Cancer: Erosion of the Epigenetic Landscape 2

Epigenetics and Aging (17)

For a complete picture of the epigenetic landscapes of normal cells and their erosion by cancers we need to consider the role of noncoding RNAs. They are called noncoding because they do not serve as templates for protein construction; their functions lie elsewhere. There are many forms of noncoding RNAs, two of which everyone learns in high school biology: ribosomal RNAs and transfer RNAs, both crucial for constructing proteins, amino acid by amino acid, from the nucleotide triplet sequences on messenger RNA, the process known as translation.

For many years after the discovery of the genetic code, decades in fact, that was the extent of our knowledge of noncoding RNAs. Things have changed dramatically. Over the last three decades the other noncoding RNAs have had quite a run. Two types of noncoding RNAs are epigenetic actors, one quite small (short), one quite large (long). The short ones are aptly called micro RNAs (miRNA); equally apt, the long ones are called long noncoding RNAs (lncRNA).

Micro RNAs and Cancer

Small first. Micro RNAs are manufactured from longer RNA strands, through a process that need not concern us here. Each is only 21-23 nucleotides long. For perspective, if they were protein coding, that would be seven amino acids max, which is on the small side even for the smallest proteins, called peptides. Though tiny, microRNAs pack a punch, especially when working in packs. When a cell senses too many messenger RNAs of a certain species, bound for translation into a specific species of protein, the process is aborted; they are set upon by miRNAs the attachment of which deactivates and ultimately marks the messenger RNA for destruction.

There are nearly 1,000 known miRNA species, each of which can attach to several different (protein coding) species of messenger RNA. Moreover, several species of miRNA can attach to a single messenger RNA species, though at different sites. A division of labor. Though versatile they are not promiscuous. Each miRNA is limited to a certain set of messenger RNA species, which overlaps with one or more of the species sets of other miRNAs.

We can think of miRNAs as a backup form of inhibitory epigenetic regulation on gene expression, when things aren’t taken care of at the source (transcription level). This is sometimes called RNA interference. The array of miRNAs varies from cell type to cell type. This variation is an important dimension of the epigenetic landscape for any given cell type, and hence a source of cell identity (24179605). MicroRNAs are crucial for normal cell functioning, involved in the most basic cellular pathways, including proliferation. Which brings us to cancers.

Dysregulation of miRNAs is a feature of every known cancer (25960691). Some miRNAs increase in abundance in specific cancers, others virtually disappear. Moreover, miRNAs reciprocally influence both DNA Methylation and histones in pathological ways during carcinogenesis.

Consider first, miRNAs and DNA methylation. DNA methylation at miRNA genes inhibits the production of miRNAs (https://doi.org/10.1016/j.molonc.2012.07.007). On the other hand DNA methylation is partly controlled by miRNAs. Here’s how that works. DNA methylation requires a family of enzymes, as does demethylation (different family). By interfering with the messenger RNAs for those enzymes, miRNAs influence both methylation and demethylation (https://doi.org/10.1038/s41417-020-00210-7).

Same goes for histone-controlled chromatin states. When the histones bind tightly to miRNA coding regions, no miRNAs are produced. When things loosen up miRNAs can be expressed, if they aren’t methylated. It works both ways. Recall that the effect of any canonical histone on chromatin, and hence gene expression, depends on its auxiliary attachments, acetyl, methyl, phosphate groups and so forth. These attachments don’t occur spontaneously; they require specific enzymes to attach and remove, different enzymes for each auxiliary. By attacking the messenger RNA of the attackers or removers, miRNAs influence the chromatin state (https://doi.org/10.1155/2022/4889807). These reciprocal influences occur in any cell; they go off kilter in cancerous cells, dysregulating all. Any single off kilter component –DNA methylation, histone regulation, or miRNA regulation, can be diagnostic in particular cases.

Long non-coding RNAs (lncRNAs) and Cancer

With respect to cancer long noncoding RNAs are the new kids on the block. But not when it comes to epigenetics. In fact, one lncRNA was instrumental in getting the science of epigenetics off the ground. It is called Xist, (X-inactive specific transcript) which I love (the acronym, not the fully fleshed out name). The X-inactive feature is where I will focus. It refers to the fact that these lncRNAs (mostly) shut down one of the X-chromosomes in female mammals, including humans. This is extremely important in keeping males and females on the same page gene dosage-wise. The X-chromosome is by far the largest and most gene rich in the mammal genome. Females of course have two of them, males only one. If females got twice the amount as males of whatever is coded for on the X-chromosome, evolution couldn’t handle it; that would create an evolutionary state that is called sexually antagonistic—on a scale far larger than anything that has yet evolved-- in which what’s good for the goose is bad for the gander and vice versa. And the middle ground is no man’s/woman’s land. By way of resolution evolution has arrived at an elegant solution: disable one of those female X-chromosomes, even things up.

Which is where Xist comes in. By binding to one of the X-chromosomes, Xist (mostly) inactivates it, leveling the physiological playing field. But which X-chromosome? The one inherited from mom or the one inherited from dad? For non-Marsupial mammals, evolution decided on a coin flip. Not a coin flip in the zygote, nor a coin flip in every adult cell, but a coin flip that occurs between implantation and the end of the first month of embryogenesis. As a result, in some cell lineages the maternal X-chromosome is inactivated, in others the paternal X-chromosome is inactivated. It works out to about half and half in adult mammals.

Cancer, though, often wreaks havoc on X-inactivation in the effected cells, increasingly so as it progresses. For women, this means a lot of genes are unleashed that should be confined. More is not better. This is dysregulation on a massive scale. For males, on the other hand, cancer can induce X-inactivation, which is also problematic ( 32040694). This seems to occur late in cancer development, often after the cancerous cells become XXY, a condition called aneuploidy (additional or subtracted chromosomes). Aneuploidy is a sign of genomic chaos caused by upstream epigenetic and genetic dysregulation in cancerous cells.

Long after Xist was discovered it remained an outlier in the realm of noncoding RNA. A unicorn of sorts, the longest of them all. Now it is but one of a vast herd of noncoding RNAs. Thousands have been identified to date. It seems another is discovered monthly. The herd of long noncoding RNAs encompasses diverse functions, some epigenetic, some not.

There are three distinct epigenetic activities for lncRNAS. Some, like Xist, bind directly to DNA, shutting down gene activity at its source, in the manner of DNA methylation. Many, like micro RNAs, operate at the next level of gene expression, in binding to messenger RNAs, thereby inactivating them. Still others act as miRNA sponges, soaking up excess miRNAs (10.1007/978-1-0716-1697-0_21). As a result, gene expression is enhanced. This role as miRNA regulators is the most recently discovered.

As with the other epigenetic processes, lncRNAs are instrumental in establishing the cell-specific landscapes and hence cell identity. Each cell type has its own unique suite of lncRNAs. During carcinogenesis this element of the epigenetic landscape is also eroded (PMC8330958).

These are early days in the study of lncRNAs as they relate to cancer. Most generally, there are reports that a number of lncRNAs are up or downregulated in specific cancers (28701486). Their sponge function as miRNA absorbents seems to be particularly relevant in some cancers https://doi.org/10.1111/cas.13342. One example, in which the sponginess goes off the rails occurs in lung cancer. It involves the long noncoding RNA whimsically called HOTAIR, which absorbs an important tumor suppressive microRNA less whimsically called miR-613. In cancerous cells too much miR-613 gets absorbed by HOTAIR, reducing its tumor suppressive activities (29187267). This dysregulation is associated with not only tumorigenesis but also metastasis in the most common form of lung cancer.

This includes our brief survey of the four dimensional erosion of the epigenetic landscape during carcinogenesis, generally. I now turn to some specific cancers, beginning with breast cancer.