1] If this turns out to be the case, promyelocytic leukemia would serve as an example of a genetic mutation (fusion gene) that employs epigenomic alterations to sustain a neoplastic phenotype.

In later posts, we'll discuss the pure (or nearly pure) genetic cancers.

1. Nouzova M, Holtan N, Oshiro MM, Isett RB, Munoz-Rodriguez JL, List AF, et al. Epigenomic changes during leukemia cell differentiation: analysis of histone acetylation and cytosine methylation using CpG island microarrays. J Pharmacol Exp Ther 311:968-981, 2004.
   
2. Flynn PJ, Miller WJ, Weisdorf DJ, Arthur DC, Brunning R, and Branda RF. Retinoic acid treatment of acute promyelocytic leukemia: in vitro and in vivo observations. Blood 62:1211-1217, 1983.
   

-© 2010 Jules J. Berman

key words: cancer, neoplasia, epigenome, epigenetics, cytogenetics, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, epigenomics, acute promyelocytic leukaemia, differentiation

Another epigenomic tumor

In yesterday's blog, we discussed malignant rhabdoid tumors as the prototypical epigenomic cancer (a malignant tumor characterized by massive alterations of the epigenomic landscape with preservation of chromosomal diploidy and genetic (DNA sequence) stability).

Malignant rhabdoid tumors are not the only example of an epigenomic cancer. Germ cell cancers also belong in this group. In the case of germ cell cancers, cell totipotency (ability to differentiate towards any embryonic or extra-embryonic lineage) is achieved after chromosomes lose their epigenetic programming (modifications) through a normal cellular process unique to germ cells: erasure.

I have prepared a separate web page that discusses some of the biological mysteries surrounding germ cell tumors; mysteries that have puzzled cancer researchers and pathologists for decades.

http://www.julesberm.info/factoids/germcell.htm

Jump to tomorrow's post

-© 2010 Jules J. Berman

key words: cancer, neoplasia, epigenome, epigenetics, cytogenetics, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, epigenomics

The prototypical epigenomic cancer

This post continues the discussion of rhabdoid tumors found in yesterday's blog.

Rhabdoid tumors seem to be the prototypical example of an epigenomic cancer; a cancer caused by destabilization of the epigenome.

It is impossible to underestimate the importance of this assertion, if true. For decades, cancer research has been focused on the genetic changes in cancers. It was more or less assumed that every advanced cancer has numerous genetic (DNA sequence) defects that contribute to the phenotype (morphology and biological behavior) of cancer cells. To be confronted with a tumor fueled by epigenetic (i.e., non-sequence non-DNA) alterations, is profoundly shocking.

To be sure, this alleged epigenomic tumor is rare, with only a few dozen reported in the U.S. (compared with about a million new cases of "gene-based" cancers, or two million cases if you count the common skin cancers).

Still, in cancer, we tend to learn the most important lessons from the rarest of cancers.

Observations on the development of rhabdoid tumors in experimental systems, suggest that these tumors can arise incredibly rapidly in mice (11 days). This rapid development would suggest that rhabdoid tumors do not follow the long developmental process that occurs in gene-based cancers.

With an 11-day development time, it seems likely that rhabdoid tumors have no precancerous lesions (lesions that precede the development of gene-based cancers).

How can a cancer NOT have a precancer? A precancer is a sort of natural experiment conducted during carcinogenesis. Basically, the precancer provides a way for developing cancer cells to find an epigenomic pattern (i.e., differentiation phenotype) that can support the cancer that emerges. A single precancer can give rise to several epigenomic patterns available to its lineage (e.g., bronchial dysplasia can give rise to squamous carcinoma or small cell carcinoma or adenocarcinoma). Precancers sometimes regress, and this regression presumably happens when a viable "fit" into an invasive cancer cell type (i.e., epigenomic tumor pattern) fails to occur.

In the case of rhabdoid tumors, there is no "epigenomic pattern" into which the rhabdoid tumor "fits." These tumors can arise from neuroectoderm or nephric mesoderm, or mesenchyme. They don't need to fit into any differentiated lineage. They simply grow, aggressively, when their epigenome is scrambled into a viable cell. If there's no necessity to find a differentiated lieage for the rhabdoid tumor cells, then there is no need for a precancerous stage.

Because there is global alteration of normal gene transcription in rhabdoid tumors, you might expect that a chemotherapeutic intervention aimed at restoring a semblance of normal epigenomic control, might inhibit the growth of this rare tumor.

At present, one of the most promising drugs for epigenomic dysregulation are the
histone deacetylase inhibitors (HDAC inhibitors). These inhibitors block normal remodeling of chromatin structure, and thus change the way that genes are expressed.

HDAC inhibitors, and a few other agents, will probably be studied to see if they are effective agents against tumors whose epigenomes contribute to their malignant phenotypes.

Are there tumors, other than rhabdoid tumors, that are fueled by epigenomic dysregulation? This will be the topic of future posts.

Jump to tomorrow's post

-© 2010 Jules J. Berman

key words: cancer, neoplasia, epigenome, epigenetics, cytogenetics, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, pre-cancer, precancerous lesions, pre-malignant lesions, gene synonyms: Snf5 Ini1 Baf47 SmarcB1, epigenomics

Rhabdoid tumors: a reasonable hypothesis

This blog continues a discussion of malignant rhabdoid tumors, found in yesterday's post.

Most malignant rhabdoid tumors, regardless of their tissues of origin, have bi-allelic loss of the INI1 gene, also known as SNF5, SMARCB1, andBAF47. This gene is a core component of the SWI/SNF complex that plays a crucial role in chromatin remodeling.[1]

Bi-allelic loss of INI1 may cause cancer through a purely epigenomic route, whereby disruption of normal gene transcription, causing alterations in the expression of thousands of genes, and leading to a characteristic malignant phenotype (i.e., malignant rhabdoid cells) in cells of several different developmental lineages. Essenetially, bi-allelic loss of INI1 hijacks cells to create malignant phenotype that circumvents the usual process of cancer development observed in virtually every other type of cancer.

The resulting rhabdoid tumors are epigenomic cancers, marked by aberrant gene transcription without genomic abnormalities (e.g., aneuploidy, chromosome abnormalities, gene mutations, genetic instability).

Here are some of the experimental observations that support this hypothesis:

1. The only characteristic gene alteration in rhabdoid tumors is bi-allelic loss of INI1. Rhabdoid tumors lack gene amplifications, gene deletions and, aside
   from INI1 loss, are indistinguishable from normal cells on single-nucleotide polymorphism arrays.[1]
   
   
2. Oncogenic pathways often found in common cancers were found to be upregulated or amplified in rhabdoid tumors, indicating that the molecular phenotype of cancer can be achieved through an epigenomic mechanism. [1]
   
   
3. In a mouse tumor model, loss of INI1 results in cancers in 100% of INI1-deficient mice. The cancer arise, on average, in just 11 days.[2] This is much shorter than the time required to develop chemically-induced cancers in mice. Furthermore, very few chemical carcinogen protocols produce cancers in 100% of the treated animals.
   

The exceedingly rapid emergence of cancer in 100% of INI1-deficient mice would suggest that the mechanism of tumor development bypasses the protracted steps of carcinogenesis exhibited by tumors caused by the accumulation of specific genetic mutations (i.e., virtually every other kind of observed cancer).

In tomorrow's post, we'll discuss the implications of these observations on our general approach to cancer classification, cancer treatment, and on our understanding of precancers.

REFERENCES

1. McKenna ES, Sansam CG, Cho YJ, Greulich H, Evans JA, Thom CS, Moreau LA, Biegel JA, Pomeroy SL, Roberts CW. Loss of the epigenetic tumor suppressor SNF5 leads to cancer without genomic instability. Mol Cell Biol 28:6223-6333, 2008.
   
   
2. Roberts CW, Leroux MM, Fleming MD, Orkin SH.
   Highly penetrant, rapid tumorigenesis through conditional inversion of the
   Cancer Cell 2:415-425, 2002.
   

Jump to tomorrow's post

-© 2010 Jules J. Berman

key words: cancer, neoplasia, epigenome, epigenetics, cytogenetics, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, pre-cancer, precancerous lesions, pre-malignant lesions, gene synonyms: Snf5 Ini1 Baf47 SmarcB1, epigenomics

Rhabdoid tumors: the first clue

This post continues a discussion of rhabdoid tumors, begun in yesterday's blog.

One of the most important clues concerning the fundemantal nature of rhabdoid tumors is revealed when we compare its nuclear morphology with its genetic structure. When we make this comparison, we discover that much of what we thought we knew about cancer cells, in general, is simply incorrect.

Here is a photomicroscopic image of a rhabdoid tumor, available in the public domain, from wikipedia:


Look closely at all the nuclei in this very typical image of a rhabdoid tumor.
What can say about the nuclei?

1. They vary in shape and size from one cancer cell to another.
   
   
2. The nuclei are mottled, with areas of light and of dark, and the sizes and locations of these light and dark areas varies from nucleus to nucleus.
   
   
3. Some nuclei have large nucleoli; the number of nucleoli vary from nucleus to nucleus, ranging from 0 to 3; the nucleoli are not uniform in size or contour, or color.
   
   
4. The nuclei are outlined by a highly irregular membrane, with thick, dark borders in some areas, alternating with light borders elsewhere, with no constant pattern from nucleus to nucleus.
   
   
5. The nuclei are not consistently round, or oval or bean-shaped. Numerous sharp puckerings are seen on the nuclear boder; some nuclei seem to have tails and protrusions; others have focally straight edges.
   

Though you can't judge it from this image, which exclusively shows malignant cells, the nuclei of the rhabdoid tumor cells are larger than the nuclei of normal cells.

What do all of these nuclear observations indicate? The nuclear morphology of rhabdoid cells conforms to the classic cytologic features that characterize malignant cells.

Pathologists assume, when they see highly abnormal nuclei, that the observed abnormalities indicate genetic alterations in chromosome number (aneuploidy) and arrangements (cytogenetic abnormalities) resulting from genetic instability; all hallmarks of cancer.

But this is not the case with rhabdoid tumors.

Rhabdoid tumors are genetically stable, and have a normal karyotype (i.e., are diploid, not aneuploid).[1] The morphologic nuclear changes in malignant rhabdoid cells do not correlate with cytogenetic defects.

How is this remotely possible? Is everything we think we know about tumor cytology incorrect?

Yes and no. The nuclei of rhabdoid cells are highly abnormal, but the abnormalities are not the result of genetic alterations. The abnormalities are due to epigenetic alterations. Histologic stains do not bind to DNA. The routine nuclear stains used by pathologists will bind to chromosomal and nuclear proteins. Basically, routine nuclear stains provide pathologists with a view of the epigenome, not the genome. In most types of cancer, cells have profound changes in both the epigenome and the genome, so the kinds of nuclear abnormalities associated with malignant cells will almost always correlate with aneuploidy and genetic instability. But not so in the case of rhabdoid tumors.

When we look at the nuclei of a rhaboid tumor, using standard histologic stains, we are seeing the ravages of profound epigenomic alterations in a unique type of cancer.

In the next blog, we'll explain the biologic basis for rhabdoid tumors, and we will show that malignant rhabdoid tumor is the prototype for epigenomic (not genomic) carcinogenesis. In subsequent blogs, we will discuss the consequences of this finding for tumor classification and for cancer treatment, in general.

1. McKenna ES, Sansam CG, Cho YJ, Greulich H, Evans JA, Thom CS, Moreau LA, Biegel JA, Pomeroy SL, Roberts CW. Loss of the epigenetic tumor suppressor SNF5 leads to cancer without genomic instability. Mol Cell Biol 28:6223-6333, 2008.
   

Jump to tomorrow's post

-© 2010 Jules J. Berman

key words: cancer, neoplasia, epigenome, epigenetics, cytogenetics, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, pre-cancer, precancerous lesions, pre-malignant lesions, gene synonyms: Snf5 Ini1 Baf47 SmarcB1

Rhabdoid tumors.

This post continues a discussion of neoplasia begun yesterday.

The basic premise of modern tumor classification is that each type of tumor derives from a progenitor cell that belongs to a specific developmental lineage (e.g., mesoderm, endoderm, ectoderm, etc.). The resulting tumor usually looks and behaves somewhat like a normal differentiated cell type. Hence, we have tumor names like fibroscarcoma, adenocarcinoma, squamous cell carcinoma, etc. These tumors arise from the same tissues that produce normal cells (e.g., epidermis gives rise to normal squamous cells and can also be the site of squamous cell carcinoma; glandular tissues give rise to adenocarcinoma). We would not expect to see adenocarcinomas arising from muscle or pheochromocytomas arising from epidermis.

But there is one notable exception to this rule: the rhabdoid tumor.

Rhabdoid tumors are very rare tumors, that occur in children. Only a few dozen occur in the United States, each year. These aggressive tumors can arise from brain (neuroectoderm), or from kidney (mesoderm), thus breaking the one-tumor one-lineage rule of tumor development. This dual origin of the rhabdoid tumors was so unusual, that it was formerly believed that rhabdoid tumors arising in the brain must be fundamentally different from rhabdoid tumors arising in the kidneys. The biological divide between brain and kidney rhabdoid tumors began to disintegrate when it was found that 10% of patients with rhabdoid tumors of the kidney also developed rhabdoid tumors of the brain. More recently, biallelic loss of the tumor suppressor INI1 was found in both tumors. This more or less clinched the idea that rhabdoid tumors arising from the brain are the same tumor as those arising in the kidney.

Here is a list of summarizing the biological oddities of rhabdoid tumors.

1. Rhabdoid tumors are lineage nonspecific and can arise from neuroectodermal cells or mesodermal cells. All other tumors of somatic cells (nongerm cells) arise from a single germ lineage.
   
2. Cells within a single rhabdoid tumor seem to have differentiated along several developmental lineages (ectodermal, endodermal, neuroectodermal, and mesodermal). This phenomenon is otherwise restricted to pluripotent germ cell tumors.
   
3. Cells within a single rhabdoid tumor may include primitive cells indistinguishable from PNET tumors (primitive neuroectodermal tumors). PNET tumors are typically monomorphic tumors.
   
4. Rhabdoid tumors are all associated with a specific phenotypic cell (the rhabdoid cell) regardless of the developmental origin of the tumor (neuroectoderm or mesoderm).
   
5. The rhabdoid cells contain several different intermediate filaments. In normal cells, only one type of intermediate filaments is found in any single cell, and that filament is specific for the lineage of the cell.
   
6. Almost all tumors (other than rhabdoid tumors) arise from cells that resemble an observable normal cell. For example, a squamous cell carcinoma is composed of cells that resemble normal squamous cells biochemically, ultrastructurally, and by light microscopic examination. The rhabdoid cell has no known counterpart in any adult tissue or in any stage of development.
   
7. Currently, there is no other known genetic mutation that produces a specific phenotype akin to the association between INI1 and rhabdoid cells. INI1 loss produces rhabdoid tumors in mice. Eight of 125 mice with germline haploid complement of INI (Snf5+/– mice) developed INI1-negative tumors of soft tissue origin and rhabdoid cell morphology.[1] The mouse tumor is morphologically and genetically identical to the human tumor. Despite the phenotype and genotypic similarities between murine and human rhabdoid tumors, the mouse tumor arises from the branchial arch soft tissue, a tissue of origin not observed in human rhabdoid tumors.
   

Is the rhabdoid tumor just an example of tumor trivia? No. When you have a rare tumor that seems to break the rules of tumor classification, you can bet that the explanation will have fundamental importance in our understanding of tumor biology. It might even lead to a new and effective approach toward eradicating cancers.

There is an answer to the mysteries of rhabdoid tumors, and we'll approach that answer in the next few blog entries.

REFERENCE

1. Roberts CW, Galusha SA, McMenamin ME, Fletcher CD, Orkin SH. Haploinsufficiency of Snf5 (integrase interactor 1)predisposes to malignant rhabdoid tumors in mice. Proc Natl Acad Sci USA 97:13796-13800, 2000.
   

Jump to tomorrow's blog

-© 2010 Jules J. Berman

key words: cancer, neoplasia, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, pre-cancer, precancerous lesions, pre-malignant lesions, gene synonyms: Snf5 Ini1 Baf47 SmarcB1, atypical teratoid tumor, malignant rhabdoid atypical teratoid tumor

Overview of neoplastic development

Modern tumor biology is pre-occupied with oncogenes and tumor suppressors. For the past several decades, cancer researchers have attemptedt o understand how specific genetic alterations contribute to the malignant phenotype. Certainly, research in tumor genetics has led to great advances in our understanding of cancer.

Still, tumor genetics has not really helped us understand tumor speciation. Tumor speciation is an area that gets very little attention from most cancer researchers, but if you don't have a grip on the general phenomenon of tumor speciation, you're not going to fully understand tumor classification, or carcinogenesis (the necessary biological steps leading to the development of cancer), or the biology of precancers (lesions that precede the development of cancer), or the class-based approach to precancer treatment.

Much of my book Neoplasms: Principles of Development and Diversity, is devoted to the subjects of tumor speciation and classification. I've also written about the phenomenon in prior blogs. Here's a discussion of tumor speciation, condensed to a few paragraphs.

Tumor genetics tells us that most of the common cancers occurring in adult humans are genetically complex, with thousands of mutations in tumor cells, and enormous genetic heterogeneity among the different cancer cells within the same tumor. Just as every human is genetically unique, so is every cancer that has ever occurred. If this is the case, why are the different types of cancers restricted to just a few hundred species of tumors (e.g., Wilms' tumor, chronic myelogenous leukemia, pheochromocytoma, hepatocellular carcinoma, malignant melanoma). The individual specimens within each tumor species look about the same under a microscope (that's how we diagnose them) and behave within somewhat predictable biologic paramaters. If cancers are genetically diverse, and if genetics is responsible for the behavior of cancers, wouldn't you expect to find a near-infinite number of tumor species?

The hypothesis developed in my book is that while cancer is caused by genes, tumors are speciated by epigenes, the non-sequence DNA changes that determine cellular differentiation in normal cells. Just as normal cells of the human body are restricted to a few hundred differentiated cell types [by the epigenome], cancer cells are restricted to a finite number of cell types [by the epigenome].

During carcinogeneis, cumulative genetic changes occur, but these changes must somehow fit into an epigenomic pattern that supports the continued growth of the tumor. In short, carcinogenesis is a genetic phenomenon, but tumor speciation is an epigenetic phenomenon.

Precancers are a late-stage epigenomic adjustment. If the adjustment is successful, from the perspective of the cancer cell, the precancer can develop into a cancer of a specific tumor species. Otherwise, the precancer regresses [spontaneously dies]. Precancers can be thought of as biological experiments conducted by cells undergoing carcinogenesis. Sometimes the experiment fails (and the precancer regresses). You're never sure of the result of the experiment (i.e., a precancer may develop into any one of several different tumor species, as in the case of bronchogenic precancers).

I admit that this all seems vague, but that's what happens when you try to cover a complex subject in a few paragraphs, without the aid of examples.

In the next blog, I'll introduce malignant rhabdoid tumors, and I'll try to explain how this unique tumor provides perspective to the whole problem of speciation.

Jump to tomorrow's post
Visit yesterday's post

-© 2010 Jules J. Berman

key words: cancer, neoplasia, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, pre-cancer, precancerous lesions, pre-malignant lesions, neoplastic development

Neoplasms: new blog site

This is the first blog of my third blog site. My other two blog sites are: Specified Life and Machiavelli's Laboratory . The Specified Life blog deals with information retrieval, organization, indexing, annotation, classification and analysis. Machiavelli's Laboratory is a blog devoted to issues covered in my satiric online book, Machiavelli's Laboratory, a free ebook on scientific ethics, explained from the perspective of an unethical scientist.

I've written two books about cancer: Neoplasms: Principles of Development and Diversity, and Precancer: The Beginning and the End of Cancer.

My philosophy, in a nutshell, is this: Malignant cancers are very hard to treat, but they are preceded by precancers, that are much easier to treat. If we successfully treated all of the precancers, we would stop all new cancers from developing. Though there are many different neoplasms (i.e., cancers and precancers), we can classify neoplasms into groups with shared properties. These shared properties can be exploited to develop effective treatments that target all of the members of the same class. By dividing neoplasms into sensible classes, and by developing targeted treatments for the precancers within each class, we can put an end to human cancer.

There are many important cancer topics that could not be fully explored in my two published cancer books. I've been writing about cancer in my Specified Life blog, but it makes a poor fit. It's probably best to have a dedicated blog site written for cancer reserachers, pathologists, and healthcare professionals who might be interested in my approach to neoplasia.

In the next few weeks, I plan to expand my discussion of germ cell tumors, started in the Specified Life blog, and to begin a series on epigenomic neoplasia; malignant rhabdoid tumor serving as the prototype.

Jump to tomorrow's blog

-© 2010 Jules J. Berman

key words: cancer, neoplasia, neoplasms, precancer, tumor biology, tumour biology, carcinogenesis, cancer development, pre-cancer, precancerous lesions, pre-malignant lesions, neoplastic development