Quantitative study of the selectivity of human sirtuins to acylated nucleosomes

Visit the full literature paper here: https://www.nature.com/articles/s41598-018-21060-2.pdf

As we have learned for essentially our entire college experience, chromatin is regulated by post-translational modifications (PTMs) of the histones in nucleosomes. Most commonly, we hear about the acetylation of histones in regards to PTMs and how it is essential in the activation of gene expression. This is accomplished by the loss of charge on lysine residues, the promotion of chromatin coming out of its compact state, and therefore gene transcription. The acetylation of lysines in the nucleosome is dependent on the ratio of histone acetyltransferases (HATs) which put on acetyl groups and histone deacetylases (HDACs) which take off acetyl groups, but the selectivity to these lysines is unknown (Seto and Yoshida 2014).

HDACs can be zinc-dependent or NAD+ dependent, known as sirtuins. Sirtuins, as we discussed in BCM441 come in seven different human homologs which all share a catalytic domain, even though they all serve different purposes, and can regulate many important bodily functions. More specifically, they are known to be involved in the regulation of metabolic and aging related diseases (Yoshida et al. 2017). In BCM441, we discussed how activation of sirtuins, by molecules like resveratrol, resembles phenotypes of caloric restriction. We also discussed how Sirt3 deacetylates isocitrate dehydrogenase 2 leading to increased NADPH and increased glutathione, protecting the cells from oxidative stress. Their ability to impact aging has made understanding the mysterious activity of specific sirtuins an important way to develop therapeutic targets for aging related diseases (Tanabe et al. 2018).

It has been found that different isoforms of sirtuins have different preferences to substrate sequences and that sirtuin reactivity and site selectivity may differ between nucleosomal and non-nucleosomal substrates. New types of lysine acylations have been discovered following mass spectrometry advances, such as butyrylation of histones, adding a more hydrophobic and longer group than acetylation does. Butyrylation can be catalyzed by the same HATs as acetylation and activates transcription in the same way, meaning their effects on chromatin structure are similar (Goudarzi et al. 2016). Other acylation of acidic groups can change the lysine state from cationic to anionic, such as in malonylation using its anionic carboxylate group. Unlike butyrylation, malonylation has a greater effect on inter-nucleosome interactions than acetylation. It has been shown that Sirt 1-3 can remove hydrophobic acyl groups, Sirt4 and 5 can remove acidic acyl groups, and Sirt6 can remove long-chain acyl groups. This study works to characterize acyl group and site-selectivity of the sirtuins using acylated nucelosomes as substrates by using the analytical method of mass spectrometry (Tanabe et al. 2018).

Various compounds were used to develop an artificial catalyst system and acetylated, butyrylated, and malonylated nucleosomes were prepared. Their acyl group selectivity was then measured using LC-MS/MS analyses and using acetyl preference index (the ratio between deacetylation and debutyrylation efficiencies) for each lysine on H3. Only Sirt5 did a sufficient job at demalonylation on H3K18. Sirt1 and Sirt3 prefers deacetylation over debutyrylation at H3K18 and also at other residues, whereas SIrt3, Sirt6, and Sirt7 preferred butyryl groups at H3K18. Sirt4 showed no activity. Therefore, Sirt1,2,3,6, and 7, prefer to remove hydrophobic acyl groups and Sirt5 prefers to remove the negatively charged malonyl group (Tanabe et al. 2018).

Unfortunately, understanding the site-selectivity of sirtuins isn’t as cut and dry, as the deacylase activity depends on the location of the lysine residue. Interestingly, it seemed to be easier for all of the sirtuins to deacylate at H3 tails than at H4 tails. The authors found that H3K9 was more susceptible to debutyrylation by Sirt1 than H3K18. This makes sense because the amino acid sequence around H3K9 matches to the high-affinity sequence of Sirt1. Although, the butyryl group was removed at similar efficiency at H3K9 and at H3K18 by Sirt2, Sirt3, Sirt5, and Sirt6. Oddly, Sirt7 efficiently removed butyryl groups (and most likely acetyl groups due to their similarity) at H3K36 and H3K37, but not at H3K18, H3K9, or any other sites, which led to authors to direct their focus to Sirt7 and its selectivity (Tanabe et al. 2018).

Previous literature gives some controversial findings on Sirt7, where some reported that its deacetylase activity is specific to H3K18 peptide and to H3K18Ac on full length histone H3 in purified poly-nucleosomes. Another group reported Sirt7’s deacetylase activity was activated by double-stranded DNA and RNA, and unable to do this activity on H3K9 and H3K18 peptides and purified histones when dsDNA is not present. Activation by RNA aided the activity of Sirt7 at H3K18 peptide and that the RNA binding motif of Sirt7 resides on the C-terminal basic region. Though, this paper showed exceptional activity at K36 and K37 without RNA present, so the authors hypothesized that Sirt7 binds with DNA on the nucleosome via C-terminal basic region and its nucleosome-binding would enhance Sirt7’s deacetylase activity. By using purified FLAG-Sirt7 wild-type, mutant, and catalytic dead proteins, the authors showed the wild-type having strong deacetylase and debutyrylase activity near H3K36 and H3K37 and weak activity near K18, K23, and K27. Shockingly, it even removed the malonyl group at K36 and K37, even though it was weaker than its deacetylation and debutyrylation. The catalytically dead protein’s activity was completely shut down. The authors concluded that these deacylations were catalyzed by the protein itself, and not by other factors that it may be copurified with. In addition, the wild-type and catalytically dead proteins (not the mutant) were bound with nucleosomes and the mutant also reduced deacylation activity compared to the wild type. These findings support that Sirt7 binds with DNA on the nucleosomes via the C-terminal basic region, and its nucleosome binding enhances the deacylase activity toward H3K36 and H3K37 (Tanabe et al. 2018).

In regards to site selectivity, the authors found that H4 tails are poor substrates and H3K18 is generally susceptible to deacylation, though these ideas are not totally consistent with previous studies using non-nucleosomal substrates. Nucleosome structure may be a block of the accessibility to sirtuins to H4, as it is a possibility that H3 is located near the catalytic pockets of sirtuins. The authors also believe that Sirt7 activity may be spatially regulated by distinct nucleic acids in cells, where nucleosomal DNA activate Sirt7 on chromatin and ribosome RNAs activate Sirt7 in the nucleolus (Tanabe et al. 2018).

The regulation of acylation in histones is crucial for understanding chromatin epigenetics. The idea that the ratio of acetylated versus butyrylated histones could be affected by the ratio of the sirtuins that prefer acetyl groups verus butyryl groups is extremely useful. This information can be used to address whether the concentration of each sirtuin present has an effect on the acylation and therefore a specific disease. In future work, the authors should address if diacylation of H3K36 and H3K37 by Sirt7 is physiologically relevant. It should also be studied if Sirt7 deacetylation at K36 and K37 is a prerequisite for H3K36 methylation (Tanabe et al. 2018). Another interesting area to explore would be hydroxylated acyl groups by identifying a way to make the nucleosome furnish these groups to study sirtuin selectivity. The authors should also work to understand the acyl selectivity of zinc dependent HDACs. Overall, this study helped to further understand how histone acylations are regulated, which can then provide insight into epigenetic mechanisms and developing therapeutic targets to age related diseases.



Goudarzi, A., Zhang, D., Huang, H., Barral, S., Kwon, O.K., Qi, S., Tang, Z., Buchou, T., Vitte, A.-L., He, T., Cheng, Z., Montellier, E., Gaucher, J., Curtet, S., Debernardi, A., Charbonnier, G., Puthier, D., Petosa, C., Panne, D., Rousseaux, S., Roeder, R.G., Zhao, Y., and Khochbin, S. 2016. Dynamic Competing Histone H4 K5K8 Acetylation and Butyrylation Are Hallmarks of Highly Active Gene Promoters. Mol. Cell 62(2): 169–180. doi:10.1016/j.molcel.2016.03.014.

Seto, E., and Yoshida, M. 2014. Erasers of Histone Acetylation: The Histone Deacetylase Enzymes. Cold Spring Harb. Perspect. Biol. 6(4): a018713. doi:10.1101/cshperspect.a018713.

Tanabe, K., Liu, J., Kato, D., Kurumizaka, H., Yamatsugu, K., Kanai, M., and Kawashima, S.A. 2018. LC–MS/MS-based quantitative study of the acyl group- and site-selectivity of human sirtuins to acylated nucleosomes. Sci. Rep. 8(1): 2656. doi:10.1038/s41598-018-21060-2.

Yoshida, M., Kudo, N., Kosono, S., and Ito, A. 2017. Chemical and structural biology of protein lysine deacetylases. Proc. Jpn. Acad. Ser. B 93(5): 297–321. doi:10.2183/pjab.93.019.


14 thoughts on “Quantitative study of the selectivity of human sirtuins to acylated nucleosomes

  1. Hi Brittany,

    Thank you for your review, it was really helpful. This study is excellent in helping us better understand epigenetic mechanisms and the different preferences in types of lysine acylations. In your review, you mention that butyrylation can be catalyzed by the same HATs as acetylation and therefore influences transcription in the same way. Therefore, in your opinion, why would certain Sirt isoforms preferentially remove the acetyl group over the butyryl group (or vice versa) if they ultimately have the same impact? Is it just a matter of accessibility, or are there other factors that may contribute to this selectivity?

    1. It seems as if the sirtuins that remove butryl and acetyl groups overlapped the most, though yes some will prefer one over the other. It mostly seems as if it is a matter of accessibility. Its going to depend on which parts of which histones are in the catalytic pockets of which sirtuins. It is also going to depend on the amino acid sequence of the histone and its affinity for the sirtuin amino acid sequence. There are many factors that are going to come into play, so it is difficult to pin point one specific thing.

  2. This was interesting to further explore the roles of the various Sirtuins in histone deacetylation. The specificity of Sirt-7 is interesting to see, as it poses the thought that it must have special in vivo targets to deacetylate, while waiting on a regulatory signal from the dsDNA or dsRNA. Further exploration of Sirt-7’s natural targets would elucidate its specific epigenetic role and explore targeted gene regulation. Also of some interest is Sirt-4’s apparent lack of function. Is it possible that this sirtuin also is dependent on an activator molecule like Sirt-7?

    1. Yes Sirt7 is very interesting throughout this paper as its results were unexpected. A lot more research needs to be done with Sirt7 and see if it and its targets have any relevance at all. Also, yes Sirt4 peaks my interest as well and sure it seems very possible that there may be some kind of activator that wasn’t present in this specific study. Maybe it has a more specific, less common function.

  3. This review was very interesting, especially given what we have been learning recently about the importance of sirtuins. One result I found incredibly interesting that you mentioned was the spatial regulation of Sirt7 by nucleic acids. What that sounds like to me is that these distant nucleic must fold back in some way to alter the activity of Sirt7 at nucleic acids further down the molecule. I’m interested to know if you and/or the literature think there might be something akin to a “binding pocket” that the DNA is forming? If so, that could inspire a lot of interesting research into the mechanisms of structures that bind DNA.

    1. I haven’t seen anything in the literature about it, but that does seem like a valid explanation. There really doesn’t seem like a better explanation than one based on secondary structure and folding, as it can fold a certain way and activate Sirt7. DNA probably can form something like a binding pocket, in order to activate. I am sure more research will soon come to light on this topic as I agree, it is very interesting.

  4. Hi Brittany,

    This review was very well written and you did a great job taking what we learned in class to the next level. Although the role of sirtuins was generally understood, now many more practical applications can be revealed since the specific roles and regulation of the individual sirtuins have been presented. Moving forward, how big of a factor is location of the acetylated lysine in determining if deacylation is possible? Is it still possible for deacylation to occur enough to influence gene expression on H4 even though it is not near the catalytic pocket like H3?

    1. Exactly, general applications of sirtuins were understood, but not of specific sirtuins. One of the factors to consider is if the lysine is in the head of the tail of the histone. It seems it would be fairly important, due to accessibility seeming like the most important factor in deacylation. I am sure it may be possible if H4 is accessible enough, or maybe it just relies on another form of deacylation, not deacetylation. More research needs to be done to determine this.

  5. Hi Brittany, I really liked the article! I completely agree that hydroxylated acyl groups seem interesting to study, as the hydrogen bonding could make for some interesting chemistry. However, when I was reading the paper, the authors were pretty vague about how their model “cannot adopt such acyl groups at this stage” and how “New methods are required to address this issue.” Could you shed some light on why their model is insufficient to do these types of analyses?

    1. The authors say, “We anticipate that sirtuins’ selectivity toward hydroxylated acyl groups can be addressed by identifying a method to synthesize the nucleosomes furnishing such acyl groups.” Therefore, it is a matter of developing a new way to make nucleosomes with hydroxylated acyl groups, which their current method does not allow.

  6. Hey Brittany, thank you for your review, it was really helpful! This paper is very topical to the last chapter of our class. I really enjoyed seeing someone use analytical chemistry for a biological purpose; I am even more shocked to see analytical chemistry being presented in nature. I have not seen a lot of analytical chemistry articles in this journal. With this being said, I thought it was interesting that the authors were able to quantify various types of PTMS; even types we did not hear of in class. Do you think other scientists will be able to confirm these results through other methods such as cyro-EM? Additionally, the authors used LC-MS/MS, do you know why they did not use HPLC-MS/MS?

    1. Thanks for the question Calvin, you make a good point. I am sure it can be possible to use other methods such as cryo-EM and HPLC-MS/MS. LC-MS/MS probably would show the best results with the method that the authors developed to analyze sirtuins. Though, I am not totally sure why they chose one method over another.

  7. Hi Brittany, great review. It was interesting to read more about sirtuins since we talked about them pretty in depth in class. You mentioned that understanding histone acetylation regulation can provide insight into developing new therapeutic targets. My question would be, is there something novel and feasible from this research that would allow for the creation of new drugs for treatments? Can you propose a mechanism that pharma companies would use to create these drugs using this research?

    1. Understanding the epigenetic mechanisms, the preferences of sirtuins, the concentrations of sirtuins, what regulates sirtuns, etc. would be important in understanding age related diseases, though I don’t think there is one specific drug target that came out of this research. I think the broader understanding is what can be translated forward, as not enough research as been conducted to develop a therapeutic target to a specific disease.

Leave a Reply

Your email address will not be published.