Featured Paper Interview
Essential Science IndicatorsSM from
Reuters, the paper ranked at #5 in the field of
Biology & Biochemistry is the August 2000
Science paper by Krzysztof Palczewski and his team,
"Crystal structure of rhodopsin: a G protein-coupled
receptor," (289: 739-45, 4 August 2000), with
2,467 cites up to June 30th of this
Dr. Palczewski's record in the database includes 171 papers, the
majority of which are classified in Biology & Biochemistry, cited a
total of 8,609 times. He is the John H. Hord Professor and Chair of the
Department of Pharmacology at the Case Western Reserve University School of
Medicine in Cleveland, OH.
In the interview below,
ScienceWatch.com correspondent Gary Taubes talks
with Dr. Palczewski about this paper and his subsequent
work with rhodopsin.
What were you working on in the years leading up
to your landmark 2000 publication in Science on the crystal
structure of rhodopsin?
We were always interested in molecular detail of vision, namely the
understanding of proteins, protein interactions, and signal transduction
involved in eyesight at a structural level. We'd been working for many
years on different proteins of visual signal transduction. It wasn't until
a very talented post-doc from Japan came to my lab—Tetsuji
Okada—that we put a lot of resources into studying rhodopsin, which
is the key protein of visual signal transduction. He came from Japan with
several ideas and a significant understanding of rhodopsin. We put this
together with resources in the laboratory to crystallize rhodopsin and
identify its crystal structure.
Considering rhodopsin's key role in visual processing,
why hadn't this been done before?
It had been tried by several laboratories. But the prevailing understanding
of membrane protein crystallography was that these are very unstable
proteins. They do not crystallize well, they don't form a proper crystal
lattice, and they're just too difficult to handle. And all this is
absolutely true. Because of these difficulties, it's still the case that
only a very small number of integral membrane proteins have been
Once you know how these proteins behave in different detergents, they're
not that different than soluble proteins. I think this was a major
breakthrough at that time: realizing that some of these G protein-coupled
receptors (GPCRs) are stable enough for crystallization. Since then, there
have been more efforts to crystallize other proteins. Today five or six
other GPCRs have been crystallized, most in the last year.
"Our work on vision now extends to retinal diseases
and to other signaling processes in the
So I think the original reason rhodopsin hadn't been crystallized was a
combination of a lack of faith that it could be accomplished, and the
difficulty in obtaining the quality and quantity of highly purified
rhodopsin that was needed. With a little bit of luck and a tremendous
amount of work—there were many collaborators on that 2000
Science paper—all this was possible.
Was there one particular obstacle you had to overcome to
get rhodopsin crystallized and elucidate the structure?
Yes, it was the purification. The technique for doing that was invented by
Tetsuji Okada. It was critical to obtain high-purity material. After our
work was published, three groups came out with different protocols for
purification that were also successful, so it wasn't absolutely necessary
to follow ours. So why did we get there first? I think it was the
combination of a very careful production, isolation, and purification of
the protein and a very careful analysis of the crystals that led to that
What is it that makes rhodopsin itself such an important
receptor, so much so that your paper has been averaging a few hundred
citations a year for almost a decade?
Rhodopsin is part of this family of receptors called G protein coupled
receptors—or "GPCRs," as I've mentioned. There are about 800 members
of the same family and they're very critical for virtually all
physiological functions. They're involved in sensory transduction, for
example, in smell and taste as well as vision, and also in many other
biological processes. Virtually all such processes are modulated by this
class of receptors.
A cell has to communicate with the external world; it has to receive clues
about what the neighbors are doing and it has to receive long-distance
signals from the brain and elsewhere, and the mechanisms by which those
signals are transmitted from the outside of the cell to the inside requires
some sort of receptor. Most of these receptors, although by no means all,
So there are hundreds of other GPCRs and a great need in the pharmaceutical
industry to understand their structures for rational drug design. Today
about 50% of all drugs on the market influence receptors of signaling
proteins. That stimulates a tremendous interest from a pharmaceutical
perspective. I think the work on rhodopsin from my lab and from others
really opened the door to the belief that it's possible to crystallize
these receptors. It also allowed us to begin understanding the function of
these receptors at atomic resolution rather than at just a crude
So only five or six of the 800 have been crystallized
and most of those in the past year. That doesn't seem like a lot. Why
isn't that number significantly larger?
Yes, it's not a lot. The problem is that each of these other membrane
proteins exists in very small quantities in nature and they require very
special handling. We need to know the properties of each of these
particular receptors very, very well in order to be successful. The next
step will be to develop a method that is truly generic and can be applied
to all GPCRs. It will have to start with the expression, purification, and
crystallization. That's an effort of ours as well—to get all three of
these steps lined up in a very generic way.
I think if we can do that we'll advance the field as much and as rapidly as
we did with the rhodopsin structure alone nine years ago—if we can
come back and say, here's the method, take your receptor of interest, and
it will work! And I think it's worthwhile putting a lot of energy to make
just that happen. Every time we get another structure, it's a wonderful
accomplishment. But the true goal will be a method that is robust and
applicable to all GPCRs, not just one.
What's the major problem that has to be solved to make
this a reality?
The main roadblock that still prohibits fast resolution of these structures
is expression of the protein, and that's going to be a problem until more
effort is given to cell biology. We’ve made a lot of progress in the
past 10 years making the construct to drive expression rapidly, but not the
folding machinery to provide the platform for proper folding of these
GPCRs. This is still a major problem.
When a GPCR is synthesized in the cell, it's post-translationally modified
and transported from the cell's endoplasmic reticulum to the plasma
membrane where proof-reading machinery checks to see if it's properly
folded. Without this process, you get a protein that's incorrectly folded.
Then you end up purifying a protein that's in the wrong conformation for
crystallization. So separation of properly folded and improperly folded
protein is a big issue. We've made a lot of progress since 2000, but this
problem still has to be solved.
Are you making progress understanding these receptors
even without having the structures?
"Rhodopsin is part of this family of
receptors called G protein coupled
receptors...There are about 800 members of
the same family and they're very critical for
virtually all physiological
Well, this year, for example, we had a paper in Molecular
Pharmacology (Mustafi D, Palczewski K, "Topology of class A G
protein-coupled receptors: Insights gained from crystal structures of
rhodopsins, adrenergic and adenosine receptors," 75: 1-12, January 2009)
where we showed that all those structures are really very homologous to
each other, so we are making some progress. We also stress the differences,
which account for the specific Actions of these proteins.
But the overall topology, the overall design, like that of a house, is very
similar. Between all these GPCRs, some 800 sequences, we can say with some
confidence that the overall topology, the three-dimensional structure is
going to be very similar. Although obviously each will have some subtle
changes to react specifically with one hormone and not another.
We talked seven years ago and at the time you said you
had the structure of rhodopsin but not the structure of the form it
takes in the cell membranes. Did you ever get that structure figured
In order to do that we needed different technologies than just
crystallography. But we did it. In 2003, we published the membrane
organization of rhodopsin in Nature (Fotiadis D, et al.,
Atomic-force microscopy: Rhodopsin dimmers in native disc membranes,"
421: 127-8, 9 January 2003) and the Journal of Biological
Chemistry (Liang Y, et al., "Organization of the G
protein-coupled receptors rhodopsin and opsin in native membranes,"
278: 21655-62, 13 June 2003).
Those results showed that rhodopsin exists as a dimeric structure. There
are two receptors that form a unit. And this appears to be the prevailing
finding shown by many other methods as well—all GPCRs may function as
a dimer, as a couple. That work, which was done in collaboration with
Andreas Engel, is also very highly cited, although not as much as the 2000
How much do you still work on vision and how much on G
protein-coupled receptors in general?
We have an active research program in both. Our work on vision now extends
to retinal diseases and to other signaling processes in the retina. That's
a large fraction of my lab's interest. In 2006, for instance, we published
an article in PNAS on the photo-activated structure of rhodopsin
(Salom D, et al., "Crystal structure of a photoactivated
deprotonated intermediate of rhodopsin," 103: 16123-8, 31 October
2006). This is what rhodopsin looks like when it's activated by light. It
took six years to generate those crystals.
So I would say my lab is now working on three areas—one is retinal
processes and diseases and another is the rhodopsin structure and its
interactions with other proteins. The third is our work on other GPCRs as
Has there been any element of serendipity in your
research—some accomplishment which came about because you just
In my case, I can you tell I've just been extremely lucky to be given the
opportunity to work in science in this country. I cannot stress enough how
US science is open to immigrants and how much those of us who have come
from another country appreciate that and how much we want to contribute.
The support we receive—in terms of positions, grants,
training—is really phenomenal!
Krzysztof Palczewski, Ph.D.
Department of Pharmacology
School of Medicine
Case Western Reserve University
Cleveland, OH, USA
Krzysztof Palczewski's current
most-cited paper in
Indicators, with 2,467
Palczewski K, et al., "Crystal structure of rhodopsin:
a G protein-coupled receptor," Science 289(5480):
739-45, 4 August 2000. 2,467 cites. Source:
Essential Science Indicators from
Read a classic Science
Watch® interview with Krzysztof
KEYWORDS: RHODOPSIN, CRYSTAL STRUCTURE, G PROTEIN-COUPLED
RECEPTOR, GPCR, PROTEIN INTERACTIONS, VISUAL SIGNAL TRANSDUCTION,
INTEGRAL MEMBRANE PROTEINS, PURIFICATION TECHNIQUE, SENSORY
TRANSDUCTION, RATIONAL DRUG DESIGN, FOLDING MACHINERY, MEMBRANE
ORGANIZATION, RETINAL PROCESSES, RETINAL DISEASES.