DIGE: Past and Future - Difference Gel Electrophoresis (DIGE): Methods and Protocols - page 6

(Cy) dyes. Alan and I went to lunch one day and I recounted the
four criteria for designing DIGE dyes. He said “oh, that's easy”
and immediately drew two structures on a napkin. True to CMU's
open-door, interdisciplinary policy, I recruited a fi rst year graduate
student, Mustafa Ünlü, from the Chemistry Department. The fi rst
pair of DIGE dyes, Cy3-NHS and Cy5-NHS, were synthesized in
a few months. These are lysine-reactive dyes that have a net posi-
tive charge and differ in molecular weight by 2 Da.

Since there were no fl uorescent-gel imagers that detected Cy3 and
Cy5 at that time, we had to build our own imager. The very fi rst
imager was constructed from a 12-bit, cooled CCD camera
mounted on a darkroom enlarger stand. The light source was a
35-mm slide projector mounted on a repurposed rail from another
darkroom enlarger stand, and the fl uorescent fi lters were mounted
in a manual fi lter turret. This primitive imager was housed in a
darkened room. The initial images were encouraging, but needless
to say this was not a very light-tight arrangement. The next-generation
imager was made more light-tight by building the imager around
an IKEA cabinet with holes cut in the top and sides for the CCD
camera and illuminator, respectively. This imager was suffi cient
to begin optimizing the protein labeling reaction. Our initial plan
was to saturation label all lysine residues of all proteins in a cellular
extract. This gave very poor results since the proteins tended to
precipitate. In retrospect, it was not a good idea to saturation
label lysine as this addition would increase the protein mass by
about 25%. Instead, substoichiometric labeling turned out to be
the best way to maintain protein solubility and avoid size hetero-
geneity due to multiple dyes molecules bound. In substoichiometric
(or minimal) labeling, about 2-3% of all lysine residues are labeled.
This translates to about 5% of all proteins having a single fl uorescent
dye bound, while the rest are unlabeled. This set of innovations led
to the fi rst published report on DIGE ( 1 ).

2.5. The First DIGE
Experiments

2.6. Taking DIGE to
Commercialization

A great deal of my lab's efforts was assisted by Lans Taylor and
Alan Waggoner's Center for Light Microscopy and Biotechnology,
an NSF Science Technology Center (STC). Lans and Alan had
spun off several of their inventions into companies that were even-
tually purchased by larger companies. Amersham, plc (now part of
GE Healthcare) was interested in licensing the rights to the STC's
suite of cyanine dyes, including the DIGE dyes, which Lans and Alan
negotiated masterfully. Working with Amersham to commercialize
DIGE was an important learning experience. They were meticulous
in establishing a robust, reliable protocol. Since the initial develop-
ment of Cy3-NHS and Cy5-NHS, Amersham/GE has introduced
Cy2-NHS for three-color minimal labeling and Cy3-mal and
Cy5-mal for saturation labeling of cysteines. Amersham/GE also
developed fl uorescent-gel imaging systems and image analysis