Note to users. If you're seeing this message, it means that your browser cannot find this page's style/presentation instructions -- or possibly that you are using a browser that does not support current Web standards. Find out more about why this message is appearing, and what you can do to make your experience of our site the best it can be.

Science 2 June 2006:
Vol. 312. no. 5778, p. 1312
DOI: 10.1126/science.1124767
Prev | Table of Contents | Next

Technical Comments

Response to Comment on "HST2 Mediates SIR2-Independent Life-Span Extension by Calorie Restriction"

Dudley W. Lamming,1 Magda Latorre-Esteves,1 Oliver Medvedik,1 Stacy N. Wong,2 Felicia A. Tsang,2 Chen Wang,2 Su-Ju Lin,2* David A. Sinclair1*

Our two labs and others have shown that SIR2 controls the life span of diverse species, including Saccharomyces cerevisiae and Drosophila melanogaster, and that deleting SIR2 blocks life-span extension by calorie restriction. The methods of Kaeberlein et al. allow yeast to bypass the requirement for SIR2 and its homologs, which brings into question their suitability for modeling the physiology of more complex organisms.

1 Paul F. Glenn Laboratories, Department of Pathology, Harvard Medical School, 77 Avenue Louis Pasteur, Boston, MA 02115, USA.
2 Center for Genetics and Development and Section of Microbiology, University of California, Davis, 351 Briggs Hall, Davis, CA 95616, USA.

* To whom correspondence should be addressed. E-mail: slin{at}ucdavis.edu (S.-J.L.); david_sinclair{at}hms.harvard.edu (D.A.S.)

We believe it is important to understand why experiments performed by Kaeberlein et al. (1, 2) allow yeast to bypass the requirement for SIR2, HST1, and HST2 during calorie restriction (CR) and how relevant this observation is to complex organisms. The simplest explanation is that there are at least two pathways for life-span extension in Saccharomyces cerevisiae: a sirtuin-dependent pathway and a sirtuin-independent pathway, the latter of which is invoked by the methods of Kaeberlein et al., which induce more intense nutrient restriction.

The life-span extensions shown by Kaeberlein et al. (1) were obtained using a protocol that differs substantially from ours (3–5). Although the authors state that reducing glucose by 97.5% is standard methodology for inducing CR, we are unaware of any other lab that uses this protocol. While we can debate the merits of reducing glucose to this extent, and have done so previously (5, 6), the fact that Kaeberlein et al. do not observe sirtuin-dependent effects of CR suggests that their methods may be unsuitable for understanding the role sirtuins play in mediating CR in more complex organisms (7, 8).

In our report (9), we performed the key experiments in at least two yeast strains, and data from our two laboratories were, and continue to be, consistent. In contradiction to Kaeberlein et al., we find that 0.5% glucose produces a negligible life-span extension in BY4742 sir2{Delta} fob1{Delta} hst1{Delta} hst2{Delta} (Fig. 1A). A clue as to why the life-span data of Kaeberlein et al. contradict not only our two labs but also data from Guarente and Goldfarb (3, 10) may be found in the observation that W303 does not live longer when subjected to severe nutrient limitation (Fig. 1B). The inability of Kaeberlein et al. to observe CR-mediated life-span extension in W303 indicates that their protocol induces greater nutrient stress on cells than does ours. Another clue is that even on 2% glucose medium, BY4742 cells live an average of 13 to 20% longer in the hands of Kaeberlein et al. than in ours [compare our Fig. 1A with figure 1, A and B, in (1)], which indicates that their protocol induces a greater degree of nutrient stress and induces longevity pathways even on 2% glucose.


Figure 1 Fig. 1. Effects of CR on yeast life span and rDNA recombination. (A) Calorie restriction (0.5% glucose) does not extend life span in BY4742 sir2{Delta} fob1{Delta} hst2{Delta} hst1{Delta} strain, one-tenth of that glucose concentration does. Life-span analyses were performed as previously described (8). Average life span: 2% glucose, 22.4; 0.5% glucose, 21.1; 0.05% glucose, 27.5. (B) W303 responds to CR but is sensitive to more intense glucose restriction. Average life spans: W303AR5 2% (w/v) glucose, 23.9; 0.1% glucose, 24.9; W303AR5 sir2{Delta} fob1{Delta} 2% glucose, 26.1; 0.1% glucose, 28.0. (C) Sir2 and Hst2 are recruited to the rDNA during CR (16). (D) CR suppresses rDNA recombination in the absence of SIR2. Ribosomal rDNA recombination rates were determined as previously described using a half-sector–based colony assay (8). Cells were maintained in log phase for 8 hours in the indicated medium before plating to YPD medium. [View Larger Version of this Image (36K GIF file)]
 

We thought it self-evident that to make assertions about the effect of CR on ribosomal DNA (rDNA) recombination, one would need to measure it. Despite the authors' strong claims about a variety of genes being relevant or irrelevant to rDNA recombination (11), they do not measure either rDNA recombination or extrachromosomal rDNA circles (ERCs). Measures of telomeric silencing cited by Kaeberlein et al. (1) are misleading because of known locus-specific differences in how Sir2 activity is regulated (12) and the increasing number of examples where silencing of a marker gene at the rDNA does not correlate with recombination (13, 14). By measuring rDNA recombination, we consistently observe that it is reduced by half in response to 0.5% glucose, and that this reduction is abrogated by deleting SIR2 and HST2 and completely eliminated by nicotinamide, a sirtuin inhibitor (9). Moreover, there is a twofold increase in the abundance of Sir2 at the NTS2 region of the rDNA during CR, and in the absence of Sir2, Hst2 is recruited there (Fig. 1C), consistent with our model (9).

Kaeberlein et al. ask why Hst1 and Hst2 do not offset a loss of SIR2. Loss of SIR2 results in hyper-recombination, which is countered by Hst1 and Hst2 during CR, but not to a level that extends life span significantly (Fig. 1D). When both SIR2 and FOB1 are deleted, thus lowering the amount of rDNA recombination to near wild-type levels, the effects of Hst1 and Hst2 are more apparent. As we stated, Hst1 and Hst2 are thought to augment the activity of Sir2 during CR, and their activity is more evident when SIR2 and FOB1 are deleted.

Kaeberlein et al. state that our work contradicts the work of the Gasser lab (15), but the two studies are not directly comparable. The Gasser lab did not study CR, did not measure the effects of HST2 on life span, and used a less sensitive assay for measuring rDNA recombination.

Lastly, Kaeberlein et al. have found conditions under which life span can be extended by CR in the absence of sirtuins and have concluded that they play no role. However, it is an accepted rule of genetics that if one finds that a phenotype can occur in the absence of a gene, it does not mean that said gene plays no role. A more apt interpretation is that there is an alternative pathway that can be activated under certain conditions. We look forward to future discoveries about the complexities and redundancies that underlie life-span extension by CR in yeast and multicellular organisms.


References and Notes

Received for publication 6 February 2006. Accepted for publication 8 May 2006.

To Advertise     Find Products

ADVERTISEMENT

Featured Jobs

Science. ISSN 0036-8075 (print), 1095-9203 (online)