We undertook this study to assess the possible association of changes in the transcriptional expression of individual sirtuin genes with breast cancer. Firstly, we investigated sirtuin gene expression in both primary and immortalised cells with a view to determining how cellular ageing and immortalisation influenced their transcription profile. Consequently, we observed SIRT3 and SIRT7 transcription to increase in primary cells as they approach senescence (), concurrent with an increase in p16INK4a and p21CIP1 expression () and the accumulation of SA-β-Gal-containing cells (). Secondly, we demonstrated a significant in vivo association between altered SIRT3 and SIRT7 transcription and malignant breast disease (, and ).
A link between the transcriptional expression of genes involved in biological ageing and cancer is pertinent. Individual sirtuins are respondents to intracellular redox changes and cell stress and hence biological ageing. In yeast, SIR2
provides a functional link between the telomere and ribosomal DNA cluster integrity, thus impacting on ribosome production and enabling the cell to respond to stress induced redox state changes in the mitochondrion (Guarente, 2000
). Consequently, damage responses in the cell can be modulated to match the degree of insult. In man, individual sirtuins may provide separate functional components of this process (Shiels and Davies, 2003
). This is intriguing, in the light of the observations that SIRT7
localises to the nucleolus and SIRT3
localises to the mitochondrion (Onyango et al, 2002
; Michishita et al, 2005
). These subcellular locations, along with the telomere nucleoprotein complex, provide the components of a functional trinity, we have termed the MTR, that senses, assesses and signals damage (Shiels and Davies, 2003
; Shiels and Jardine, 2003
). Our observations that SIRT3
expression is increased as primary mammary epithelial cells approach senescence and are also increased in node-positive breast cancer is in keeping with the MTR hypothesis.
expression is invariably elevated in thyroid cancer cell lines and biopsies, increased expression of SIRT7
may represent an important step in malignant transformation (de Nigris et al, 2002
; Frye, 2002
). Indeed, SIRT7
has recently been shown to activate Pol I encouraging growth and proliferation (Ford et al, 2006
). In this study, we have shown that increased SIRT7
expression is also observed in breast cancer, and may therefore be postulated, in a similar manner, to be required for malignant transformation (). Furthermore, SIRT7
expression is also associated with nodal invasion and therefore locally aggressive disease (). Whether this increased expression contributes to the increased proliferative potential of mammary cancer cells or reflects a molecular change necessary for breast cancer tumorigenesis is undetermined. Alternatively, it may reflect a compensatory mechanism as a consequence of the increased proliferative state. This is supported by observations on the activity of the yeast sirtuin orthologue, SIR2
. Nucleolar SIR2
activity is a limiting factor for determining when the cell undergoes senescence (Guarente, 1999
). The high proliferative rate in many tumours often results in them having shortened telomeres as a consequence of rapid cell turnover outstripping telomerase-mediated telomere repair. This may, in analogy with the situation in yeast (Guarente, 1999
), lead to an upregulation of sirtuin activity in an attempt to stabilise the nucleolus. The increased SIRT3
transcription in primary mammary epithelial cells, as a consequence of increased oxidant load with in vitro
growth, is supportive of such a scenario. The association of elevated SIRT7
expression in node-positive tumours, which have a greater recurrence and poorer survival, suggests that this gene may prove to be a good marker of disease progression and tumour behaviour.
There was no overall difference observed in SIRT3
expression between breast cancer and normal breast tissue. However, on subdividing the cancer biopsies by nodal status, SIRT3
expression, like that of SIRT7
, was greater in node-positive breast cancer compared to normal breast tissue (). SIRT3
has been shown to specifically targeted and converted into its active form within the mitochondria (Dryden et al, 2003
; North et al, 2003
). Cumulative mitochondrial damage contributes to a fall in relative nicotinamide adenine dinucleotide (NAD) levels, and concomitant fall in SIRT3
activity, and is associated with growth arrest, senescence and apoptosis (Shiels and Davies, 2003
). Analogous to SIRT1
may also function to provide a growth and survival advantage. Indeed, variability of the SIRT3
gene has been linked to survival in the elderly (Rose et al, 2003
). The increased expression of SIRT3
seen in lymph node-positive tumours may, therefore, contribute to survival of these more aggressive tumours.
Our data showed that the expression of SIRT1
did not differ between normal and malignant breast biopsies. This suggests that transcriptional changes for SIRT1 may not be implicated in breast cancer pathogenesis. This observation may be pertinent to the p53
status of breast cancers, as p53
is a substrate for SIRT1
(Luo et al, 2001
). Unlike the upregulation of telomerase, abnormalities in p53
are seen in only 20–40% of breast cancers, suggesting that disruption of p53
pathways may be of lesser importance in breast cancer pathogenesis (Mokbel et al, 2000
; Borresen-Dale, 2003
). These data do not, however, negate a role for SIRT1
in breast cancer pathogenesis, as we looked at overall sirtuin transcriptional expression levels and not at any putative post-transcriptional regulation.
The lack of transcriptional change observed for SIRT2
in this study may be a reflection of a cell cycle dependency in its expression (Dryden et al, 2003
; North et al, 2003
). It may be that more dynamic evaluation of changes in SIRT2
expression at specific stages of the cell cycle may reveal abnormal patterns of SIRT2
expression in breast cancer.
The clinical significance of the differences we have observed in sirtuin expression needs further investigation. Our study has indicated the potential utility of sirtuins as prognostic markers in breast cancer. The observation that SIRT7 is increased in breast cancer tissue compared to normal breast tissue suggests that it may be related to breast cancer tumorigenesis. The discovery of a greater expression of SIRT3 and SIRT7, within the prognostically poorer lymph node-positive breast cancers, suggests that overexpression of these sirtuins may be related to dissemination.
In summary, the heterogeneity of breast cancer behaviour is paralleled and may be explained by sirtuin expression. Specifically, molecular changes in SIRT3 and SIRT7 expression may contribute to tumour development and disease progression. This study shows that the study of sirtuins has a potential application in breast cancer diagnosis and prognosis, as well as in understanding disease biology.