In the Media
Aberrant Heterochromatin, Gene Expression Inflame Old Cells
Research by Sam Beck, Ph.D., finds age-related gene misregulation could create an inflammatory milieu that contributes to Alzheimer's disease risk
Aging being the biggest risk factor for neurodegenerative disease, scientists urgently want to know what exactly goes wrong in old cells. In the December 15 Science Advances, researchers led by Samuel Beck, MDI Biological Laboratory, Bar Harbor, Maine, reported age-related gene-expression changes in DNA segments that are devoid of the long stretches of cytosine-guanine repeats called CpG islands. Various types of aged human and mouse cells upregulated these “islandless” sections. This provoked production of inflammatory cytokines they encode, such as interleukin-6. Tissue from people with age-related conditions, including Alzheimer’s disease brains, also showed evidence of overexpressed genes lacking CpG islands. This suggests that age-related gene misregulation could create an inflammatory milieu that contributes to AD risk.
- The lamina, aka inner nuclear membrane layer, loosens as cells age.
- The attached heterochromatin unwinds, exposing otherwise silent genes lacking C-G repeats for transcription.
- Many of these genes encode secretory, inflammatory proteins.
“This exciting paper is relevant to AD and related tauopathies, where we see a general loss of heterochromatin and alterations in the lamin nucleoskeleton,” Bess Frost, University of Texas Health, San Antonio, wrote to Alzforum. Beck was surprised how ubiquitous CpG islandless gene misexpression was. “It occurred in pretty much all types of tissue during normal aging, age-related diseases, and premature aging disease—this pervasiveness was unknown,” he told Alzforum.
DNA segments that lack CpG islands typically lie dormant inside tightly wound heterochromatin, which latches onto lamin proteins that build the inner nuclear membrane. As cells age, this lining weakens and frees the heterochromatin, which, in turn, loosens to expose the previously silenced genes inside it for transcription (see image below). Beck and colleagues had previously found that heterochromatin regulation only affects expression of genes devoid of CpG islands (Beck et al., 2018). Could their transcriptional dysregulation contribute to cellular problems in aging or age-related disease?
To find out, first author Jun-Yeong Lee and colleagues used genetically diverse wild-type mice with known, variable rates of physiological aging, called diversity outbred (DO) mice (Nov 2015 conference news). The researchers sequenced RNA from kidneys and hearts of 192 DO mice at 6, 12, and 18 months of age, 64 in each age group, then divided genes by CpG island status. Expression of genes lacking these islands drove age-related gene-expression changes. For example, in kidney tissue from 18-month-old mice compared to 6-month-olds, one-third of islandless genes doubled their expression versus less than 10 percent of genes containing CpG islands.
Among the 64 older mice, 17 had significantly higher expression of CpG islandless genes in their kidneys. Protein levels in their urine were abnormally high, a sign of kidney trouble, linking this type of gene overexpression with deteriorating function.
Kidney cells from those mice showed signs of nuclear membrane and chromatin breakdown. The cells had less nuclear lamina protein lamin B1, inner nuclear membrane protein lamin B receptor, and heterochromatin protein HP1α than kidney cells from young animals or old mice with normal gene expression (see image below).
To explore whether chromatin disruption enabled this aberrant gene expression, the scientists used 4-week-old mice lacking the lamin B receptor, as it tethers heterochromatin to the nuclear envelope to help keep the DNA condensed. They sequenced RNA and ran western blots on proteins from the kidneys of three knockouts and three wild-type mice. Indeed, the knockout mice had higher expression of genes lacking CpG islands and fewer heterochromatin markers than wild-type.
Many of the same genes emerged as upregulated when the scientists meta-analyzed published RNA-Seq data from other mouse lines with genetic damage to their nuclear lamina, as well as from senescent mouse cells and from mice with a lamin A mutation that causes progeria, the disease of accelerated aging.
The scientists report that one in four of the misexpressed genes encode secreted proteins, including the inflammatory cytokine IL-6 and chemokine Cxcl13. Aging cells are known to release a cocktail of pro-inflammatory molecules, including cytokines and chemokines, known as the senescence-associated secretory phenotype (Sep 2015 news). Suspecting abnormal CpG-islandless gene expression in this phenotype, the researchers used gene ontology to annotate the abnormally expressed genes in aged DO mouse kidney or heart cells. Of the hundreds that encode secretory proteins, 72 and 88 percent, respectively, lack CpG islands. Such aged tissue showed signs of inflammation, containing white blood cells and cytokines. Islandless genes predominated among overexpressed secretory genes in senescent or progeria mouse cells.
Age-related inflammation may not be limited to tissues. The researchers’ meta-analysis of published human plasma proteome data showed that most proteins abundant in blood from old people arise from CGI-negative genes. Given that aged plasma can evoke systemic inflammation, localized CGI-negative gene misexpression may spread throughout the body through the blood (May 2019 news).
To look for CpG islandless genes in the “inflammaging” signature of normal aging, the scientists combed through published RNA-Seq data from various tissues of people aged 10 to 30 versus people over 50, as well as 1- to 6-month-old mice versus older than 12-month-old mice. Compared to young tissue, genes lacking CpG islands were upregulated in eight of 14 old tissues, including human fat and bone, and mouse brain and heart.
What might this mean for neurodegenerative diseases? The researchers parsed published RNA-Seq data on about 13,600 tissue samples from people with age-related diseases, including Alzheimer’s. Diseased tissue expressed more CpG islandless genes than tissue from age-matched controls (see image below). Many genes upregulated in diseased cells were also up in mouse cells with disrupted nuclear or chromatin structure. Together, the authors concluded that nuclear membrane and chromatin disruption underlie normal aging and age-related diseases alike.
Frost thought that, although 18-month-old mice are not that old, these analyses in old human and age-related disorder tissues do support the idea that genes lacking CpG islands are expressed more with age.
Curiously, disease-associated proteins, including tau, C9ORF72, and TDP-43, and the anti-aging protein REST alter the nuclear lamina (Jan 2019 news; Feb 2019 news; Jan 2018 news; Feb 2019 news). Frost was intrigued at the possibility that unraveling heterochromatin and aberrant genes may spur post-mitotic neurons to express cell-cycle-related proteins in Alzheimer’s disease. “Is tau-induced loss of heterochromatin causing cells to lose their mature neuronal identity?” Frost wondered.
In a computational biology exploration of a potential treatment angle of their work, Beck’s team ran a meta-analysis on RNA-Seq data from aged mice given anti-aging treatments, such as rapamycin or calorie restriction. The scientists found that treated mice expressed fewer CpG islandless genes than untreated mice. Beck told Alzforum he is testing whether suppressing expression of these islandless genes could slow aging and delay disease onset. He identified a molecule that reduced such gene expression and made C. elegans worms live longer. The researchers are currently testing the compound in aging wild-type mice and hope to do so in age-related disease models, as well. He would not say what the compound is.
View Original Article