Daniel C. Christophe
Institut de Biologie et de Médecine Moléculaires (IBMM)
IRIBHN-Faculté de Médecine, Université Libre de Bruxelles
Rue Profs. Jeener et Brachet, 12
B-6041 GOSSELIES, BELGIUM
telephone: +32-2-650 9828
fax: +32-2-650 9820
presenter: D. Christophe
D. Christophe1, B. Pichon1, D. Mercan2, V.
Pouillon1, C. Christophe-Hobertus1
1Institut de Biologie et de Médecine Moléculaires (IBMM-IRIBHN), Université Libre de Bruxelles, B-6041 Gosselies and
2Service de Chimie, Hôpital Erasme, B-1070 Bruxelles, Belgium.
Determining the subcellular localization of a gene product is instrumental in the elucidation of its function in the cell. We have developped an experimental strategy allowing the direct identification of DNA sequences encoding nuclear proteins by transient transfection of cDNA libraries in cultured cells.
Briefly, our method involves the following steps: (i) the construction of a green fluorescent protein (GFP) fusion proteins library in an expression vector harboring the SV40 origin of replication and the kanamycin resistance gene; (ii) the dilution of the library DNA in a 100-fold excess of unrelated plasmid DNA prior to transfection into COS- 7 cells, which results in the expression of individual library constructs in most of transfected cells; (iii) the permeabilization of the transfected cells by saponin, which results in the release of most of the cytoplasmic GFP fusions proteins, and immediate sorting by flow cytometry of those cells that still contain large amounts of (essentially nuclear) GFP fusions proteins; and (iv) the recovery of the transfected library DNA by extraction of the DNA from the pool of sorted cells, transformation of bacteria and selection of kanamycin-resistant clones. The experimental conditions used in these steps were optimized using a mixture of DNA constructs encoding nuclear and cytoplasmic GFP fusion proteins.
The technique was then used to screen a dog thyroid cDNA library. Fluorescence
microscopy observation of cells transfected with the library DNA revealed that
fluorescence was essentially nuclear in only 1% of the GFP-expressing cells
. Starting from 1,000,000 transfected cells (containing about 100,000 GFP-expressing
cells), we isolated by FACSing a pool of 116 cells, and 560 bacterial clones
were obtained after transformation with the extracted DNA. Individual transfection
of 110 out of these clones and fluorescence microscopy observation showed that
73% of them encoded nuclear GFP-fusion proteins. Partial DNA sequencing of the
clones revealed that 45 out of the 63 different sequences isolated corresponded
to known cDNAs, more than half of these (25) encoding known nuclear proteins.
Thirteen other clones harbored uncharacterized DNA sequences. Further characterization
of 4 of the cDNAs led to the clear identification of 2 novel nuclear proteins.
The method described here is easy to apply and may be scaled up, which opens the way to a wide search for novel nuclear proteins.
Abstracts * Speakers * Organizers * Home