Thursday, 1 March 2012

Removing epigenetic memory blockages in Alzheimer's disease

The progressive deterioration of one’s social and cognitive functioning is often thought of as being synonymous with the normal aging process. After all we all forget names, misplace our keys and stumble over our words from time to time. Hell, sometimes it even happens in the absence of that second glass of wine. Yet you only need to look at Christopher Plummer's recent Oscar acceptance speech to realise that a deteriorating mind is not an inherent part of growing old. After all Plummer is 82 and appears to more on the ball than me most weekday mornings. Instead more often than not impairments to cognition in the elderly, and at times the not so elderly, are attributable to the presence of an underlying neurodegenerative dementia known as Alzheimer's disease. Whilst the causes leading to these impairments remain poorly understood, and largely contentious, new research published in Nature suggests that cognitive capacities in the deteriorating brain may arise due to 'epigenetic blockages in gene transcription', which, if the results are anything to go by, just might be reversible.

Some facts about dementia you'd probably rather forget

But before we sink our teeth into the recent study by Johannes Graff, let's take a moment to get some perspective on just how big a problem Alzheimer's really is. As most of you will already know, Alzheimer's disease (AD) is a highly varied and progressive disease accounting for up to 60% of dementia diagnoses worldwide (mostly in those over the age of 65). In fact it's estimated that five percent of people over the age of 65 will develop dementia, with the figure rising to 20% over the age of 80. However, that's not to say that those of us finding ourselves on the riper side of 65 are exempt from this insidious disease, as genetic forms of AD frequently strike people in their 50s, their 40s and occasionally even in their 30s. As of 2010, the number of people diagnosed with AD exceeded 35 million, bringing with them an annual cost of over $605 billion in care, treatment and lost productivity. That's right six hundred and five billion dollars! To put that in perspective the revenue of ExxonMobil, the highest ranking company in the world (with regard to revenue at least), is only $486 billion per annum and if we consider gross domestic product (GDP) then AD would rank at number 25, right above South Africa. And it doesn't stop there. The aging global population means that the prevalence of AD will increase dramatically, over the years to come. In fact it is expected to almost double every 20 years, reaching 65 million in 2030 and 115 million by 2050. This increase in prevalence will of course be accompanied by significant increases in the costs associated with the care and treatment of AD sufferers and highlights the need for more accurate and efficient diagnostic and therapeutic measures to be developed. And that's potentially why these new findings are so exciting.

I'll take my histones acetylated

In their study, Graff and co focused on the knowledge that our epigenetic processes help to orchestrate the stable changes in gene expression enabling memory formation. Of the modifications identified to date it is the acetylation of histones which is consistently associated with learning and memory, in such a way that histone acetylation is frequently reduced in animal models of neurodegeneration. If you're scratching your head at this stage wondering exactly what all this means then here's some basic background knowledge you mighht find useful. Histones are essentially protein spools which DNA wraps around to gain its structure and order. Without histones DNA would be an unruly mess akin to a dropped pot of spaghetti (if the pot had been full of only one really long piece of spaghetti). Acetylation on the other hand simply denotes introducing an acetyl group into a chemical compound, which in this case are the histones. Histone acetylation is important as it reduces 'the electrostatic affinity between neighbouring histones and DNA', thus allowing more space for the transcription of those all important memory-related genes. So histone acetylation means more readily transcribable genes, and conversely histone deacetylation means too much affinity between histones and DNA. To use the spool analogy again the DNA is wound to tight around the histones and thus a reduction in the ability to transcribe those memory genes is the result.

But what exactly is causing this reduction in histone acetylation in the AD-affected system? Well according to Graff it's all down to histone deacetylase (HDAC) levels, or more specifically a class I HDAC, known ironically as HDAC2 levels. Graff analysed the levels of HDAC2 in the neuronal tissue of two separate strains of AD transgenic mice (CK-p25 and 5XFAD for those of you playing at home) and found that HDAC2 levels were significantly elevated in the brain regions associated with neuronal degeneration (such as the prefrontal cortex and the CA1 area of the hippocampus, although curiously not in the CA3 area or the dentate gyrus). Interestingly, levels of HDAC2 in areas not affected by the neurodegeneration, such as the amygdala, remained normal and no differences in the structurally similar HDAC1 or HDAC3 were observed regardless of the brain region investigated.

Next the authors set out to ascertain whether HDAC2 was associating with genes in the mice hippocampi implicated in learning and memory, such as Homer1 and Cdk5, as well as genes involved in synaptic plasticity, such as Syp and Syt1. And not surprisingly it was! And yet it didn't seem to be associating more with housekeeping genes, such as β-actin, β-globulin or β-tubulin. In other words the tightening of the DNA around the histone spool led to a reduction in the transcription and subsequent expression of genes involved in memory and neural plasticity. Perhaps more excitingly though, was the finding that when the mouse CA1 hippocampal areas were injected with short-hairpin RNAs (obviously not actual hairpins just a means of knocking down genetic expression) to knock down the expression of HDAC2 they found that not only did itmanage to reduce HDAC2 levels to those indistinguishable from control animals but they also managed to alleviate the memory deficits associated with elevated HDAC2.

Of mice and men

Whilst promising these findings wouldn't be of much significance unless they could be tied back to the AD in humans. After all, unless specifically bred to do so, mice don't tend to make a habit of ending up with Alzheimer's related memory impairments. To do this the researchers analysed HDAC2 levels in the hippocampal CA1 tissue from 7 controls and 19 ADs. Like in the mouse tissue the researchers found that levels of HDAC2, but not HDAC1 or HDAC3, were elevated within AD brain tissue, suggesting that 'elevated levels of HDAC2 may also accompany cognitive decline' in the human brain.

But what is about the AD brain that leads to a buildup of HDAC2? Well like most features of AD it all comes back to β-amyloid (Aβ). That's right the researchers found that the pathological hallmark of AD, the peptide thought to be responsible for it all may also be responsible for the increases in HDAC2 observed in the study. This was determined by exposing primary hippocampal neurons to aggregated forms of Aβ, known as oligomers, with either the neurotoxic wild-type sequence (Aβ1-42) or the non-active reversed sequence (Aβ42-1). As expected this lead to increased levels of HDAC2 in the primary neurons exposed to wild-type oligomers but not the reverse sequenced ones. These results are perhaps best summarised by lead investigator Dr Li-Huei Tsai who said;
'We think beta-amyloid triggers a cascade of damaging reactions. One of these is to activate HDAC2, which in turn blocks the expression of genes needed for brain plasticity. Once this blockade is in place, it may have a more systemic, chronic effect on the brain.'

So where does it all lead from here? Well not surprisingly the authors of the study are currently working on identifying HDAC2-specific inhibitors to use in further drug development, a far cry from the Aβ-centric agents currently under investigation. However even if they are successful it is unlikely that such a drug will be the silver bullet the AD field is hoping for, as despite the fact that reducing HDAC2 appears to restore cognition it doesn't actually appear to stop neuronal death. Instead it is thought that HDAC2 reduction acts to enhance the neuroplasticity of the neurons which remain, strengthening their ability to connect and communicate with those around them. Either way, we'll more than likely find out just how effective an HDAC2-inhibitor can really be in the years to come (or the decades to come if you know your drug development) and in the meantime we get to be amazed by the new research that will no doubt be spurred on by these results. After all it was only a few days ago that the functionality of damaged neurons in the AD brain were thought to be irrevocably lost, but today. Well today there's just that little bit of hope.


  • Gräff, J., Rei, D., Guan, J., Wang, W., Seo, J., Hennig, K., Nieland, T., Fass, D., Kao, P., Kahn, M., Su, S., Samiei, A., Joseph, N., Haggarty, S., Delalle, I., & Tsai, L. (2012). An epigenetic blockade of cognitive functions in the neurodegenerating brain Nature DOI: 10.1038/nature10849
  • Karran E, Mercken M, & De Strooper B (2011). The amyloid cascade hypothesis for Alzheimer's disease: an appraisal for the development of therapeutics. Nature reviews. Drug discovery, 10 (9), 698-712 PMID: 21852788


This post was written by Andrew Watt for A Hippo on Campus.


  1. This is a very beautiful and interesting research
    The most educating one i have read today!

    GED Online

  2. As always, the bigger question is what causes gradual Abeta accumulation in sporadic AD. The best evidence points strongly to neuronal lipid membrane peroxidation, which inhibits Abeta degrading enzymes. Cerebral lipid peroxidation can be induced quite easily with common steam-treated polyunsaturated food oils.