Failure of RB1 to reverse the malignant phenotype of human tumor cell lines.

In addition to retinoblastoma and osteosarcoma, mutation of both alleles of the RB1 gene occurs frequently in several other types of tumors. In order to evaluate the role of RB1 in cancer, the wild type RB1 gene was introduced into the RB1-deleted breast cancer cell line MDA-468-S4 and retinoblastoma cell lines WERI-Rb1 and Y-79. The RB1 complementary DNA was under control of the inducible murine metallothionein promoter in MDA-468-S4 and the thymidine kinase promoter in the retinoblastoma lines. The protein, p110RB1, produced from the exogenously introduced gene appeared normal by immunoprecipitation, Western blot analysis, and nuclear localization and also showed normal cell cycle-dependent phosphorylation and an ability to bind to E1a protein. No changes in growth rate or morphology were observed in either of the reconstituted cell types. Expression of p110RB1 in MDA-468-S4 did not affect anchorage-independent growth when measured by colony formation in soft agar. Although the ability of WERI-Rb1 cells expressing p110RB1 to form colonies in methylcellulose was reduced, the reconstituted retinoblastoma cell lines formed intraocular tumors in immunodeficient mice with the same efficiency as the RB1-negative parent cell lines and the tumors produced by the RB1-reconstituted cells continued to express p110RB1. These experimental results suggest that the malignant phenotype is little affected by the replacement of p110RB1 and that RB1 is a relatively weak tumor suppressor gene.

Evidence for retinoblastoma protein (RB) dependent and independent IFN-gamma responses: RB coordinately rescues IFN-gamma induction of MHC class II gene transcription in noninducible breast carcinoma cells.

The class II major histocompatibility (MHC) genes encode cell surface heterodimers that present processed antigen to CD4 positive T-cells. The class II genes are expressed constitutively on B-cells and can be induced by IFN-gamma on a variety of other cell types. Because the class II genes are aberrantly expressed on many mesenchymal tumors, which are frequently defective for the retinoblastoma tumor suppressor protein (RB), we investigated the role of RB in the regulation of HLA-DR and -DP. The RB defective breast carcinomas cell line, MDA-468-S4 (S4), as well as S4 subclones reconstituted with RB coding sequences under the control of a zinc inducible promoter, were treated with IFN-gamma and examined for DR and DP expression. Surface DR is not inducible in S4 cells, but inducibility is rescued by RB. DP is only slightly inducible in S4, but inducible to a much higher level in the RB positive subclones of S4. IFN-gamma induction of DR and DP mRNAs are correspondingly dependent on RB. IFN-gamma receptors are present on S4 cells, and the guanylate binding protein and ICAM-1 genes respond to IFN-gamma, ruling out the possibility that all IFN-gamma signal transduction pathways are defective in S4 cells. These data indicate RB regulates the coordinate response of class II genes to IFN-gamma. Possible roles for RB in this process are discussed, as well as the role of the class II-noninducible phenotype in tumor rejection.

Speculations on the roles of RB1 in tissue-specific differentiation, tumor initiation, and tumor progression.

Studies of retinoblastoma clearly identify mutation of the RB1 gene on chromosome 13 as the primary cause of this cancer. However, all retinoblastoma tumors have an abnormal karyotype (1, 2) indicating the presence of additional mutations and suggesting that mutation of both RB1 alleles is insufficient for development of retinoblastoma. In addition, analysis of RB1 expression and of RB1 mutations in different tumors leads to the following dilemma: while the RB1 gene product, p110RB1, is expressed in most dividing cells, germline mutations inactivating the function of p110RB1 predispose primarily to retinoblastoma and to a lesser extent to osteosarcoma, but do not predispose to cancer in general. However, many tumors contain somatic mutations that disrupt RB1 function. Thus, we are faced with the unusual situation in which germline mutations in the RB1 gene predispose to a very limited set of cancers, but somatic mutations in RB1 appear to contribute to malignancy in many tissues. We propose that the role of the RB1 gene is to maintain the cells in a stable, quiescent state required for terminal differentiation and that the effect of RB1 mutations in different tissues depends on the pattern of differentiation in that tissue. In tissues where differentiation follows a linear process from undifferentiated precursors to fully differentiated cells, loss of RB1 function during early stages of differentiation may lead to uncontrolled growth and the development of cancer. On the other hand, in cell renewal systems where cell number is usually maintained by a process of programmed cell death (PCD) or apoptosis, loss of RB1 function may lead to cell death.

The retinoblastoma protein and the cell cycle.

Although mutations of the retinoblastoma gene (RB1) contribute to malignant progression in many types of tumor, the role of RB1 mutation in cancer initiation is highly restricted to the rare embryonic tumor, retinoblastoma. However, RB1 is expressed and its product, p110RB1, is regulated through the cell cycle in many tissues. This apparent paradox may be clarified by the emerging evidence that p110RB1 functions as a regulator of transcription of cell cycle genes. Tissue specific effects may be determined by the cellular proteins that interact with p110RB1, or by cell type-specific expression of target genes regulated by p110RB1.

How do retinoblastoma tumours form?

The causes of retinoblastoma (RB) can now be described with considerable accuracy, although many details are still unclear. Understanding the genetic changes leading to RB has provided an awareness of general mechanisms of cancer development and progression, previously only suspected. From the basic understanding have come new diagnostic technologies that are now ready to be applied directly to RB patients and their families, and a rational approach, based on this understanding, will help us to develop new therapies that avoid the severe complications of conventional treatment.