Virginia Cornish
Helena Rubinstein Professor
Department of Chemistry

work : +1 212-854-5209

Cornish Group

Research: In Vivo Imaging


Green fluorescent protein (GFP) and its derivatives have emerged as invaluable tools for tagging and visualizing proteins in vivo.  GFP can be appended to the protein of interest simply by engineering the GFP fusion at the DNA level using standard molecular biology techniques.  The localization, interaction and fate of the labeled protein can then be monitored in real time in intact cells using live cell imaging.  While in theory the fluorescent properties of GFP can be modulated by mutating the protein sequence, in practice only limited changes and improvements to GFP have been possible because the chromophore is limited by the protein sequence.  The difficulty in engineering well-behaved GFP variants with distinct absorption and emission maxima has hindered multi-colored tagging and FRET applications.  Long before GFP was employed for protein labeling, proteins were labeled in vitro with small molecule fluorophores.  For example, a unique Cys residue could be engineered on the surface of a protein and then labeled with a thiol-reactive small molecule.  The advantage of labeling a protein with a small molecule is that the fluorescent properties of the small molecule can be readily varied and different types of labels, such as a photoaffinity label, can be employed.  The problem, however, in extending this small molecule approach to in vivo labeling is that it is difficult to selectively label the protein of interest in the sea of proteins and other reactive species present in the cell.  What is needed are approaches that combine the ability to genetically encode the label as for GFP with the flexibility of small molecule labels.  In collaboration with the Sheetz laboratory in the Biological Sciences Department at Columbia, we are exploiting the high-affinity interaction between Mtx and DHFR to label proteins in vivo by fusing the protein of interest to DHFR and then labeling the protein with small-molecule Mtx conjugates.  Recently, we have demonstrated the feasibility of this approach, labeling both a plasma membrane and a nuclear protein fused to DHFR with Mtx-Texas Red in Chinese Hamster Ovary (CHO) cells.  This work has been published in Angew. Chem.  Currently, we are focused on engineering orthogonal Mtx/DHFR variants for selectively labeling multiple proteins in the same cell and using the Mtx conjugates for applications beyond simple fluorescent labeling.

Back to Top  

Methotrexate-Texas Red as an RFP Surrogate

[Published in L. Miller, J. Sable, P. Goelet, M. Sheetz, V.W. Cornish.  "Methotrexate Conjugates: A Molecular In Vivo Protein Tag."  Angew. Chem. Int. Ed., 43, 1672-1675 (2004).]

The goal of this work was to show that a fluorescent conjugate of Mtx could be used to label an E. coli DHFR fusion protein in cultured mammalian cells.  DHFR (-/-) Chinese Hamster Ovary (CHO) cells were grown on coverslips, transfected with DNA encoding either plasma membrane (PM) or nuclear-localized DHFR, incubated in a medium containing Mtx-Texas Red (Mtx-TR), washed and imaged with a confocal fluorescent microscope.  Mtx-TR efficiently labeled the fusion proteins, and the labeling could be competed out with an excess of free Mtx.  This work establishes the feasibility of using the noncovalent interaction between Mtx and DHFR to label proteins in vivo with a wide variety of Mtx small molecule conjugates.

Back to Top  

Orthogonal Methotrexate Tag

[L. Miller, M. Sheetz, V.W. Cornish, unpublished data.]

To allow labeling in a wide variety of cell types, a methotrexate-DHFR variant is needed that
has no cross-reactivity with the natural cellular machinery.  Toward this end, we are using a “bump-hole” strategy to engineer an orthogonal methotrexate-DHFR pair.

Back to Top  

Home > Research > In Vivo Labeling