Cyclin D1 is a key regulator of cell proliferation and its expression is subject to both transcriptional and post-transcriptional regulation. In different cellular contexts, different pathways assume a dominant role in regulating its expression, whereas their disregulation can contribute to overexpression of cyclin D1 in tumorigenesis. Here, we discuss the ability of the NF-κB (nuclear factor κB)/IKK [IκB (inhibitor of NF-κB) kinase] pathways to regulate cyclin D1 gene transcription and also consider the newly discovered role of the SNARP (SNIP1/SkIP-associated RNA processing) complex as a co-transcriptional regulator of cyclin D1 RNA stability.
The CCND1 gene-encoded cyclin D1 protein is tightly regulated by a variety of signalling pathways at the transcriptional level as well as post-transcriptionally, through modulation of RNA splicing and stability, nucleocytoplasmic transport, protein stability and degradation (Figure 1). Cyclin D1 differs from other cyclins by being induced by mitogens such as growth factors and integrins as a delayed early gene in G1 phase [1,2]. For DNA synthesis to occur during S-phase, cyclin D1 needs to be significantly down-regulated, which can be achieved, in part, through GSK3β (glycogen synthase kinase 3β)-mediated phosphorylation of cyclin D1 on Thr286, allowing its nuclear export followed by polyubiquitination and proteasomal degradation .
Regulation of cyclin D1
Cyclin D1 promotes cell cycle progression through G1-phase by forming active holoenzymes with CDK (cyclin-dependent kinase) 4 and CDK6, which phosphorylate the pRb (retinoblastoma protein) . This causes pRb to release the E2F transcription factor, which can then activate genes essential for G1–S transition and S-phase . However, cyclin D1 can also regulate cellular functions in a CDK-independent fashion through interaction with transcriptional regulators and nuclear receptors [6,7]. The deregulation and overexpression of cyclin D1 are frequently linked to various types of human cancer . With increased understanding of its functions and regulation, cyclin D1 is now an interesting potential therapeutic target for cancer and other human diseases [8,9].
Our laboratory first developed an interest in cyclin D1 expression after its identification as an NF-κB (nuclear factor κB) target gene [10–12]. More recently, through investigating the function of SNIP1 (Smad nuclear interacting protein 1), we have also explored post-transcriptional regulation of cyclin D1 expression. Here, we will briefly review the regulation of CCND1 gene expression, with an emphasis on the function of NF-κB and the newly discovered role of SNIP1.
The cyclin D1 promoter
The human, mouse and rat cyclin D1 promoter sequences are homologous and transcription factor-binding sites have been mapped to their regulatory regions [13,14]. Significantly, these studies have revealed enormous diversity in the transcriptional programmes controlling cyclin D1 transcription, with a wide variety of signalling pathways converging on this promoter, dependent on the cell type or context in which cyclin D1 expression needs to be induced. These include the PI3K (phosphoinositide 3-kinase)/protein kinase B (also called Akt) signalling pathway, Wnt/β-catenin signalling pathway, nuclear hormones, inflammatory signalling through the NF-κB/IKK [IκB (inhibitor of NF-κB) kinase] pathway and cell cycle regulators. Concomitant with this, binding sites for a wide range of transcription factors have been described, including AP-1 (activator protein 1), STAT (signal transducer and activator of transcription), TCF (T-cell factor), E2F, Sp (specificity protein) transcription factors, ER-α (oestrogen receptor-α), c-Myc and NF-κB (reviewed in ).
NF-κB and the regulation of cyclin D1 transcription
The family of NF-κB transcription factors comprises five members: RelA (p65), c-Rel, RelB, NF-κB1 (p50/p105) and NF-κB2 (p52/p100) . These proteins can homodimerize or heterodimerize to form functional transcriptional complexes. In the inactive state, NF-κB is kept in the cytoplasm by members of the IκB family. On activation of the classical NF-κB pathway, the IKK complex, which is composed of the catalytic subunits IKKα (IKK1), IKKβ (IKK2) and the regulatory subunit NEMO (NF-κB essential modulator; IKKγ), phosphorylates IκBα, leading to its degradation. This allows the NF-κB dimer to translocate into the nucleus, resulting in the activation or repression of different target genes, depending on the signal the cell receives . IKKβ is responsible for phosphorylating IκBα in the classical pathway. By contrast, IKKα regulates the alternative (or non-canonical) pathway by phosphorylating the NF-κB2 p100 precursor protein, thus stimulating its proteolytic processing to p52 .
Although the first reports of NF-κB regulation of the cyclin D1 promoter focused on the most commonly observed heterodimer RelA–p50 [10,12], it is now clear that many other NF-κB complexes can regulate its expression and have a dominant role in some contexts [11,17,18]. For example, p52, together with its co-activator Bcl-3, an IκB-like protein that functions as a nuclear transcriptional co-activator, has also been shown to be important in the regulation of cyclin D1 [11,19,20]. Although much earlier work on NF-κB regulation of cyclin D1 involved in vitro studies, such as DNA-binding EMSAs (electrophoretic mobility-shift assays) or promoter reporter assays, later studies were able to exploit the development of ChIP (chromatin immunoprecipitation) analysis, which confirmed NF-κB subunit binding to the cyclin D1 promoter region ex vivo.
NF-κB regulation of cyclin D1 occurs in a variety of cell types and contexts. For example, in keratinocytes, the tumour suppressor CYLD (cylindromatosis) has been demonstrated to bind and deubiquitinate Bcl-3, thus preventing its nuclear accumulation, association with p50 and p52 and consequent induction of cyclin D1 . In mammary epithelial cells, p50 occupies the cyclin D1 promoter constitutively where it represses transcription. However, after TNFα (tumour necrosis factor α) induction, p52 and RelB bind, resulting in an increase in cyclin D1 expression . c-Rel has also been implicated in the regulation of cyclin D1 in a model for mammary tumorigenesis. The disregulation and overexpression of c-Rel correlated to an increase in cyclin D1 expression and contributed to the formation of mammary tumours .
Interestingly, IKKα has been reported to regulate cyclin D1 through NF-κB-independent pathways, which include the phosphorylation and activation of ER-α transcription factor together with its co-activator protein SRC3 (steroid receptor coactivator-3) [22,23]. In addition, IKKα is involved in the activation  and stabilization  of β-catenin, a transcriptional co-activator for the TCF family of transcription factors, which can bind the cyclin D1 promoter .
Other transcription factors can influence NF-κB-dependent regulation of the cyclin D1 promoter
The interaction of NF-κB with other transcription factors can also have an impact on the regulation of cyclin D1. For example, induction of cyclin D1 through RelA–p50, in response to Rac activation, is dependent on both the NF-κB and the ATF-2 (activating transcription factor-2) binding sites, suggesting that the combinatorial interactions of NF-κB and other transcription factors are crucial . Moreover, AP-1 and Ets transcription factors have both been shown to regulate cyclin D1 transcription  and also physically associate with NF-κB [28,29]. Consistent with this, Toualbi-Abed et al.  demonstrated that recruitment of JunD to the κB2 site of the cyclin D1 promoter was dependent on RelA–p50. Therefore it is likely that these and other transcription factors have a role in determining NF-κB's role in cyclin D1 transcription, although this aspect of its regulation is still largely undetermined.
The p53 tumour suppressor can also regulate cyclin D1 expression. However, the cyclin D1 promoter appears to lack a p53-binding site. Furthermore, no recruitment of p53 to the proximal promoter was observed by ChIP analysis, suggesting that p53 is not a direct transcriptional regulator of cyclin D1 . Instead, induction of p53 was shown to result in the repression of the cyclin D1 promoter by inducing a switch from p52–Bcl3 activator complexes to p52–HDAC1 (histone deacetylase 1) repressor complexes .
Cell-cycle-dependent NF-κB-dependent regulation of the cyclin D1 promoter
In early passage U2OS cells, cyclin D1 mRNA levels were found to fluctuate during the cell cycle, with significantly higher levels being found in G1-phase relative to G2-phase . After prolonged growth in culture, this effect was lost, suggesting that a regulatory mechanism was in place that could be down-regulated, possibly due to selective pressure. Using centrifugal elutriation to enrich different populations of cells at different cell cycle stages, ChIP assays revealed differential NF-κB subunit binding to the cyclin D1 promoter in G1-, S- and G2-phase . In G1-phase, p52–RelA heterodimers bound to CBP [CREB (cAMP-response-element-binding protein)-binding protein] and p52 homodimers bound to Bcl-3 were detected. In S-phase, where cyclin D1 expression starts to be repressed in these cells, p52–RelA heterodimers bound to HDAC1 were present, whereas in G2-phase, all NF-κB subunits were found to bind the promoter, and these recruited transcriptional co-repressor complexes. In part, this phased change in NF-κB subunit activity at different cell cycle stages results from differential phosphorylation. Differentially phosphorylated RelA was seen to be recruited to the cyclin D1 promoter at different stages of the cell cycle and this segregated with co-activator and co-repressor binding: Ser468-phosphorylated RelA associated with CBP and was recruited to the cyclin D1 promoter during G1-phase; Thr505-phosphorylated RelA associated with HDAC1 and was recruited in S-phase; and Ser536-phosphorylated RelA associated with HDAC1 and was seen in G2-phase . Interestingly, putative cell cycle phosphorylation of p52 and RelB has also been observed , but it is not yet known whether these also participate in the regulation of CCND1 gene expression.
This pattern of promoter recruitment was also observed at the c-Myc and Skp2 promoters but not at other NF-κB target genes . So, while this mechanism of regulation is not restricted to cyclin D1, other mechanisms, such as interactions with heterologous transcription factors discussed above, must provide promoter specificity to NF-κB recruitment.
Post-transcriptional regulation of cyclin D1
In addition to transcriptional regulation, cells employ a vast array of post-transcriptional mechanisms that can rapidly modulate cyclin D1 activity, to allow appropriate cell cycle progression in response to the cellular context. These regulate cyclin D1 co- and post-transcriptionally, affecting its processing and stability. Translation of cyclin D1 mRNA is also regulated while post-translational regulation can determine its association with CDK4/6, nucleocytoplasmic transport and degradation.
Accumulating evidence has shown that DNA transcription, mRNA processing and translation are all closely interconnected with regulatory information flowing bi-directionally . A key player that facilitates communication between transcriptional and post-transcriptional regulators is RNA polymerase II, via its C-terminal heptad repeat domain [CTD (C-terminal domain)]. We have been studying a nuclear protein that has been implicated in regulating cyclin D1 transcription as well as mRNA processing and which co-immunoprecipitates with RNA polymerase II, the Smad nuclear interacting protein, SNIP1.
SNIP1 was initially identified as a Smad transcription factor interactor and regulator of TGF-β (transforming growth factor-β) and NF-κB RelA signalling by binding to the transcriptional co-activators CBP/p300 [35,36]. SNIP1 has been shown to play a regulatory role in various processes, including cell growth, differentiation and development , the DNA damage response  and miRNA (microRNA) biogenesis . Like cyclin D1, SNIP1 is up-regulated in several human cancers and enhances the oncogenic properties of the proteins c-Myc and H-Ras .
SNIP1 and cyclin D1
Down-regulation of SNIP1 by RNA interference results in G1 cell cycle arrest at least in part as a consequence of reduced cyclin D1 mRNA and protein levels . These effects were independent of the p53 and pRb tumour suppressor proteins . Based on these results and its known role as a transcription factor, SNIP1 was initially hypothesized to regulate the cyclin D1 promoter. However, further studies, sparked by the observed homology between SNIP1 and known RNA-binding and splicing factors, revealed that SNIP1 specifically affects post-spliced cyclin D1 mRNA, promoting nascent mRNA stability . SNIP1 did not affect polyadenylation, nucleocytoplasmic transport or alternative splicing of cyclin D1 mRNA, nor did it have any effect on cyclin D2.
Previously, most nucleoplasmic SNIP1 was found to be present in an unknown high-molecular-mass complex . By affinity purification and mass spectrometry, components of this protein complex were identified as THRAP3 (thyroid-hormone-receptor-associated protein 150 kDa; also called TRAP150) and its related protein BTF (Bcl-2 associated transcription factor; also called BCLAF for Bcl2-associated transcription factor 1), Pinin and the SkIP (Ski-interacting protein) . Given that all these factors have known or suspected roles in RNA processing or splicing, we termed this complex SNARP (SNIP1/SkIP-associated RNA processing) complex. Interestingly, elutriation studies from U2OS cells showed SNIP1 and SNARP levels to be highest in G1-phase of the cell cycle, suggesting that co-ordination between mRNA transcription and processing leads to the patterns of cell cycle expression of cyclin D1 seen in these cells [32,41].
ChIP analysis demonstrated that the SNARP components are co-recruited to the CCND1 gene downstream of its promoter and showed that they assemble independently of RNA or known spliceosome complexes . SNARP also binds to the 3′-UTR (untranslated region) of cyclin D1 mRNA, putatively through evolutionarily conserved regions. SNIP1 was shown to be essential for the recruitment of the RNA processing factor U2AF65 (U2 small auxiliary factor 65), although no direct interaction between SNARP and U2AF65 was observed. This factor binds the same regions as SNARP along the cyclin D1 3′-UTR and co-localizes with SNIP1 in nuclear speckles. U2AF65 is important in 3′ splice site selection and is believed to have further RNA processing roles. In fact, U2AF65 and components of the SNARP complex besides SNIP1 are known to form part of other RNA processing complexes [42,43]. We proposed that the failure to recruit U2AF65 in the absence of SNIP1 destabilizes the nascent cyclin D1 mRNA, resulting in a failure of processing and its premature degradation.
Alternative splicing of cyclin D1
The CCND1 gene is known to produce two alternative splice isoforms: the canonical CD1a isoform with all five exons and the CD1b transcript, which only includes exons 1–4 and part of intron 4 . The lack of exon 5 makes CD1b a constitutively nuclear protein, as it lacks a PEST (Pro-Glu-Ser-Thr) motif and the Thr286 residue, which are required for the nuclear export and degradation of cyclin D1 . CD1b can still bind CDK4/6 to form an active holoenzyme, but it is less efficient in phosphorylating pRb when compared with CD1a . Overexpression of CD1a alone is not thought to be sufficient for cellular transformation. Rather, the constitutive nuclear accumulation of cyclin D1 is associated with its oncogenic potential, as this can lead to DNA re-replication, genomic instability  and evasion of contact inhibition . In contrast with CD1a, CD1b can cause cellular transformation and has been linked to human cancers [45,47]. Disregulation of cyclin D1 phosphorylation, its nucleocytoplasmic export or degradation may also lead to nuclear build-up of the protein .
There have been over 100 SNPs (single nucleotide polymorphisms) identified in the CCND1 locus, although none alter the amino acid sequence of the protein . However, the SNP at G/A870 (Pro241) located at the splice donor site of intron 4 has been linked to alternative splicing . The A allele is thought to form a poor splice site, allowing translation into intron 4 to produce the truncated CD1b protein, although both genotypes can still produce CD1a and CD1b . Despite the effect of the G/A870 SNP on cyclin D1 splicing, the increase in cancer risk linked to this genotype is low. The polymorphism is common (~42%A, 58%G) and therefore does not appear to hold much value as a clinical marker on its own [48,49].
As a critical regulator of cell proliferation, cyclin D1 expression needs to be kept under tight control. Here we have focused specifically on the role of NF-κB and SNIP1, but it is clear that almost all aspects of its synthesis are subject to regulation, including gene transcription, RNA processing, splicing, modification, protein stability and cellular localization (Figure 1). How these are co-ordinated in different cellular contexts is likely to be a key aspect controlling cyclin D1 levels in cells. Moreover, disregulation of any of these steps can potentially lead to cyclin D1 overexpression, which is known to promote tumorigenesis in many cell types.
Gene Expression in Development and Disease: 13th Tenovus-Scotland Symposium, an Independent Meeting held at University of Glasgow, Glasgow, U.K., 16–17 April 2009. Organized and Edited by Sheila Graham (Glasgow, U.K.).
activator protein 1
CREB (cAMP-response-element-binding protein)-binding protein
glycogen synthase kinase 3β
histone deacetylase 1
nuclear factor κB
inhibitor of NF-κB
Smad nuclear interacting protein 1
SNIP1/SkIP-associated RNA processing
single nucleotide polymorphism
U2 small auxiliary factor 65
Neil Perkins is the recipient of the 2009 Tenovus Medal. We thank all the members of the Perkins Laboratory for their help and assistance. We apologize to those of our colleagues whose work we were not able to cite in this paper due to space limitations.
I.-I.W. is funded by a Cancer Research UK Ph.D. studentship and L.F.K. is a member of the ADP (A*STAR-University of Dundee Partnership Ph.D. Programme).
These authors contributed equally to this work.