| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department Molecular Genetics, German Cancer Research Center (DKFZ), Heidelberg, Germany; and
Department of Internal Medicine III, Medical School Ulm, Ulm, Germany
1Correspondence: Abteilung "Molekulare Genetik" (B060), Deutsches Krebsforschungszentrum (DKFZ), Im Neuenheimer Feld 280, Heidelberg D-69120, Germany. E-mail: m.macleod{at}dkfz.de
SPECIFIC AIMS
A broad range of malignant diseases, such as mantle cell lymphoma (MCL), is associated with complex genomic alterations, demanding multimodal functional testing of candidate genes. The purpose of this study was to develop a single cellular tool for the functional analysis of candidate genes in a cellular disease system, suited for testing activated as well as inactivated genes. A novel bidirectional targeted transgenesis system was developed and applied to MCL cells, which allowed single copy integration of transgenes in an isogenic background. The tested transgenes antagonized the expresssion of originally activated or inactivated genes.
PRINCIPAL FINDINGS
1. Construction of a cellular MCL model system
Mantle cell lymphoma (MCL), a non-Hodgkin B-cell lymphoma, is characterized by a block in differentiation and deregulation of cell proliferation, as well as insufficient induction of apoptosis. The characteristic translocation t(11;14)(q31;q32) results in a strong overexpression of the cyclin D1 gene (CCND1). The cell line Granta-519, originally established from a MCL patient, carries most of the MCL typical genomic aberrations, namely t(11;14), del 1p22, del 9p21, del 13q14, del 17p13, +18q21. A targeting vector with a selection cassette was integrated into the genome of Granta-519 cells to create an isogenic cell system for bidirectional recombinase-mediated cassette exchange (RMCE). The cassette consists of three elements: a neomycin gene for positive selection, a Herpes simplex virus thymidine kinase (HSV-Tk) gene for negative selection, as well as the lox-sites loxP and loxP511 flanking both selection markers and allowing for sequence-specific Cre-mediated recombination (see 
Fig. 3
). This RMCE-system permits i) knocking down activated oncogenes via RNAi, as well as ii) complementing inactivated tumor suppressor genes via knock-in of genomic DNA. Complementation by genomic DNA with the respective BAC-clones allows wild-type promoter regulated gene expression. Alternatively, to counteract overexpression of activated genes, we stably integrated a modified RNAi-plasmid (pSUPERdL) coding for gene-specific shRNA. Following cotransfection of a Cre-expression plasmid, the selection markers HSV-Tk and neo are substituted by a single copy of the cloned insert, and the new subclones are negatively selectable via ganciclovir (GCV).
|
|
|
2. Site-specific and single copy genomic integration of transgenes
The genes CDKN2A/B, biallelically deleted in Granta-519, were complemented using the respective BAC-clone RP11–149I2. RMCE-mediated site-specific integration resulted in new sublines with a single BAC-clone insert. B-cell lymphocytes, like Granta-519, are notoriously difficult to transfect, in particular using BAC DNA. The high specificity of the RMCE system shown in the present study overcomes this limitation, since modulation of the cells was obtained even with low transfection efficiencies. To knock down the CCND1 overexpression, an insert coding for CCND1-shRNA was stably integrated into the Granta-519 subline. Further sublines were generated by integration of an empty pSUPERdL vector and an insert coding for unspecific firefly luciferase-shRNA. These clones served as additional controls for locus or position dependent effects and unspecific shRNA-effects like IFN response.
3. Antagonizing gene activity of the oncogene CCND1 and the tumor suppressor gene CDKN2 in an isogenic system
To validate our strategy for the specific modulation of gene activity, we performed expression studies on both RNA and protein concentration. CDNK2A/B transcripts derived from stably integrated BAC DNA were detected by RT-PCR. Three different fragments were amplified, two of which were specific for CDKN2A and one was specific for CDKN2B. Based on the distance of the exons of both genes, this finding indicates integration of the complete insert of the original BAC clone. Additionally, restoration of the expression of protein coded by the CDNK2A/B-BAC clone was detected by immunocytochemical staining of CDKN2B, which was absent in unmodified Granta-519 cells (Fig. 1
A). Analyzing the stable CCND1-shRNA cells, we observed significant decrease in CCND1 protein concentration (Fig. 1)
. This was not observed in subclones stably transfected with the empty vector pSUPERdL, the selection cassette or an unspecific shRNA. Thus, we achieved a highly efficient knock-down of an overexpressed target gene from a single pol III-shRNA transcription unit.
4. Prevention of RNAi induced IFN response
Since RNAi reflects an evolutionary conserved, endogenous defense mechanism against virus infection in mammalian cells, the potentially costimulated IFN-response might be difficult to distinguish from effects by the gene knock-down. However, the stable CCND1-shRNA modulated cell clones did not show any up-regulation of IFN-stimulated genes like OAS1 and STAT1. In contrast, transient expression of CCND1-shRNA induced OAS1 expression. Since stable integration of the same shRNA coding insert did not induce OAS1 expression, this effect cannot be due to the CCND1 short hairpin sequence. OAS1 or STAT1 activation had been reported in other studies applying lentiviral shRNA- or synthetic siRNA techniques.
5. Proliferation of the modulated cells is inhibited
The functional consequences of modified Granta-519 subclones were assessed using proliferation assays. The RMCE modulated cell sublines complementing loss of CDKN2A/B and knocking down CCND1, respectively, were each derived from two different parental lines with integration of the selection cassette on chromosome 18q (G18) or chromosome 11q (G11). The two control sublines modified with empty vector or unspecific shRNA (shRNA-Luc) were derived from the parental line G18. As shown in Fig. 2
, all four sublines modulating CDKN2A/B- or CCND1-gene activity (BAC-CDKN2 in G11 and G18, shRNA-CCND1 in G11 and G18) displayed a decrease in proliferation after selection to monoclonality, whereas all five control lines (+shRNA-Luc, +empty vector, G18, G11, Granta-519) display the typical doubling time characteristic of Granta-519 cells. Thus, restoration of CDKN2A/B expression resulted in an inhibition of cell proliferation and down-regulation of CCND1 yielded a decrease of cycling cells. With respect to cell proliferation, knock down of CCND1 was sufficient to antagonize a tumor-specific gene activation. Due to the vital role of CCND1, such a test would be impossible in a knock out system. The inhibition of proliferation by restored CDKN2A/B activity was so efficient that it was difficult to maintain the respective cell clones in culture, highlighting the dramatic effect of CDKN2A/B in the pathogenesis of MCL.
CONCLUSIONS AND SIGNIFICANCE
In conclusion, we here report for the first time a stable modulation and correction strategy for gene expression regardless of whether the gene is inactivated, deleted or overexpressed. We document the establishment and application of a novel and flexible RMCE targeted transgenesis system, allowing for either knocking down up-regulated genes via RNA interference or complementing inactivated genes via knock-in of genomic DNA in an isogenic system (Fig. 3)
. By using this birectional site-specific cellular tool, we can analyze candidate genes in a specific cellular disease system, to mediate single copy integration events of transgenes, and to restore gene expression of a formerly inactivated candidate gene. Reconstitution of gene defects regulated via endogenous promoters, enhancers, alternative transcription start sites, and splice variants of a specific gene is of distinct advantage, compared to a cDNA transfer. Therefore, our strategy of targeted integration of transgenes into a disease-background via RMCE to compensate for genomic deletion or inactivation of genes is based on the utilization of BACs containing the entire genomic information including regulatory elements. These BACs became widely available through the Human Genome Project (http://www.ensembl.org).
Notably, the presented site-specific shRNA-strategy circumvents interference by IFN-response induced when using other RNAi gene knock-down methods. Thus, our system results in a highly specific gene knock-down without unspecific effects due to the IFN system.
The versatile transgenesis system presented in this study can be used as a screening tool exploiting BAC or shRNA plasmid libraries. It is a well-suited system to analyze the relevance of a single genomic aberration in the context of multiple genomic alterations. In animal models, genomic changes characteristic for human tumors are often introduced to mimic tumorigenesis and to test for therapeutic strategies. In contrast, the efficient RMCE-based modulation of human cells derived from disease tissue, carrying different genetic alterations, provides a powerful alternative to study pathogenic gene function and to test interfering therapeutic measures. In conclusion, we here demonstrate that adequate restoration of a range of different gene activities yields in a desired antiproliferative effect in MCL-derived cells. By antagonizing inactivated tumor suppressor genes or activated oncogenes, this approach can be readily used for the functional analysis of a broad range of disease-related genetic defects.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.05-4854fje
This article has been cited by other articles:
|
|
 |
|
  S. Wang, Y. Zhao, M. A. Leiby, and J. Zhu Studying human telomerase gene transcription by a chromatinized reporter generated by recombinase-mediated targeting of a bacterial artificial chromosome Nucleic Acids Res., September 1, 2009; 37(17): e111 - e111. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Heck, J. Rom, V. Thewes, N. Becker, B. Blume, H. P. Sinn, U. Deuschle, C. Sohn, A. Schneeweiss, and P. Lichter Estrogen-Related Receptor {alpha} Expression and Function Is Associated with the Transcriptional Coregulator AIB1 in Breast Carcinoma Cancer Res., June 15, 2009; 69(12): 5186 - 5193. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Ortega-Paino, J. Fransson, S. Ek, and C. A. K. Borrebaeck Functionally associated targets in mantle cell lymphoma as defined by DNA microarrays and RNA interference Blood, February 1, 2008; 111(3): 1617 - 1624. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
