<b>Assay Of Cellular HDAC And Sirtuin Activities With Fluorogenic Acetyllysine Substrates</b><br/><br/>

Assay Of Cellular HDAC And Sirtuin Activities With Fluorogenic Acetyllysine Substrates

by Konrad T. Howitz, Ph.D.

Abstract

Histone Deacetylases (HDACs), including the NAD⁺-dependent sirtuins (class III HDACs), are subjects of much current research interest. As the name implies, HDACs deacetylate histone acetyllysines, principally on their N-terminal 'tails'. This activity is an essential part of the epigenetic machinery that controls chromatin structure and through it, gene expression. However, regulatory acetylation/deacetylation of lysine epsilon-amino groups extends well beyond histones to include, for example, α-tubulin and a wide variety of transcription factors. HDACs and sirtuins show promise as therapeutic targets for a number of pathological conditions, including cancer, heart disease, diabetes, neurodegenerative and inflammation-related disorders. In the past, measurement of HDAC activity in a cellular context has been a lengthy process, typically involving [³H]acetate pulse labeling, isolation of histones, gel electrophoresis and autoradiography. A simple, fluorescence method for measuring cellular lysine deacetylase activity is described in which cells are cultured, experimentally treated and assayed in the same 96-well plate.

Introduction

Figure 1. Method for Assaying Intracellular HDAC Activity with the Fluorogenic, Cell-permeable Substrate, Fluor de Lys. Fluor de Lys is added to cell growth medium and enters the cells. Deacetylation by class I/II HDACs or sirtuins yields the deacetylated form of Fluor de Lys, some of which may exit the cell. Lysis of the cells with detergent allows contact between the non-cell permeable Fluor de Lys Developer and both intra- and extracellular deacetylated substrate, thus producing a fluorescent signal (symbol). Along with detergent, a strong class I/II HDAC inhibitor (e.g. trichostatin A) is included in the lysis buffer to insure that no deacetylation occurs after cell lysis. Click to enlarge.

Local remodeling of chromatin structure is an important factor in the control of gene expression and consequently affects many cellular functions including proliferation and differentiation. ^{(1, 2)} Histone hyperacetylation correlates with an open, decondensed chromatin structure and transcriptional activation, whereas hypoacetylation correlates with a condensed chromatin structure and transcriptional repression. Acetylation of epsilon-amino groups of specific lysines is catalyzed by histone acetyltransferases (HATs), whereas removal of these acetyl groups is catalyzed by HDACs.^{(1, 2)}

There are two types of HDAC reaction mechanism. Class I enzymes (e.g. human HDACs 1-3, & 8) and class II enzymes (HDACs 4-7, 9 & 10) remove acetyl groups by hydrolysis and are sensitive to hydroxamate inhibitors such as trichostatin A (TSA).⁽³⁾ Class III HDACs, also known as sirtuins (e.g. yeast Sir2 or human SIRT1), employ a unique NAD+-dependent mechanism⁽⁴⁾ and are inhibited by the reaction product nicotinamide⁽⁵⁾ (See Fig. 1).

Eighteen human lysine deacetylases have been identified to date.^{(6, 7)} Class I enzymes (HDAC1-3, and HDAC8) are almost exclusively nuclear proteins and associate with a number of transcription factors and co-repressor complexes.^{(1, 2, 8)} Class II enzymes (HDAC4-7 and HDAC9-11) also associate with a number of transcription factors, but in contrast to class I enzymes, class II deacetylases shuttle between the nucleus and cytoplasm.⁽⁹⁾ The sirtuins (SIRT1-7) ⁽⁷⁾occur in the nucleus, cytoplasmic and mitochondrial compartments.⁽¹⁰⁾ As a class, the SIRTs are implicated particularly in the transduction signals that connect metabolism, cell stress and survival responses and gene expression. ⁽¹¹⁾ At least one of the human sirtuins, SIRT4, functions not as a deacetylase, but rather as an ADP-ribosyl transferase.⁽¹²⁾

Although enzymes from each class can deacetylate histones, they act on non-histone substrates as well. SIRT1 deacetylates the tumor suppressor p53^{,(13, 14)} several class I HDACs deacetylate NF-kappaB,^{(15, 16)} and HDAC6 deacetylates alpha-tubulin. ^{(17, 18)} The existence of many non-histone, acetylated proteins, including transcription factors and nuclear import factors, ⁽¹⁹⁾ plus the localization of deacetylases to the cytoplasm⁽⁹⁾ and mitochondria, ⁽¹⁰⁾ two cellular compartments that are devoid of histones, suggest that deacetylases may have diverse functions throughout the cell.

Small-molecule modulators of HDAC activity are currently the focus of intense research and drug discovery efforts. Several inhibitors of class I and II (non-sirtuin) HDACs have shown promise as cancer therapeutics and are currently in clinical trials.^{(20, 21)} Other potential targets for class I/II HDAC inhibitors include neurodegenerative diseases such as amyotrophic lateral sclerosis, Huntington's disease and Alzheimer's disease.^(22-24) Activation of SIRT1 may represent a strategy to treat various age-related diseases,⁽²⁵⁾ including neurodegenerative disorders^{(26, 27)} and type II diabetes.⁽²⁸⁾

HDACs are typically found in large multi-protein complexes and are tightly regulated by subcellular localization, phosphorylation, and likely by other mechanisms. Therefore, a cellular assay that allows the determination of deacetylase activity within an undisturbed cellular environment is likely to provide more accurate activity information. Also, a cellular assay allows the study of the effects of upstream regulators on deacetylase activity and the detection of inhibitors or activators that act indirectly to affect deacetylase activity. Traditional methods for measuring cellular HDAC activity have required pulse labeling of cells with [³H]acetate, purification of histones and analysis by gel electrophoresis and autoradiography. ⁽²⁹⁾ Given the length and complexity of this procedure the development of convenient, quantitative, high-throughput-friendly assays for measuring cellular deacetylase activity could contribute significantly to progress in HDAC research.

Figure 2. Time Course of Fluor de Lys Substrate Deacetylation by HeLa Cells. HeLa cells were seeded at 4 × 104 cells per well in and grown overnight to 80% confluence. Cells were then incubated with 200 µM Fluor de Lys Substrate, µM Trichostatin A, and fluorescence determined as described in the text (AFU=Arbitrary Fluorescence units, CytoFluor II, Perseptive Biosystems, Ex. 360 nm, Em. 460 nm, gain 70). Each point represents the mean of three determinations, with error bars representing ± standard deviation.Click to enlarge.

While the advantages of the in situ cellular HDAC activity assays just discussed apply primarily to basic and pharmacological HDAC research, cellular HDAC activity assays may also have clinical diagnostic applications. For example, in extracts of lung and bronchial biopsies and alveolar macrophages from patients with chronic obstructive pulmonary disease (COPD), total HDAC activity decreases progressively with increased severity of the disease. ^{(30, 31)} Decreased HDAC activity was also observed in extracts of alveolar macrophages from asthma patients^{(31, 32)} and in those from healthy smokers relative to healthy non-smokers. ^{(30, 31)} All of these studies of HDAC activity in nuclear extracts of alveolar macrophages employed the fluorogenic acetyllysine substrate Fluor de Lys (BIOMOL International, Plymouth Meeting, PA), the substrate used in the cellular HDAC assay method described below. Since, in this technique, intact cells are incubated directly with the Fluor de Lys Substrate, the same results could presumably have been obtained without the necessity for nuclear extract preparation.

Method

The Fluor de Lys Substrate exhibits a number of characteristics that make it a useful reagent for assaying cellular deacetylase activity. It is a relatively small (a single acetyllysine labeled with a fluorophore), uncharged molecule, making it cell-permeable. It can be deacetylated by a broad range of deacetylases.^{(5, 25, 33-35)} Unlike histone-based assays, deacetylation of the Fluor de Lys Substrate can reveal activity of enzymes that act on non-histone substrates as well. Deacetylase assays with the Fluor de Lys Substrate are performed in two steps. First, the substrate is exposed to a source of deacetylase activity (cells, extracts, purified enzymes). Deacetylation sensitizes the substrate so that, in the second step, treatment with the Fluor de Lys Developer (BIOMOL International) produces a fluorophore (Fig. 1).

The design of the cellular assay (BIOMOL International) is based on two considerations, first that the Fluor de Lys Substrate is cell-permeable and second that Fluor de Lys Developer is not (see Figure 1). The simplest version of the assay therefore consists of adding substrate to cells, in culture medium, incubating for a desired length of time to allow substrate to be taken up and deacetylated, and then adding a buffer which contains detergent (1% NP-40) to lyse the cells, developer to generate a fluorescent signal from the deacetylated substrate, and inhibitor(s) to stop deacetylase activity (Figures 2 and 3).

Results and discussion

Figure 3. Fluor de Ly Substrate Deacetylation at Two Cell Densities. HeLa cells were seeded at either 0.5 × 104 or 2 × 104 cells per well and grown two days to the indicated confluences. Cells were then incubated with 200 µM Fluor de Ly Substrate, ? µM Trichostatin A, and fluorescence determined as described in the text (AFU=Arbitrary Fluorescence units, CytoFluor II, Perseptive Biosystems, Ex. 360 nm, Em. 460 nm, gain 85). Click to enlarge.

The HeLa cellular HDAC activity measured with this technique is time-dependent (Figures 2 and 3), with a signal that is linear for at least four hours (Figure 2), and proportional to the cell number (Figure 3). Approximately 95% of the total deacetylation rate, the fraction of the signal due to class I/II HDACs (non-sirtuins), is sensitive to 1 μM TSA.

Variations of this cellular assay technique have been used to assess the effects of modulators of both class I/II HDACs⁽³⁶⁾ and sirtuins.^{(25, 37-40)} The cultured cell types successfully assayed include HeLa,⁽²⁵⁾ H358 and H460 (non-small cell lung cancer lines), ⁽³⁸⁾ multiple neuroblastoma lines, ⁽³⁶⁾ HT29 (colon carcinoma),⁽³⁹⁾ primary vascular smooth muscle cells⁽⁴⁰⁾ and Drosophila S2 cells.⁽³⁷⁾ In two of the sirtuin studies, the single-lysine Fluor de Lys Substrate was replaced by a four amino acid substrate based on p53 residues 379-382 (Fluor de Lys-SIRT1, BIOMOL International).^{(39, 40)} This may be advantageous for the detection of total sirtuin activity, since SIRTs 1-3 deacetylate this substrate with 10-fold or more the rate of the single-lysine Fluor de Lys Substrate. On the other hand, since SIRT1 is the only human sirtuin displaying reasonable activity with the single-lysine Fluor de Lys Substrate, use of this substrate, in combination with TSA to eliminate the background from class I/II HDACs, provides an assay relatively specific for SIRT1.⁽²⁵⁾ Interestingly, SIRT1 expression levels in HT29, H358 and H460 cells are apparently high enough that addition of the SIRT1 activator resveratrol results in a significant increase in the total deacetylation signal, even in the absence of TSA.^{(38, 39)}

Conclusion

Taken together, these experiments (Figures 2 and 3) and the wide array of results from the literature demonstrate that the Fluor de Lys Substrate and Developer System provides a rapid, simple to perform assay capable of detecting deacetylase activity in a broad range of cultured cells. Importantly, this application of the system opens the door to high-throughput, cell-based inhibitor screening that, in contrast to assays with purified enzymes, has the potential for identifying novel therapeutics that act indirectly on HDAC activity.

About the author

Dr. Howitz, Director of Molecular Biology, BIOMOL International, Plymouth Meeting, PA, can be reached via e-mail at: howitzkt@biomol.com, or more information is available from:

References

1. Shahbazian, M. D. and Grunstein, M. Annu Rev Biochem (2007).
2. Ng, H. H. and Bird, A. Trends Biochem Sci 25:121-6 (2000).
3. de Ruijter, A. J. et al. Biochem J 370:737-49 (2003).
4. Imai, S. et al. Nature 403:795-800 (2000).
5. Bitterman, K. J. et al. J Biol Chem 277:45099-107 (2002).
6. Gregoretti, I. V. et al. J Mol Biol 338:17-31 (2004).
7. Frye, R. A. Biochem Biophys Res Commun 273:793-8 (2000).
8. Gray, S. G. and Ekstrom, T. J. Exp Cell Res 262:75-83 (2001).
9. Fischle, W. et al. Biochem Cell Biol 79:337-48 (2001).
10. Michishita, E. et al. Mol Biol Cell 16:4623-35 (2005).
11. Yang, T. et al. Trends Endocrinol Metab 17:186-91 (2006).
12. Haigis, M. C. et al. Cell 126:941-54 (2006).
13. Luo, J. et al. Nature 408:377-81 (2000).
14. Vaziri, H. et al. Cell 107:149-59 (2001).
15. Ashburner, B. P. et al. Mol Cell Biol 21:7065-77 (2001).
16. Chen, L. et al. Science 293:1653-7 (2001).
17. Matsuyama, A. et al. Embo J 21:6820-31 (2002).
18. Hubbert, C. et al. Nature 417:455-8 (2002).
19. Polevoda, B. and Sherman, F. Genome Biol 3, reviews0006 (2002).
20. Bolden, J. E. et al. Nat Rev Drug Discov 5:769-84 (2006).
21. Monneret, C. Anticancer Drugs 18:363-70 (2007).
22. Hockly, E. et al. Proc Natl Acad Sci 100:2041-6 (2003).
23. Corcoran, L. J. et al. Curr Biol 14:488-92 (2004).
24. Rouaux, C. et al. Biochem Pharmacol 68:1157-64 (2004).
25. Howitz, K. T. et al. Nature 425:191-6 (2003).
26. Araki, T. et al. Science 305:1010-3 (2004).
27. Anekonda, T. S. and Reddy, P. H. J Neurochem 96:305-13 (2006).
28. Picard, F. et al. Nature 429:771-6 (2004).
29. Yoshida, M. et al. J Biol Chem 265:17174-9 (1990).
30. Cosio, B. G. et al. J Exp Med 200:689-95 (2004).
31. Ito, K. et al. N Engl J Med 352:1967-76 (2005).
32. Cosio, B. G. et al. Am J Respir Crit Care Med 170:141-7 (2004).
33. Zhou, X. et al. Proc Natl Acad Sci 98:10572-7 (2001).
34. Kapustin, G. V. et al. Org Lett 5:3053-6 (2003).
35. Gurvich, N. et al. Cancer Res 64:1079-86 (2004).
36. de Ruijter, A. J. et al. Biochem Pharmacol 68:1279-88 (2004).
37. Wood, J. G. et al. Nature 430:686-9 (2004).
38. Yeung, F. et al. Embo J 23:2369-80 (2004).
39. de Boer, V. C. et al. Mech Ageing Dev 127:618-27 (2006).
40. van der Veer, E. et al. J Biol Chem (2007).