Regulation of cyclin expression and cell cycle progression in ... - Nature

Oncogene (2005) 24, 381–389

& 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00 www.nature.com/onc

Regulation of cyclin expression and cell cycle progression in breast epithelial cells by the helix–loop–helix protein Id1 Alexander Swarbrick1,2, Mia C A˚kerfeldt1, Christine SL Lee1, C Marcelo Sergio1, C Elizabeth Caldon1, Lisa-Jane K Hunter1, Robert L Sutherland1 and Elizabeth A Musgrove*,1 1

Cancer Research Program, Garvan Institute of Medical Research, St Vincent’s Hospital, 384 Victoria St, Darlinghurst, Sydney, NSW 2010, Australia

The helix–loop–helix protein Id1 has been implicated in regulating mammary epithelial cell proliferation and differentiation but the underlying molecular mechanisms are not well characterized. Under low serum conditions, ectopic expression of Id1, but not Id2, allowed continued proliferation of immortalized mammary epithelial cells and breast cancer cells. Conversely, downregulation of Id1 impaired proliferation. The effects of short interfering RNA (siRNA)-mediated downregulation of Id1 were the same as those following downregulation of c-Myc: decreased expression of cyclins D1 and E, reduced phosphorylation of pRb at Ser780 (a site targeted by cyclin D1–Cdk4) and reduced cyclin E–Cdk2 activity. Decreased cyclin D1 expression was an early response to Id1 antisense oligonucleotide treatment. Inhibition of cMyc function by siRNA, antisense oligonucleotides or a dominant repressor resulted in downregulation of Id1, while ectopic expression of c-Myc resulted in rapid induction of Id1, suggesting that Id1 may be downstream of c-Myc. These data indicate that in mammary epithelial cells, Id1 has cell cycle regulatory functions that are similar to those of c-Myc, and suggest that cyclin D1 may be involved in Id1 regulation of cell cycle progression. Oncogene (2005) 24, 381–389. doi:10.1038/sj.onc.1208188 Published online 18 October 2004 Keywords: cell cycle; Id1; breast cancer

Introduction Basic helix–loop–helix transcription (bHLH) factors are ubiquitous regulators of a range of cellular functions. Through heterodimerization and binding to ‘E box’ sequences, they control the expression of genes involved in lineage commitment and cell fate determination, proliferation and survival (Massari and Murre, 2000). *Correspondence: EA Musgrove; E-mail: [email protected] 2 Current address: GW Hooper Foundation, University of California, Box 0552, HSW 1531, 513 Parnassus, San Francisco, CA 94143-0552, USA Received 6 August 2004; revised 6 September 2004; accepted 14 September 2004; published online 18 October 2004

Members of the ‘inhibitor of differentiation’ (ID) protein family possess a helix–loop–helix domain sufficient for dimerization to a subset of bHLH factors but lack the DNA binding domain required for transcriptional activity, and thus act as dominant repressors of bHLH function. Expression of ID proteins is sensitive to extracellular stimuli such as growth factors and cytokines, and ID proteins display distinct spatiotemporal expression patterns, suggesting a requirement for ID proteins in specific developmental and physiological processes (Norton, 2000; Sikder et al., 2003). Increasing evidence implicates ID family members, particularly Id1, in breast epithelial cell proliferation, differentiation and oncogenesis (Desprez et al., 2003). In breast cancer cells, Id1 is induced by estrogen, and estrogen-induced mitogenesis is impaired when Id1 levels are decreased (Lin et al., 2000). Conversely, progesterone inhibition of proliferation is accompanied by decreased Id1 expression, and constitutive expression of Id1 is sufficient to induce progestin resistance in breast cancer cells (Lin et al., 2000) and impair differentiation of mammary epithelial cells in culture (Desprez et al., 1995). Serum-dependent Id1 expression is deregulated in aggressive breast cancer cell lines (Lin et al., 2000), suggesting that deregulation of Id1 expression may be associated with mammary oncogenesis. Consistent with this possibility, the Id1 locus at chromosome 20q11 is amplified in breast carcinomas (Guan et al., 1996; Tanner et al., 1996), and a recent study indicates that Id1 is overexpressed in breast cancers and this confers poor outcome (Schoppmann et al., 2003). In vitro, Id1 can stimulate breast epithelial cell invasiveness (Desprez et al., 1998), and expression of Id1, but not Id2, correlates with aggressiveness in a panel of breast cancer cell lines (Desprez et al., 1998). Similarly, Id1 expression increases in invasive breast cancers in comparison with ductal carcinoma in situ but Id2 expression appears to be reduced (Lin et al., 2000; Itahana et al., 2003). Id1 is required for the G0/G1–S phase transition in fibroblasts (Hara et al., 1994), but there is relatively little information on the mechanisms by which it regulates cell cycle progression. Unlike other ID proteins, Id1 does not regulate the activity of the retinoblastoma family gene products – Rb, p107 and p130 – by direct binding (Iavarone et al., 1994). However, Id1 can

Id1 effects on breast epithelial cell proliferation A Swarbrick et al

382

repress the expression of p16INKA (Alani et al., 2001; Ohtani et al., 2001). The ID target MyoD activates p21WAF1/Cip1 gene expression in myoblasts (Halevy et al., 1995) and its partner E2A positively regulates p21WAF1/Cip1 transcription in fibroblasts (Peverali et al., 1994). This regulation of p21WAF1/Cip1 by E2A is antagonized by Id1 (Prabhu et al., 1997), suggesting that Id1 may govern proliferation via antagonism of E2A-dependent p21WAF1/Cip1 expression. In this study, we have addressed the role of Id1 in the proliferation of immortalized and transformed mammary epithelial cells. We show that Id1 is required for expression of cyclin D1 and cyclin E, for activation of their associated cyclin-dependent kinases (CDKs) and for proliferation in breast cancer cells. Decreased expression of cyclin D1 is an early response to downregulation of Id1 and is thus implicated in the proliferative arrest following decreased Id1 expression. These effects are strikingly similar to the effects of decreased c-Myc expression in the same cells (Carroll et al., 2002, this manuscript). c-Myc is centrally involved in steroidal control of breast cancer proliferation (Sutherland et al., 1998). Like Id1, it is induced by estrogen (Dubik et al., 1987), and c-Myc antisense oligonucleotides impair estrogen-induced proliferation (Watson et al., 1991). The steroid regulation of c-Myc and Id1 and the commonality of their involvement in breast cancer proliferation raised the possibility that cMyc might regulate Id1 and we present evidence suggesting that Id1 may be downstream of c-Myc in these cells.

Results Id1, but not Id2, stimulates proliferation of mammary epithelial cells in low serum Mammary epithelial cells and hormone-responsive breast cancer cells are serum-dependent in culture, and Id1 expression is tightly regulated by serum (Hara et al., 1994; Tournay and Benezra, 1996); so we tested whether Id1 overexpression altered the response of these cells to serum deprivation as an initial approach to investigating Id1 regulation of proliferation in mammary epithelial cells. In 0.5% serum, cultures of vector-transfected HC11-immortalized murine mammary epithelial cells displayed low levels of Id1 expression and bromodeoxyuridine (BrdU) incorporation relative to those in 5% serum (Figure 1a). In contrast, in low serum, HC-11 cells overexpressing Id1 exhibited levels of Id1 expression and proliferation approaching those of control cells grown in high serum (Figure 1a). Similarly, in vector control MCF-7 human breast cancer cells, Id1 expression and BrdU incorporation were both decreased in low serum but cells overexpressing Id1 continued to proliferate in low serum (Figure 1b). To determine whether the related HLH protein Id2 could also stimulate serum-independent proliferation of mammary epithelial cells, we measured DNA synthesis in HC-11 cells overexpressing Id2. HC-11 cells were infected with an Id2-IRES-GFP retrovirus and cells Oncogene

Figure 1 Id1, but not Id2, is sufficient for continued proliferation after serum deprivation. (a) HC-11-Id1 (Id1) and HC-11-LXSN (Vector) cells were cultured in either high (5%, solid bars) or low (0.5%, open bars) serum for 48 h. BrdU incorporation was measured using flow cytometry and is presented as the mean and range of two independent experiments. The endogenous murine Id1 (endo) has slightly different mobility from the transfected human Id1 (exo) in Western blots. The faint band in vector-transfected cells with similar mobility to the transfected Id1 appears to be nonspecific. (b) MCF-7-Id1 (Id1) and MCF-7-Babe cells (Vector) were cultured in 0.5% serum (open bars) or 5% serum (solid bars) for 48 h. BrdU incorporation measured using flow cytometry in a representative experiment is shown. (c) HC-11 cells infected with MSCV-Id2-IRES-GFP retrovirus were FACS-sorted into GFPpositive and -negative populations and the purified cell populations used for Western blotting or for measurement of BrdU incorporation using flow cytometry. BrdU incorporation is presented as the mean7s.e.m. of four independent experiments. Open bars: 0.5% serum; solid bars: 5% serum. The Western blot is from cells cultured in 5% serum

overexpressing Id2 were separated on the basis of GFP positivity, using flow cytometry. Western blotting confirmed higher Id2 expression in the GFP-positive fraction compared with the GFP-negative fraction (Figure 1c). There was, however, no increase in BrdU incorporation in cells overexpressing Id2 in either 5 or 0.5% serum (Figure 1c). Overall, these data indicate that overexpression of Id1, but not Id2, allows continued proliferation in these cells under low-serum conditions.


383

Id1 is required for proliferation of mammary epithelial cells To investigate the requirement for Id1 expression in mammary epithelial cell proliferation, we reduced Id1 expression using small interfering RNAs (siRNAs). A number of 21-nucleotide siRNAs targeted to Id1 were designed using published criteria (Elbashir et al., 2002). After preliminary tests of efficacy, two distinct siRNAs were chosen for further experiments, one targeting the coding region and the other the 30 UTR. Transfection of MCF-7 cells with these Id1 siRNAs (Id1-12 and Id1-17) resulted in downregulation of Id1 protein within 48 h of transfection (Figure 2a). In replicate experiments, Id1 levels were reduced to an average of 47% for Id1-12 and 56% for Id1-17, relative to cells treated with control siRNAs (see Figure 6a). In initial experiments, we noticed a close similarity between the effects of reduction in Id1 expression and those we had previously reported using antisense oligonucleotides to reduce cMyc expression (Carroll et al., 2002), and therefore directly compared the effects of Id1 and c-Myc siRNAs in the experiments presented here. To decrease c-Myc expression, we used a commercial siRNA targeting the c-Myc 30 UTR (Ambion, referred to here as ‘MycUTR’) and another targeting the central portion of the c-Myc open reading frame (Myc940) (von der Lehr et al., 2003). Transfection with these c-Myc siRNAs reduced cMyc expression to 51% (MycUTR) and 54% (Myc940) of control within 48 h (Figure 2b, see also Figure 6a). In MCF-7 cells transfected with either Id1 or c-Myc siRNAs, DNA synthesis was reduced by approximately one-third within 24–48 h of transfection (Figure 2c and data not shown). Dual parameter BrdU/DNA content analysis by flow cytometry indicated that both c-Myc and Id1 siRNAs predominantly led to accumulation of cells in G1 phase, although a minor population of BrdUnegative cells with an S phase DNA content was apparent (data not shown). To compare the long-term effects of Id1 or c-Myc deprivation, we examined the effect of siRNA transfection on cell number over 5 days. Cells transfected with a GFP control siRNA proliferated with a doubling time of approximately 1.2 days until day 3, when proliferation began to slow as the cells approached confluence. In contrast, cells transfected with Id1-12, Id1-17 or Myc940 siRNAs proliferated more slowly, with a doubling time of approximately 3 days and consequent decrease in relative cell number of approximately 50% by day 3 (Figure 2d). To determine whether reductions in proliferation following Id1 ‘knockdown’ were associated with reduced CDK activity, the phosphorylation of pRb at Ser780, a site targeted exclusively by cyclin D1–Cdk4 (Kitagawa et al., 1996) and cyclin E–Cdk2 activity, was measured in lysates harvested from cultures transfected with Id1, c-Myc or control siRNAs. Transfection of MCF-7 cells with Id1 or c-Myc siRNA was associated with a significant reduction in the level of pRb-Ser780 phosphorylation and the in vitro kinase activity of cyclin E–Cdk2 complexes (Figure 3a and b). In replicate experiments, decreased expression of c-Myc reduced

Figure 2 Id1 or c-Myc downregulation inhibits proliferation in MCF-7 breast cancer cells. MCF-7 cells were transfected with siRNAs to Id1, c-Myc or controls (GFP, GAPDH). (a–c) Cells were harvested for Western blotting (a, b) or measurement of BrdU incorporation (c) 48 h after transfection. Data in (c) are the mean7s.e.m. of five experiments. (d) MCF-7 cells were replated into 96-well plates 24h after siRNA transfection and relative cell number was estimated using a colorimetric assay 1, 3 and 5 days after replating. Data are the mean7s.e.m. (where this is larger than the symbol used) of 6–12 wells in a representative experiment

pRb-Ser780 phosphorylation by approximately 65% and cyclin E–Cdk2 activity by approximately 50% (Figure 3c). Similarly, decreased expression of Id1 reduced pRb-Ser780 phosphorylation by up to 75% of control and reduced cyclin E–Cdk2 activity by approximately 40% (Figure 3c). We next turned our attention to identifying the mechanisms by which Id1 might regulate CDK activity in this context. Id1 has been proposed to regulate proliferation and senescence in other tissues by repression of the CDK inhibitors p16INK4A and p21WAF1/Cip1 (Peverali et al., 1994; Prabhu et al., 1997; Alani et al., 2001; Ohtani et al., 2001) and it might therefore be Oncogene


384

Figure 3 Id1 downregulation reduces CDK activity in MCF-7 breast cancer cells. (a, b) Whole-cell extracts were prepared 48 h following transfection of MCF-7 cells with siRNAs to Id1, c-Myc or controls (GFP, GAPDH) and phosphorylation of pRb at Ser780 was assessed by Western blot (a) or in vitro kinase activity of cyclin E immunoprecipitates was determined (b). (c) Relative pRb-Ser780 phosphorylation (open bars) and cyclin E–Cdk2 activity (closed bars). Data are the mean7s.e.m. of three experiments

expected that decreased Id1 expression would lead to increases in their expression. The INK4A gene is deleted in MCF-7 cells (Musgrove et al., 1995), indicating that in these cells induction of p16INK4A is not an essential element of the decreased proliferation following decreased Id1 expression. Examination of the expression of p21WAF1/Cip1 and the related CDK inhibitor p27Kip1 following transfection of siRNAs targeting Id1 did not reveal an increase in either p21WAF1/Cip1 or p27Kip1 expression (Figure 4a). However, a clear reduction in cyclin D1 levels and more modest reduction in cyclin E levels were apparent (Figure 4a). Transfection with cMyc siRNAs led to the same responses: no increase in either p21WAF1/Cip1 or p27Kip1, decreased cyclin D1 expression and more modest effects on cyclin E expression. Quantitation of data from replicate experiments showed that cyclin E expression decreased by up to B50% but cyclin D1 protein levels decreased by B75% after reduction in either Id1 or c-Myc levels (Figure 4b). To confirm these data using another approach and to examine more acute effects of Id1 downregulation, we used a previously characterized Id1 antisense oligonucleotide (Hara et al., 1994). Within 6 h of treatment with the Id1 antisense oligonucleotide, Id1 expression decreased by approximately 70% relative to that in MCFOncogene

Figure 4 Effects of Id1 and c-Myc siRNAs on cyclin and CDK inhibitor expression. (a, b) MCF-7 cells were transfected with siRNAs to Id1, c-Myc or controls (GFP, GAPDH) for 48 h and whole-cell extracts were analysed by Western blotting (a). (b) Cyclin D1 (open bars) and cyclin E (solid bars) protein expressions are presented as the mean7s.e.m. of four experiments

7 cells transfected with scrambled control oligonucleotides or mock-transfected cells (Figure 5a) and by 16 h of treatment DNA synthesis was decreased to 33% of controls (Figure 5b). At the early timepoint, 6 h, there was no apparent change in the levels of p21WAF1/Cip1, p27Kip1 or cyclin E, but cyclin D1 expression was decreased by B60% (Figure 5a). Phosphorylation of pRb at Ser780 was also reduced (Figure 5c). These data indicate that cyclin D1 downregulation and decreased cyclin D1–Cdk4 activity are rapid responses to decreased Id1 expression. Altered c-Myc expression is associated with altered Id1 expression Evidence implicating both c-Myc and Id1 in steroid regulation of proliferation (see Introduction) raised the


385

Figure 5 Effects of Id1 antisense oligonucleotides on cyclin and CDK inhibitor expression. (a–c) MCF-7 cells were transfected with Id1 antisense or control scrambled (Scr) oligonucleotides. Wholecell extracts were analysed by Western blotting 6 h following transfection (a, c) or BrdU incorporation was measured using flow cytometry 16 h following transfection (b, mean7range of two experiments)

possibility that they might act through similar pathways, which was confirmed by data presented in Figures 2–4 identifying close parallels between the effects of decreased Id1 expression and decreased c-Myc expression. We therefore examined Id1 levels when c-Myc expression was decreased and vice versa. Overall, siRNA-mediated downregulation of Id1 in MCF-7 cells did not result in a significant decrease in c-Myc levels, but the B50% decrease in c-Myc expression 48 h after siRNA treatment was accompanied by a similar decrease in Id1 expression (Figures 4a and 6a). This result was confirmed using c-Myc antisense oligonucleotides that we had previously demonstrated led to rapid, specific decreases of c-Myc expression in MCF-7 cells (Carroll et al., 2002). Consistent with our previous data (Carroll et al., 2002), c-Myc expression was substantially reduced within 6 h of antisense oligonucleotide treatment and BrdU incorporation was reduced to B30% of control at 18 h (Figure 6b and data not shown). Importantly, expression of Id1 was reduced within 6 h (Figure 6b), suggesting that the decrease in Id1 expression is not simply a consequence of arrested proliferation following decreased c-Myc expression. In a complementary approach, T-47D human breast cancer

Figure 6 c-Myc downregulation reduces Id1 expression. (a) Whole-cell extracts were prepared 48 h following transfection of MCF-7 cells with c-Myc, Id1 or control (GFP, GAPDH) siRNAs, analysed by Western blotting with the indicated antibodies and quantitated. Id1 expression: open bars, mean7s.e.m. of 3–5 experiments; c-Myc expression: solid bars, mean7s.e.m. of 4–5 experiments. (b) Whole-cell extracts were prepared 6 h following transfection of MCF-7 cells with c-Myc antisense (AS) or sense control (Sen) oligonucleotides and analysed by Western blotting with the indicated antibodies. (c) T-47D human breast cancer cells were infected with pLib-CD8-Maxrep retrovirus encoding the Max-rep synthetic repressor of c-Myc, or a vector control virus and 48 h later fractionated using flow cytometry, based upon expression of CD8. Whole-cell extracts were prepared from these sorted populations and subjected to Western blotting with antibodies as indicated. Max-rep is detected using an anti-c-Max polyclonal antibody

cells were infected with a retrovirus encoding Max-rep, a dominant repressor of c-Myc function (Schreiber-Agus et al., 1995) and the surface marker CD8. Cells expressing Max-rep were separated on the basis of CD8 positivity, using flow cytometry. They expressed significantly reduced levels of Id1 compared with vectorinfected, sorted control cells or CD8-negative cells from the same culture (Figure 6c). They also had reduced clonogenic potential and S phase fraction, consistent with inhibition of c-Myc function (data not shown). These data indicate that the dependence of Id1 expression on c-Myc is not confined to MCF-7 cells. To examine whether induction of c-Myc was accompanied by induction of Id1, we used MCF-7 cells engineered to inducibly express c-Myc under the control of the metallothionein promoter (MCF-7-DMT-Myc) (Prall et al., 1998). Addition of ZnSO4 to these cells elevated c-Myc mRNA within 2 h (Figure 7a; Prall et al., 1998) and this was closely followed by increased Oncogene


386

that increased c-Myc levels were associated with increases in Id1 expression using T-47D cells constitutively expressing c-Myc (Figure 7d). Together, these data suggest that alterations in c-Myc levels lead to rapid changes in Id1 levels in breast epithelial cells.

Discussion

Figure 7 c-Myc induction increases Id1 expression. (a, b) MCF7DMT-Myc (black bars) and MCF-7DMT control cells (white bars) cultured in medium containing 1% serum for 48 h were treated with 40 mM ZnSO4 to induce c-Myc, and both c-Myc (a) and Id1 (b) mRNA levels were determined by reverse transcription–real-time PCR at various timepoints. The mean7range of two experiments are presented relative to the 0 h value. (c) MCF-7-DMT-Myc (cMyc) and MCF-7DMT control (Vector) cells were cultured in medium containing 1% serum and then treated with 40 mM ZnSO4 or water control for 7 h. Whole-cell lysates were prepared and subjected to Western blotting with the indicated antibodies. NS, nonspecific band indicating equivalent loading. (d) Whole-cell lysates prepared from T-47D-Myc cells constitutively expressing cMyc and vector control cells (T-47D-Babe) cultured in 0.4% serum were Western blotted with the indicated antibodies. NS, nonspecific band indicating equivalent loading

expression of Id1 mRNA, which was apparent within 4 h of zinc addition (Figure 7b). The level of Id1 protein also increased within 7 h of c-Myc induction (2.7-fold in the experiment shown in Figure 7c). Id1 expression was not significantly altered by zinc treatment of empty vector control cultures (Figure 7b and c). We confirmed Oncogene

The discovery that Id2 binds and inactivates Rb has dominated recent thinking about regulation of proliferation by ID family members. However, Id1 cannot bind Rb, raising the question of how it interfaces with the cell cycle. We show here that downregulation of Id1 leads to downregulation of cyclin D1 and cyclin E, implicating regulation of cyclin expression in Id1 effects on proliferation in breast epithelial cells. Although in some model systems p16INK4A and p21WAF1/Cip1 are Id1 targets (Peverali et al., 1994; Prabhu et al., 1997; Alani et al., 2001; Ohtani et al., 2001), regulation of these CDK inhibitors does not appear to account for Id1 regulation of proliferation in breast epithelial cells. The INK4A gene encoding p16INK4A is deleted in MCF-7 cells (Musgrove et al., 1995) but they respond to alterations in Id1 expression, indicating that p16INK4A regulation is not required for Id1 to regulate proliferation in this context. Similarly, p21WAF1/Cip1 and p27Kip1 are unlikely to mediate decreased proliferation following decreased Id1 expression in mammary epithelial cells since the expression of these CDK inhibitors was not increased. Decreases in Id1 expression led to downregulation of cyclin D1, cyclin E and their associated kinase activities, and inhibition of proliferation. The correspondence between the magnitude of the decrease in expression of these cyclins and the decreased activity of their associated CDKs suggests regulation of cyclin expression as a likely mechanism for the decrease in CDK activity. In addition, although decreased cyclin D1 levels in breast cancer cells can lead to redistribution of CDK inhibitors and consequent decreased cyclin E–Cdk2 activity (Carroll et al., 2000; Swarbrick et al., 2000), cyclin E immunoprecipitates did not contain increased levels of either p21WAF1/Cip1 or p27Kip1 when Id1 expression was decreased, further supporting the conclusion that the decreased cyclin E–Cdk2 activity here results largely from decreased cyclin E expression. The effects of decreased Id1 expression on cyclin D1 expression and pRb-Ser780 phosphorylation were more profound than those on cyclin E expression and cyclin E–Cdk2 expression. In addition, decreased cyclin D1 expression was apparent after 6 h treatment with Id1 antisense oligonucleotides, a timepoint when no decrease in cyclin E expression was detected. Cyclin D1 is essential for cell cycle progression in breast cancer cells (Baldin et al., 1993) and we have previously shown that a specific reduction in cyclin D1 expression of as little as 50% is sufficient to inhibit proliferation in MCF-7 cells (Carroll et al., 2000). Thus, decreased cyclin D1 expression is an early response to decreased Id1 expression and a likely


387

mechanism for inhibition of proliferation. The later onset of the decrease in cyclin E expression when Id1 expression is reduced suggests that decreased cyclin E expression is possibly a consequence of the inhibition of proliferation. However, since cyclin E is essential for cell cycle progression (Ohtsubo et al., 1995), the longer-term decrease in cyclin E expression may contribute to growth inhibition after treatment with Id1 siRNA. Id2 is essential for normal mammary development and differentiation (Mori et al., 2000; Miyoshi et al., 2002). In murine mammary glands, Id1 expression is high in early pregnancy, and low in mid-late pregnancy, when Id2 expression is high (Mori et al., 2000; Parrinello et al., 2001), suggesting complementary roles for Id1 and Id2 in mammary gland proliferation and differentiation, and challenging the view of Id proteins as functioning redundantly to inhibit differentiation. The observation that overexpression of Id2 in HC-11 cells did not affect the response to serum deprivation whereas Id1 overexpression allowed continued proliferation adds further weight to this conclusion. Our studies of the effects of decreased Id1 expression identified close similarities between the effects of decreased c-Myc expression and decreased Id1 expression, that is, decreases in cyclin D1, cyclin E and their associated kinase activities but no increase in the expression of either p21WAF1/Cip1 or p27Kip1. Treatment with c-Myc antisense oligonucleotides for 6 h downregulated cyclin D1 expression in breast cancer cells with little effect on cyclin E expression (Carroll et al., 2002), similar to the effects of Id1 antisense oligonucleotides and suggesting that the decreased cyclin E expression following c-Myc siRNA treatment may require sustained downregulation of c-Myc. Together these data indicate that c-Myc and Id1 regulate proliferation in breast epithelial cells by similar pathways. Consistent with this idea, both c-Myc and Id1 are necessary and sufficient for G1–S progression (Amati et al., 1998; Norton, 2000) and induce apoptosis under certain conditions (Norton, 2000; Oster et al., 2002). They also regulate common targets including h-TERT (Wang et al., 1998; Alani et al., 1999) and thrombospondin-1 (Tikhonenko et al., 1996; Volpert et al., 2002). Several lines of evidence suggest that Id1 may not simply target similar mechanisms of cell cycle control, but may also be downstream of c-Myc in breast epithelial cells. Overexpression of c-Myc resulted in increased expression of Id1 mRNA within 2 h and, conversely, downregulation of c-Myc resulted in decreased Id1 expression within 6 h. The observation that overexpression of the Max-rep dominant negative, which blocks c-Myc binding to Max and hence prevents c-Myc association with DNA, was able to downregulate Id1 expression suggests that c-Myc transactivation is required for expression of Id1. Whether this might be a direct action of c-Myc at the Id1 promoter is not known. The Id1 promoter does not contain canonical E-boxes, but a number of noncanonical E-boxes of the sequence CA(C/T)G(T/G)G that may bind Myc–Max complexes (Blackwell et al., 1990; Grandori et al., 1996) are present within 3000 bp upstream of the transcription start site.

Since decreased c-Myc expression led to reduced Id1 expression, and mimicking the decrease in Id1 expression was sufficient to decrease cyclin D1 expression to a similar degree, Id1 may be involved in the dependence of cyclin D1 on c-Myc expression in breast cancer cells. The mechanism underlying the dependency of cyclin D1 on Id1 is unknown, although the promoter of the cyclin D1 gene contains consensus motifs for several transcription factors (Wingender et al., 2001) known to bind to or interact with ID proteins, including ETS and TCF family members (Yates et al., 1999; Parrinello et al., 2001; Kolligs et al., 2002). An important goal of future experiments will be to delineate the relationship between c-Myc, Id1 and cyclin D1 in breast cancer cells.

Materials and methods Cell culture and retroviral infection T-47D and MCF-7 human breast cancer cells were cultured as described previously (Carroll et al., 2000; Swarbrick et al., 2000). HC-11 murine mammary epithelial cells were cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal calf serum and insulin (10 mg/ml). T-47D cells expressing the murine ecotropic receptor were infected with ecotropic retroviruses packaged as described previously (Musgrove et al., 2001). HC-11-Id1 and HC-11LXSN cell lines were generated by retroviral infection, using vectors kindly provided by Dr P-Y Desprez, followed by selection in 0.5 mg/ml G418 (Gibco, Melbourne, Vic., Australia) for 7 days. HC-11 cells were infected with MSCV-Id2IRES-GFP (a gift of Dr Stephen Jane) followed by sterile flow cytometric sorting into GFP-positive and -negative cells (i.e. Id2-positive and -negative cells) as described (Holmes et al., 1999). MCF-7-Id1, MCF-7-Babe, T-47D-Myc and T-47DBabe cells were selected using 0.66 mg/ml puromycin (Gibco) after infection with pBabe-Id1 (kindly provided by Dr P-Y Desprez; Lin et al., 2000), pBabe-Myc (constructed by cloning the c-myc cDNA into pBabe) or pBabe. MCF-7-DMT-Myc is a clonal cell line containing c-Myc under the control of the metallothionein promoter (Prall et al., 1998). MCF-7-DMT is the corresponding vector-transfected control clonal cell line. The max-rep cDNA (from Drs N Schreiber-Agus and RA DePinho; Schreiber-Agus et al., 1995) was cloned into pLibCD8. This vector is based on pLib (BD Biosciences Clontech, North Ryde, NSW, Australia), and expresses the murine CD8a extracellular domain under the control of the thymidine kinase promoter (excised from pSRMSVtk, from Dr CJ Sherr) and the cDNA of interest under the control of the pLib 50 LTR. Cells infected with pLib-CD8-based viruses were stained with phycoerythrin-labelled anti-CD8a (CMC023, Cell Marque, Hot Springs, AK, USA) 72 h after infection. Small interfering RNAs and antisense oligonucleotides Several sequences within the Id1 mRNA with low identity to other ID family members were identified using the ‘Silencer’ web-based design tool (http://www.ambion.com) and examined as potential siRNA targets. Blast analysis of GenBank was used to eliminate any targets with >17 nucleotide identity to other genes. Using MCF-7 cells, five siRNAs were tested for Id1 knockdown, and two were selected for further analysis: Id1-12, directed toward a segment of the Id1 open reading frame (target sequence: AACTCGGAATCCGAAGTTGGG); and Id1-17, directed toward a portion of the 30 UTR (target Oncogene


388 sequence: AAGAGGAATTACGTGCTCTGT). siRNAs targeting GAPDH (cat# 4602, Ambion, Austin, TX, USA) and EGFP (target sequence: AAGATGAACTTCAGGGTCAGC; based on Caplen et al. (2001), referred to here as GFP, were used as controls. siRNAs targeting the c-Myc 30 UTR (Ambion, cat# 4604), referred to here as ‘MycUTR’, and the central portion of the c-Myc open reading frame (von der Lehr et al., 2003), referred to here as ‘Myc940’, were also used. siRNAs were prepared by in vitro transcription (Silencer, Ambion) and transfected (Oligofectamine, Invitrogen, Mt Waverley, Vic., Australia) at 2 nM final concentration. Transfection with fluorescinated siRNAs showed that >90% of target cells were transfected using this protocol. Antisense and control oligonucleotides were manufactured (Geneworks Australia, Adelaide, SA, Australia) with phosphorothioate residues at the 30 and 50 nucleotides to minimize exonuclease cleavage. Id1 antisense and scrambled control oligonucleotides (Hara et al., 1994) were transfected (Cellfectin, Invitrogen) at a final concentration of 1.0–1.5 mM. c-Myc sense and antisense oligonucleotides were transfected at a final concentration of 1.2 mM (Carroll et al., 2002). Again, using this protocol, >90% of target cells were transfected, as demonstrated using fluorescinated oligonucleotides. Western blot analysis Cells were lysed for preparation of whole-cell protein extracts as previously described (Swarbrick et al., 2000). Samples were separated using SDS–PAGE and transferred to PVDF membranes, which were incubated with the following primary antibodies: Id1 (sc-488), Id2 (sc-489), Max (C-124) and cyclin E (HE12) from Santa Cruz Biotechnology, Santa Cruz, CA, USA; cyclin D1 (DCS6, Novocastra, Newcastle upon Tyne, UK); p21WAF1/Cip1 and p27Kip1 (C24420 and K25020, respectively, BD Transduction Laboratories, Lexington, KY, USA); pRb-P-Ser780 (9307, Cell Signaling Technology, Beverly, MA, USA); b-actin (AC-15, Sigma, Castle Hill, NSW, Australia). The secondary antibodies were horseradish peroxidase-conjugated sheep anti-mouse or donkey anti-rabbit (Amersham, Castle Hill, NSW, Australia) and specific proteins were visualized by chemiluminescence (Perkin-Elmer, Melbourne, Vic., Australia). Proliferation assays Incorporation of BrdU was measured by flow cytometry after incubation with 10 mM BrdU (120 min) and staining with an FITC-anti-BrdU antibody (Chemicon, Boronia, Vic., Australia). In some experiments, incorporation of BrdU was determined using an ELISA-based colorimetric assay following the manufacturer’s instructions (Cell proliferation ELISA, BrdU (colorimetric), Roche, Castle Hill, NSW, Australia). An MTS-based colorimetric assay was used as instructed by the manufacturer (CellTitre 96 MTS assay; Promega, Annan-

dale, NSW, Australia) to estimate relative cell number. Data are presented as average absorbance per sample corrected for background. Cyclin E–Cdk2 assay The activity of cyclin E immunoprecipitates (M-20 or C-19, Santa Cruz) from whole-cell extracts toward a histone H1 substrate was determined as previously described (Prall et al., 1997). Quantitative PCR Total RNA was harvested using the RNEasy mini kit (Qiagen, Clifton Hill, Vic., Australia). RNA (1 mg) was reverse transcribed using AMV reverse transcriptase with Poly-A priming (Promega). Primers to human Id1 (Fwd: caaggtgagcaaggtggagattc; Rev: gcttcagcgacacaagatgcg) and human cMyc (Fwd: TTCTGTGGAAAAGAGGCAGGC; Rev TCACGCAGGGCAAAAAAGC) were designed using MacVector. The PCR reactions were performed in a LightCycler (Roche) using 0.5 ml of cDNA, 5 pmol of primers and the FastStart DNA master SYBR Green I enzyme mix (Roche) in a 10 ml reaction volume. Relative expression was calculated using Relquant software (Roche) using GAPDH (Fwd: GACATCAAGAAGGTGGTG; Rev: TGTCATACCAGGAAATGA) primers to control for cDNA input. Data analysis Images captured by PhosphorImager (Molecular Dynamics 445SI) or digital scans of X-ray film were quantitated using IP LabGel H software (Signal Analytics, VA, USA). For siRNA experiments, data were normalized to the average of the two control siRNAs (GFP, GAPDH). Statistical analyses were performed using Microsoft Excel. Acknowledgements We thank the following colleagues: Joseph Webster (Centenary Institute of Cancer Medicine and Cell Biology, Sydney) and Dr Rebecca Newton for flow cytometric cell sorting; Dr Pierre Yves Desprez (Geraldine Brush Cancer Research Institute, San Francisco) and Dr Jason Carroll for helpful discussions, reagents and advice; and Dr Rania Kairouz and Dr Sophie Doisneau-Sixou for their contributions to some experiments. This research was supported by the National Health and Medical Research Council of Australia, The Cancer Council NSW, The Cure Cancer Australia Foundation and The Association of International Cancer Research. AS was a recipient of an Australian Postgraduate Award and a Fatemeh Moradzadeh Fellowship. CEC is a recipient of an Australian Postgraduate Award, the LH Ainsworth Scholarship for Cancer Research and the Beth Yarrow Scholarship in Medical Science.

References Alani RM, Hasskarl J, Grace M, Hernandez M-C, Israel MA and Munger K. (1999). Proc. Natl. Acad. Sci. USA, 96, 9637–9641. Alani RM, Young AZ and Shifflett CB. (2001). Proc. Natl. Acad. Sci. USA, 98, 7812–7816. Amati B, Alevizopoulos K and Vlach J. (1998). Front. Biosci., 3, d250–d268. Baldin V, Lukas J, Marcote MJ, Pagano M and Draetta G. (1993). Genes Dev., 7, 812–821. Blackwell TK, Kretzner L, Blackwood EM, Eisenman RN and Weintraub H. (1990). Science, 250, 1149–1151. Oncogene

Caplen NJ, Parrish S, Imani F, Fire A and Morgan RA. (2001). Proc. Natl. Acad. Sci. USA, 98, 9742–9747. Carroll JS, Prall OWJ, Musgrove EA and Sutherland RL. (2000). J. Biol. Chem., 275, 38221–38229. Carroll JS, Swarbrick A, Musgrove EA and Sutherland RL. (2002). Cancer Res., 62, 3126–3131. Desprez PY, Hara E, Bissell MJ and Campisi J. (1995). Mol. Cell. Biol., 15, 3398–3404. Desprez PY, Lin CQ, Thomasset N, Sympson CJ, Bissell MJ and Campisi J. (1998). Mol. Cell. Biol., 18, 4577–4588.


389 Desprez PY, Sumida T and Coppe JP. (2003). J. Mammary Gland Biol. Neoplasia, 8, 225–239. Dubik D, Dembinski TC and Shiu RPC. (1987). Cancer Res., 47, 6517–6521. Elbashir SM, Harborth J, Weber K and Tuschl T. (2002). Methods, 26, 199–213. Grandori C, Mac J, Siebelt F, Ayer DE and Eisenmann RN. (1996). EMBO J., 15, 4344–4357. Guan XY, Xu J, Anzick SL, Zhang H, Trent JM and Meltzer PS. (1996). Cancer Res., 56, 3446–3450. Halevy O, Novitch BG, Spicer DB, Skapek SX, Rhee J, Hannon GJ, Beach D and Lassar AB. (1995). Science, 267, 1018–1021. Hara E, Yamaguchi T, Nojima H, Ide T, Campisi J, Okayama H and Oda K. (1994). J. Biol. Chem., 269, 2139–2145. Holmes ML, Haley JD, Cerruti L, Zhou WL, Zogos H, Smith DE, Cunningham JM and Jane SM. (1999). Mol. Cell. Biol., 19, 4182–4190. Iavarone A, Garg P, Lasorella A, Hsu J and Israel M. (1994). Genes Dev., 8, 1270–1284. Itahana Y, Singh J, Sumida T, Coppe JP, Parrinello S, Bennington JL and Desprez PY. (2003). Cancer Res., 63, 7098–7105. Kitagawa M, Higashi H, Jung H-K, Suzuki-Takahashi I, Ikeda M, Tamai K, Kato J-Y, Segawa K, Yoshida E, Nishimura S and Taya Y. (1996). EMBO J., 15, 7060–7069. Kolligs FT, Nieman MT, Winer I, Hu G, Van Mater D, Feng Y, Smith IM, Wu R, Zhai Y, Cho KR and Fearon ER. (2002). Cancer Cell, 1, 145–155. Lin CQ, Singh J, Murata K, Itahana Y, Parrinello S, Liang SH, Gillett CE, Campisi J and Desprez PY. (2000). Cancer Res., 60, 1332–1340. Massari ME and Murre C. (2000). Mol. Cell. Biol., 20, 429–440. Miyoshi K, Meyer B, Gruss P, Cui Y, Renou J-P, Morgan FV, Smith GH, Reichenstein M, Shani M, Hennighausen L and Robinson GW. (2002). Mol. Endocrinol., 16, 2892–2901. Mori S, Nishikawa S-I and Yokota Y. (2000). EMBO J., 19, 5772–5781. Musgrove EA, Hunter LK, Lee CS, Swarbrick A, Hui R and Sutherland RL. (2001). J. Biol. Chem., 276, 47675–47683. Musgrove EA, Lilischkis R, Cornish AL, Lee CS, Setlur V, Seshadri R and Sutherland RL. (1995). Int. J. Cancer, 63, 584–591. Norton J. (2000). J. Cell Sci., 113, 3897–3905. Ohtani N, Zebedee Z, Huot TJG, Stinson JA, Sugimoto M, Ohashi Y, Sharrocks AD, Peters G and Hara E. (2001). Nature, 409, 1067–1070. Ohtsubo M, Theodoras AM, Schumacher J, Roberts JM and Pagano M. (1995). Mol. Cell. Biol., 15, 2612–2624.

Oster SK, Ho CSW, Soucie EL and Penn LZ. (2002). Adv. Cancer Res., 84, 81–154. Parrinello S, Lin CQ, Murata K, Itahana Y, Singh J, Krtolica A, Campisi J and Desprez PY. (2001). J. Biol. Chem., 276, 39213–39219. Peverali F, Ramqvist T, Saffrich R, Pepperkok R, Barone M and Philipson L. (1994). EMBO J., 13, 4291–4301. Prabhu S, Ignatova A, Park S and Sun X. (1997). Mol. Cell. Biol., 17, 5888–5896. Prall OWJ, Rogan EM, Musgrove EA, Watts CK and Sutherland RL. (1998). Mol. Cell. Biol., 18, 4499–4508. Prall OWJ, Sarcevic B, Musgrove EA, Watts CKW and Sutherland RL. (1997). J. Biol. Chem., 272, 10882–10894. Schoppmann SF, Schindl M, Bayer G, Aumayr K, Dienes J, Horvat R, Rudas M, Gnant M, Jakesz R and Birner P. (2003). Int. J. Cancer, 104, 677–682. Schreiber-Agus N, Chin L, Chen K, Torres R, Rao G, Guida P, Skoultchi AI and DePinho RA. (1995). Cell, 80, 777–786. Sikder HA, Devlin MK, Dunlap S, Ryu B and Alani RM. (2003). Cancer Cell, 3, 525–530. Sutherland RL, Prall OW, Watts CK and Musgrove EA. (1998). J. Mammary Gland Biol. Neoplasia, 3, 63–72. Swarbrick A, Lee CS, Sutherland RL and Musgrove EA. (2000). Mol. Cell. Biol., 20, 2581–2591. Tanner MM, Tirkkonen M, Kallioniemi A, Isola J, Kuukasjarvi T, Collins C, Kowbel D, Guan XY, Trent J, Gray JW, Meltzer P and Kallioniemi OP. (1996). Cancer Res., 56, 3441–3445. Tikhonenko AT, Black DJ and Linial ML. (1996). J. Biol. Chem., 271, 30741–30747. Tournay O and Benezra R. (1996). Mol. Cell. Biol., 16, 2418–2430. Volpert OV, Pilli R, Sikder HA, Nelius T, Zaichuk T, Morris C, Shiflett CB, Devlin MK, Conant K and Alani RM. (2002). Cancer Cell, 2, 473–483. von der Lehr N, Johansson S, Wu S, Bahram F, Castell A, Cetinkaya C, Hydbring P, Weidung I, Nakayama K, Nakayama KI, Soderberg O, Kerppola TK and Larsson LG. (2003). Mol Cell, 11, 1189–1200. Wang J, Xie LY, Allan S, Beach D and Hannon GJ. (1998). Genes Dev., 12, 1769–1774. Watson PH, Pon RT and Shiu RPC. (1991). Cancer Res., 51, 3996–4000. Wingender E, Chen X, Fricke E, Geffers R, Hehl R, Liebich I, Krull M, Matys V, Michael H, Ohnhauser R, Pruss M, Schacherer F, Thiele S and Urbach S. (2001). Nucleic Acids Res., 29, 281–283. Yates PR, Atherton GT, Deed RW, Norton JD and Sharrocks AD. (1999). EMBO J., 18, 968–976.

Oncogene