Advertisement
English transcript
My name is Bob Lefkowitz, I'm an investigator of the Howard Hughes Medical Institute, I have been for many years,
and a professor of medicine and biochemistry at Duke University Medical Center, in Durham, North Carolina.
Today, I'm going to be talking to you about my favorite scientific subject, receptors for hormones and drugs.
I'm going to give two lectures. The first one will review some of the history of this field over the past forty or so years,
that I've personally been involved, which has brought us to our current understanding of drug and hormone receptors.
Then, in the second lecture, I'm going to try to bring you up to date on some recent, exciting research
going on in my laboratory and in other laboratories around the world,
which is changing in some very fundamental ways our understanding of how the receptors function and how they're regulated.
Moreover, I'll tell you in that second lecture about how this new information can be harnessed
to develop entirely new classes of pharmacological agents.
The receptors that I'll be talking about today are known as the seven transmembrane receptors,
or also known as G-protein coupled receptors.
In this first lecture, I'm going to review some of the history
of how we have come to understand this remarkable family of plasma membrane receptors.
Let me begin though, with a few definitions.
A receptor is a molecule on or in a cell with which a drug or hormone or a neurotransmitter initially interacts.
A ligand is a compound which interacts specifically with the receptor.
There are two main types of ligands: agonists, which stimulate the receptor,
and antagonists, which bind to, but do not stimulate, the receptor.
They are sometimes known as "blockers."
This little animation, which you see here, depicts an agonist interacting with a receptor
and changing its conformation, or shape, so that it can signal.
Two examples of agonists which work through this particular family of seven transmembrane receptors
are adrenaline and morphine.
In this animation, an antagonist is shown, in the form of this capsule, binding to the receptor,
but not changing its shape, and thereby, not leading to biological signals.
Several common examples of blockers, or antagonists, that you may be aware of are beta-blockers,
which are simply short-hand for beta adrenergic receptor blockers or antagonists,
anti-histamines or angiotensin receptor blockers, also known by the shorthand ARBs.
Now, today, I'm going to be talking about this remarkable family of seven transmembrane receptors,
or G-protein coupled receptors, as they're also known.
The name seven-transmembrane receptor, of course, derives from their highly conserved structure,
which we'll talk more about in just a few moments,
which features a poly-peptide chain which crisscrosses the plasma membrane seven times.
This remarkable family of receptors contains about 800 to a thousand members
and regulates virtually all known physiological processes in mammals and humans.
In particular, many of the sensory receptors, such as those that mediate taste and smell and vision,
are members of this family, as are so many drug receptors,
such as those for beta blockers, anti-histamines, glucagon, to mention just a few.
But what makes this family of receptors even more important
is their central place in modern pharmacotherapy of human illnesses.
It's said that somewhere north of fifty percent of all prescription drugs sold in the world today
are drugs which function as agonists or antagonists for various members of this family of receptors.
Let's take a moment to review some of the basic characteristics of how the seven transmembrane,
or G-protein coupled receptors, function.
I'm sure many of you are familiar with this information,
but I thought it would be a good idea to make sure that everybody is on the same page.
So, when one of these receptors, for example the so-called beta adrenergic receptor,
for adrenaline or noradrenaline, is stimulated by its agonist, it changes shape, or conformation.
This allows it to interact with a heterotrimeric G-protein, in this case, Gs.
These proteins, of course, are referred to as heterotrimeric because they consist of three distinct subunits,
in this case, termed alpha, beta and gamma.
The interaction of the agonist activated receptor with the G-protein is thought to dissociate it,
or at least loosen the association of the alpha subunit from the beta-gamma subunit dimer.
Then, both the alpha subunit and the beta-gamma subunits can independently function by activating various downstream signaling systems.
Of course, the classical pathway is for activation of second messenger generating enzymes,
such as adenylate cyclase or phospholipase C, which would generate cyclic AMP or diacyl glycerol.
Now, although receptors have been an interest of pharmacologists for more than a century,
quite remarkably, there remained great skepticism as to whether receptors existed in specific biophysical terms
as specific molecules in the cell that could be studied, until quite recently.
In fact, I've put up here a quote from a very influential and very well-known pharmacologist, of the twentieth century,
Raymond Ahlquist. Now, ironically, Raymond Ahlquist, a very decorated scientist, and a winner of the Lasker prize,
amongst others, was the man who first suggested that there might be two distinct types of receptor
for adrenaline and adrenaline-like molecules, which he referred to as alpha and beta adrenergic receptors.
This seminal contribution would ultimately lay the basis for the development of beta blockers by Dr. James Black.
However, as late as 1973, Ahlquist himself wrote the following
about the idea that receptors might actually represent specific molecules. He wrote:
"This would be true if I were so presumptuous as to believe that alpha and beta receptors really did exist.
There are those that think so and even propose to describe their intimate structure.
To me they are an abstract concept conceived to explain observed responses of tissues
produced by chemicals of various structure."
Well, this quote, from 1973, corresponds to approximately the time that I began my own, independent work
in my own laboratory at Duke
And it was clear to me and to others in the field at the time
that if we wanted to bring these mythical receptors to life, to show that they really did exist as specific molecules,
it would be necessary to develop a whole new set of technologies,
and the first of these would have to be some form of radioligand binding.
Now, by radioligand binding, we mean the idea that one could take a ligand,
be it agonist or antagonist, radioactively label it, and physically demonstrate the binding of that radioligand to the receptor.
If we could do this, we would have a way of studying the receptor's properties,
studying how they were regulated, and perhaps, a way of discovering new drugs,
and most importantly, a means for following the receptors during purification efforts,
which might allow us to isolate the receptors and thereby learn about their molecular properties.
Let me give you an example of a radioligand binding experiment,
the kinds of experiments that we did some years ago,
and which led to some very interesting new information about how the receptors function.
Shown in this top panel is a typical example of what's called a radioligand displacement curve.
In this experiment, a radioactively labeled beta-blocker was bound to beta adrenergic receptors,
which are coupled to adenylate cyclase, in the plasma membranes of cells.
Then, an unlabeled antagonist, such as alprenolol or propoanolol, is allowed to compete at increasing concentrations,
and to block the binding of the radioligand. As you can see, this leads to a classical sigmoidal dose response curve;
we refer to this as being uni-phasic, it has a Hill slope of 1, it is Michealean.
In striking contrast, as shown in the bottom panel, if the same experiment is done with a beta-adrenergic agonist,
such as isoproterenol, one gets a much shallower curve, with a much lower Hill slope,
which can be resolved by computer fitting to two distinct affinity states: one of high affinity and one of low affinity.
If one adds a guanine nucleotide, which will bind to the alpha subunit of the G-protein,
then as you can see, the curve is shifted to the right, and steepened.
And in fact, the affinity of this now steep, or uniphasic curve, corresponds to the lower of the two affinities, shown here.
A great deal of analysis of these types of curves taught us that agonists are able to stabilize a high affinity form of the receptor.
This was set forth many years ago in the so-called ternary complex model.
In this model, there's an initial step, where the agonist, in this case depicted as 'H' for hormone,
binds to the receptor to form a low affinity initial hormone-receptor, or agonist-receptor complex.
When H is an agonist, a presumed conformational change in the receptor occurs
which at the time was hypothesized to lead to the binding of the HR complex to a third component,
hence the name ternary complex, leading to the formation of the high affinity HRX complex.
When a guanine nucleotide, such as GTP binds to this complex, it leads to its dissociation back to the low affinity state,
and hence the marked shift in that curve. At the time the ternary complex was originally developed,
it was not clear what X was, but the pronounced effect of guanine nucleotides on the stability of this complex
immediately suggested that X might be the recently, at that time, discovered GTP binding, or guanine nucleotide binding, G-protein.
In fact, this model proved extremely useful over the years in dissecting such ligand binding curves
and led to many insights. For example, the stronger the agonist, in terms of its so called efficacy,
and ability to promote signaling, the greater the stability of that complex, and the higher the affinity of the high-affinity complex.
So, this type of modeling proved very, very useful in developing models for how receptors were working together with G-proteins.
Now, once one had in hand such radioligand binding methods,
it became possible to approach the difficult issue of trying to solubilize the receptors from the plasma membranes,
where they exist, and beginning the difficult work of purifying them.
It's important to understand that members of this family,
the so-called G-protein coupled, or seven membrane spanning receptors,
are present at extraordinarily low concentrations in the plasma membranes of cells
and would require several hundred thousand-fold purification in order to actually isolate them away from the other proteins.
This is obviously an extremely difficult task and one which occupied my laboratory for many years.
As our models, we used the adrenergic receptors, the receptors for adrenaline and related molecules,
and focused on one in particular, the so-called beta 2 adrenergic receptors.
The technique that proved up to the job, so to speak, and which was so difficult,
was the technique of affinity chromatography.
Now, in this technique, if one is trying to isolate a molecule, in this case the receptors,
and one has a specific ligand, we had several, for that receptor, one takes the ligand,
in this case that would be a beta adrenergic antagonist, such as alprenolol, and covalently couples it to a solid support,
such as a sepharose bead. One can then make a column of these beads, containing the bound ligand,
and if one has succeeded in solubilizing the receptors, which we were able to do after much trial and error,
with detergents, we were able to then apply the solubilized plasma membrane preparations, containing the receptors,
but also many, many other proteins, to the columns of these beads, and now the receptors would stick,
by virtue of their specific binding affinity for the bound ligand.
One could then wash the columns extensively to remove other proteins and then finally elute the receptors
with another ligand, such as alprenolol itself or any other beta adrenergic antagonist, or even agonist,
at high concentration. In doing this, one could achieve very, very high purification in a single step.
One could then combine this with a variety of other more conventional chromatographic methods to hopefully isolate the receptors.
We were able ultimately to succeed in purifying, to homogeneity, essentially all four of the then-known subtypes of adrenergic receptors,
called alpha 1, alpha 2, beta 1 and beta 2. And we did this by virtue of creating specific affinity chromatography columns,
in which we covalently linked either beta or alpha adrenergic antagonists to these sepharose beads.
This effort took about a dozen years in my laboratory and devoted work of many students and post-docs.
Ultimately, we succeeded in purifying all four of these subtypes to homogeneity.
This is a slide I used to enjoy showing back in the eighties because it essentially reduced to a single frame
the work of many, many students and fellows over more than ten years.
And what it is is an SDS polyacrylamide gel on which we have run three of the four subtypes of highly purified adrenergic receptors.
As you can see, each seems to migrate through the gel as a single polypeptide chain of about 60,000 Daltons.
In each of these cases, the purified protein, the isolated molecule, appeared to bind radioligands
and other adrenergic compounds with exactly the specificity and the stereospecificty
that one would expect for the individual adrenergic receptor subtype.
At this point, and this would have been the early 1980's, I would say,
we thought that these kinds of conclusive findings would convince the skeptics that the receptors really did exist
as specific isolated molecules, But in fact, the skepticism persisted in some quarters because, it was said,
well, how do you know that this isolated molecule, even though it can bind your drugs with the right specificity,
how do you know that it's really the receptor, that it's really capable of conveying responsiveness to an isolated system.
And to take that final step, we turned to the technique of receptor reconstitution.
So, in this method, we took our purified receptors, and we constituted them into phospholipid vesicles.
We wanted to now take a system which was unresponsive to adrenaline
and show that if we could insert this isolated molecule,
the putative receptor, into that system, we could convey responsiveness to the adrenaline.
So, what we needed was a cell that didn't have the receptors. Well, that turned out to be a difficult problem.
It turns out virtually all mammalian cells contain beta 2 adrenergic receptors.
However, by trial and error and searching, we found that erythrocytes from an amphibian, the African clawed toad,
Xenopus laevis, lacked beta adrenergic receptors and hence did not respond to adrenaline,
although it did contain the other components of the adenylate cyclase system,
the G-protein and the adenylate cyclase itself.
It also contained prostaglandin receptors,
and hence the adenylate cyclase of this cell could be stimulated to make cyclic-AMP
by prostaglandins, but not adrenaline.
So, we took these cells lacking the beta receptor and we mixed them with these phospholipid vesicles containing the receptor.
The vesicles fused with the plasma membrane of the cell, thereby inserting our putative receptors into the membrane.
And, lo and behold, the cell then became responsive to adrenaline
and its rate of cyclic AMP production increased then, in response to adrenaline.
And so, the last skeptics finally yielded and agreed that this molecule,
which could both bind beta-adrenergic compounds with appropriate specificity,
and to convey to a previously unresponsive cell responsiveness to adrenaline, was in fact the beta adrenergic receptor.
Once we had validated that the molecules we had isolated, and again in these days, back in the eighties,
we never had more than about 25 to 50 micrograms of purified receptor in hand at any one time,
but once we had rigorously validated that these molecules were the receptors,
we were ready to begin the difficult work of cloning the genes and cDNAs for these molecules,
to try to find out about their primary amino acid sequence.
And after several years of difficult work, in 1986, we reported the cloning of the beta-2 adrenergic receptor.
Now this was a bit of a landmark in the field, because to that point in time, no other G-protein coupled receptor had been cloned.
The only exception was the molecule rhodopsin.
Now, let's talk a little bit about rhodopsin. Rhodopsin, the so-called visual pigment,
is, if you will, a receptor for photons of light. When photons of light are captured by retinal,
bound to the opsin protein, it isomerizes, changes the configuration, or conformation of the rhodopsin,
and signals are emitted which actually activate a G-protein in the retina called transducin.
Now, back in the mid-eighties, most of us working in the signaling field were aware of the fact
that there was a great deal of functional analogy between the visual system
and the beta-adrenergic receptor stimulated adenylate cyclase.
Both worked through a G-protein, GS in one case, transducin in the other,
and both worked through an enzyme, either a cyclic GMP phosphodiesterase in the retina,
or an adenylate cyclase in other mammalian tissues, which responded to the G-protein.
But, nobody really thought that there was any structural analogy between rhodopsin
and all the other members of the G-protein coupled receptor family.
However, when we succeeded in 1986 in cloning the beta-2 adrenergic receptor,
we immediately could see its structural relationship to rhodopsin.
Like rhodopsin, it had seven stretches of hydrophobic amino acids, presumably representing helical transmembrane domains.
As shown in blue, many of the residues in the transmembrane domains were identical, homologous, with rhodopsin.
There were, of course, connecting external and internal loops.
Many sites for regulatory phosphorylation on the C-terminus, as in rhodopsin,
and toward the amino terminus, consensus sites for N-linked glycosylation.
When I lecture today about the discovery, more than 20 years ago, of the close structural relationship
between the beta-2 adrenergic receptor and rhodopsin,
people are always amazed that we were all so surprised to find this.
How could you not be thinking along those lines? At the time. You knew about the functional analogies,
this is true. Yet, as the contemporaneous literature reflects,
none of us ever expected that there would be this close structural relationship.
I think perhaps part of the reason for that had to do with even some prior history.
The structure of rhodopsin, which is a remarkably abundant protein,
it's about 90% or more of all the protein in isolated retinal rod outer segments,
Rhodopsin was available in large quantities and when its structure was determined in 1982,
it was originally, it's sequence was determined, not by cloning, but rather, by classical Edmond degradation,
protein sequencing, because there was so much protein available.
It was found to have a structure consistent with, suggestive of, seven transmembrane domains.
At the time only one other such protein was known, that is one other protein which had these seven transmembrane domains,
and that was bacteriorhodopsin. Now, bacteriorhodopsin is a protein present in very primitive archaebacteria,
which is a light driven proton pump. Now, when the structure of retinal rhodopsin was determined,
and it too was found to have seven-membrane spans,
it was very reasonably speculated by both groups who had succeeded in getting that sequence,
that the common feature of bacteriorhodopsin and rhodopsin was that they were light sensitive proteins.
And so, it was put forth as an hypothesis that the seven membrane spans must be a signature structural feature of light-sensitive proteins,
which seemed very reasonable at the time.
Only with the
…
and a professor of medicine and biochemistry at Duke University Medical Center, in Durham, North Carolina.
Today, I'm going to be talking to you about my favorite scientific subject, receptors for hormones and drugs.
I'm going to give two lectures. The first one will review some of the history of this field over the past forty or so years,
that I've personally been involved, which has brought us to our current understanding of drug and hormone receptors.
Then, in the second lecture, I'm going to try to bring you up to date on some recent, exciting research
going on in my laboratory and in other laboratories around the world,
which is changing in some very fundamental ways our understanding of how the receptors function and how they're regulated.
Moreover, I'll tell you in that second lecture about how this new information can be harnessed
to develop entirely new classes of pharmacological agents.
The receptors that I'll be talking about today are known as the seven transmembrane receptors,
or also known as G-protein coupled receptors.
In this first lecture, I'm going to review some of the history
of how we have come to understand this remarkable family of plasma membrane receptors.
Let me begin though, with a few definitions.
A receptor is a molecule on or in a cell with which a drug or hormone or a neurotransmitter initially interacts.
A ligand is a compound which interacts specifically with the receptor.
There are two main types of ligands: agonists, which stimulate the receptor,
and antagonists, which bind to, but do not stimulate, the receptor.
They are sometimes known as "blockers."
This little animation, which you see here, depicts an agonist interacting with a receptor
and changing its conformation, or shape, so that it can signal.
Two examples of agonists which work through this particular family of seven transmembrane receptors
are adrenaline and morphine.
In this animation, an antagonist is shown, in the form of this capsule, binding to the receptor,
but not changing its shape, and thereby, not leading to biological signals.
Several common examples of blockers, or antagonists, that you may be aware of are beta-blockers,
which are simply short-hand for beta adrenergic receptor blockers or antagonists,
anti-histamines or angiotensin receptor blockers, also known by the shorthand ARBs.
Now, today, I'm going to be talking about this remarkable family of seven transmembrane receptors,
or G-protein coupled receptors, as they're also known.
The name seven-transmembrane receptor, of course, derives from their highly conserved structure,
which we'll talk more about in just a few moments,
which features a poly-peptide chain which crisscrosses the plasma membrane seven times.
This remarkable family of receptors contains about 800 to a thousand members
and regulates virtually all known physiological processes in mammals and humans.
In particular, many of the sensory receptors, such as those that mediate taste and smell and vision,
are members of this family, as are so many drug receptors,
such as those for beta blockers, anti-histamines, glucagon, to mention just a few.
But what makes this family of receptors even more important
is their central place in modern pharmacotherapy of human illnesses.
It's said that somewhere north of fifty percent of all prescription drugs sold in the world today
are drugs which function as agonists or antagonists for various members of this family of receptors.
Let's take a moment to review some of the basic characteristics of how the seven transmembrane,
or G-protein coupled receptors, function.
I'm sure many of you are familiar with this information,
but I thought it would be a good idea to make sure that everybody is on the same page.
So, when one of these receptors, for example the so-called beta adrenergic receptor,
for adrenaline or noradrenaline, is stimulated by its agonist, it changes shape, or conformation.
This allows it to interact with a heterotrimeric G-protein, in this case, Gs.
These proteins, of course, are referred to as heterotrimeric because they consist of three distinct subunits,
in this case, termed alpha, beta and gamma.
The interaction of the agonist activated receptor with the G-protein is thought to dissociate it,
or at least loosen the association of the alpha subunit from the beta-gamma subunit dimer.
Then, both the alpha subunit and the beta-gamma subunits can independently function by activating various downstream signaling systems.
Of course, the classical pathway is for activation of second messenger generating enzymes,
such as adenylate cyclase or phospholipase C, which would generate cyclic AMP or diacyl glycerol.
Now, although receptors have been an interest of pharmacologists for more than a century,
quite remarkably, there remained great skepticism as to whether receptors existed in specific biophysical terms
as specific molecules in the cell that could be studied, until quite recently.
In fact, I've put up here a quote from a very influential and very well-known pharmacologist, of the twentieth century,
Raymond Ahlquist. Now, ironically, Raymond Ahlquist, a very decorated scientist, and a winner of the Lasker prize,
amongst others, was the man who first suggested that there might be two distinct types of receptor
for adrenaline and adrenaline-like molecules, which he referred to as alpha and beta adrenergic receptors.
This seminal contribution would ultimately lay the basis for the development of beta blockers by Dr. James Black.
However, as late as 1973, Ahlquist himself wrote the following
about the idea that receptors might actually represent specific molecules. He wrote:
"This would be true if I were so presumptuous as to believe that alpha and beta receptors really did exist.
There are those that think so and even propose to describe their intimate structure.
To me they are an abstract concept conceived to explain observed responses of tissues
produced by chemicals of various structure."
Well, this quote, from 1973, corresponds to approximately the time that I began my own, independent work
in my own laboratory at Duke
And it was clear to me and to others in the field at the time
that if we wanted to bring these mythical receptors to life, to show that they really did exist as specific molecules,
it would be necessary to develop a whole new set of technologies,
and the first of these would have to be some form of radioligand binding.
Now, by radioligand binding, we mean the idea that one could take a ligand,
be it agonist or antagonist, radioactively label it, and physically demonstrate the binding of that radioligand to the receptor.
If we could do this, we would have a way of studying the receptor's properties,
studying how they were regulated, and perhaps, a way of discovering new drugs,
and most importantly, a means for following the receptors during purification efforts,
which might allow us to isolate the receptors and thereby learn about their molecular properties.
Let me give you an example of a radioligand binding experiment,
the kinds of experiments that we did some years ago,
and which led to some very interesting new information about how the receptors function.
Shown in this top panel is a typical example of what's called a radioligand displacement curve.
In this experiment, a radioactively labeled beta-blocker was bound to beta adrenergic receptors,
which are coupled to adenylate cyclase, in the plasma membranes of cells.
Then, an unlabeled antagonist, such as alprenolol or propoanolol, is allowed to compete at increasing concentrations,
and to block the binding of the radioligand. As you can see, this leads to a classical sigmoidal dose response curve;
we refer to this as being uni-phasic, it has a Hill slope of 1, it is Michealean.
In striking contrast, as shown in the bottom panel, if the same experiment is done with a beta-adrenergic agonist,
such as isoproterenol, one gets a much shallower curve, with a much lower Hill slope,
which can be resolved by computer fitting to two distinct affinity states: one of high affinity and one of low affinity.
If one adds a guanine nucleotide, which will bind to the alpha subunit of the G-protein,
then as you can see, the curve is shifted to the right, and steepened.
And in fact, the affinity of this now steep, or uniphasic curve, corresponds to the lower of the two affinities, shown here.
A great deal of analysis of these types of curves taught us that agonists are able to stabilize a high affinity form of the receptor.
This was set forth many years ago in the so-called ternary complex model.
In this model, there's an initial step, where the agonist, in this case depicted as 'H' for hormone,
binds to the receptor to form a low affinity initial hormone-receptor, or agonist-receptor complex.
When H is an agonist, a presumed conformational change in the receptor occurs
which at the time was hypothesized to lead to the binding of the HR complex to a third component,
hence the name ternary complex, leading to the formation of the high affinity HRX complex.
When a guanine nucleotide, such as GTP binds to this complex, it leads to its dissociation back to the low affinity state,
and hence the marked shift in that curve. At the time the ternary complex was originally developed,
it was not clear what X was, but the pronounced effect of guanine nucleotides on the stability of this complex
immediately suggested that X might be the recently, at that time, discovered GTP binding, or guanine nucleotide binding, G-protein.
In fact, this model proved extremely useful over the years in dissecting such ligand binding curves
and led to many insights. For example, the stronger the agonist, in terms of its so called efficacy,
and ability to promote signaling, the greater the stability of that complex, and the higher the affinity of the high-affinity complex.
So, this type of modeling proved very, very useful in developing models for how receptors were working together with G-proteins.
Now, once one had in hand such radioligand binding methods,
it became possible to approach the difficult issue of trying to solubilize the receptors from the plasma membranes,
where they exist, and beginning the difficult work of purifying them.
It's important to understand that members of this family,
the so-called G-protein coupled, or seven membrane spanning receptors,
are present at extraordinarily low concentrations in the plasma membranes of cells
and would require several hundred thousand-fold purification in order to actually isolate them away from the other proteins.
This is obviously an extremely difficult task and one which occupied my laboratory for many years.
As our models, we used the adrenergic receptors, the receptors for adrenaline and related molecules,
and focused on one in particular, the so-called beta 2 adrenergic receptors.
The technique that proved up to the job, so to speak, and which was so difficult,
was the technique of affinity chromatography.
Now, in this technique, if one is trying to isolate a molecule, in this case the receptors,
and one has a specific ligand, we had several, for that receptor, one takes the ligand,
in this case that would be a beta adrenergic antagonist, such as alprenolol, and covalently couples it to a solid support,
such as a sepharose bead. One can then make a column of these beads, containing the bound ligand,
and if one has succeeded in solubilizing the receptors, which we were able to do after much trial and error,
with detergents, we were able to then apply the solubilized plasma membrane preparations, containing the receptors,
but also many, many other proteins, to the columns of these beads, and now the receptors would stick,
by virtue of their specific binding affinity for the bound ligand.
One could then wash the columns extensively to remove other proteins and then finally elute the receptors
with another ligand, such as alprenolol itself or any other beta adrenergic antagonist, or even agonist,
at high concentration. In doing this, one could achieve very, very high purification in a single step.
One could then combine this with a variety of other more conventional chromatographic methods to hopefully isolate the receptors.
We were able ultimately to succeed in purifying, to homogeneity, essentially all four of the then-known subtypes of adrenergic receptors,
called alpha 1, alpha 2, beta 1 and beta 2. And we did this by virtue of creating specific affinity chromatography columns,
in which we covalently linked either beta or alpha adrenergic antagonists to these sepharose beads.
This effort took about a dozen years in my laboratory and devoted work of many students and post-docs.
Ultimately, we succeeded in purifying all four of these subtypes to homogeneity.
This is a slide I used to enjoy showing back in the eighties because it essentially reduced to a single frame
the work of many, many students and fellows over more than ten years.
And what it is is an SDS polyacrylamide gel on which we have run three of the four subtypes of highly purified adrenergic receptors.
As you can see, each seems to migrate through the gel as a single polypeptide chain of about 60,000 Daltons.
In each of these cases, the purified protein, the isolated molecule, appeared to bind radioligands
and other adrenergic compounds with exactly the specificity and the stereospecificty
that one would expect for the individual adrenergic receptor subtype.
At this point, and this would have been the early 1980's, I would say,
we thought that these kinds of conclusive findings would convince the skeptics that the receptors really did exist
as specific isolated molecules, But in fact, the skepticism persisted in some quarters because, it was said,
well, how do you know that this isolated molecule, even though it can bind your drugs with the right specificity,
how do you know that it's really the receptor, that it's really capable of conveying responsiveness to an isolated system.
And to take that final step, we turned to the technique of receptor reconstitution.
So, in this method, we took our purified receptors, and we constituted them into phospholipid vesicles.
We wanted to now take a system which was unresponsive to adrenaline
and show that if we could insert this isolated molecule,
the putative receptor, into that system, we could convey responsiveness to the adrenaline.
So, what we needed was a cell that didn't have the receptors. Well, that turned out to be a difficult problem.
It turns out virtually all mammalian cells contain beta 2 adrenergic receptors.
However, by trial and error and searching, we found that erythrocytes from an amphibian, the African clawed toad,
Xenopus laevis, lacked beta adrenergic receptors and hence did not respond to adrenaline,
although it did contain the other components of the adenylate cyclase system,
the G-protein and the adenylate cyclase itself.
It also contained prostaglandin receptors,
and hence the adenylate cyclase of this cell could be stimulated to make cyclic-AMP
by prostaglandins, but not adrenaline.
So, we took these cells lacking the beta receptor and we mixed them with these phospholipid vesicles containing the receptor.
The vesicles fused with the plasma membrane of the cell, thereby inserting our putative receptors into the membrane.
And, lo and behold, the cell then became responsive to adrenaline
and its rate of cyclic AMP production increased then, in response to adrenaline.
And so, the last skeptics finally yielded and agreed that this molecule,
which could both bind beta-adrenergic compounds with appropriate specificity,
and to convey to a previously unresponsive cell responsiveness to adrenaline, was in fact the beta adrenergic receptor.
Once we had validated that the molecules we had isolated, and again in these days, back in the eighties,
we never had more than about 25 to 50 micrograms of purified receptor in hand at any one time,
but once we had rigorously validated that these molecules were the receptors,
we were ready to begin the difficult work of cloning the genes and cDNAs for these molecules,
to try to find out about their primary amino acid sequence.
And after several years of difficult work, in 1986, we reported the cloning of the beta-2 adrenergic receptor.
Now this was a bit of a landmark in the field, because to that point in time, no other G-protein coupled receptor had been cloned.
The only exception was the molecule rhodopsin.
Now, let's talk a little bit about rhodopsin. Rhodopsin, the so-called visual pigment,
is, if you will, a receptor for photons of light. When photons of light are captured by retinal,
bound to the opsin protein, it isomerizes, changes the configuration, or conformation of the rhodopsin,
and signals are emitted which actually activate a G-protein in the retina called transducin.
Now, back in the mid-eighties, most of us working in the signaling field were aware of the fact
that there was a great deal of functional analogy between the visual system
and the beta-adrenergic receptor stimulated adenylate cyclase.
Both worked through a G-protein, GS in one case, transducin in the other,
and both worked through an enzyme, either a cyclic GMP phosphodiesterase in the retina,
or an adenylate cyclase in other mammalian tissues, which responded to the G-protein.
But, nobody really thought that there was any structural analogy between rhodopsin
and all the other members of the G-protein coupled receptor family.
However, when we succeeded in 1986 in cloning the beta-2 adrenergic receptor,
we immediately could see its structural relationship to rhodopsin.
Like rhodopsin, it had seven stretches of hydrophobic amino acids, presumably representing helical transmembrane domains.
As shown in blue, many of the residues in the transmembrane domains were identical, homologous, with rhodopsin.
There were, of course, connecting external and internal loops.
Many sites for regulatory phosphorylation on the C-terminus, as in rhodopsin,
and toward the amino terminus, consensus sites for N-linked glycosylation.
When I lecture today about the discovery, more than 20 years ago, of the close structural relationship
between the beta-2 adrenergic receptor and rhodopsin,
people are always amazed that we were all so surprised to find this.
How could you not be thinking along those lines? At the time. You knew about the functional analogies,
this is true. Yet, as the contemporaneous literature reflects,
none of us ever expected that there would be this close structural relationship.
I think perhaps part of the reason for that had to do with even some prior history.
The structure of rhodopsin, which is a remarkably abundant protein,
it's about 90% or more of all the protein in isolated retinal rod outer segments,
Rhodopsin was available in large quantities and when its structure was determined in 1982,
it was originally, it's sequence was determined, not by cloning, but rather, by classical Edmond degradation,
protein sequencing, because there was so much protein available.
It was found to have a structure consistent with, suggestive of, seven transmembrane domains.
At the time only one other such protein was known, that is one other protein which had these seven transmembrane domains,
and that was bacteriorhodopsin. Now, bacteriorhodopsin is a protein present in very primitive archaebacteria,
which is a light driven proton pump. Now, when the structure of retinal rhodopsin was determined,
and it too was found to have seven-membrane spans,
it was very reasonably speculated by both groups who had succeeded in getting that sequence,
that the common feature of bacteriorhodopsin and rhodopsin was that they were light sensitive proteins.
And so, it was put forth as an hypothesis that the seven membrane spans must be a signature structural feature of light-sensitive proteins,
which seemed very reasonable at the time.
Only with the
…