- Top of page
- Patients and methods
Cyclin D1, encoded by the CCND1 gene, is immunohistochemically detectable in up to one-third of cases of multiple myeloma (MM). To examine the mechanism of cyclin D1 overexpression, we compared cyclin D1 immunoreactivity with the results of conventional cytogenetics to determine if the t(11;14)(q13;q32) or other abnormalities of 11q11–14 explained cyclin D1 overexpression. Karyotypic abnormalities were found in 45 out of 67 (67%) MM cases; the t(11;14) was present in seven cases (10%). Additional 11q11–14 abnormalities were not identified. The t(11;14) correlated with cyclin D1 upregulation in low to intermediately proliferative MM, but was not present in highly proliferative tumours (assessed using bromodeoxyuridine labelling index). Cyclin D1 indirectly activates the transcription factor E2F-1. In the bone marrow biopsy specimens of MM cases, E2F-1 was concurrently expressed with cyclin D1 (P = 0·001), indicating that cyclin D1 is functional. However, as neither E2F-1 nor cyclin D1 expression correlated with proliferative activity, the speculation that t(11;14) upregulates the CCND1 gene to induce higher proliferation and possibly more aggressive disease is not supported. We conclude that in low to intermediately proliferative MM cases, cyclin D1 is probably upregulated by t(11;14), but an alternative mechanism is more probable in highly proliferative MM.
Defects of normal cell cycle regulation, particularly at the G1 to S phase transition, play a pivotal role in tumour development (Sherr, 1996; Reed, 1997). Cyclin D1 is one of the most important of the many activating and inhibitory proteins that affect G1 to S phase progression (Hunter & Pines, 1994). Cyclin D1 activates cyclin-dependent kinases (CDKs) to trigger phosphorylation of retinoblastoma protein (Weinberg, 1995). Phosphorylation causes release of the transcription factor E2F-1, which is normally bound to retinoblastoma protein (Chellappan et al, 1991). E2F-1, in its free and active form, is thought to be the ultimate factor that promotes transition into S phase and mitosis (Johnson et al, 1993).
Cyclin D1 protein is overexpressed in mantle cell lymphoma as a result of the t(11;14) translocation, in which the CCND1 (also known as PRAD1) gene on chromosome 11q13 is juxtaposed with the immunoglobulin heavy chain gene at 14q32 (Seto et al, 1992). Deregulated expression of cyclin D1 results in these cells having a proliferative advantage. In multiple myeloma (MM), the t(11;14) is identified in approximately 4–10% of bone marrow (BM) specimens when examined using conventional cytogenetics and in 12–16% of BM specimens assessed using fluorescent in situ hybridization (FISH) methods (Lai et al, 1995; Sawyer et al, 1995; Nishida et al, 1997; Avet-Loiseau et al, 1998). In contrast, cyclin D1 protein is immunohistochemically detectable in up to one-third of MM cases and may be associated with greater tumour burden (Vasef et al, 1997; Lai et al, 1998). The relationship between cyclin D1 expression and t(11;14) is unclear in these cases (Kobayashi et al, 1995). If cyclin D1 is functional and involved in the pathogenesis of MM, cytogenetic studies are either insufficiently sensitive to detect the t(11;14), or cyclin D1 expression occurs independently of t(11;14) and may exert its activity in collaboration with other oncogene products, as occurs in many other tumour types (Motokura & Arnold, 1993).
In this study, we attempt to address three questions. First, what is the mechanism of cyclin D1 expression in MM and, in particular, can it be explained by cytogenetic abnormalities? To answer this question, we examined the relationship between cyclin D1 protein expression and t(11;14) or other chromosome 11q abnormalities detectable using cytogenetic studies in BM specimens from MM patients. Second, is cyclin D1 functional in MM? As cyclin D1 overexpression should result in increased free E2F-1 protein, we assessed the cases immunohistochemically with an antibody specific for E2F-1. Finally, to determine if cyclin D1 overexpression results in increased tumour proliferation, either independently or by activating E2F-1, trephine biopsy sections of MM with predetermined bromodeoxyuridine labelling index (BrdU LI) were selected for evaluation of both cyclin D1 and E2F-1 immunostaining. The BrdU LI determines the percentage of plasma cells in S phase and is used in plasma cell dyscrasias to estimate the proliferative activity of BM plasma cells (Ffrench et al, 1995; Joshua et al, 1996).
- Top of page
- Patients and methods
Cyclin D1 immunoreactivity was identified in 31% of BM biopsy specimens of MM, a frequency similar to previous studies (Vasef et al, 1997; Lai et al, 1998). Abnormal karyotypic abnormalities were found in 67% of cases and were related to the proliferative activity of the tumours. All highly proliferative MM cases, as measured by BrdU LI, had cytogenetic abnormalities, compared with less than 30% of cases with low proliferative disease. The t(11;14), detected in 10% of all BMs, was found exclusively in cyclin D1-positive cases. Among low to intermediately proliferative MM with abnormal karyotypes, 79% of cyclin D1-positive but no cyclin D1-negative tumours carried the t(11;14), suggesting a significant relationship. The t(11;14) was probably underrepresented using cytogenetic analysis in this population, given the presence of aneuploid populations in 80% of low proliferative MM determined by flow cytometric evaluation.
The t(11;14) was not detected in MM with high proliferative activity, despite the finding of cyclin D1 positivity in 37% of cases. Other abnormalities of 11q11–14 that could account for overexpression of CCND1 were also not identified. While clonal plasma cell subpopulations with t(11;14) may have been missed (Hallek et al, 1998), attempts to better define genetic defects in MM with multicolour spectral karyotyping (SKY) and FISH analysis do not support this possibility. We performed SKY on 50 MM cases with complex chromosomal aberrations that were not fully characterized by G-banding, but had no evidence of t(11;14) (Sawyer et al, 1998). The t(11;14) was identified in only one case, involving 14q21 rather than the 14q32. Avet-Loiseau et al (1999) used metaphase FISH analysis to identify t(11;14) in only one of 32 MM negative for the t(11;14) using G-banding techniques. Similarly, Nishida et al (1997) applying dual-colour FISH on metaphase spreads and interphase nuclei, detected the t(11;14) in 5 out of 47 (10%) plasma cell malignancies, four of which had normal G-banded karyotypes. The single patient with an abnormal (complex) karyotype and detection of t(11;14) using FISH had plasma cell leukaemia. Extrapolation of these findings suggests that t(11;14) may be missed by cytogenetic studies in low to intermediately proliferative MM with normal karyotypes, but adequate detection of t(11;14) usually occurs in MM with abnormal karyotypes. Possible alternative mechanisms for cyclin D1 expression in the highly proliferative tumours include abnormalities in CDK inhibitors that normally compete with cyclin D1 for binding to CDKs (Kawano et al, 1997; Ng et al, 1997; Tasaka et al, 1997; Urashima et al, 1997) and truncation with increased stabilization of CCND1 mRNA (Hirama & Koeffler, 1995). Regardless of the exact mechanism, cyclin D1 in highly proliferative MM may act as a protooncogene in concert with other oncogenes to transform cells (Motokura & Arnold, 1993; Hunter & Pines, 1994).
The t(11;14) has been suggested to be a poor prognostic indicator in some studies of patients with MM (Fonseca et al, 1999), but in other studies it had no prognostic significance (Avet-Loiseau et al, 1998; Sonoki et al, 1999). In this study, the cyclin D1 protein appeared to be functionally active because expression strongly correlated with E2F-1 immunoreactivity. However, a significant relationship between cyclin D1 or E2F-1 expression and increased proliferative activity of MM was not found, despite the role cyclin D1 and E2F-1 play in promoting cells from G1 to S phase of the cell cycle. Evaluating proliferative activity using the BrdU LI, which measures the percentage of plasma cells in S phase in BM aspirate material (BrdU LI), correlated strongly with Ki-67 immunostaining of BM biopsy sections, which discriminates between cycling (G1, S and G2 + M) and resting (G0) plasma cells (Falini et al, 1988; Tsurusawa et al, 1995). Therefore, factors other than cyclin D1 and E2F-1 must be affecting these tumours. For example, variations in cycle times in the heterogenous plasma cell populations (Shackney & Shankey et al 1999), or a second oncogene, similar to the putative transforming gene myeov found in a subset of MM cell lines with t(11;14) (Janssen et al, 2000), may be altering proliferative activity.
The highly proliferative MM in this series had other poor prognostic features including an increased frequency of monosomy 13 and/or deletion 13q chromosomal abnormalities and moderate to poorly differentiated tumour histology (Bartl grade II–III) (Tricot et al, 1995; Waldron et al, 1997). One of the surprising findings among the highly proliferative, cyclin D1-positive cases was the high incidence of women with IgA-secreting MM. Unfavourable cytogenetic karyotypes are known to correlate with the IgA isotype (Tricot et al, 1997). However, the specific risk for women in this group of tumours with cyclin D1 expression needs to be determined in a standardized clinical trial. In conclusion, the probable mechanism for cyclin D1 upregulation in MM with low or intermediate proliferative activity is through t(11;14). Cyclin D1 overexpression in highly proliferative tumours occurs independently of t(11;14). The data suggests that cyclin D1 upregulation does not appear to be directly linked to proliferation.