723 6.

3
Receptor and Binding Studies
Peter Hein, Martin C. Michel, Kirsten Leineweber, Thomas Wieland,
Nina Wettschureck and Stefan Offermanns

6.3.1
Radioligand Binding Studies in Cardiovascular Research
(Saturation and Competition Binding Studies)
Peter Hein and Martin C. Michel

Introduction
Radioligand binding studies are an important tool to quantify and qualitatively char-
acterize receptors and hence have become a universal tool of cardiovascular research.
While radioligand binding experiments are relatively simple to perform, interpreta-
tion of the results and optimisation of the experimental settings are more complex.
Several criteria must be met to consider a labelled binding site a receptor: Binding
should be of high affinity, saturable, stereospecific and have a ligand recognition pro-
file similar to that of a functional response in that tissue. A hilariously funny descrip-
tion of the fundamental principles of radioligand binding has been presented by Guth
(1982).
Radioligand binding experiments can be divided into two major types of experi-
ments, kinetic and equilibrium studies. Kinetic (i.e. association and dissociation)
studies are required to define optimal assay conditions. On theoretical grounds they
also allow the most reliable determination of the affinity of a radioligand for its cog-
nate receptor. However, kinetic studies are cumbersome to perform and hence rarely
used in routine studies. Their most frequent application is in the phase of character-
ization of novel radioligands. Once radioligands have been characterized and proper
assay conditions have been established, the translation of such assays to cardiovas-
cular tissues can be done in most cases without repeating the kinetic studies.
The vast majority of applications of radioligand binding studies are based on
equilibrium binding studies. The key molecular principle underlying such experi-
ments is the idea that association of a radioligand with its cognate receptor and dis-
sociation of the resulting ligand-receptor complex are reversible processes that occur
concomitantly until equilibrium has been reached. Key factors in determining the
 6 724 Biochemical and Analytical Techniques

Figure 1
A saturation binding experi-
ment allows measuring Kd and
Bmax of a ligand/receptor-pair
In-Vitro Techniques

development of equilibrium conditions are incubation time, incubation temperature

and assay volume (the latter as a measure of amount of radioligand relative to num-
ber of binding sites in the assay). Equilibrium binding studies can be divided into two
main types of experiments, i.e. saturation and competition studies (with the latter
often but incorrectly referred to as displacement studies). Saturation binding studies
allow determination of the affinity of a receptor for a radioligand and, more impor-
tantly, receptor densities. Competition binding studies allow a pharmacological char-
acterization of receptors and, under proper experimental conditions, identification of
the receptor subtype distribution within a tissue or cell line.
Saturation binding studies are performed using several concentrations of a
radioligand and allow determination of the dissociation constant (Kd) and number of
binding sites in the assay (Bmax) for the radioligand (Fig. 1). The Kd-value is a measure
of the affinity of the ligand for its binding sites. Since the Kd is the concentration at
which the ligand binds to half its binding sites, the ligand’s affinity for the receptor is
small when its Kd is large. Likewise, ligands with a high affinity have small Kds. Some-
times the affinity is given as pKd, i.e. the negative logarithm of Kd. Kd is a constant and,
unless a new radioligand or receptor is characterized, is largely used as an internal
quality control of a binding assay. The receptor density in a tissue or cell line is called
Bmax. Bmax values are typically adjusted to account for tissue content in the assay, e.g. for
protein content or tissue wet weight and expressed as fmol/mg protein or fmol/mg
wet weight, respectively. Saturation binding experiments can also be used to deter-
mine whether a given drug indeed acts as a competitive antagonist at a receptor of
interest. Thus, concomitant saturation binding experiments in the absence and pres-
ence of a competitor should yield identical Bmax values but distinct apparent Kd values
if the competing drug acts in a competitive manner. In this case the shift in Kd value
can even be used to determine antagonist affinity in a manner analogous to that used
when analyzing the shift of an agonist concentration-response curve by an antagonist
in organ bath experiments with isolated tissues.
 Receptor and Binding Studies 725 6.3

Figure 2
C50, the concentration of the
ligand causing 50% inhibition
of radioligand binding, is de-
termined by a competition
binding experiment

Competition binding studies, sometimes also referred to as displacement studies,

use a single concentration of a radioligand that is competed for by multiple concen-
trations of another, not radioactively labelled compound (Fig. 2). They allow the phar-
macological characterization of binding sites. Competition studies primarily yield the
concentration of the inhibitor causing 50% inhibition of radioligand binding (IC50),
which can be converted into the inhibitor constant of the inhibitor (Ki) using appro-
priate equations. If a tissue contains multiple binding sites for a given radioligand and
the competing compound differentiates them, biphasic or even multiphasic competi-
tion curves are obtained which allow not only determination of Ki values at the vari-
ous sites but also an estimate of the relative quantitative contributions of each site.
The following considerations are primarily developed for radioligand binding
studies with G-protein-coupled receptors, which are the most frequent subjects of
studies in cardiovascular research. While radioligand binding studies to other recep-
tor classes such as ligand-gated ion channels, receptors with intrinsic enzyme activ-
ity or ligand-activated transcription factors (e.g. glucocorticoid receptors) follow
similar principles, not all details given below apply to these other receptor classes. On
the other hand, many of the approaches described hereafter can also be applied to
non-receptor proteins such as uptake carriers using minor variations of the protocols
given below.

Descriptions of the Methods and Practical Approach

Receptor Theory

Binding of the ligand to the receptor follows the law of mass action. The ligand and
receptor associate and form a ligand-receptor-complex, or they may dissociate again.
 6 726 Biochemical and Analytical Techniques

receptor + ligand ↔ receptor · ligand (1)

The association rate is called Kon, and the amount of newly associated receptor-ligand
complex depends on Kon, the concentration of the free ligand and the concentration of
free receptors. Accordingly, the dissociation rate Koff, and the concentration of ligand-
receptor complexes describe the number of dissociations. At equilibrium concentra-
tions, the rate at which new receptor-ligand complexes are formed equals the rate at
which they dissociate.

[ligand] × [receptor] × Kon = [receptor · ligand] × Koff (2)

This may be rearranged to

[ligand] × [receptor] Koff
= = Kd (3)
[receptor ⋅ ligand] Kon

The quotient Koff/Kon is named the equilibrium dissociation constant Kd. It has the
units of a concentration and, more specifically, is the concentration at which half the
In-Vitro Techniques

receptors are occupied by the ligand or free, respectively. Note that there is a specific
Kd value for each ligand/receptor-pair, and since in different tissues different receptor-
subtypes (or even conformational different states of the same subtype) may be
present, there may be different Kd values for the same ligand and receptor in different
tissues.
Koff has a major effect on the time required to reach equilibrium and hence usu-
ally governs the choice of incubation time for saturation and competition equilibrium
binding studies. Equilibrium is often reached after 45–60 min, but with some ligands
or receptors, longer time periods may be necessary. It should also be noted that Koff is
affected by the incubation temperature, but this has little effect on kinetic determina-
tion of Kd as long as Kon and Koff are measured at the same temperature (for criteria for
choice of temperature see below).

Radioligands

Chemically, radioligands can be derived from agonists or antagonists, and it should be

noted that the two classes of drugs label different receptor populations. Thus, antago-
nists theoretically should label all receptors while agonists typically only label a frac-
tion thereof with high affinity. Therefore, antagonists are usually preferred as
radioligand unless labelling the receptor population in a high affinity state for ago-
nists is specifically attempted. However, for some receptor classes, e.g. for many pep-
tide receptors, only agonist radioligands are available, which are typically radio-
labelled forms of the endogenous ligand. Moreover, it should be noted that in some
cases antagonist radioligands from distinct chemical classes may label different num-
bers of receptors, even if it is guaranteed that only one receptor subtype exists within
the assay (Hein et al. 2000).
 Receptor and Binding Studies 727 6.3
In some cases a given receptor can be labelled with either a more lipophilic or a
more hydrophilic radioligand, e.g. for muscarinic acetylcholine receptors with
[3H]-quinuclidinyl-benzylate or [3H]-N-methyl-scopolamine, respectively. While more
lipophilic radioligands tend to have more non-specific binding (see below), they also
tend to label greater receptor numbers. This is due to the fact that, with intact cells,
some receptors are located intracellularly or, with membrane preparations, in spon-
taneously formed inside-out vesicles, and hence are inaccessible to hydrophilic li-
gands.
Different radioisotopes can be used to label ligands, with tritium ([3H]) and iodine
([ I]) being used most frequently. Since [3H] and [125I] emit distinct types of radia-
125

tion, different methods for detection and hence different hardware are required. The
key advantage of [3H] is that it does not change the molecular structure of the ligand
because a hydrogen atom is exchanged with its radioactive homologue. A practical
advantage is the long half-life of [3H] which allows storage of the radioligand for
extended periods of time. On the other hand, tritiated radioligands have rather low
specific activity (usually less than 100 Ci/mmol), which may lead to weak radioactive
signals in some cases. This is mainly a problem with radioligands with very high, i.e.
subnanomolar, affinity. The specific activity of the tritiated compounds slowly de-
clines over time, as tritiated molecules become untritiated. Since this does not per se
destroy the chemical integrity of the radioligand, the fraction of labelled molecules
will become less. Because of the long half-life of tritium, this is mainly a problem
when you store the radioligand for a long time (years), as long as reasonably fresh
stocks are used, the decay of tritium does not have to be corrected for. In contrast,
high specific radioactivity (typically 2200 Ci/mmol) is the key advantage of iodinated
radioligands. This is counterbalanced by possible disadvantages. Introduction of io-
dine into a molecule may change its affinity for the target receptor, particularly if it is
a low molecular weight compound (it seems to be less of a problem with larger
ligands such as peptides). Moreover, based on the shorter half-life of [125I], iodinated
radioligands cannot easily be stored over extended time periods without marked loss
of activity; hence, it is usually recommended to use fresh stock every month. Finally,
it should be remembered that radioactive decay of [125I] tends to destroy the ligand
chemically. Hence, correction for radioactive decay is typically not necessary since the
amount of ligand and the amount of radioactivity decline in parallel.
While the radioactivity emitted by [125I] can be measured directly using appro-
priate hardware, the counting of [3H] usually requires the use of scintillation cocktails
and counters. The efficiency of counting depends on a variety of factors including
hardware-specific ones and, in the case of [3H] also depends on the scintillation fluid
being used. The choice of scintillation fluid depends on the type of sample (e.g. rela-
tive amount of aqueous fluid contained therein) but can also involve factors such as
environmental safety.
Counting efficiency must be taken into consideration when experimental results
(usually obtained as counts per minute, cpm) are converted into disintegrations per
minute (dpm) to ultimately allow conversion into molar units. Data-sheets for
radioligands often state the specific activity in Ci/mmol. However, at the end of the
experiment you are measuring dpm or cpm which can be transformed to molar con-
 6 728 Biochemical and Analytical Techniques

centrations, with intermediary conversion of cpm to dpm not always being necessary.
By definition, 1 Ci equals 2.22×1012 disintegrations per minute. To convert Z Ci/mmol
to Y cpm/fmol, use the following equation:

Y = Z × 2.22 × counter efficiency (4)

If you count C cpm in a volume of V ml, you can compute the concentration accord-
ing to:

C C C (5)
÷V= =
Y Y × V Z × 2.22 × e × V

with e being the relative counting efficiency (e.g. 0.4 for 40%) and Z the specific activ-
ity in Ci/mmol.

Buffer and Incubation Conditions

In-Vitro Techniques

Fortunately, most radioligand binding assays are not very sensitive to changes in
buffer conditions (with experiments involving receptor agonists being an exception)
and buffer systems established for well-characterized radioligands can easily be
picked from the literature. The choice of buffer depends on the receptor type under
investigation. While radioligand binding experiments are usually carried out in cell or
tissue homogenates, it is also possible to perform them with intact cells. Although it
is not necessary to establish a physiological environment for binding studies with
intact cells, a somewhat isotonic buffer should be used to avoid shrinkage or bloating
of cells due to buffer hyper- or hypo-osmolarity, respectively. For binding studies with
cell or tissue homogenates hypotonic buffers are used in most cases. Buffering is
mostly achieved using Tris, HEPES or phosphate buffer. While Tris is widely used, it
has the disadvantage of having a pKA value (>10) far from physiological pH. Moreover,
the pH of a Tris-buffered solution is quite temperature-dependent and hence its pH
must be titrated at the desired incubation temperature.
Frequently, chelating substances such as EDTA are added to prevent endogenous
enzymes from degrading membranes and/or proteins. In some cases ascorbic acid or
other anti-oxidants are added to prevent oxidation of the radioligand or competing
agents, but this may also be harmful since e.g. the specific binding of [125I]-BE 2254
(also known as [125I]-HEAT) to α1-adrenoceptors is markedly inhibited by ascorbic
acid. When an agonist is used as radioligand or competing ligand, the ionic compo-
sition of the buffer can have additional requirements such as the inclusion of magne-
sium. Moreover, agonist binding to some receptors is sensitive to sodium.
The choice of temperature affects the kinetics of radioligand binding to its recep-
tor. In most cases 25 °C or 37 °C are used, with the higher temperature yielding equi-
librium faster and hence requiring shorter incubation times. In some cases
incubations for long times are carried out at 4 °C, e.g. in order to look at total vs. cell
surface receptors differentially in intact cells.
 Receptor and Binding Studies 729 6.3

Specific and Non-Specific Binding

Within the incubation mixture only a small fraction of the radioligand binds to the
receptor of interest, and this fraction is designated “specific” binding. Another fraction
binds to sites distinct from the receptor of interest and is called “non-specific” bind-
ing. A third fraction of the radioligand does not bind at all but remains free in solu-
tion and is referred to as “free” radioligand. The analysis of radioligand binding
studies assumes that only a small fraction of the radioligand binds, and hence that the
concentration of the free radioligand is approximately equal to the total concentration
added. If the percentage of bound (specific and non-specific together) exceeds 10% of
total radioligand, ligand depletion occurs, which prevents equilibrium conditions and
hence invalidates analysis of the data. Thus, it is necessary that free radioligand is
always in great excess of bound radioligand, i.e. that at least 90% of the total amount
of radioligand in the assay remains unbound, with a greater percentage being even
better. Ligand depletion is frequently a problem when the assay contains a very high
receptor number, e.g. when transfected cells, and a tritiated radioligand with
subnanomolar affinity are used. Such problems can be avoided by increasing the as-
say volume, thereby increasing the amount of radioligand but not that of receptors.
Non-specific binding may involve binding to receptors other than the target recep-
tor (if the radioligand is not sufficiently specific) as well as truly non-specific binding,
e.g. to cell membrane lipids or the wall of the test tube. To discriminate specific and
non-specific binding, a number of samples within each assay (typically one for each
radioligand concentration being tested) are incubated in the additional presence of a
high concentration of an agent to define non-specific binding. Specific binding is then
calculated as the difference between total binding and non-specific binding. To define
non-specific binding a competing compound should be chosen that also binds to the
receptor of interest with high specificity but, if possible, belongs to a different chemi-
cal class. Using the non-labelled form of the radioligand to define non-specific bind-
ing should be avoided if possible, but this may not be feasible e.g. for some peptide
receptors where specific agents apart from the native agonist peptide are frequently
not available. The concentration of the competing agent used to define non-specific
binding should be so high that it practically occupies all receptors of interest, i.e. typi-
cally about 1000× its Kd at that receptor. Before starting a major project with a new
radioligand it is useful to determine the appropriate concentration of the compound
used for definition of non-specific binding experimentally in a competition-binding
assay. From such a competition curve a concentration can be picked which is high
enough to yield complete competition but not so high that it causes non-specific
effects in its own right.
Pseudo-non-specific binding, i.e. binding to other receptors due to insufficient
pharmacological specificity of the radioligand, can become a serious problem when
the definition of non-specific binding involves a compound that also lacks pharma-
cological specificity and also binds to this additional receptor. For example, both the
β-adrenoceptor radioligand [3H]-dihydroalprenolol and the antagonist propranolol
can also bind to certain serotonin receptors at high concentrations. Hence, attempts
to label β-adrenoceptors using too high a concentration of these two compounds
 6 730 Biochemical and Analytical Techniques

Figure 3
Specific binding is obtained by
subtracting non-specific bind-
ing from total. Note that non-
specific binding increases li-
nearly with the concentration
of the radioligand, while speci-
fic binding becomes saturated
at Bmax

may also label serotonin receptors, and this may masquerade as “specific” binding.
Whether the “specific” binding of a given radioligand is indeed specific for the target
receptor can be judged on the basis of its affinity (Kd value) as well as upon the results
In-Vitro Techniques

of competition studies with a range of chemically distinct compounds with known

affinities for the receptor of interest.
When reporting results of radioligand binding experiments, the typical degree of
non-specific binding should be mentioned. In this regard specific binding is by defi-
nition saturable, whereas non-specific increases linearly with the concentration of the
radioligand for all practical purposes (Fig. 3). Hence, the percentage of non-specific
binding within a given experiment always depends on the concentration of the
radioligand. Thus, it is customary to report the percentage of non-specific binding for
a radioligand concentration close to its Kd value. High levels of non-specific binding
may make interpretation of the results dubious. In that case assay conditions should
be carefully revised to minimize non-specific binding.

General Protocol Considerations

A radioligand binding experiment can be divided into five steps: preparing the assay,
preparing the cell or membrane suspension, incubation, separating bound from free
ligand, and data analysis. While the first and the last steps are somewhat different for
saturation and competition binding assays, the three steps in between are typically the
same for both types of experiments. Hence, we will first discuss these three steps and
then the specific aspects of assay preparation and data analysis in saturation and com-
petition binding experiments.

Preparing the Cell or Membrane Suspension

Radioligand binding experiments can be performed with cell and tissue homoge-
nates, intact cells, tissue sections (autoradiography) or solubilised receptors. Each of
these approaches has different advantages and disadvantages. Homogenates are used
 Receptor and Binding Studies 731 6.3
most frequently, and the other techniques are mainly employed for specific questions
that cannot be easily answered using homogenates. Thus, we will focus on homo-
genates here. A detailed discussion of the pros and cons of the other techniques has
been provided elsewhere (Hulme and Buckley 1992).
To avoid degradation of the receptors within a cell or tissue, it is paramount that
tissues are rapidly excised and that the entire membrane preparation is performed
under well cooled conditions, i.e. at 4 °C. For most applications it is possible to snap
freeze tissues and to perform the homogenization later with thawed tissue or to freeze
the freshly prepared membrane preparation for later use in a binding experiment. If
such storage is planned, it is good to store at temperatures not exceeding -70 °C. Tis-
sues are usually minced with scissors prior to homogenization. For the latter step, a
variety of devices are commercially available. For soft tissues Potter-Elvehjem glass/
glass or glass/Teflon homogenizers are most frequently used, whereas for tougher tis-
sues Polytron-type devices are preferred (Hulme and Buckley 1992). The type and
intensity of homogenization also depend on the desired outcome, i.e. whether crude
homogenates or specific subcellular fractions are desired. It should be kept in mind
that the homogeneity of the membrane preparation is a key determinant of data scat-
ter in the assay.
Depending on the experimental aims, subcellular fractions are frequently pre-
pared by centrifugation, sometimes involving density gradient centrifugations. The
use of more centrifugation steps increases the purity of the preparation at the cost of
yield and time. The most desirable balance between these factors depends on the cell
or tissue and receptor under investigation.
For example, β-adrenoceptor binding assays have been performed on human
heart biopsies weighing no more than 5 mg (Brodde et al. 1991). Since the mechani-
cal forces of homogenization will generate heat, the vessel in which it is performed
should always be cooled; a short break between homogenization bursts will poten-
tially be helpful in keeping the temperature low.

Incubation

Incubation conditions depend on the tissue or cell line and/or radioligand under in-
vestigation. In principle, equilibrium binding has to occur during the incubation time.
Since adrenoceptors and muscarinic receptors are studied most frequently in the car-
diovascular system, Table 1 gives an overview of some commonly used radioligand
and incubation conditions. For other receptor classes and radioligands, readers
should consult recent publications for reasonable conditions.

Separating Bound from Free Radioligand

At the end of the incubation period bound and unbound radioligand must be sepa-
rated. The two main methods for this are vacuum filtration and centrifugation. Since
vacuum filtration is used most frequently, we will not discuss centrifugation tech-
niques here but rather refer to an in-depth description of such techniques (Hulme
1992a). In specific instances, e.g. with solubilised receptors, other separation methods
may become relevant including gel filtration (Hulme 1992b) or charcoal adsorption
 6 732 Biochemical and Analytical Techniques

Table 1
Incubation conditions for some frequently used radioligands for adrenergic and muscarinic acetylcho-
line receptors

Receptor Ligand Assay Incubation Incubation

volume time temperature
α1-adrenoceptors [3H]-prazosin 1000 µl 45 min 25°C

α2-adrenoceptors 3
[ H]-RX 821002 250 µl 60 min 25°C

β-adrenoceptors [125I]-iodocyano- 250 µl 90 min 37°C

pindolol
Muscarinic [3H]-quinuclidinyl- 1000 µl 60 min 37°C
receptors benzylate

(Strange 1992). For filtration assays, a variety of devices and filter types are offered.
The filtration devices range from simple pots hooked up to a vacuum pump to elabo-
In-Vitro Techniques

rate harvesters, with the former being the cheapest and the latter being most conve-
nient, particularly if large numbers of samples are processed on a routine basis. Glass
fibre filters are used most frequently, particularly the GF/C type. Filtration plates that
allow direct filtering of samples from a 96 well plate into a filter plate are being offered
by some manufacturers.
After the filtration, it is helpful to rinse the filters a few times (with a total of about
10 ml incubation buffer) to minimize non-specific binding. This should be done with
ice-cold buffer to minimize dissociation of bound ligand.

Saturation Binding Experiments

As stated above, saturation binding experiments are used to determine the affinity of
a radioligand to a receptor (Kd), as well as the total number of receptors present in the
assay (Bmax). To obtain reliable estimates of these parameters, at least 6 different con-
centrations of radioligand must be tested, but particularly when a receptor is first
detected in a given tissue or cell type, an even greater number of concentrations is
helpful.

Preparing the Assay

To minimize experimental error all assays are performed at least in duplicate. This can
be done in polypropylene vials but also in 96 well plates. The former approach is more
appropriate for smaller scale experiments, whereas the latter is more economical
when large numbers of samples are to be processed.
Radioligand dilutions should be prepared to cover the desired concentration
range. Optimally these concentrations should cover both the high range (correspond-
ing to 5–10×Kd and hence saturation of the receptor) and the low range (around Kd)
 Receptor and Binding Studies 733 6.3
in order to allow reliable estimates of both Kd and Bmax. If an assay is well established
and the number of data points has to be constrained to 6 within an experiment for
technical reasons, an emphasis on higher concentrations is advisable if the primary
parameter of interest is Bmax whereas it should be on the lower concentrations if the
interest is primarily in precise Kd estimates. Duplicates of total binding and non-spe-
cific binding should be tested for each concentration of the radioligand.

Data Analysis

To derive Kd and Bmax from the experimental data several steps are required. After
having calculated the mean value of the duplicate, triplicate or quadruplicate mea-
surements for total binding, non-specific binding and totals (i.e. the concentrations of
the radioligand used which become the x-axis values in the subsequent analysis), it is
necessary to calculate specific binding by subtracting non-specific binding from to-
tal binding at each concentration of the radioligand. It may be helpful to graph total
and non-specific binding separately to get an impression of the raw data (see Fig. 3).
Thereafter, two steps are required but their sequence is not relevant. In one step ex-
perimental results in cpm must be converted into molar units based on the specific
activity of the radioligand as described above. In the other step, affinity and receptor
number are calculated from the data. Moreover, it is customary to then correct mo-
lar units of Bmax for tissue content in the assay. The most frequently used denomina-
tor for this is mg protein, but cell number, mg wet weight or DNA content are equally
acceptable and may even be preferable e.g. when receptor number in a hypertrophied
or scarred tissue is compared to a control tissue.
To obtain estimates of Kd and Bmax from saturation binding experiments, it was
customary to use the Rosenthal plot, frequently but incorrectly referred to as a
Scatchard plot. This plot was very popular twenty years ago at a time when computer-
assisted data analysis was not yet widely available. However, the transformations in-
volved in the Rosenthal analysis may cause problems due to inappropriate weighing
of certain data points, and the assumptions of linear regression are no longer met
since x and y are no longer independent of each other. This eventually leads to inad-
equate results (for review see Burgisser 1984). Since saturation binding isotherms
should follow rectangular hyperbolic functions based on theoretical considerations,
they are best analysed by fitting such functions to the experimental data using com-
puter-assisted non-linear iterative curve fitting programs. A variety of such programs
are commercially available, with Prism (Graphpad Software, San Diego, CA, USA) and
Sigmaplot (SPSS Inc., Chicago, IL, USA) being most popular.
Any of these computer programs requires that the concentration of radioligand be
entered for the x-axis (the amount of radioactivity in the assay measured as explained
below is more reliable than calculated theoretical concentrations). The amount of
specific binding is entered for the y-axis values (if desired this can also be done for
total and non-specific binding). The programs will fit a rectangular hyperbola to your
data according to
Bmax × x
y= (6)
Kd + x
 6 734 Biochemical and Analytical Techniques

This equation is derived as follows: When you substitute [ligand] with x and [re-
ceptor · ligand] with y in equation 2 and rearrange it, you obtain
Kon 1
y = [receptor]× x × = [receptor]× x × (7)
Koff Kd

[receptor] is the concentration of the free receptors and usually unknown. However,
free receptors are all receptors minus the occupied ones

[receptor] = Bmax – [receptor · ligand] = Bmax–y (8)

Inserting and rearranging leads to

(Bmax – y) × x Bmax × x – y × x
y= = (9)
Kd Kd

y × Kd = Bmax × x – y × x (10)
In-Vitro Techniques

y × Kd + y × x = Bmax × x (11)

y × (Kd + x) = Bmax × x (12)

Bmax × x
y= (13)
Kd + x

You will get Kd and Bmax as results. Note that, when the concentration of the ligand (the
x-value) equals Kd, you will get
Bmax × Kd Bmax × Kd Bmax
= = (14)
Kd + Kd 2Kd 2

Figure 4
A Rosenthal plot was used to
determine Kd and Bmax before
nonlinear regression methods
were widely available. How-
ever, this analysis may yield
unreliable results (see text)
 Receptor and Binding Studies 735 6.3
In other words, when the concentration of the ligand equals Kd, half of the receptors
(Bmax/2) are occupied as it corresponds to the definition of Kd.
Although the Rosenthal (a.k.a Scatchard) plot is no longer considered appropriate
for quantitative data analysis (Burgisser 1984), it remains helpful to gauge the experi-
mental data visually. The principle of the Rosenthal plot is that the amount of specifi-
cally bound radioligand on the x-axis is plotted against the ratio of bound/free
(unbound) ligand on the y-axis (Fig. 4). In this analysis the x-intercept corresponds
to Bmax, while the negative reciprocal of the slope corresponds to Kd. A deviation of the
regression line of the data points in this representation indicates some form of lack of
equilibrium in the assay.

Practical Approach (Protocol)

Note that two sets of tubes are necessary: one to perform the incubation of receptor
and ligand, and one (this is actually done directly in vials) to quantify the radioactiv-
ity, and hence the concentration of radioligand, being used. Following the separation
step of the experiment, the radioactivity is transferred from the incubation tubes to
the vials. The vials have to be compatible with your radiation counter, too.
Not all of these steps have to be taken in the indicated order. Many of them can be
done in parallel or in an intercalated way, e.g. tubes can be set up and solutions be
prepared during the centrifugation steps of the membrane preparation.
1. Prepare a number of incubation tubes (or a microtitre plate) according to the
planned number of data points (2 each for total and non-specific binding (in the
example presented below, designated ‘TB’ and ‘NS’, respectively) for each concen-
tration of the radioligand).
2. Prepare the same number of vials for use in your radiation counter (alternatively
a corresponding filter plate which can be placed directly into the counter). Addi-
tionally, prepare 2 vials for each concentration of the radioligand to allow quan-
tification of the amount of radioligand used in the assay (total radioactivity, ‘T’ in
the example below) and hence its true concentration. This is necessary because the
radioligand is not diluted to a certain molar concentration; rather, the concentra-
tion is calculated from the counts obtained in these vials. Place a filter in the vials
to make sure that they are treated like the assay samples.
3. Prepare sufficient amounts of incubation buffer, check its pH at the desired incu-
bation temperature, and cool it to 4 °C in an ice-bath.
4. Prepare the working solution for determining the non-specific binding.
5. Prepare the working solution of the radioligand from the stock solution provided
by the radiochemical manufacturer. Dilute the working solution to obtain a sepa-
rate working solution for each desired concentration of the radioligand.
6. Add the working solution for non-specific binding or corresponding amounts of
incubation buffer to all incubation tubes (or wells).
7. Add the working solutions of the various concentrations of the radioligand to the
respective tubes (or wells).
Don’t forget to add it to the tubes used for determination of total amount of
radioligand in the assay as well (see 2).
 6 736 Biochemical and Analytical Techniques

8. Resuspend the membrane preparation (the pellet from the last centrifugation
step) thoroughly in incubation buffer to ensure a homogenous suspension. Add
aliquots of membrane suspension to all incubation tubes (or wells).
9. Incubate tubes (or plate) at the chosen temperature for the chosen period of time.
A gently shaking water bath may be helpful to ensure constant agitation during
the incubation. Since the incubation time should allow for equilibrium condi-
tions, minor extensions of the incubation are not usually harmful to the assay.
10. Incubations are stopped by vacuum filtration of the samples either manually or
with a harvester. Rinse the incubation tubes (or wells) with approximately 10 ml
of ice-cold incubation buffer each and filter that as well. Apply the filtration
vacuum until the filters appear dry.
11. Put the filters in the corresponding scintillation vials (not necessary in the case of
filtration plates). If the radioligand is labelled with [3H], dry the filters (e.g. for 1 h
at 60 °C or overnight at room temperature) and then add an adequate volume of
scintillation fluid to each vial.
12. Seal the vials, shake gently. Count the probes in a counter.
In-Vitro Techniques

Competition Binding Experiments

Competition binding experiments (sometimes incorrectly referred to as displace-

ments) can serve multiple purposes. Firstly, they characterize the ligand recognition
properties of a binding site. Comparison of this pharmacological specificity with that
of a given functional response allows one to determine whether the binding site in-
deed corresponds to a functional receptor.
Secondly, in many case a cell or tissues expresses more than one subtype of a re-
ceptor, e.g. β1- and β2-adrenoceptors in the heart, but measurements of each of them
individually may be cumbersome or, due to the lack of proper subtype-selective
radioligands, even impossible. Competition studies with suitable subtype-selective
agents can reveal the relative proportions of receptor subtypes within a tissue or cell.
Together with information on the total receptor density this will allow indirect deter-
mination of the individual numbers of each subtype (for an example see Brodde et al.
1991).
Thirdly, competition studies with agonist in suitable buffer conditions will also
allow determination of the relative percentages of the receptor in the high and low
affinity state for agonists.
Finally, competition studies can substitute for saturation binding studies in cases
where true saturation curves are not feasible, e.g. for economical reasons with very
expensive radioligands as is the case with some peptide receptors. This is called ho-
mologous competition binding and based on the assumption that radioligand and its
unlabelled analogue have identical affinities for the receptor of interest (for details see
DeBlasi et al. 1989).
The overall design of a competition binding experiment is quite similar to that
of a saturation binding experiment, and hence we will focus here on its specific as-
pects.
 Receptor and Binding Studies 737 6.3

Preparing the Assay

In contrast to saturation binding experiments, competition-binding assays employ

only a single concentration of radioligand. The concentration of the radioligand in
competition studies should be about 2–3× its Kd value as determined in saturation
binding. It is also necessary to determine the total amount of radioligand in the assay
and total and non-specific binding within a competition binding experiment. Ad-
ditional data points are generated in which radioligand binding is quantified in the
presence of various concentrations of an unlabelled compound of interest. If possible
the concentrations of the competitor should be chosen so that the lowest concentra-
tion causes no inhibition (i.e. binding equals total binding) and the highest concen-
tration causes complete inhibition of specific binding (i.e. binding equals non-specific
binding).
In most cases it will be necessary to test at least 2 competitor concentrations per
log unit of concentration increment. If one is attempting to look for the presence of
multiple binding sites (e.g. receptor subtypes or agonist high affinity states), 4 concen-
trations per log unit are advisable.

Data Analysis

If a compound competes for a homogeneous class of binding sites, the competition

curve should theoretically follow a sigmoidal function with a Hill-slope of unity. Thus,
experiments should primarily be analyzed by fitting a sigmoidal curve to the data
with a floating Hill-slope.

Top – Bottom (15)

y = Bottom + log (IC –x)* HillSlope
1 + 10 50

If it deviates consistently from unity, the data do not support interaction with a homo-
geneous class of sites under equilibrium conditions. If a cell or tissue contains two
subtypes of a receptor and the radioligand binds to both with similar affinity, com-
petitors that are selective for one of them will yield shallow and biphasic competition
curves, with each of the two components having a Hill-slope of unity.
To test this assumption, both a monophasic and a biphasic model may be fitted to
the data. The biphasic model should only be accepted if the fit is statistically superior
to a monophasic fit in an F-test. A monophasic curve is characterized by its top (cor-
responding to total binding), its bottom (corresponding to non-specific binding) and
the position of half-maximal inhibition. The latter is referred to as the IC50 value. A
biphasic curve is defined by top, bottom, distinct IC50 values for each of the two sites
and the percent contribution of each site. Theoretically, it is also possible to distin-
guish more than two receptors but practically this is rarely feasible due to insuffi-
ciently selective competitors and/or an insufficient number of data points to reach
statistical superiority over a two-site fit. While performing the curve fitting, there is
no need to transform the y-values (which have the unit of cpm) to molar concentra-
tions.
 6 738 Biochemical and Analytical Techniques

However, it may sometimes be useful to normalize them (setting top to 100% and
bottom to 0%) in order to compare different experiments, or to show them on a single
graph.
If radioligand and competitor interact at the receptor in a competitive manner, the
IC50 value depends in part on the concentration of radioligand in the assay and thus
gives only a very crude estimate of the competitor affinity (Ki value) at this receptor.
The Ki value is computed from the IC50 using the Cheng-Prusoff (1973) equation

IC50
Ki = (16)
1 + [ligand]
K d, ligand

This calculation requires knowledge of the concentration of the radioligand in the

assay, which is determined in the same way as in a saturation binding experiment.

Practical Approach (Protocol)

In-Vitro Techniques

The practical approach to competition binding is very similar to that used in satura-
tion binding experiments.
1. Prepare a number of incubation tubes (or a microtitre plate) according to the
planned number of data points (2 each for total and non-specific binding and for
each concentration of the competitor).
2. Prepare the same number of vials for use in your radiation counter (alternatively
a corresponding filter plate which can directly be placed into the counter). Addi-
tionally, prepare 2 vials for the radioligand to allow quantification of the amount
of radioligand used in the assay and hence its true concentration. The latter vials
should also contain a filter to make sure that they are treated like the assay
samples.
3. Prepare sufficient amounts of incubation buffer, check its pH at the desired incu-
bation temperature, and cool it to 4 °C in an ice-bath.
4. Prepare the working solution for determining the non-specific binding.
5. Prepare the working solution of the radioligand from the stock solution provided
by the radiochemical manufacturer.
6. Prepare the working solutions of the competitor and its dilutions.
7. Add the working solution for non-specific binding, corresponding amounts of in-
cubation buffer or corresponding amounts of competitor solution to all incuba-
tion tubes (or wells).
8. Add the working solution of the radioligand to all tubes (or wells). Don’t forget to
add it as well to the tubes used for determination of total amount of radioligand
in the assay (see item 2).

All subsequent steps are identical to those for saturation binding (see above items
9–13).
 Receptor and Binding Studies 739 6.3

Examples

Saturation Binding Experiment

A saturation binding experiment in rat heart was performed, using [3H]-prazosin as

the radioligand to label α1-adrenoceptors. A rat heart (700 mg wet weight) was pre-
pared, and 120 µg protein per assay were used.
Counts were reported as follows:

Counts for total radioactivity (T) in the assay at radioligand concentrations 1–6:

T1 T2 T3 T4 T5 T6
30300 17813 16117 7619 3417 1720
31472 17804 14251 7084 3536 1963

Remember that this is actually the amount of radioligand used at the different dilu-
tions. These values will be used to calculate the concentration of the ligand according
to equation 5.

Counts for total (TB) and non-specific (NS) binding at radioligand concentrations 1–6:

TB1 TB2 TB3 TB4 TB5 TB6 NS1 NS2 NS3 NS4 NS5 NS6
513 440 451 374 289 168 154 94 51 32 22 15
541 477 410 392 279 161 171 87 62 37 22 21

Figure 5 shows these data with the averaged total values on the x-axis and total and
non-specific binding shown separately on the y-axis.

Figure 5
Total and non-specific binding,
expressed in cpm
 6 740 Biochemical and Analytical Techniques

Figure 6
Specific binding, expressed in
cpm

It may be helpful to look at your data this way in order to detect outliers or entry
errors. However, to evaluate the experiment, total counts are averaged, and the aver-
aged non-specific counts are subtracted from the averaged total counts:
In-Vitro Techniques

radioligand used, cpm 30886.0 17808.0 15184.0 7351.0 3476.0 1841.0

specific binding, cpm 364.5 368.0 374.0 348.5 262.0 146.5

The resulting graph of specific saturation binding is shown in Fig. 6.

The nonlinear regression estimated a Kd of 2430 and a Bmax of 421 cpm, which can
then be converted into molar concentrations using the specific activity of the
radioligand and the assay volume. Alternatively, these factors can be used to convert
cpm into molar concentrations prior to curve fitting to obtain the following mean
values:

radioligand used, pM 435 251 214 104 48.9 25.9

radioligand bound, fmol 5.13 5.18 5.26 4.91 3.69 2.06

When these values are used in curve fitting, Fig. 7 is obtained, which is basically iden-
tical to Fig. 6, except that the units have changed. Note that in this particular experi-
ment two data points are close to the estimated Kd and the three points with the
highest x-values define a plateau. This indicates that the experiment is indeed a satu-
ration experiment, and that binding to the receptor is saturable (i.e. that there are a
finite number of binding sites). If the smallest x-value were larger, the experiment
should be repeated with less radioligand, in order to get more reliable results from the
nonlinear fit.
In molar units, the Kd of prazosin for rat heart is 34.2 pM. As a last step, the recep-
tor density is calculated relative to a denominator such as protein content. Here a Bmax
of 5.93 fmol was determined using 120 µg protein per assay, and hence receptor den-
sity was 49.4 fmol receptors/mg protein used. Note that total binding is less than 9%
of the radioligand used (therefore no ligand-depletion). Non-specific binding
amounts to less than 1% of total radioligand used.
 Receptor and Binding Studies 741 6.3

Figure 7
Specific binding, expressed
in molar units. Compared
with Fig. 6, only the axes have
changed

Rosenthal Plot

Rosenthal plots were used before nonlinear regression by computer programs became
widely available. The data was transformed in such a way that Kd and Bmax could be
calculated by linear regression. However, the step of transforming the data enhances
the error, alters the relative weighting of the data and the x- and y-values will no
longer be independent of each other. So, it is no longer advisable to calculate Kd and
Bmax using Rosenthal plots based on linear regression.
However, Rosenthal plots are often useful to simply display data and therefore the
method of transforming the data is described briefly. X-values of Rosenthal plots are
the amount of bound radioligand, and the y-values are the ratio of bound to free
radioligand. So, x- and y-values have first to be interchanged, and then, the reciprocal
of each y-value is multiplied by its corresponding x-value to yield Fig. 8. The linear re-
gression line of the data will cross the x-axis at an x-value corresponding to Bmax, and
the slope of the line will be -1/Kd.

Figure 8
The same data as in Figs. 5 to
7, transformed to generate a
Rosenthal plot
 6 742 Biochemical and Analytical Techniques

Figure 9
Linear regression (dotted line)
of the transformed data. Note
the difference from the solid
line, which was generated using
Bmax and Kd obtained by non-
linear regression of the same
data

Figure 9 illustrates the differences in Kd and Bmax estimates when linear regression
of a Rosenthal plot and nonlinear curve fitting of untransformed data are used. The
solid line was generated using the best-fit values of nonlinear regression performed
on the raw data as described above. The dotted line, however, is a best-fit linear regres-
In-Vitro Techniques

sion. In a quantitative analysis, Kd is estimated to be 34.2 pM and 50.0 pM in non-lin-

ear regression and Rosenthal plot, respectively, whereas Bmax values of 5.93 and
6.53 fmol, respectively, are obtained. Depending on the technical quality of the data
even greater differences may occur.

Competition Binding Experiment

A competition experiment was performed for rat cardiac α1-adrenoceptors with [3H]-
prazosin (Kd = 34.2 pM, taken from the above example) as the radioligand and nora-
drenaline as the competitor (concentrations ranging from 10–9 to 10–4 M). The protein
content was 1180 µg per assay. The following results were obtained (Table 2).
These counts were graphed with the logarithm of the concentrations as x- and the
counts as y-values (Fig. 10). Total binding (TB) was graphed at an x-value of 10–12, and
non-specific binding (NS) at an x-value of 10–3. Next, a sigmoidal curve with a variable
Hill-slope was fitted to the data. This was characterized by a Hill-slope of -1.14 (i.e.
close to unity), a -log IC50 of 4.75, a top of 3311 and a bottom of 273. While the data
nicely define a top plateau, they do not cover a bottom plateau. Hence, the calculated
bottom plateau is slightly lower than the averaged non-specific binding of 289. When
performing further experiments, the concentration range should be shifted to –8 to
–3; however, it should be kept in mind that non-specific binding of the competitor can
also increase with such high concentrations.
Since the curve fits the data nicely and the calculated bottom-plateau looks reli-
able, this experiment may also be evaluated further. The duplicate measurements were
averaged and normalized – 3311 cpm were defined as 100% and 273 cpm as 0%
(Fig. 11). Since top and bottom plateaus are defined by normalizing the data, nonlin-
ear regression of the normalized data should fit the curve with top and bottom set
constant to 100 and 0%, respectively, and the data points corresponding to total and
 Receptor and Binding Studies 743 6.3
Table 2
Competitor concentrations given as log values

log M Cpm log M Cpm

TB 3315 3202 –6.5 3288 3126
NS 252 271 –6.3 3400 3487
–9 3077 2998 –6 3214 3355
–8.7 2906 3370 –5.7 3123 2925
–8.5 3380 3475 –5.5 3104 2960
–8.3 3270 3100 –5.3 2798 2410
–8 3336 3176 –5 2420 2219
–7.7 3408 3439 –4.7 1598 1764
–7.5 3275 3228 –4.5 1347 1223
–7.3 3527 3284 –4.3 1054 1059
–7 3572 3271 –4 670 650
–6.7 3456 3474 NS 298 337

non-specific binding should be excluded from further analysis. This normalization

process is not part of the analysis of the experiment but may be helpful e.g. in order
to pool results from several such experiments.
As a last step, the Ki value is calculated. The total counts, i.e. the counts of the to-
tal amount of radioligand used, were 22998 and 23050 cpm, averaging 23024 cpm.
This corresponds to a concentration of the radioligand (specific activity 80 Ci/mmol)

Figure 10
Total and non-specific binding
(closed circles) and binding at
various competitor concentra-
tions (open circles). The IC50
value can already be obtained
 6 744 Biochemical and Analytical Techniques

Figure 11
Data from Fig. 10 normalized
(see text for details)

of 324 pM based on an assay volume of 1000 ml and a counter efficiency of 0.4. As

presented in the example above, the Kd of [3H]-prazosin for rat heart is estimated to
In-Vitro Techniques

be 34.2 pM. Using the equation of Cheng-Prusoff (1973), the Ki can be calculated by

–4.75
IC50 10 M
= 1.70 × 10 M = 1.70 µM
–6
Ki = = (17)
1 + [ligand]
K
1+ 324pM
32.4pM
d, ligand

This equals a pKi for noradrenaline in rat heart of -5.77.

Troubleshooting
Problem: Duplicate values are not in good agreement with each other.

Solution: Apart from sloppy pipetting, lack of homogeneity of the membrane prepa-
ration is the most frequent reason. Try to homogenize the tissue or cell preparation
more vigorously. Don’t forget that the pellet from the last centrifugation step also
needs to be re-homogenised. If all else fails, use triplicates or quadruplicates rather
than duplicates to get more robust readings.

Problem: Non-specific binding at radioligand concentrations around Kd is high.

Solution: High non-specific relative to specific binding is a signal-noise question, and

hence can result from to little signal or too much noise. Too little signal is due to too
few receptors within the assay. This can sometimes be overcome by adding more
membrane preparation to the assay. However, if receptor expression in the tissue of
interest is very low, this will not help since non-specific binding increases as well. Too
much noise can have many reasons. Some radioligands intrinsically have more non-
 Receptor and Binding Studies 745 6.3
specific binding than others (e.g. [3H]-quinuclidinyl-benzylate relative to [3H]-N-
methyl-scopolamine for muscarinic acetylcholine receptors or [3H]-rauwolscine and
[3H]-yohimbine relative to [3H]-RX 821002 for α2-adrenoceptors). Some radioligands
stick to certain surfaces, and attempts to use other types of vials or silicon coating of
vials can be helpful. Sometimes radioligands stick to the filters. Use of alternative fil-
ter types and/or pre-treatment of the filters with polyethylimine (toxic!) can be help-
ful. Alternatively, a switch to other means of separation of bound and free radioligand
can be considered.

Problem: Total binding (sum of specific and non-specific binding) exceeds 10% of
total radio-ligand in the assay.

Solution: This situation creates ligand depletion and hence prevents a valid math-
ematical assessment of the data obtained. Increase the assay volume while maintain-
ing the amount of membrane preparation in the assay. This will increase the amount
of radioligand without affecting its concentration. Additionally, the amount of mem-
brane (receptor) in the assay can be reduced to decrease binding; this can be particu-
larly helpful when working with transfected cells expressing artificially high receptor
numbers.

Problem: Confidence intervals for Bmax estimates in saturation binding are very wide.

Solution: Use more radioligand concentrations in a range expected to cover Bmax. This
will enhance the robustness of Bmax estimates.

Problem: Confidence intervals for Kd estimates in saturation binding are very wide.

Solution: Use more radioligand concentrations in a range around the expected Kd.
This will enhance the robustness of Kd estimates.

Problem: Fitting of curves to competition data does not yield a reasonable top plateau.

Solution: Use more data points in the region where you expect the top plateau, par-
ticularly at very low competitor concentrations.

Problem: Fitting of curves to competition data does not yield a reasonable bottom pla-
teau or even negative estimates.

Solution: A variety of possible explanations exists. Firstly, it is possible that the com-
puter program is attempting to fit to unreasonably low values because there are insuf-
ficient data points at high competitor concentrations, i.e. the program extrapolates far
beyond the data. This is best prevented by having more data points in the concentra-
tion range of competitor where the bottom plateau is expected. Secondly, it is also
possible that the competitor indeed competes for more sites than the agent used for
definition of non-specific binding (because the definition was chosen incorrectly in
the first place or because the solution of agent used for defining non-specific binding
 6 746 Biochemical and Analytical Techniques

does not contain what it is supposed to contain). Check whether the definition of non-
specific binding is indeed correct, e.g. by performing a competition experiment with
a range of concentrations of the agent used for definition of non-specific binding.

Problem: Confidence intervals for the estimate of IC50 in a competition binding experi-
ment are very wide.

Solution: If the overall data are good, this is usually because there are an insufficient
number of competitor concentrations being checked per log unit of concentration
range. Increase the number of competitor concentrations per log unit.

Problem: When fitting a curve with a floating Hill-slope to competition binding data,
the Hill-slope becomes considerably greater than unity (the competition curve is very
steep).

Solution: If this is reproducibly so, the most likely explanation is that the mechanism
of inhibition is not strictly competitive. This can be tested by performing saturation
binding experiments in the absence and presence of a moderate concentration of in-
In-Vitro Techniques

hibitor. A strictly competitive inhibitor should increase the apparent Kd of the

radioligand but not affect its Bmax.

Problem: When fitting a curve with a floating Hill-slope to competition binding data,
the Hill-slope becomes considerably smaller than unity (the competition curve is very
shallow).

Solution: This is the expected result if the radioligand labels two sites (e.g. subtypes of
a receptor) that are discriminated by the competing agent. Perform a two-site fit and
compare it to the results of a one-site fit (Hill-slopes must be fixed at unity for both
equations) by an F-test. However, not every shallow competition curve is due to a
biphasic inhibition curve. Other more complex reasons such as negative cooperativity
of binding may also exist.

Problem: The fitted curve obviously does not explain the data.

Solution: This can occur due to the phenomenon of local minima in iterative curve
fitting. During the iterative processes the program performs stepwise deviations from
the starting parameters to test whether they improve the fit. When the starting param-
eters are far away from the actual data, this can lead to inaccurate parameter estimates
(for a detailed discussion see Motulsky and Ransnäs 1987). Use your scientific under-
standing to “manually” estimate starting parameters for the curve fitting and start
anew.

Problem: Curve fitting does not converge.

Solution: There are two possible explanations. The starting values of the fitting pro-
cedure may be implausible. In this case select new starting parameters using your
scientific understanding. Alternatively, the mathematical model does indeed not fit
 Receptor and Binding Studies 747 6.3
the data. In this case look at the quality of your data (this problem frequently occurs
if the technical quality of the data is poor and lots of scatter exists). If data quality
seems good, reconsider whether the mathematical model is appropriate for these data.

Problem: I have a problem that is not resolved by any of the above hints.

Solution: Write an e-mail to the friendly authors of this chapter. Perhaps they can help.

References
Brodde O-E Khamssi M, Zerkowski H-R (1991) β-Adrenoceptors in the transplanted human heart: unaltered β-
adrenoceptor density, but increased proportion of β2-adrenoceptors with increasing posttransplant time. Naunyn-
Schmiedeberg’s Arch. Pharmacol. 344:430–436
Burgisser E (1984) Radioligand-receptor binding studies: What’s wrong with the Scatchard analysis? Trends Pharmacol.
Sci. 5:142–145
Cheng Y-C, Prusoff WH (1973) Relationship between the inhibition constant (Ki) and the concentration of inhibitor
which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol. 22:3099–3108
DeBlasi A, O’Reilly K, Motulsky H (1989) Calculating receptor number from binding experiments using the same com-
pound as radioligand and competitor. Trends Pharmacol. Sci. 10:227–229
Guth PS (1982) The structurally specific, stereospecific, saturable binding of pepperoni to pizza. Trends Pharmacol.
Sci. 3:467
Hein P, Goepel M, Cotecchia S, Michel MC (2000) Comparison of [3H]tamsulosin and [3H]prazosin binding to wild-
type and constitutively active α1B-adrenoceptors. Life Sci. 67:503–508
Hulme EC (1992a) Centrifugation binding assays. In: Hulme EC (ed) Receptor-Ligand Interactions. IRL Press, Oxford,
p 235–246
Hulme EC (1992b) Gel-filtration assays for solubilized receptors. In: Hulme EC (ed) Receptor-Ligand Interactions. IRL
Press, Oxford, p 255–264
Hulme EC, Buckley NJ (1992) Receptor preparations for binding studies. In: Hulme EC (ed) Receptor-Ligand Interac-
tions. IRL Press, Oxford, p 177–212
Motulsky HJ, Ransnäs LA (1987) Fitting curves to data using nonlinear regression: a practical and nonmathematical
review. FASEB J. 1:365–374
Strange PG (1992) Charcoal adsorption for separating bound and free radioligand in radioligand binding assays. In:
Hulme EC (ed) Receptor-Ligand Interactions. IRL Press, Oxford, p 247–254

Further reading
Hulme EC (1992) (ed) Receptor-Ligand Interactions. IRL Press, Oxford, p 255–264 (An affordable, in-depth text ex-
plaining many different aspects of radioligand binding experiments)
Yamamura HI, Enna SJ, Kuhar MJ (1985) Neurotransmitter Receptor Binding. 2nd edition, Raven Press, New York (The
mother of all reference books on radioligand binding)

6.3.2
Tracer Techniques for Detection of Transport
Kirsten Leineweber

Introduction
Transport can be investigated by tracer techniques. As an example for this method, in
the following chapter the investigation of noradrenaline uptake will be described.
Sympathetic nerve activity is altered and is a prognostic factor for many cardiovascu-
 6 748 Biochemical and Analytical Techniques

lar diseases such as hypertension, coronary syndromes, and congestive heart failure.
It is now generally accepted that an increase in sympathetic nerve activity is closely
associated with increased plasma noradrenaline levels and reduced myocardial cate-
cholamine stores due to decreased cardiac neuronal noradrenaline transporter activ-
ity, which is also known as uptake1, and a reduction in noradrenaline transporter
density. Therefore, uptake1 that removes more than 90% of the synaptically released
noradrenaline into noradrenergic neurons, plays a pivotal role in the inactivation of
the transmitter (for review see Esler et al. 1997).
In addition to the endogenous catecholamines, noradrenaline, dopamine and
adrenaline, uptake1 also transports indirectly acting amines such as tyramine and
the psychostimulant amphetamine, as well as the dopaminergic neurotoxin MPP+
[1-methyl-4-phenylpyridinium]. Uptake1 is inhibited selectively, with nanomolar af-
finity, by nisoxetine and the tricyclic antidepressant desipramine and less potent by
cocaine.
Accordingly, drugs acting on the noradrenaline transporter can be broadly clas-
sified as substrate analogs or as non-substrate uptake1 inhibitors. The former com-
pounds inhibit noradrenaline transport by being transported instead of
noradrenaline, whereas uptake1 inhibitors prevent noradrenaline uptake but are not
In-Vitro Techniques

themselves transported into the cell.

In the case of the noradrenaline transporter protein, three binding sites have been
identified, one for Na+, one for Cl– and one for the protonated substrate. Substrate
transport as well as inhibitor binding are absolutely dependent on Na+ and Cl– ions:
while Cl– can be replaced by Br–, and to a lesser extent by SCN– or NO2–, it is not pos-
sible to replace Na+ by other cations. The Na+/Cl-/noradrenaline stoichiometry ob-
served for the noradrenaline transporter protein is 1:1:1 (for review see Graefe and
Bönisch 1988; Esler et al. 1997).
The noradrenaline transporter is mainly located in the axonal membranes of nor-
adrenergic neurons and adrenal medullary cells. However, it has also been found na-
tively expressed in some non-neuronal tissues such as glandular epithelial cells of
rabbit uterine endometrium, rat phaeochromocytoma (PC12) cells, human placental
brush border membrane vesicles and rat placenta trophoblasts, rabbit dental pulp
fibroblasts, rabbit lingual gingiva and nasal mucosa, rat interscapular brown adipose
tissue, bovine adrenal medullary cells and pulmonary endothelial cells. Evidence has
accumulated that the noradrenaline transporter is also expressed in many com-
ponents of the cardiovascular system since the lining of arteries (e.g. the dorsal aor-
ta) and veins (e.g. subcardial vein), the epicardium and the cardiac trabeculae are
strongly noradrenaline transporter-immunoreactive (for review see Graefe and Bö-
nisch 1988; Esler et al. 1997).
The human noradrenaline transporter (hNAT, Gen Bank Acc No: M65105) was the
first noradrenaline transporter to be cloned, followed much later by cloning of the
bovine (bNAT, X79015), the rat (rNAT, Y13223), the mouse (mNAT, U76306) and the
monkey (Macaca mulatta) noradrenaline transporter (mmNAT; AF286026). At the
nucleotide level, hNAT and rNAT share 87% identity in the coding region, while hNAT
and bNAT show 90% and hNAT and mmNAT show 98.9% identity. However, studies
investigating these cloned transporters transfected in different cell types revealed
naturally occurring variations in the properties of the various noradrenaline trans-
porters in different species. The hNAT has higher affinities than the rNAT for sub-
 Receptor and Binding Studies 749 6.3
strates, such as noradrenaline and MPP+, with no differences between affinities for
tricyclic and non-tricyclic antidepressants, such as nisoxetine and desipramine. On
the other hand, the hNAT has a higher affinity for cocaine than the rNAT.
The gene encoding hNAT has been assigned to 16q12. It consists of 14 exons cod-
ing for a 617 amino acid protein that shows sequence similarity to the dopamine
transporter, the serotonin transporter and the GABA transporter. Recently, thirteen
naturally occurring single nucleotide polymorphisms within the hNAT gene have
been identified: Five intron variants in flanking intron sequences that do not alter
splice site acceptor/donor consensus sequences, three silent mutations (955T/C,
1287G/A, 1779G/A) and five missense variants (Val69Ile, Thr99Ile, Val245Ile, Val449Ile,
Gly478Ser) located at TMDs 1, 2, 4, 9 and 10, respectively (for review see Brüss et al.
1997). Runkel et al. (2000), who investigated the functional importance of the five
amino acid substitutions within the noradrenaline transporter protein in vitro found,
however, that none of these variants affected the potency of desipramine for inhibition
of uptake1. Furthermore, there were no differences in the kinetic constants (KM, Vmax)
for the hNAT Val69Ile, Thr99Ile, Val245Ile and Val449Ile variants. The hNAT Gly478Ser
variant, on the other hand, exhibited a four-fold reduced substrate affinity for nora-
drenaline (without change in maximum rate of transport) in comparison to the wild-
type hNAT. Whether these polymorphisms are of functional importance in vivo is still
a matter of debate.
The noradrenaline transporter protein, with a molecular mass of about 80 kDa, is
characterized by 12 potential transmembrane domains (TMDs) and cytoplasmic
amino and carboxyl termini. A large extracellular loop, containing 2-4 potential N-
glycosylation sites, is located between TMD-3 and TMD-4 and potential sites for phos-
phorylation by protein kinases (two for protein kinase C, and one for casein kinase II)
are found at intracellular domains. Structure-function analysis using recombinant
chimeras of the dopamine and noradrenaline transporter, and serotonin and norad-
renaline transporter identified several functional domains within the monoamine
transporter proteins where TMD-1 has been identified as being critically involved in
ligand binding and substrate transport.
The process by which the noradrenaline transporter specifically binds and trans-
fers its substrate across the membrane resembles a typical enzyme-substrate reaction
and follows the saturation kinetics of the Michaelis-Menten type (Graefe and Bönisch
1988): 1. When the carrier protein is saturated (that is when all binding sites are oc-
cupied), the rate of transport is maximal (Vmax) and 2. The noradrenaline transporter
has a characteristic binding constant for its substrate (KM, concentration of substrate
when the transport rate is half its maximum value). Therefore, from the number of
transporter sites estimated from inhibitor binding (Bmax) and from the maximal trans-
port rate of noradrenaline (Vmax), the duration of the transport cycle for a substrate at
Vmax can be calculated (Vmax/Bmax). It should be noted that Vmax for uptake1 is substrate-
concentration-dependent and by this is the length of the transport cycle. The trans-
porter undergoes a conformational change in order to transfer the substrate across
the membrane. This reaction includes:
1. binding of the substrate (i.e. the transported ligand) to a specific binding site
2. translocation
3. dissociation and release of the substrate to the other side of the membrane and
4. return to the unloaded state.
 6 750 Biochemical and Analytical Techniques

General characteristics of substrate transport mediated by uptake1 include:

1. temperature-dependence of transport
2. requirement of Na+ and Cl– ions
3. saturability of transport
4. structural substrate specificity and/or stereoselectivity (uptake1 operates at low
concentrations (< 3 µmol/ l) and favours the accumulation of (-)-noradrenaline
that has more than twice the affinity of (–)-adrenaline)
5. sensitivity to specific competitive inhibitors (which compete for the same binding
site and may or may not be transported by the carrier) or non-competitive inhibi-
tors (which bind elsewhere and specifically alter the structure of the carrier)
which differ chemically from the solute under study and
6. transport occurs via a limited number of transporter proteins.

The properties determined by the noradrenaline transporter protein permits a vecto-

rial reaction, i.e., substrate movement from one side of the membrane to the other
side. To test the characteristics of uptake1 it is therefore useful to assess:
1. the time course of uptake of a labeled substrate at different temperatures (4 °C to
37 °C) to determine whether the transported substrate is accumulated and
In-Vitro Techniques

whether substrate uptake increases linearly with time (indicating a unidirectional

inward transport),
2. the dependence of transport-rate or its kinetic determinants KM and Vmax on Na+
and Cl– concentration,
3. KM and Vmax for a given substrate by incubating the material under investigation
(vesicles, cells, tissue slices) with various concentrations (below and above the
expected KM), of the radiolabeled substrate for a short time period (initial rate)
and a defined temperature,
4. whether [3H]-noradrenaline uptake is inhibited by standard inhibitors to deter-
mine the IC50-values (i.e., the concentration which causes half-maximum inhibi-
tion of either uptake or binding of a labeled ligand at a given concentration) for
the unlabeled substrates,
5. whether [3H]-noradrenaline uptake is inhibited by standard inhibitors of the
transport system to determine the IC50-values for the unlabeled antagonists,
6. Bmax (as mentioned above the value reflects the density (number) of carrier sites)
and KD (value reflects the concentration by which half of the carriers are occupied
by the radiolabeled ligand) for the transporter,
7. whether the accumulation of the substrate is dependent on the protein or tissue
amount that is used.

Description of Methods and Practical Approach

The measurement of noradrenaline transporter activity and density has become rou-
tine now that many radioactively labelled ligands, substrates like [3H]-noradrenaline
or metaiodobenzylguanidine (MIBG) radioiodinated with 131I or 123I or antagonists
like [3H]-desipramine, [3H]-mazindol and [3H]-nisoxetine, are commercially available.
 Receptor and Binding Studies 751 6.3
The use of the noradrenaline analog [131I/123I]-MIBG, to determine uptake1 in vivo,
was first developed and utilized clinically as an agent for scintigraphic imaging of
neuroendocrine tumors, such as pheochromocytomas and neuroblastomas, but has
also proven to be useful for imaging of cardiac sympathetic nerves (for review see
Kiyono et al. 2001). In the myocardium, MIBG is carried like noradrenaline to the
sympathetic nervous terminals by myocardial blood flow; it is taken up by cardiac
sympathetic neurons via uptake1 and shares noradrenaline storage in intraneuronal
vesicles. Like noradrenaline, MIBG is released from synaptic vesicles into the synap-
tic cleft by neuronal stimulation. However, it should be noted that MIBG also shares
noradrenaline extraneuronal uptake, which is not inhibited by desipramine. There-
fore, a desipramine pretreatment should be considered to differentiate specific neu-
ronal MIBG uptake by uptake1 from non-specific extraneuronal uptake. Additional
radiopharmaceuticals useful for PET imaging of neuroendocrine tumors or sympa-
thetic nerves are [11C]-hydroxyephedrine (Allman et al. 1993; Malizia et al. 2000) and
[18F]-6-fluorodopamine (FDA).
Desipramine and substrates share a common binding site and like noradrenaline
the binding of desipramine is also Na+ and Cl– dependent with a coupling ratio for
Na+/desipramine of 2 or 3 and for the Cl–/desipramine ratio of 1. Both Na+ and Cl–
enhance the binding affinity of desipramine to the noradrenaline transporter. The
radiolabeled noradrenaline transporter antagonist [3H]-desipramine was reported by
several investigators (for review see Lee and Snyder 1981) to bind uptake1 sites for NA
in rat brain. However, [3H]-desipramine exhibits a high amount of non-specific bind-
ing due to its highly-lipid solubility and binds in addition to a high affinity site to a
low affinity site not related to the noradrenaline transporter. If non-specific binding
is properly defined so that only the high-affinity site (transporter) is measured, the
percentage of specific binding is found nevertheless to be low (40% at twice the KD
concentration). Therefore, although desipramine has a high affinity for uptake1 and
can be used in ex vivo and in vitro experiments to locate the noradrenaline trans-
porter, its usefulness in vivo is limited due to the high level of non-specific binding.
Furthermore, it should be noted that the pharmacokinetics of desipramine are non-
linear and related to cytochrom P450-2D6 activity, making single subject pharmaco-
kinetics prediction (different plasma levels) difficult.
[3H]-Mazindol has also been used to label uptake1 sites in the rat brain or the fail-
ing human heart (for review see Böhm et al. 1995). However, [3H]-mazindol binds as
well to dopamine uptake sites with high affinity, necessitating the use of selective in-
hibitors of the dopamine uptake when measuring uptake1 sites for noradrenaline.
Additionally, to displace [3H]-mazindol, extremely high concentrations of desipra-
mine (30–100 mmol/l) are necessary. Thus, it is possible that desipramine also dis-
places [3H]-mazindol from non-specific binding sites resulting in a lower than actual
noradrenaline transporter density.
On the other hand, [3H]-nisoxetine ([N-methyl-3-(o-methoxyphenoxy)-3-phenyl-
propylamine]) is a potent and selective inhibitor of uptake1, being approximately
1000-fold and 400-fold more potent in blocking the uptake of NA than that of seroto-
nin or dopamine, respectively (for review see Tejani-Butt 1992; Jayanthi et al. 1993).
Furthermore, [3H]-nisoxetine has little or no affinity for many neurotransmitter re-
ceptors. In contrast to [3H]-desipramine, [3H]-nisoxetine binds to a single class of
 6 752 Biochemical and Analytical Techniques

binding sites and the percentage of specific binding is about 80% at twice its KD con-
centration. It is suggested that the binding site for noradrenaline and nisoxetine are
either identical or exhibit a considerable steric overlap and that [3H]-nisoxetine does
not bind to intracellular non-adrenergic sites as do the mentioned ligands above.
Binding of [3H]-nisoxetine is pH-dependent with a pKa of 5.7, with maximal binding
in the pH range of 7.0–8.5, and shows an absolute requirement for Na+ and Cl–. The
presence of Na+ increases the binding affinity of [3H]-nisoxetine in a concentration-
dependent manner.
The relationship between the binding and the Na+ concentration is sigmoidal,
suggesting a coupling ratio of Na+/nisoxetine of 2. On the other hand, the relationship
between the binding and the Cl– concentration is hyperbolic, suggesting that the cou-
pling ratio of Cl–/nisoxetine is one.
Although Tris-containing buffers can be used with this radioligand, the best re-
sults with regard to total and non-specific binding can be obtained with a buffer
consisting of 10 mM Na2HPO4, 120 mM NaCl and 5 mM KCl (pH 7.4). Binding of [3H]-
nisoxetine is also temperature-dependent. Maximal binding occurs within 15 min of
incubation at 22 °C, whereas, a 3–4 h incubation is required to obtain maximal bind-
ing at 4 °C.
In-Vitro Techniques

On the other hand, the dissociation at 22 °C is very rapid, the time required for
50% dissociation is less than 5 min. Although nisoxetine is an appropriate radio-
ligand for labelling and blocking uptake1 sites in in vitro and ex vivo experiments,
the [11C] analog of nisoxetine, created for PET imaging in vivo was found unsuit-
able.

Examples
Detection and Quantification of Uptake1 Sites in Crude Cardiac Membranes

Protocol 1: Crude Membrane Preparation

1. Thaw 100–150 mg frozen tissue (after excision immediately frozen in liquid N2

and stored at –80 °C) on ice in 5 ml ice-cold incubation buffer (10 mM Na2HPO4,
120 mM NaCl, 5 mM KCl; pH 7.4) containing 0.25 M sucrose.
2. Mince the tissue with scissors and gradually homogenize it with an Ultra Turrax
(Ultra-Turrax T25, Jahne andKunkel IKA Labortechnik, Germany) at 24,000 rpm
for 10 s. and twice at 17,500 rpm for 20 s. with 1 min intervals on ice.
3. Bring the homogenate up to 20 ml with incubation buffer (+ 0.25 M sucrose) and
centrifuge it at 1,200×g for 10 min at 0 °C.
4. Filter the supernatant through four layers of cheese cloth and centrifuge it at
20,000×g for 20 min at 4 °C.
5. Re-suspend the resulting pellet in 20 ml incubation buffer (+ 0.25 M sucrose) and
re-centrifuge at 20,000×g for 20 min at 4 °C.
6. Re-suspend the final crude membrane pellet in 1 ml ice-cold incubation buffer
(without sucrose!).
 Receptor and Binding Studies 753 6.3
7. Freeze the protein suspension immediately in liquid N2 and store it at –80 °C (nor-
adrenaline transporter protein in suspension is stable for up to 4 days at –80 °C).

Protocol 2: [3H]-Nisoxetine Binding

1. Thaw the protein suspension obtained in Protocol 1 on ice.

2. Determine the protein content according to Bradford (1976) using bovine γ-glo-
bulin as a standard.
3. Dilute the protein suspension with incubation buffer (without sucrose) to reach a
protein concentration of 100 µg protein/400 µl incubation buffer (without su-
crose) and store it on ice until further use.
4. Dilute [3H]-nisoxetine ([N-methyl-3H]-Nisoxetine; specific activity 80 Ci/mmol;
NENTM Life science Products Inc., Boston, MA, USA) with incubation buffer to
reach 6 different concentrations ranging from 100 nM down to 3.125 nM.
5. Prepare a solution of 1 mM desipramine by dissolving the antagonist in distilled
water and dilute it by adding 50 µl of this solution to 4.95 ml incubation buffer
(without sucrose).
6. Place 12×75 mm polypropylene tubes in a tube rack standing in an ice-water bath
and add 50 µl either of incubation buffer (total binding) or 50 µl desipramine
(final concentration 1 µM /assay; non-specific binding) and 400 µl of the protein
solution.
7. Pre-incubate the compounds for at least 15 min on ice.
8. Add to the tubes (+/– desipramine) 50 µl of the corresponding [3H]-nisoxe-
tine dilution (final assay volume 500 µl) and incubate the compounds for 3 h at
4 °C.
9. Terminate the incubation by adding 5 ml of ice-cold incubation buffer (without
sucrose) to the reaction mixture and rapidly filter them through Whatman GF/C
filters using a 12-well harvester. Wash the filters twice each time with 5 ml incuba-
tion buffer (without sucrose).
10. Transfer the filters into scintillation vials and allow them to dry over night.
11. Add 4 ml scintillation cocktail (LumasafeTM Plus, Lumac, Groningen, Nether-
lands) to each vial and shake them for at least 10 min at room temperature.
12. Determine the radioactivity with a scintillation counter (e.g. Packard TRI-Carb
2250CA) that has a minimum efficiency of 60% for [3H]-labeled ligands.

Protocol 2 should be performed in duplicates for each [3H]-nisoxetine dilution. Ac-

cording to this protocol “non-specific” binding of [3H]-nisoxetine is defined as bind-
ing in the presence of 1 µM desipramine. Specific binding is defined as total
binding minus non-specific binding and should amount 80% at 2.5 nM [3H]-nisoxe-
tine.
To estimate the equilibrium dissociation constant (KD) and the maximal number
of binding sites (Bmax) for [3H]-nisoxetine, we use the iterative curve-fitting program
Prism 2 from GraphPad Software (San Diego, CA, USA), to analyze the data by non-
linear regression (hyperbolic function). The non-linear regression analysis of the
saturation experiment should reveal that [3H]-nisoxetine binds to a single population
of binding sites.
 6 754 Biochemical and Analytical Techniques

Protocol 3: Drug Competition Experiments

1. Follow steps 1 to 3 of Protocol 2.

2. Dilute [3H]-nisoxetine ([N-methyl-3H]-Nisoxetine; specific activity 80 Ci/mmol;
NENTM Life science Products Inc., Boston, MA, USA) with incubation buffer to
reach a concentration of 20 nM.
3. Prepare a solution of 1 mM desipramine, nisoxetine, cocaine or GBR 12909 (do-
pamine transporter antagonist) by dissolving each in distilled water and dilute
them to get at least 8–12 different concentrations ranging from 1 mM down to
0.01 nM.
4. Place 14×12 ml polypropylene tubes in a tube rack standing in an ice-water bath
and add in one tube 50 µl incubation buffer (to determine total binding), in one
tube 50 µl 1 µM desipramine (final concentration 1 µM /assay; to determine non-
specific binding), and in the other tubes 50 µl of the corresponding antagonist di-
lution. Add in each tube 400 µl of the protein solution.
5. Pre-incubate the compounds for at least 15 min on ice.
6. Add to all tubes 50 µl of the 20 nM [3H]-NIS dilution (final assay volume 500 µl)
and incubate the compounds for 3 h at 4 °C.
In-Vitro Techniques

7. Follow steps 9 to 12 of Protocol 2.

Protocol 3 should be performed in duplicates for each antagonist dilution. According

to this protocol “non-specific” binding of [3H]-nisoxetine is defined as binding in the
presence of 1 µM desipramine. Specific binding is defined as total binding minus non-
specific binding. The competition binding data can be analyzed by computer-assisted
non-linear regression analysis (Prism 2 from GraphPad Software) to fit a one-site
competition curve to estimate the IC50-value (concentration of the competitor to in-
hibit specific [3H]-nisoxetine binding by 50%).

Determination of Uptake1 Activity in Cardiac Tissue Slices

Protocol 4: [3H]-Noradrenaline Uptake1 Activity

1. Place the tissue immediately after excision in ice-cold oxygenated (5% CO2/95%O2)
Krebs-Henseleit solution (118 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl2, 0.54 mM
MgCl2, 25 mM NaHCO3, 1 mM NaH2PO4, 11 mM glucose, 0.094 mM EDTA, 1.14 mM
ascorbic acid, 0.067 mM nialamide, MAO-blocker).
2. Mince about 120 mg tissue into small pieces using scissors and cut them first into
500×500 µm and then into 250×250 µm slices covered with about 500 µl ice-cold
oxygenated Krebs-Henseleit solution using a tissue chopper (McIllwain; Bach-
hofer, Reutlingen, Germany).
3. Wash the slices twice in fresh ice-cold oxygenated Krebs-Henseleit solution.
4. Re-suspend the tissue slices in ice-cold oxygenated Krebs-Henseleit solution, con-
taining 40 µM corticosterone (stock solution solved in 0.1% DMSO) to block
extraneuronal uptake of noradrenaline), to reach a dilution of 10 mg slices per
400 µl buffer.
 Receptor and Binding Studies 755 6.3
5. Because the KM value is in the high nanomolar range it is necessary to dilute the
labelled with unlabelled substrate to reveal high noradrenaline concentration.
Therefore, prepare a solution of 250 nM noradrenaline (diluted in ice-cold oxy-
genated Krebs-Henseleit solution). Take 2.5 ml of the solution and add 25 µl [3H]-
noradrenaline (D,L-(7-3H[N])-Noradrenaline hydrochloride; specific activity
13.5 Ci/mmol; NENTM Life science Products Inc., Boston, MA, USA).
6. Dilute the 250 nM [3H]-noradrenaline stock solution to get at least 5 different con-
centrations ranging from 250 nM down to 15.6 nM.
7. Place 10×6 ml polypropylene tubes (two tubes per noradrenaline dilution) in a
tube rack standing at room temperature and 10×6 ml polypropylene tubes in a
tube rack standing in an ice-water bath (to determine non-specific accumulation
of radioactivity) and add in all tubes 50 µl ice-cold oxygenated Krebs-Henseleit
solution. Add in each tube 400 µl (=10 mg) tissue slices.
8. Equilibrate the compounds to each temperature for at least for 15 min.
9. Add to the tubes (room temperature/4 °C) 50 µl of the corresponding [3H]-nora-
drenaline dilution (final assay volume 500 µl) and incubate the compounds for
15 min either at 37 °C or 4 °C.
10. Terminate the incubation by adding 3 ml of ice-cold ice-cold oxygenated Krebs-
Henseleit solution to the reaction mixture and rapidly filter them through
Whatman GF/C filters, that had been pre-soaked for 15 min in ice-cold Krebs-
Henseleit solution containing 0.05% polyethylenimine, using a 12-well harvester.
Wash the filters twice each time with 5 ml incubation buffer (without sucrose).
11. Transfer the filters into scintillation vials. Add 500 µl of ice-cold 3% trichloroace-
tic acid (TCA) to each filter and incubate them (digestion of the slices and release
of [3H]-noradrenaline in the liquid phase) overnight at 4 °C.
12. Add 4 ml scintillation cocktail (LumasafeTM Plus, Lumac, Groningen, Nether-
lands) to each vial and shake them for at least 10 min at room temperature.
13. Determine the radioactivity with a scintillation counter as mentioned in Protocol
2, step 12.

According to this protocol, which determines the initial rate of [3H]-noradrenaline

uptake at low concentrations (1.56–25 nM), specific uptake1 is defined as total uptake
(37 °C) minus non-specific uptake (4 °C). To determine uptake1 kinetic parameters
the slices have to be incubated as described above with [3H]-noradrenaline concentra-
tions leading to the saturation of uptake1 (final concentration about 250 nM). Again,
specific uptake1 must be determined by incubating the slices in parallel at 37 °C and
4 °C. Competition studies with desipramine, nisoxetine, cocaine and GBR 12909 have
to be conducted as described above by incubating tissue slices with a fixed concentra-
tion of [3H]-noradrenaline (25 nM) in the presence of 8–12 concentrations of each
competitive drug. As described in Protocol 3 the tissue slices have also to be pre-in-
cubated with the antagonists for at least 15 min. (the interaction between the norad-
renaline transporter and inhibitors requires more time than this for substrate,
implying that the binding of inhibitors might involve a significant conformational
change in the transporter protein), therefore the volume of 50 µl of Krebs-Henseleit
solution can just be replaced by the same volume of the antagonist dilution. Again,
non-specific accumulation of radioactivity is determined by parallel incubation at
 6 756 Biochemical and Analytical Techniques

4 °C. To ensure that only uptake1 is active under the experimental conditions describ-
ed above, tissue slices have to be incubated with [3H]-noradrenaline in the presence
of 40 µM of the uptake2-inhibitor corticosterone. To estimate the kinetic parameters
KM and Vmax of the noradrenaline transporter we use the iterative curve-fitting pro-
gram Prism 2 from GraphPad Software to analyze the data by non-linear regression
(hyperbolic function). The competition data can be analyzed by computer-assisted
non-linear regression analysis (Prism 2 from GraphPad Software) to fit a one-site
competition curve to estimate the IC50-value (concentration of the competitor to in-
hibit specific [3H]-noradrenaline uptake by 50%).

Indirect Determination of Uptake1 Activity in Isolated Trabecular Strips

Human trabecular strips can be obtained from e.g. right atrial appendages that are re-
moved during the installation of a cardiopulmonary bypass. Each muscle prepara-
tion should be used for only one concentration-response curve to exclude β-adreno-
ceptor desensitization. An uptake1 blocker should shift the concentration-response
curve of noradrenaline to the left. As a result, an increase in the potency of noradrena-
In-Vitro Techniques

line is able to reflect a reduced inactivation of noradrenaline via uptake1 and this repre-
sents indirectly the activity of the noradrenaline transporter (Leineweber et al. 2002).
To verify that the effect is clearly dependent on the uptake1 activity the experi-
ments should be repeated in the presence of isoprenaline instead of noradrenaline,
since isoprenaline is not a substrate for uptake1. To block a-adrenoceptors, phentola-
mine (1 µM) should be present throughout the experiments.

Protocol 5: Preparation and Organ-Bath Incubation of Trabecular Strips

1. Immediately after excision, transfer the cardiac tissue into carbogenated (95% O2/
5% CO2) Tyrode-solution (mmol/l: NaCl 119.8; KCl 5.4; CaCl2 1.8; MgSO4 1.05;
NaH2PO4 0.42; NaHCO3 23.6; glucose 5.05; EDTA 0.05; ascorbic acid 0.28; at room
temperature). The time that it needs to reach the laboratory where the experiment
is conducted should not exceed 15–30 min.
2. Prepare the tissue in continuously aerated Tyrode-solution (see above) at room
temperature to minimize inadequate oxygenation. Remove the connective tissue
carefully without endocardial damage.
3. Dissect trabecular strips of uniform size (4–5 mm in length and 1 mm or less dia-
meter) with muscle fibres running approximately parallel to the length of the
strips in aerated Tyrode solution under microscopic control using scissors. (Iso-
lated cardiac tissue should only be used if at least two functioning trabecular
strips can be obtained).
4. Suspend the muscles in an organ bath maintained at 37 °C containing continuous-
ly aerated Tyrode solution (between 10 to 70 ml with respect to organ bath size).
5. Stimulate the muscles by two platinum electrodes using field stimulation from a
stimulator (frequency 1 Hz, impulse duration 5 ms, intensity 10 to 20% [e.g. be-
tween 3–12 V] greater than threshold).
6. Stretch each muscle to a length at which force of contraction is maximal (the force
of rest has to be kept constant throughout the experiment).
 Receptor and Binding Studies 757 6.3
7. Measure the resultant tension isometrically with an inductive force transducer at-
tached to a recorder.
8. To determine the functional evidence of uptake1 on the potency of noradrenaline
to increase force of contraction mediated by b-adrenoceptors, pre-incubate the
preparations with an uptake1 blocker (e.g. 1 µM desipramine) or vehicle for at least
30 min and then estimate cumulative concentration-response curves for norad-
renaline within a concentration range of 10 nM to 10 µM.

Experimental data of the contraction studies can be fitted and analyzed by non-lin-
ear regression analysis (sigmoidal concentration-response curve: y = bottom + [top
– bottom]/[1 + 10(logEC50–x)], with x = logarithm of agonist concentration and y = re-
sponse), using Prism 2 from GraphPad Software.

Troubleshooting
Great Variations in IC50 Values

The affinity of the uptake antagonists desipramine and nisoxetine (both specific up-
take1 blockers) as well as cocaine (monoamine-transporter blocker) and GBR12909
(dopamine uptake blocker) for the noradrenaline transporter is dependent on the cell
type and species that is used. Nevertheless, for uptake1 the order of potency should be
in general desipramine ≥ nisoxetine >> cocaine ≥ GBR12909 independent whether
noradrenaline activity or density is investigated (Leineweber et al. 2000; 2002). The
concentration of desipramine that inhibits 50% of uptake1 activity (IC50 values in nM)
ranges between 1.2 to 2.0 in hNAT- or mmNAT-transfected cells, 7.0 to 13 in rat car-
diac slices or natively expressing rat PC12 cells and 450 nM in human placental brush
border membrane vesicles. The IC50 values (nM) of nisoxetine ranges between 1.7 to
4.2 in bNAT- or mmNAT-transfected cells, 11.8 in rat cardiac slices and 17.2 in na-
tively expressing rat PC12 cells; that of cocaine ranges between 380 in cardiac synap-
tosomes, 2300 in human placental brush border membrane vesicles and 3817 in rat
cardiac slices. Finally, that of GBR12909 ranges for all above 5000. Equally, the concen-
tration of desipramine that inhibits 50% of [3H]-nisoxetine binding (IC50 values
in nM) ranges between 0.73 in rat heart, 2.66 in in rat palcenta, 31 in human right atria
and 450 in human placenta. That of nisoxetine ranges between 3.75 in rat placenta,
10.7 in rat heart, 82 in human right atria and 600 in human placenta. For cocaine, the
IC50 values (nM) are above 2000 and above 3000 for GBR 12909.

Stimulating Agents on Noradrenaline Transporter Activity and Increase In Density

In general, treatment with compounds that increase noradrenaline content (e.g.

monoamine oxidase inhibitors) positively regulate mRNA expression and protein
density of the noradrenaline transporter. Further modulatory/stimulating agents
implicated in regulating transporter-density and -activity in vivo include angiotensin
II, atrial natriuretic peptide, insulin, ATP, Ca2+ and nitric oxide. In general, protein
 6 758 Biochemical and Analytical Techniques

kinase A (PKA)-dependent pathways produce an increase in Vmax and up-regulation

of the noradrenaline transporter. Furthermore, (patho-) physiological conditions, like
dietary restriction or chronic hypoxia, seems to up-regulate the noradrenaline trans-
porter-activity and density.

Inhibiting Agents on NAT Activity and Reduction in NAT Density

In general, treatment with compounds that deplete noradrenaline stores (e.g. reser-
pine) negatively regulate the noradrenaline transporter. Uptake1 is further found to be
blocked by low temperature, sodium azid and ouabain. Activation of protein kinase C
(PKC) by phorbol esters results in phosphorylation of the transporter protein and
decreases Vmax of noradrenaline transport. Long-term treatment in vitro and in vivo
with INF-α, nerve growth factor or desipramine leads to a down-regulation of
transporter density and therefore to a decrease in uptake1. On the other hand, am-
phetamine and metamphetamine, haloperidol, procaine, tetracaine and dibucaine,
long-term treatment with propofol, thiamylal and diazepam and N-ethylmaleimide
are acute inhibitors and inhibiting uptake1 to a great extent (>40%). (Patho-)physi-
In-Vitro Techniques

ological conditions like age, cold (4 °C) exposure; long-term pressure overload, dia-
betic cardiomyopathy, congestive heart failure, ischemia-induced arrhythmia and
orthostatic intolerance are accompanied by noradrenaline transporter malfunction.

Anesthetics

It is now commonly accepted that anesthesia depresses the activity of the sympathetic
system. Enflurane and isoflurane were shown to be associated with decreases in nor-
adrenaline spillover rates leading to results which showed lower than actual norad-
renaline concentrations. Ketamine, on the other hand, has a dual effect on the
noradrenaline transporter (for review see Hara et al. 2000). Whereas short-term treat-
ment causes inhibition, long-term-treatment leads to an increase of uptake1 by up-
regulation of transporter-density. Chloralose induces large decreases (about –38%) in
the noradrenaline clearance rate, whereas diethyl ether induces a stimulation of
sympathetic nerve activity with an increase in noradrenaline spillover rate and
plasma concentration while the uptake1 mechanism is unchanged. On the other hand,
pentobarbital, a typical general anesthetic agent, does not affect the uptake of
monoamines and should therefore be the first choice for anaesthesia of animals.

References
Böhm M, LaRosee K, Schwinger RHG, Erdmann E (1995) Evidence for reduction of norepinephrine uptake sites in the
failing human heart. J. Am. Coll. Cardiol. 25:146–153
Bradford MM (1976) A rapid and sensitive method for the quantification of microgram quantities of protein utiliz-
ing the principle of protein-dye binding. Anal. Biochem. 72:248–254
Brüss M, Pörzgen P, Bryan-Lluka LJ, Bönisch H (1997) The rat norepinephrine transporter: molecular cloning from
PC12 cells and functional expression. Mol. Brain Res. 52:257–262
Esler M, Kaye D, Lambert D, Jennings G (1997) Adrenergic nervous system in heart failure. Am. J. Cardiol. 80:7L–14L
Graefe KH and Bönisch H (1988) The transport of amines across the axonal membranes. In: Trendelenburg U, Weiner
N (eds) Catecholamines I. Springer-Verlag, Berlin, p 193–245
 Receptor and Binding Studies 759 6.3
Hara K, Yanagihara N, Min ami K, Hirano H, Sata T, Shigematsu A, Izumi F (2000) Dual effects of intraveneous anes-
thetics on the function of norepinephrine transporters. Anestesiol. 93:1329–35
Jayanthi LD, Prasad PD, Ramamoorthy V, Mahesh VB, Leibach FH, Ganapathy V (1993) Sodium- and chloride-depen-
dent, cocaine-sensitive, high-affinity binding of nisoxetine to the human placental norepinephrine transporter.
Biochem. 32: 12178–12185
Kiyono Y, Lida Y, Kawashima H, Tamaki N, Nishimura H, Saji H (2001) Regional alterations of myocardial norepineph-
rine transporter density in streptozotocin-induced diabetic-rats: implications for heterogeneous cardiac accu-
mulation of MIBG in diabetes. Eur. J. Nucl. Med. 28:894–899
Lee CM and Snyder SH (1981) Norepinephrine neuronal uptake binding sites in rat brain membranes labeled with
[3H]desipramine. Proc. Natl. Acad. Sci. USA 78:5250–5254
Leineweber K, Seyfarth T, Brodde O-E (2000) Chamber-specific alterations of noradrenaline uptake (uptake1) in right
ventricles of monocrotaline-treated rats. Brit. J. Pharmacol. 131:1438–1444
Leineweber K, Wangemann T, Giessler Ch, Bruck H, Dhein S, Kostelka M, Mohr F-W, Silber R-E, Brodde O-E (2002)
Age-dependent changes of cardiac neuronal reuptake transporter (uptake1) in the human heart. J. Am. Coll.
Cardiol. 40:1459–1465
Runkel F, Brüss M, Nöthen MM, Stöber G, Propping P, Bönisch h (2000) Pharmacological properties of naturally
occuring variants of the human norepinephrine transporter. Pharmacogenetics 10:397–405
Tejani-Butt SM (1992) [3H]-Nisoxetine: A Radioligand for quantification of norepinephrine uptake sites by autorad-
iography or by homogenate binding. J. Pharmacol. Exp. Therap. 260:427–436

6.3.3
Study of G proteins in Cardiovascular Research
Thomas Wieland, Nina Wettschureck and Stefan Offermanns

Introduction
Many neurotransmitters, hormones and growth factors which play important roles in
the regulation of the cardiovascular system, act through seven-transmembrane-span-
ning G protein coupled receptors (GPCRs). Of these, adrenergic and muscarinic cho-
linergic receptors mediate the control of cardiovascular functions by the autonomic
nervous system. Other receptors, such as those for endothelins, angiotensin II or
eicosanoids are involved in the local or systemic regulation of cardiovascular func-
tions. The cellular effects induced through these receptors depend on the type of G
protein activated by the receptor (see Table 3).
There are basically four families of heterotrimeric G proteins: Gs, which couples
receptors in a stimulatory fashion to adenylyl cyclase; Gi/o, which couples receptors in
an inhibitory fashion to adenylyl cyclase and which mediates the activation of potas-
sium channels as well as the inhibition of voltage-dependent Ca2+ channels; Gq/11,
which couples receptors to the stimulation of β-isoforms of phospholipase C; and
finally the G12/13 family, which mediates the activation of the small GTPase Rho. These
families of heterotrimeric G proteins play central roles in transmembrane signalling
by processing and sorting of signals conveyed by GPCRs. They are also able to adjust
the sensitivities of the system to certain stimuli, a function involving various mecha-
nisms including the action of “regulators of G protein signalling” (RGS) proteins,
which are able to “switch off ” the active state of the G protein.
Multiple excellent reviews and book chapters have been written on the analysis
of receptor G protein coupling and on the determination of G protein activity [e.g.
Methods in Enzymology Vols. 237 (1994) and 344 (2002)] which can be applied to
 6 760 Biochemical and Analytical Techniques

Table 3
G protein coupled receptors (GPCRs) in the cardiovascular system.

Endogenous ligand(s) Receptor Coupling to

G protein subclass(es)
Acetylcholine M2 Gi/o
Adrenaline, Noradrenaline α1A, α1B, a1D, Gq/11, (Gi/o?)
α2A, α2B, α2C Gi/o
β1 Gs
β2, β3 Gs, Gi/o
Prostacyclin (PGI2) IP Gs
Thromboxane A2 (TXA2) TP Gq/11, G12/13
Angiotensin II AT1 Gq/11, G12/13, Gi/o
AT2 (Gi/o?)
Bradykinin B1, B2 Gq/11
Endothelin- 1, -2, -3 ETA (ET-1, ET-2) Gq/11, G12/13
In-Vitro Techniques

ETB (ET-1, -2, -3) Gq/11, Gi

Urotensin II UT Gq/11
Vasopressin V1a,V1b Gq/11
V2 Gs
Thrombin and others PAR-1, PAR-3, PAR-4 Gq/11, G12/13, Gi/o
Trypsin and others PAR-2 Gq/11

study the cellular and sub-cellular role of heterotrimeric G proteins in the cardiovas-
cular system. To study G protein mediated signalling in the living organism, molecular
biological and genetic methods are required. This chapter therefore will focus on re-
cent progress in applying adenoviral vectors and genetic approaches to the study of
heterotrimeric G protein functions in the cardiovascular system.

Adenovirus Mediated Overexpression of Negative

Regulators of Heterotrimeric G Proteins to Analyze
Signalling Pathways in Cardiac Myocytes
A variety of G protein coupled receptors (GPCRs) in cardiac myocytes, for example
endothelin-1 receptors, angiotensin II receptors and protease activated receptors, si-
multaneously activate more than one subtype of heterotrimeric G proteins. They
thereby regulate several effector molecules and induce a variety of cellular responses.
Therefore, the contribution of a distinct G protein subtype or its subunits to the regu-
 Receptor and Binding Studies 761 6.3
Table 4
Adenoviral constructs suitable for analyzing G protein mediated pathways in cardiac myocytesa

Recombinant Expressed Function Obser ved effect after

adenovirus protein overexpression
Ad5RGS2 RGS2 Gαq/11- >> Inhibition of Gαq/11-
Gαi/o-GAP mediated effects
Ad5RGS4 RGS4 Gαq/11- ≈ Inhibition of Gαi/o plus
Gαi/o-GAP Gαq/11-mediated effects
Ad5Lsc Lsc-RGS Gα12/13-GAP Inhibition of Gα12/13-
domain mediated effects
Ad5KIAA380 PDZRhoGEF- Gα12/13-GAP Inhibition of Gα12/13-
RGS domain mediated effects
Ad5TDα Transducin Gβγ scavenger Inhibition of Gβγ-
α-subunit mediated effects
a
An adenoviral construct expressing the RGS-homology domain of RGS-PX1 should theoretically be suitable to
suppress Gas-mediated signals (Zheng et al. 2001). This concept has however not been approved in our laboratories

lation of a specific signalling pathway is often not readily evident and additionally
may vary between different receptors and their subtypes. The (over)-expression of
specific negative regulators of heterotrimeric G protein function, for example GTPase-
activating proteins (GAPs) for Gα subunits or scavenging proteins for activated Gα
subunits or free Gβγ dimers, is often used to dissect the involvement of G proteins in
distinct signalling pathways in living cells. Besides the fact that isolated cardiac
myocytes are difficult to culture, the transfection of cultured cardiomyocytes by stan-
dard methods such as liposomes, electroporation or calcium phosphate precipitation
is known to have an extremely low efficiency. Thus, the use of these methods for the
above mentioned expression of negative regulators of G protein function in cardiac
myocytes is very limited.
An approach which is however widely used to express recombinant proteins in
cardiac myocytes is the infection of the cells with recombinant adenoviruses. More-
over, this method can additionally be applied to express recombinant proteins in en-
gineered heart tissue (El-Armouche et al. 2003), heart muscle strips (Janssen et al.
2002), perfused Langendorff hearts (Wiechert et al. 2003) in vitro and even in the liv-
ing heart of animal models (Neyroud et al. 2002). We will describe herein the use
of recombinant adenoviruses encoding the RGS homology domains of regulators of
G protein signalling proteins or the α subunit of the retinal G protein transducin
(Table 4) to analyze G protein coupling of several GPCRs in neonatal rat and adult
rabbit cardiac myocytes. Besides the over-expression of the aforementioned nega-
tive regulators to analyze the role of endogenous G proteins, adenoviral mediated
over-expression of Gai2, Gaq and Gas has been used to study the role of these G pro-
tein a subunits in cardiomyocytes, heart muscle and the hearts of living animals
(Table 5).
 6 762 Biochemical and Analytical Techniques

Table 5
Adenoviral over-expression of wild type or mutant G protein α subunits in the cardiovascular system

Transgene Tissue Phenotype Reference

Gαs Neonatal rat Enhanced activation Novotny et al. (1994),
cardiomyo- of adenylyl cyclase, Biochem Mol Biol Int
cy tes induction of hyper- 34:993-1001
trophy
Gαi2 Rabbit Impaired inotropic Janssen et al. (2002),
trabeculae response to cate- Cardiovasc Res 55:300-308
cholamines
Gαi2 Catheterized Reduction of heart Donahue et al. (2000),
porcine rate during atrial Nature Medicine 6:1395-
hearts fibrillation 1397
Gαi2 Adult rat Induction of the Gong et al. (2002),
cardiomyo negative inotropic Circulation 105:2497-2503
cy tes effect of β2-
In-Vitro Techniques

adrenoceptors
Gαi2 Neonatal Suppression of iso- Rau et al. (2003),
and adult prenaline-stimulated FASEB J. 17:523-525
rat cardio- adenylyl cyclase and
myocy tes inotropic response
Gαq(w t) Neonatal rat Enhanced PGF2α- Adams et al. (1998)
cardiomyo- and phenylephrine- Proc Natl Acad Sci USA
cy tes induced PLC activa- 95:10140-10145,
tion and apoptosis Adams et al. (2000)
Circulation Res 87:1180-1187
Gαq Neonatal rat Activation of PLC, Adams et al. (1998)
(Q209L) cardiomyoc- induction of apop- Proc Natl Acad Sci USA
(const. y tes tosis 95:10140-10145,
active) Adams et al. (2000)
Circulation Res 87:1180-1187
const. active constitutively active mutant; wt wild type form.

Description of Methods and Practical Approach

Isolation and Culture of Neonatal Rat Cardiac Myocytes

Buffers and Media.

▬ CBFHH buffer, pH 7.4 (mM)
NaCl 136, Hepes 20, glucose 5.55, KCl 5.36, MgSO4 0.81,
KH2PO4 0.44, NaH2PO4
 Receptor and Binding Studies 763 6.3
▬ Dnase II solution
Dnase II (Sigma) 0.02 mg/ml, foetal calf serum (Invitrogen) 0.65% (v/v), penicil-
lin 100 U/ml, streptomycin 0.1 mg/ml, in CBFHH buffer
▬ Trypsin solution
Trypsin (Sigma) 0.3% (w/v), Dnase II (Sigma) 0.02 mg/ml, penicillin 100 U/ml,
streptomycin 0.1 mg/ml, in CBFHH buffer
▬ MEM
Foetal calf serum (FCS), inactivated, 10% (v/v), penicillin 100 U/ml, streptomycin
0.1 mg/ml, 5-bromo-2-desoxyuridine (BrdU, Sigma) 1 mM, in Minimum Essential
Medium (Invitrogen).

Procedure. Newborn (60–80 animals) Wistar rats (1–3 d old) are decapitated and the
heart is extracted from the thorax. The atria and tissues from surrounding organs are
excised. The hearts are minced with scissors in CBFHH buffer and poured into a ster-
ile 50 ml Falcon tube. After sedimentation of the tissue, the supernatant is discarded
and the pellet is incubated in 10 ml freshly prepared trypsin solution for 20 min at
4 °C. Thereafter, the supernatant is discarded and the pellet is resuspended in 10 ml
fresh trypsin solution. After an incubation of 10 min at room temperature with gentle
shaking, the supernatant is removed and added to 2.5 ml FCS (not inactivated) and
stored on ice. The pellet is resuspended in 9 ml of DNase II solution and the suspen-
sion is triturated by repeated suction through a pipette with a 3 mm wide opening
27×.
After sedimentation of the pellet, the supernatant is removed and added to the al-
ready stored supernatant. This procedure has to be repeated as long as the digest pro-
duces a detectable clouding. Thereafter, the cells in the stored supernatants are
pelleted by centrifugation for 15 min at 60×g in a swingout rotor. The supernatant is
discarded, the cells are resuspended in 20 ml of MEM without BrdU and 0.4 mg DNase
II is added. After trituration with the pipette, the cells are again pelleted by centrifu-
gation for 15 min at 60×g. Thereafter, the cells are resuspended in 32 ml of MEM with-
out BrdU and plated through a 60 mm mesh filter on to four culture dishes (∅ 10 cm).
The dishes are kept for about 2 h in an incubator at 37 °C. This time is sufficient to let
non-myocytes adhere to the dishes whereas the cardiac myocytes are still floating.
Thus, the myocytes are harvested with the supernatant and the cells are plated on new
culture dishes (∅ 6 cm) in MEM, now complemented with the antimetabolite BrdU, at
a density of 1–3×106 cells per dish. The cardiac myocytes adhere to the dishes and
form a monolayer after 1 day of culture. Regular autonomous contractions are ob-
served usually after 2–3 days of culture.

Isolation of Adult Rabbit Cardiac Myocytes

Buffers and Media.

▬ Tyrode solution, pH 7.55 (mM)
NaCl 137, Hepes 20, glucose 15, KCl 5.4, MgSO4 1.2, NaH2PO4 1.2, saturated with O2
▬ Enzyme Solution
Collagenase CLS II (Biochrom) 1 mg/ml, protease fraction XIV (Sigma), 0.04 mg/
ml, CaCl2 0.025 mM in Tyrode solution
 6 764 Biochemical and Analytical Techniques

▬ M199 (mM)
Creatine 5.0, D,L-carnitine 5.0, taurine 5.0, L-glutamine 2, penicillin 100 U/ml,
streptomycin 0.1 mg/ml, in medium 199 (Invitrogen)

Adult New Zealand White rabbits receive heparin anticoagulation (1,000 units i.v.)
prior to pentobarbital (50 mg/kg i.v.). The heart is extracted and rinsed twice in ice-
cold Tyrode solution. Next, the aorta is cannulated, and the heart is suspended in an
insulated chamber at 35–37 °C. Langendorff perfusion is performed by retrograde
flow from the cannula in the ascending aorta to the coronary arteries. The heart is
perfused with Tyrode containing 1 mM Ca2+ at 20 ml/min until the heart beats regu-
larly at about 130 beats per minute. Thereafter, the perfusion is switched to calcium-
free Tyrode for 10 min at 20 ml/min, before perfusion with the enzyme solution for an
additional 10–15 min. The digestion is stopped by perfusion with 100 ml Tyrode con-
taining 0.05 mM Ca2+ and 20 mM 2,3-butanedione-monoxime (BDM). The presence
of BDM is crucial as it binds reversibly to cardiac myofilaments and thereby prevents
hypercontractions that might lead to sarcolemmal injury and cell death. The ven-
tricles are excised, minced in Tyrode containing 0.05 mM Ca2+ and 20 mM BDM, and
agitated by repeated suction through a Pasteur pipette. The resulting cell suspension
In-Vitro Techniques

is strained through a 200 mm nylon mesh filter. In general, the ventricles are uni-
formly digested, and the residue remaining on the filter is from the valvular and ten-
don structures, with only minimal myocardium. The Ca2+ concentration is gradually
restored in several steps (0.05 mM, 0.078 mM, 0.156 mM, 0.3125 mM, 0.625 mM) be-
fore the cells are finally resuspended in M199. Viable myocytes are enriched by sedi-
mentation through 6% (w/v) albumin fraction V (Sigma) in M199. The cells are plated
at a density of 50,000 cells per dish in laminin-coated dishes (∅ 3.5 cm, 20 µg/ml
laminin (Sigma) in M199 for 1 h at 37 °C) and cultured in a 37 °C incubator for a
maximum of 48 h.

Construction and Amplification of Recombinant Adenoviruses

Recombinant adenoviruses are constructed using the AdEasy system developed by the
group of Bert Vogelstein (He et al. 1998). As detailed information and protocols for
that system are available in the World Wide Web at www.coloncancer.org/adeasy.htm,
only a brief description of the system will be given here. The cDNA encoding the gene
of interest is first cloned into a shuttle vector. The most commonly used pAdTrack-
CMV vector additionally encodes enhanced green fluorescent protein (EGFP) under
the control of an independent CMV promotor. E. Coli BJ 5183 cells are co-transformed
with purified recombinant pAdTrack-CMV linearized by a PmeI digest and the aden-
oviral backbone vector pAdEasy-1. By homologous recombination, sequences of the
pAdTrack-CMV are inserted into the pAdEasy-1. As by the homologous recombi-
nation the gene encoding resistance against ampicillin is also switched to the gene
encoding resistance to kanamycin, positive recombinants can be selected on agar
plates containing kanamycin. Note however that the homologous recombination has
to be verified by a digest with restriction endonucleases. Thereafter, adenovirally
transformed human embryonic kidney (HEK 293) cells are transfected with the posi-
 Receptor and Binding Studies 765 6.3
tive plasmid linearized by digestion with PacI. If pAdTrack-CMV was used as shuttle
vector originally, the transfected cells can be visualized 12–24 h post transfection by
fluorescence microscopy. Usually the recombinant viruses are harvested 5–6 days
post transfection. After several rounds of amplification to increase the numbers of in-
fected cells, the viruses are purified from the culture medium either by centrifugation
through a CsCl2 gradient or using commercially available purification kits. Viral
stocks can be stored in aliquots in a glycerol enriched storage buffer for several
months. To determine the titres of the viral stocks an endpoint dilution assay (see
below) can be recommended for EGFP encoding adenoviruses.

Endpoint Dilution Assay to Determine the Viral Titre of EGFP-encoding Adenoviruses

HEK293 cells are harvested from a culture dish (∅ 10 cm) and resuspended in 5 ml
DMEM (Dulbecco’s minimal essential medium (Invitrogen) supplemented with 2%
(v/v) inactivated FCS, 2 mM glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomy-
cin at a density of 20,000 cells/90 µl. 90 µl of this suspension is added into each well of
a 96 well plate. For controls, the first 6 wells of the first row of the 96 well plate re-
main uninfected (negative control) whereas the remaining 6 wells of this row are
infected with 1 µl of original virus stock (positive control). The vast majority of the
cells in these six wells should become infected and can be seen by fluorescence mi-
croscopy.

Virus Dilutions

From the original virus stock a 100-fold dilution (1 µl stock + +99 µl DMEM) is
prepared. This dilution is further diluted with DMEM to 10–4, 10–5, 10–6, 10–7, 10–8 and
10–9. From these dilutions 10 µl are added to the 90 µl cell suspension (additional 1:10
dilution, see below) and mixed with the pipette. 12 wells = 1 row are used per dilution.
The cells are incubated for 4 days in the 37 °C incubator. Thereafter the number of
wells that contain EGFP expressing cells is determined by fluorescence micros-
copy. One single green cell per well scores as a positive count (infected). A well with-
out EGFP expression is negative (uninfected). The results are noted in a result table
(see under Examples).

Examples

See Table 6

Calculation of Proportional Distance PD

Use dilutions surrounding the 50% infection value: A effectivity (%) lower dilution, B
effectivity (%) higher dilution
▬ PD = A-50/(A-B)
▬ Example from above (57.8-50)/(57.8-27.3) = 0.256
 6 766 Biochemical and Analytical Techniques

Table 6
Results

Final Infected Uninfected Total Total %

dilution infected* uninfected* Infected*
10-5 12 0 34 0 100.0
-6
10 11 1 22 1 95.7
10-7 5 7 11 8 57.8
-8
10 4 8 6 16 27.3
-9
10 2 10 2 26 7.1
10-10 0 12 0 38 0.0
Sum 34 38
*This calculation assumes that all wells which have been infected by the respective virus dilution would also have
been infected by the higher dilutions which caused an infection. On the other hand wells, not infected at a specific
dilution would also not have been infected by any lower dilution. In the example given, at a 10–5 dilution all 12 wells
are infected. Thus, a 10–5 dilution would also infected the 11 positive wells at 10–6, 5 at 10–7, 4 at 10–8 and 2 at 10–9, to-
In-Vitro Techniques

gether 34 wells. On the other hand, 10 wells are not infected at a 10–9. Thus a 10–9 dilution would also not infect the
uninfected wells at 10–8, 10–7 and 10–6, in sum 26 wells. +At each dilution the percentage of total number infected is
calculated, e.g. 10–7. 11/(11+8) = 57.8% infected.

Calculation of TCID50 (Dilution causing 50% infection)

▬ Log TCID50 = Log of lower dilution -PD

▬ Example: Log TCID50 = Log 10–7–0.256 = –7–0.256 = –7.256 ⇒ TCID50 = 5.546 × 10–8

Calculation of Viral Titre

Titre (pfu/µl) = 1/TCID50 per dilution volume used × ln 2

Example: 1/(5.546 × 10–8)/10(µl) × 0.69 = 1.244 × 106 pfu/µl = 1.244 ×109 pfu/ml

Infection of Cardiac Myocytes

Neonatal Rat Cardiac Myocytes. After 4 days of culture, dishes with morphologically
well-shaped and contracting myocytes are selected for viral infection. The medium is
removed and 1 ml fresh MEM (with 2% FCS) containing the desired amount of virus
is added. For example to achieve a multiplicity of infection (MOI) of 8, 8×106 pfu of
the virus are given to 1×106 myocytes. The dishes are placed in an incubator for
20 min at 37 °C. Thereafter 4 ml of fresh medium are added and the cells are cultured,
if not otherwise required, for an additional 48 h before assay. Usually 12–24 h after
infection, the efficiency of the transduction can be monitored by the expression of
EGFP (see Fig. 12a).

Adult Rabbit Cardiac Myocytes. Directly after preparation of the myocytes, the re-
quired amount of adenoviral stock (e.g. 50,000 pfu for a MOI of 1) is added to 50,000
cells in the M199 medium. The cells are plated and incubated for 2 h in the incubator.
 Receptor and Binding Studies 767 6.3

Figure 12a,b
Adenoviral expression of EGFP in neonatal rat and adult rabbit cardiac myocytes. Neonatal rat
cardiomyocytes (a) and adult rabbit (b) cardiac myocytes were infected with an adenovirus encoding
EGFP and RGS4 at a MOI of 10. The expression of EGFP was visualized by fluorescence microscopy 24 h
(b) and 48 h (a) post infection

Thereafter, the virus-containing medium is removed and replaced by fresh M199. The
efficiency of the transduction can be monitored 12–24 h after infection by fluores-
cence microscopy (see Fig. 12b).

Differential Activation of Neonatal Cardiac Myocyte Phospholipase D by Protease

Activated Receptors, Endothelin-1 Receptors and a1-Adrenoceptors. Activation of
the phosphatidylcholine-specific phospholipase D (PLD) in isolated cardiomyocytes
has been demonstrated to occur for example by stimulation of endo-thelin-1 recep-
tor subtype A (ETAR), thrombin activated protease receptor type 1 (PAR1) and α1-
adrenoceptors (α1AR). PLD and its reaction product, phosphatidic acid, are assumed
to participate in cardiac hypertrophy. Therefore we analyzed the coupling of these
receptors to PLD in neonatal rat cardiac myocytes. As the stimu-lation of PLD by these
GPCRs is pertussis toxin insensitive in cardiac myocytes, we discriminated between
the contributions of the pertussis toxin insensitive Gq/11- and G12/13-types of
heterotrimeric G proteins by adenovirus-mediated overexpression of the Gq/11-inhib-
iting RGS4 and the G12/13-inhibiting LscRGS, respectively. The cultured myocytes
(3×106 cells per culture dish, ∅ 6 cm) were infected either with a control adenovirus,
e.g. an adenovirus expressing only EGFP, or Ad5-RGS4 or Ad5-LscRGS 48 h prior to
the assay. As judged by co-expression of EGFP, the efficiency of adenoviral gene trans-
fer (at a MOI of 10) was ≥75% (see Fig. 12a).
For labelling of the myocytes 2.5 µCi/ml [3H]oleic acid was added to the MEM
medium 24 h before the assay. Thereafter, the cells were washed once with Hank’s
balanced salt solution (HBSS, 118 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM MgCl2,
5 mM D-glucose, 15 mM HEPES, pH 7.4) and then allowed to equilibrate for 30 min at
37 °C. The equilibration medium was replaced by fresh HBSS containing 2% ethanol
and the indicated stimuli, and incubation was continued for 30 min at 37 °C. Thereaf-
 6 768 Biochemical and Analytical Techniques
In-Vitro Techniques

Figure 13a–c
Involvement of distinct G proteins in neonatal rat cardiac myocyte PLD stimulation by endothelin-1,
thrombin and noradrenaline. Neonatal rat cardiomyocytes were infected with adenoviruses encoding
bacterial galactosidase (Ad5Control, open bars) (a,b), Lsc-RGS (Ad5Lsc, grey bars) (a–c) or RGS4
(Ad5RGS4, closed bars) (a) 48 h prior to the assay if not otherwise indicated at a MOI of 10. The forma-
tion of the PLD-specific reaction product [3H]phosphatidylethanol was determined in the absence
(Basal) and presence of endothelin-1 (ET-1, 1 µM), thrombin (Thr, if not otherwise indicated 1 U/ml)
or noradrenaline (NA, 10 mM in the presence of 1 µM nadolol). Shown are the means ± S.E.M. of n =
6–9 (*p<0.05 vs. Ad5Control). The over-expression of Lsc-RGS is shown (c, inset)

ter, the incubation medium was replaced by 500 µl ice-cold methanol, cells were
scraped off and transferred into a 3 ml tube. The culture dish was washed with an
additional 500 µl methanol which were added to the 3 ml tube. Chloroform (1 ml) plus
water (0.5 ml) were added and the mixture was vortexed for 30 sec. For phase ex-
traction, the tubes were centrifuged for 10 min at 1180×g. The lower phase was placed
in a 1.5 ml Eppendorf tube and the solvent was removed in a vacuum concentrator.
The pellet was dissolved in 25 µl chloroform:methanol (1:1). Ten µl were spotted on
LK6D silica gel 60 A thin layer chromatography plates (Whatman) and the phospho-
lipids were resolved with a mobile phase consisting of ethyl acetate:isooctane:ace-
 Receptor and Binding Studies 769 6.3
tic acid:water (13:2:3:10). A mixture of phosphatidic acid and phosphatidylethanol
(50 µg each) was used as standard. The spots containing phosphatidylethanol were
scraped off and 10 ml of a scintillation cocktail were added. The formation of the PLD
specific product [3H]phosphatidylethanol was quantified in a liquid scintillation
counter.
As shown in Fig. 13 endothelin-1, thrombin and noradrenaline (in the presence of
the β-adrenoceptor antagonist nadolol) stimulated [3H]phosphatidylethanol forma-
tion in the myocytes to a similar extent The over-expression of RGS4 did not alter
PLD stimulation by thrombin but blunted the ETAR- and α1AR-induced PLD activity
by about 50–60% (Fig. 13a). In contrast, thrombin-induced PLD stimulation was sig-
nificantly blunted, by about 50% in cardiomyocytes over-expressing LscRGS, whereas
the stimulation by endothelin-1 and noradrenaline was not affected. The over-expres-
sion of LscRGS reduced the maximal extent of thrombin-induced PLD stimulation,
without altering the concentration dependency (Fig. 13b). Infection of cardio-
myocytes with an increasing MOI of Ad5Lsc resulted in an increased expression of the
recombinant protein and a concomitant reduction in thrombin-induced PLD stimu-
lation. At a MOI of 50 the PAR1-induced stimulation of the PLD activity was com-
pletely suppressed (Fig. 13c). The data therefore demonstrate that the PAR-1 induced
stimulation of cardiac myocytes’ PLD is mediated by the G12/13-type of heterotrimeric
G proteins. In contrast, the ETAR- and α1AR-induced PLD stimulation is mediated by
the Gq/11 – phospholipase C – protein kinase C pathway (Fahimi-Vahid et al. 2002;
Gosau et al. 2002).

Figure 14a,b
Agonist-induced fractional shortening of adult rabbit ventricular cardiac myocytes. Adult rabbit cardiac
myocytes were infected with Ad5GFP or Ad5RGS4 at a MOI of 1 24 h before the assay. Fractional short-
ening was measured at increasing concentrations of ET-1 (10–10–10–7 M) (a). At the end of the experi-
ment isoprenaline (0.1 µM) was added (b). Basal fractional shortening did not differ: Ad-GFP: 4.1 ±
0.7%, Ad-RGS4: 3.6 ± 0.4%. The over-expression of RGS4 was monitored by immunoblotting
 6 770 Biochemical and Analytical Techniques

Over-expression of RGS4 Inhibits the ETAR-induced but not b -Adrenoceptor-induced Contractility

in Adult Rabbit Cardiac Myocytes

Adult rabbit cardiomyocytes were infected with Ad5-RGS4 or Ad5-EGFP as a control

at a MOI of 1 and cultured for 48 h. About 30% of cells showed fluorescence by EGFP
(see Fig. 12) and only EGFP positive cells were used. In Ad5-RGS4 infected cells the
average over-expression was estimated to reach about 3–6-fold. Myocyte shortening
was measured by an edge-detection system (Crescent Electronics) at a stimulation
frequency of 0.5 Hz and a sampling rate of 240 Hz. Online and offline analysis was
performed with custom-designed Labview software (National Instruments). Cumu-
lative concentration response curves were made with increasing concentrations of
endothelin-1 (10–10–10–7 M) in both groups. The agonist increased the fractional short-
ening concentration dependently by about 60–70% in Ad5-EGFP infected cells. No
effect was seen in Ad5-RGS4 infected cells (Fig. 14a). In contrast, the β-adrenoceptor
agonist isoprenaline (10–7 M) given on top of the highest endothelin-1 concentration
increased fractional shortening in both groups without significant difference
(Fig. 14b). These data demonstrate that the overexpression of RGS4 completely sup-
presses the ETAR-induced and Gq/11-mediated increase in myocytes contractility (Mitt-
In-Vitro Techniques

mann et al. 2002). The Gs-mediated contractile response induced by β-adrenoceptors

remains unaffected.

Troubleshooting

Preparation of Neonatal Rat Cardiac Myocytes

1. Sterility of cultures: All materials have to be sterilized and all work has to be per-
formed under aseptic conditions. The extraction of the hearts is most critical. If
the cultures become contaminated with fungi, the most likely reason is a contami-
nation of the CBHFF buffer during the chopping up of the hearts.
2. Over-digestion with trypsin: One likely reason for a low number of viable cells
from the preparation is an over-digestion with trypsin, which might for example
occur after changing the supplier or with a new batch of the enzyme. It is very im-
portant to test every new batch of the enzyme before general use. Another possi-
bility for an over-digestion is the use of heated (inactivated) instead of original
FCS for trypsin inactivation during the preparation. Additionally, it is a good idea
to gradually reduce the volume and exposure time during the repeated digests in
the preparation.
3. Mechanical disruption of the tissue: It is very important to perform the tritura-
tions with a pipette with a 3 mm wide opening. Regular pipettes have a narrower
opening. Thus, the shear forces will probably break the myocytes. It is similarly
important to use 40–80 animals for one preparation. With lower animal numbers
the number of viable cells obtained from one heart decreases. With higher num-
bers the efficiency of the procedure decreases. Thus, more animals do not neces-
sarily provide more cells.
 Receptor and Binding Studies 771 6.3
4. Non-myocyte contamination: The preparation usually contains less than 5% non-
myocytes. The separation relies on the faster adherence of the non-myocytes to
the plastic material of the culture dishes. Usually the non-myocytes adhere within
2 h of culture, whereas the myocytes are still floating. Nevertheless, this is a dy-
namic process and the adherence should be frequently checked during the 60 and
120 min of culture. On the one hand, floating non-myocytes will increase the
number of these cells in the final myocyte culture. On the other hand, if the
myocytes start to adhere early, they are lost to the planned experiments. It is essen-
tial that the myocytes be cultured in the presence of the antimetabolite BrdU. Oth-
erwise the dividing non-myocytes will overgrow the myocytes in the culture and
this might cause false results.
5. Plating of the myocytes: The counting of the myocytes is a critical step. If a Neu-
bauer chamber is used, note that the glass plate may crush some myocytes. Thus,
too many cells must not be put into the chamber. Otherwise, the number of viable
myocytes is underestimated due to mechanical destruction. Cardiac myocytes do
not adhere to plastic dishes of all brands. Thus, it is not a good idea to buy dishes
from a new supplier without having them tested.

Preparation of Adult Rabbit Myocytes

1. Sterility of cultures: As the rabbit hearts are excised in the open, every material
and solution used during the preparation has to be sterile. Otherwise bacterial or
fungal contaminations occur frequently.
2. Low number of viable cells: The most sensitive part of this protocol is the enzyme
solution. It has to be prepared freshly before use and the pH has to be exactly ad-
justed. Collagenase, even if bought from the same supplier, differs largely in its
activity and every new batch has to be tested. In most cases of a low efficiency of
the preparation the collagenase is not working adequately. Another reason for low
efficiency is too long a period between extraction of the heart and the start of the
perfusion. The perfusion should start within 90–120 s after extraction. If too many
cells in hypercontracture appear during the preparation, the Ca2+ concentrations
have been increased too rapidly. The cells should be allowed to equilibrate at each
concentration for at least 5 min.
Additionally, it is a good idea to perform more steps in increasing the Ca2+ concen-
tration, e.g. 0.05 mM, 0.0625 mM, 0.078 mM, 0.117 mM, 0.156 mM, 0.235 mM,
0.3125 mM, 0.469 mM, 0.6125 mM.

Amplification and Storage of Recombinant Adenoviruses

1. Unintended viral contamination: During changing of infected cells’ medium dur-

ing the amplification procedure, the virus may contaminate the pipette used and
thereby also the stock medium used for feeding. If the same stock medium is used
to feed other HEK 293 cells, the virus may contaminate all HEK 293 cells currently
in culture. Thus, it is a good idea to prepare an aliquot of the stock medium for
feeding infected cells and discard any remnant. If possible, virally infected cells
 6 772 Biochemical and Analytical Techniques

should be kept in a separate incubator. We also highly recommend the use of an

inverse fluorescence microscope to look at virally infected cells. If a regular micro-
scope is used to look at the cells, the covers of the culture dishes have to be re-
moved to focus on the monolayer. The objective of the microscope may get
contaminated and spread the virus from dish to dish.
2. Loss of viral titre during storage: Even if the viruses are stored in the recom-
mended high glycerol containing storage buffer at –20 °C, the amount of bio-
active virus decreases with time. Therefore repeating the endpoint-dilution assay
from time to time to check the virus titre is recommended. Note that freezing
and thawing as well as storage at higher temperatures will further reduce the
amount of bioactive virus. Therefore the virus stock should be stored in small
aliquots adjusted to the volume used in one experiment. If the virus titre falls to
less than 50% of its original value, a re-amplification of the virus should be per-
formed.

Infection of Cells and Assay

1. Low number of infected cells: The most common reason is loss of the bioactivity
In-Vitro Techniques

of the virus (see above). Another cause is inhibition of viral infection by high
amounts of FCS in the medium used for the infection. A concentration of 2% FCS
can be recommended. An alternative procedure that is also tolerated by the cells
is addition of the virus in medium without FCS followed by filling up with
fresh FCS containing medium after 20 min in the incubator. Viral infection is
also dependent on the density of the cells. If more than 3´106 neonatal myocytes
are plated on a culture dish (∅ 6 cm) the percentage of infected cells will de
crease.
2. Consequences of EGFP expression: The expression of EGFP in the infected cells is
not independent of the second transgene in the adenovirus. Thus, a lower amount
of EGFP expression (for example compared to Ad5-EGFP) does not necessarily
indicate a low expression of the protein of interest. The (over)-expression of each
protein of interest should therefore be monitored by immunoblotting. Analysis of
the time dependence of the transgene expression at different MOIs is recom-
mended to obtain optimal conditions.
3. It has been found that the expression of high amounts of EGFP can impair func-
tions in cardiac myocytes. Therefore, in each experimental setting it must be
checked that Ad5-EGFP infected cardiomyocytes do not differ largely in their re-
sponses from uninfected cells. In our hands, expression of EGFP under the condi-
tions shown herein did not influence neonatal rat cardiac myocytes’ PLD nor
single cell fractional shortening in adult rabbit myocytes.
4. “Side effects” of adenoviral gene transfer: To account for side effects of adenovi-
ral infection, the myocytes for the control experiments are usually infected with a
recombinant adenovirus encoding only EGFP. Alternatively, a virus leading to the
expression of bacterial galactosidase is often used. Nevertheless, the protein of
interest itself may cause effects, e.g. induction of apoptosis, which are not readily
evident. Such events might however feign an inhibitory effect on the parameter
under study. Therefore, appropriate positive controls which are definitely not af-
fected by the (over)-expressed protein are required.
 Receptor and Binding Studies 773 6.3

Genetic Analysis of G Protein Function in the

Cardiovascular System of the Mouse
Genetic manipulation of the mammalian genome using transgenic or gene targeting
techniques provides a powerful tool to characterize the molecular mechanisms un-
derlying cardiovascular function under the conditions of a living organism. The
mouse organism has been the principle system where these techniques have been
established and refined. Recent advances in the miniaturization of techniques used to
study cardiovascular functions in whole animals are now allowing detailed analyses
of the effects produced by defined genetic manipulations in the mouse organism
(Christensen et al. 1997; Fitzgerald et al. 2003). Several G protein α subunits have been
over-expressed in a cardiomyocyte-specific manner, both as wild type or constitu-
tively active forms (Table 7). Also various classical and conditional knockouts of genes
encoding murine G protein α subunits have revealed phenotypes in the cardiovascu-
lar system (Table 8).
General methods for the generation of transgenic or gene targeted mice have been
published in a variety of excellent reviews and books. In the following we would like
to give an overview of the different approaches, which have been used to alter gene
expression in the cardiovascular system of the mouse and we will give some detailed
descriptions of the problems, which may arise when G protein mediated signalling
pathways have to be analyzed in mice on a cellular level.
In order to express a transgene constitutively or conditionally in the cardiovascu-
lar system various promoters have been used. For cardiomyocyte-specific expression,
the α-myosin heavy chain (aMHC) promoter has been used most widely since it re-
sults in high level and homogenous expression, both in adult atria and ventricles in a
highly cardiac specific manner (Subramaniam et al. 1991). However, other cardio-
myocyte-specific promoters, which allow transgene expression before birth have also
been successfully used (MLC2a, MLC2v etc.) Several promoters can be used to drive
transgene expression in vascular smooth muscle cells. Some promoter fragments of
the SM22α-gene have been reported to direct transgene expression to specific parts
of the vascular system (Li et al. 1996), and the SMMHC-promoter allows expression in
all smooth muscle cells starting from embryonic day 15 (Regan et al. 2000). Specific
expression of transgenes in endothelial cells has been achieved using promoter frag-
ments of the tie1 and tie2 genes (Korhonen et al. 1995; Schlaeger et al. 1999).
When using plasmid-based transgenes containing short promoter fragments, re-
searchers often observe problems such as integration site dependent effects on the
level and specificity of transgene expression. Random integration of foreign DNA can
result in position effects that influence the expression of the transgene itself as well as
of adjacent endogenous genes. A major improvement in transgenic expression was
achieved when large artificial chromosomes containing all the regulatory sequences
of a given gene were used for the generation of transgenic vectors. These constructs,
which are based on yeast artificial chromosomes (YACs) or P1-derived artificial chro-
mosomes (BACs/PACs) allow position-independent and copy number-dependent
transgene expression which guarantees faithful expression of transgenes (Heintz et al.
2001; Giraldo and Montoliu 2001). To better control transgene expression levels, in-
ducible systems have been developed which allow control of the onset, the duration
 6 774 Biochemical and Analytical Techniques

Table 7
Transgenic expression of wild type or mutant G protein subunits in the cardiovascular system

Transgene Promoter Tissue Phenotype Reference

Gγ2(C68S) αMHC cardio- reduction of Gehrmann et al. (2002)
myocy tes Gβγ concen- Am J Physiol 282:
tration; para- H445–456
sy mpathetic
heart rate
control↓
Gαs (w t) αMHC cardio- increased ino- Gaudin et al. (1995)
myocy tes tropic and J Clin Invest 95: 1676
chronotropic –1683,
responses to Iwase et al. (1996)
catecholamines Circ Res 78:517–524
Gαq (w t) αMHC cardio- cardiac hyper- D'Angelo et al. (1997)
myocy tes trophy, cardio- Proc Natl Acad Sci
myopathy USA 94: 8121–8126,
In-Vitro Techniques

Adams et al. (1998)

Proc Natl Acad Sci
USA 95: 10140–10145
Gαq αMHC cardio- cardiac hyper- Mende et al. (1998)
(Q209L) myocy tes trophy and Proc Natl Acad Sci
(const. dilatation USA 95: 13893–13898
active)
Gαq- αMHC cardio- pressure over- Akhter et al. (1998)
peptide myocy tes load induced Science 280: 574–577
(305–359) ventricular
hypertrophy ↓
Gαq- SM22α smooth phenylephrine-, Keys et al. (2002)
peptide muscle serotonin-, and Hypertension 40:
(305–359) angiotensin- 660–666
induced blood
pressure and
cardiac hyper-
trophy ↓
MHC myosin heavy chain; const. active constitutively active mutant; wt wild type form. For details, see text.

and the level of transgene expression. Another advantage of these techniques lies in
the possibility of studying an animal before, during and after transient transgene
expression, thus allowing comparison of the effect of the expressed transgene with the
wild type situation within the same animal. Various systems for inducible and tissue-
specific gene expression have been reported. The tetracycline-controlled transacti-
 Receptor and Binding Studies 775 6.3
Table 8
Cardiovascular abnormalities in mice lacking α-subunits of heterotrimeric G proteins

Subtype Expression Effector Phenotype Reference

Gαi2 w ide AC↓; 1
Defective mus- Nagata et al. (2000)
GIRK2↑ carinic regula- Circ Res 87: 903–909;
tion of cardiac Chen et al. (2001)
L-type Ca2+ Am J Physiol 280:
channels; nor- H1989–1995,
mal basal car- Jain et al. (2001) Am J
diac function Physiol 280: H569–575
Gαi3 w ide AC↓1; Defective mus- Nagata et al. 2000;
GIRK↑2- carinic regula- Jain et al. 2001 (see
tion of cardiac above)
L-type Ca2+
channels; nor-
mal basal car-
diac function
Gαo neuronal, VDCC↓, Defective mus- Valenzuela et al. (1997)
heart, GIRK↑ carinic regula- Proc Natl Acad Sci
neuro- tion of cardiac USA 94: 1727–1732
endocrine L-type Ca2+
channels
Gαq + Gα11 ubiquitous PLC-β↑ myocardial Offermanns et al.
hypoplasia (1998)
(lethal e11) EMBO J 17: 4304–4312
Gαq + Gα11 pressure over- Wettschureck et al.
(cardio- load induced (2001)
myocy te- hypertrophy ↓ Nat Med 7: 1236–1240
restricted)
a
AC adenylyl cyclase types I,V,VI; bGIRK1-GIRK4 (Kir3.1-Kir3.4), effector is regulated through βγ-subunits; cVDCC,
voltage dependent calcium channel, P/Q-R-type (Cav2.1-Cav2.3), effector is regulated through βγ-subunits; PLC-β,
β-isoforms of phospholipase C.

vation (tTA) system is based on the tetracycline-responsive transcriptional regulatory

element of E. coli (tetO) and a mutant bacterial tet repressor protein fused to the
C-terminal domain of the VP16 protein of the herpes simplex virus (Gossen et al.
1995). Tetracycline-regulated expression of transgenes has been reported in the heart
(Yu et al. 1996; Valencik and McDonald 2001), smooth muscle (Ju et al. 2001) and en-
dothelial cells (Sarao and Dumont 1998).
The traditional gene “knock-out“ approach results in the loss of gene function in
all tissues from early embryonic stages onwards. Since cardiovascular functions are
already of great importance before birth, deletion of the genes involved in the devel-
 6 776 Biochemical and Analytical Techniques

Figure 15
Tissue-specific gene inactivation. Mice carrying the cre transgene under the control of a tissue-speci-
fic promoter (e.g. cardiomyocyte-specific promoter) are crossed with mice in which an exon of a parti-
cular gene is flanked by loxP-sites (red triangles). LoxP-sites were introduced by gene targeting in ES-
cells. In mice that carry the cre transgene and the “floxed“ exon, Cre is expressed in tissues de-fined by
the promoter. This will lead to tissue-specific excision of the exon resulting in a deletion of the gene. In
tissues that do not express Cre, the gene remains intact
In-Vitro Techniques

opment or function of the cardiovascular system often results in embryonic lethality

(Copp 1995; Conway et al. 2003) preventing the analysis of gene function in adult-
hood. In addition, lack of a gene product during development or in the adult animal
may be offset by the up-regulation of a related gene or may be counterbalanced by
more complex compensatory mechanisms. In this case, the phenotype resulting from
the deletion of a single gene reflects the result of a complex compensatory reaction of
the whole organism and cannot be attributed simply to the missing gene function. To
overcome these problems the gene targeting approach has been further developed to
restrict gene deletion to specific tissues or to define the onset of gene deletion. The
currently favoured techniques rely on the introduction of loxP (locus of crossing
over) recognition sites for the recombinase Cre (causes recombination) which recog-
nizes these sites and excises the DNA sequence flanked by two loxP sites leaving one
site behind (Le and Sauer 2000). If the targeted gene is flanked by two loxP sites, re-
combination and deletion of the gene occurs depending on the site and time of Cre
expression. Cre can be transgenically expressed from a specific promoter or after plac-
ing its gene downstream of an endogenous promoter (Fig. 15) (Rajewsky et al. 1996).
Tissue-specific gene targeting using Cre/loxP technology has been extended by
combining Cre expression with an inducible system which relies on the tet-system or
which uses Cre versions whose activity can be regulated by synthetic ligands. In the
latter case, Cre recombinase was expressed transgenically under the control of a de-
fined promoter as a fusion protein with a mutated binding domain of the oestrogen
or progesterone receptor. The mutation in the binding domain allows only the syn-
thetic compounds tamoxifen or mifepristone to bind with high affinity. In the absence
of ligand, the Cre fusion protein is inactive due to binding to the Hsp90 complex.
Binding of ligand disinhibits Cre and results in Cre-mediated recombination (Feil et
al. 1997; Kellendonk et al. 1999).
 Receptor and Binding Studies 777 6.3
Table 9
Mouse lines expressing the recombinase Cre in the cardiovascular system

Promoter driving
Expression Comments Reference
Cre expression
SM22α smooth muscle constitutive, Holtw ick et al.
(Li et al. (1996) some preference (2002) Proc Natl
J Cell Biol 132: for vascular Acad Sci USA
849–859) SMCs 99:7142-7147
SM22α smooth muscle knock-in; Kuhbandner et
inducible al. (2000)
(tamoxifen) Genesis 28: 15–22
SMMHC smooth muscle transgenic; Regan et al.
(Madsen et al. (1998) constitutive (2000) Circ Res
Circ Res 82:908–917) 87: 363–369
MLC2a cardiomyocy tes knock-in; Wettschureck et
constitutive, al. (2001) Nat
e8.5–10.5 Med 7:1236–1240
MLC2v cardiomyocy tes knock-in; Chen et al. (1998)
constitutive, J Biol Chem 273:
e8.75 1252–1256
αMHC cardiomyocy tes transgenic; Minamino et al.
(Subramaniam et al. inducible (2001) Circ Res
(1991) J Biol Chem (RU486) 88: 587–592
266:24613-24620)
αMHC cardiomyocy tes transgenic; Sohal et al. (2001)
(Palermo et al. (1996) inducible Circ Res 89:20–25
Circ Res 78: 504–509) (tamoxifen)

αMHC cardiomyocy tes constitutive/ Agah et al. (1997)

(Subramaniam et al. postnatal J Clin Invest 100:
1991, see above) 169–179, Abel et
al. (1999) J Clin
Invest 104: 1703
–1714
Ckmm muscle cells transgenic; Wang et al. (1999)
constitutive; Nat Genet 21:
skeletal/heart 133–137
muscle
Tie-1 endothelial cells constitutive, e8 Gustafsson et al.
(Iljin et al. (1999) (2001) J Cell Sci
FASEB J 13:377–386) 114:671–676
 6 778 Biochemical and Analytical Techniques

Table 9
Continued

Promoter driving
Expression Comments Reference
Cre expression
Tie-2 endothelial cells transgenic; Kisanuki et al.
constitutive (2001) Dev Biol
230: 230–242
Tie-2 endothelial cells transgenic; Forde et al.
(Schläger et al. (1997) inducible (2002) Genesis
Proc Natl Acad Sci (tamoxifen) 33:191–197
USA 94:3058-3063)

A list of published constitutive or inducible Cre mouse lines, which allow the con-
ditional inactivation of genes in the cardiovascular system, is shown in Table 9.
In-Vitro Techniques

When analyzing genetic mouse models that carry mutations in genes coding for
components of the G protein mediated signalling system, a major goal is to relate
possible changes in cellular signalling to phenotypical changes of the cardiovascular
system. This requires the application of methods for the in vivo analysis of cardiovas-
cular functions as well as for the analysis of processes at the cellular or molecular
level. Especially when tissue-specific mutants are studied, studies on the cellular level
rely on the specificity of the genetic approach as well as on the purity of the cell prepa-
ration. A particular problem has been the analysis of cellular functions in adult
cardiomyocytes from cardiomyocyte-specific mutants. Only 14% of adult mouse ven-
tricular cells are cardiomyocytes, and 90–95% of these cells are binucleated. This
means that 24% of the DNA content of the whole murine ventricle originates from
cardiomyocytes (Soonpaa et al., 1996). Thus, methods to isolate and purify adult
murine cardiomyocytes are needed to allow the study of cellular processes in a ho-
mogenous cell population.

Description of Methods and Practical Approach

Isolation and Purification of Adult Mouse Cardiac Myocytes

Cardiomyocytes from the adult mouse heart are delicate cells and even small alter-
ations in temperature, oxygen levels and nutrient and ion concentrations can affect
their viability. Hallmarks of viable cardiomyocytes are their rod-shaped morphology,
visible cross-striation and – at least in the first hours after preparation – spontaneous,
slow contractions. Several protocols for the isolation of adult mouse cardiomyocytes
have been published (Pawloski-Dahm et al. 1998; Kim et al. 1999; Hilal-Dandan et al.
2000) and all are based on the retrograde perfusion of the heart via the aorta. Because
retrograde perfusion of the aorta leads to closure of the aortic valves, the perfusate is
 Receptor and Binding Studies 779 6.3
forced into the coronary arteries that branch off from the aorta ascendens immedi-
ately above the aortic valves. Distributed via coronary arteries, capillaries and veins
the perfusate can reach a maximal number of cardiomyocytes, leading to maximal
dissociation of cells.
The procedure described here usually results in a quantity of 2–4×106 cardio-
myocytes per heart, with about 70% being rod-shaped, cross-striated and partly
contracting cardiomyocytes. The preparations usually contain less than 5% non-
cardiomyocytes.

Buffers and Media.

▬ Perfusion buffer, pH 7.2 at 37 °C (mM)
NaCl 113, KCl 4.7, KH2PO4 0.6, Na2HPO4, 0.6, MgSO4 1.2, NaHCO3 12, KHCO3 12,
glucose 20, Na+-HEPES 10, taurine 30, creatine 2, carnitine 2
▬ Digestion buffer
Collagenase II (Worthington) 0.08%, trypsin (Invitrogen) 0.1% (w/v), fatty acid
free BSA (Sigma) 0.1% (w/v), BDM 10 mM, CaCl2 0.025 mM in perfusion buffer
▬ Mincing buffer
BDM 10 mM in perfusion buffer
▬ M199 (mM)
Taurine 30, creatine 2, carnitine 2, heat inactivated FCS 4% (v/v), penicillin 100 U/ml,
streptomycin 0.1 mg/ml, in medium 199 (Invitrogen)

Procedure (mod. from Hilal-Dandan et al. 2000).

▬ Anaesthesia: The mouse is anaesthetized by intraperitoneal injection of ketamine
(150 mg/kg) and xylazine (24 mg/kg) in saline. Coagulation is prevented with
100 U heparin i.p.
▬ Preparation of the heart: After superficial disinfection with 70% ethanol, thoraco-
tomy is performed and the heart is excised with at least 3 mm of aorta ascendens
attached to it. The still beating heart is briefly washed in ice-cold Ca2+ free perfu-
sion buffer and non-cardiac tissue is removed. The aorta is mounted on a blunted
20 G needle connected to a roller pump and fixed with a thread tied around aorta
and needle. During perfusion the heart is suspended in prewarmed perfusion
buffer. All perfusion solutions are kept at 37 °C and are bubbled with carbo-
gen (95%O2; 5%CO2). Gas bubbles must not enter the tubing. The delay between
excision of the heart and the start of perfusion should be kept as short as pos-
sible.
▬ Chemical dissociation of myocardial cells: The heart is first perfused for 4 min
with prewarmed, gassed perfusion buffer (3.5–4 ml/min), then for 15 min with
prewarmed, gassed digestion buffer at the same speed. After successful perfusion
the consistency of the myocardium should be flaccid.
▬ Mechanical dissociation of cells: The heart is removed from the perfusion appa-
ratus and suspended in perfusion buffer containing 10 mM BDM. Atria are re-
moved and ventricles are minced with scissors or blades. To dissociate cell clumps
the suspension is gently pipetted up and down with a pipette tip cut off to give a
diameter of about 3 mm. Undissociated clumps are removed by filtration through
a 200 µm nylon mesh.
 6 780 Biochemical and Analytical Techniques

Figure 16
Abrogation of endothelin-1 induced inositol
phosphate production in Gαq/α11-double defi-
cient mouse cardiomyocytes. Inositol phosphate
production in purified, cultured cardiomyocytes
after stimulation with 1 µM endothelin-1 (ET-1).
In contrast to wild type cardiomyocytes Gαq/
Gα11-deficient cardiomyocytes are unresponsive
towards ET-1 stimulation. Basal inositol phos-
phate production does not differ between the
genotypes

▬ Differential centrifugation to enrich cardiomyocytes: The resulting cell suspen-

sion is washed twice in mincing buffer and centrifuged at 50×g for 1 min at room
temperature. After the last centrifugation, the pellet is resuspend in 10 ml mincing
In-Vitro Techniques

buffer containing 0.1% fatty acid free BSA

▬ Recalcification of cardiomyocyte suspension: The calcium concentration is slowly
increased to 50, 100, 200, 500, 1000 µM by adding CaCl2 every 4 min from a 1 M
stock with gentle shaking.
▬ Cardiomyocyte culture: The cells are spun down (50×g for 1 min), resuspended in
M199, counted and seeded onto laminin-coated dishes (2 µg laminin/cm2) at a
density of 3×104/cm2. Attached cells retain their cardiomyocytic morphology for
12–24 h and then start to detach.

Example

Abrogation of Endothelin-1 Induced Inositol Phosphate Production in Gαq/Gα11-double

Deficient Mouse Cardiomyocytes

Cardiomyocytes from a wild type mouse and an age matched mouse with cardio-
myocyte-specific inactivation of the G protein α subunits Gαq and Gα11 (Wett-
schureck et al. 2001) were isolated. The yield per heart was 2.7×106 for the wild type
and 2.4×106 cells for the knockout, about 70% of cells showed the typical rod-shaped
morphology with cross striation. Cells were seeded at a density of 106 cells/5 ml me-
dium on to laminin coated 6 cm dishes (2 plates for each heart) and after 1 h about
70% of cells had adhered. The mouse cardiomyocytes were incubated for 16–20 h with
inositol-free medium (Invitrogen) containing 10 µCi/ml myo-[3H]inositol. Note that
regular medium contains considerable amounts of myo-inositol that would result in
poor labelling. Thereafter the myo-[3H]inositol containing medium was replaced by
inositol-free medium containing 10 mM LiCl to block inositol phosphatase. After
10 min of incubation, one plate of each heart was stimulated with 1 µM endothelin-1
for another 30 min. Thereafter, the medium was replaced by ice cold 10 mM formic
 Receptor and Binding Studies 781 6.3
acid and water-soluble lipids were extracted during 2 h of shaking at 4 °C. In the
meantime ion exchange columns were prepared by loading each column with 0.5 g AG
1–X8 matrix (BioRad) suspended in 3 ml 1 M formic acid. Columns were washed
three times with 5 ml H20. The formic acid was then extracted three times with ether
and once with chloroform to enrich inositol phosphates. The pH of the aqueous ex-
tract was adjusted with 25% NH4OH to pH 7–8 and debris was removed by centrifu-
gation at 12.000 ×g at 4 °C for 2 min. Columns were loaded with the supernatant,
washed twice with 10 ml H20 and inositol phosphates were eluted with 2×2 ml 2 M
ammonium formate pH 5.2. After addition of 10 ml scintillation cocktail the produc-
tion of radioactive inositol phosphates was determined (Fig. 16). In the Gαq/Gα11-
double deficient mouse cardiomyocytes the ET-1 stimulated inositol phosphate
production was completely abrogated.

Troubleshooting

Preparation of Adult Mouse Cardiac Myocytes

1. Poor digestion: Air bubbles or precipitations within the tubing of the perfusion
system can lead to occlusion of coronary arteries which results in poor digestion.
The tip of the blunted needle must not penetrate (or even touch) the aortic valves,
for only intact, closed valves will allow proper perfusion of coronary arteries. Dif-
ferent batches of collagenase can have different efficiencies; it might therefore be
useful to test several batches. Collagenase B (Roche) or Collagenase II (Worthing-
ton) usually work well.
2. Loss of typical morphology: Adult mouse cardiomyocytes easily lose their typical
rod-shaped, cross-striated morphology, especially when exposed to high calcium
concentrations. Therefore it is crucial to increase calcium slowly during the recal-
cification step and use gentle shaking. Addition of BDM protects cardiomyocytes
from hypercontractions and is therefore essential during enzymatic digestion, cell
dissociation and recalcification. Strong shear force during the mechanical disso-
ciation of cardiomyocytes might break cardiomyocytes; this can be prevented by
cutting off the tip of the pipette to give a diameter of 3–4 mm.
3. Poor adhesion: In our hands freshly isolated mouse cardiomyocytes adhere well
to laminin 10 µg/ml, but only poorly to poly-L-lysine (10 µg/ml), FCS 100%,
fibronectin (10 µg/ml) or collagen IV (10 µg/ml).
4. Contamination: To avoid fungal or bacterial contamination the tubing of the per-
fusion apparatus, instruments and nylon meshes should be rinsed with 70% etha-
nol and sterile water before and after using. Solutions should be filtered through
0.2 µm sterile filters.

References
Christensen G, Wang Y, Chien KR (1997) Physiological assessment of complex cardiac phenotypes in genetically en-
gineered mice. Am J Physiol 272:H2513–H2524
Conway SJ, Kruzynska-Frejtag A, Kneer PL, Machnicki M, Koushik SV (2003) What cardiovascular defect does my
prenatal mouse mutant have, and why? Genesis 35:1–21
Copp AJ (1995) Death before birth: clues from gene knockouts and mutations. Trends Genet 11:87–93
 6 782 Biochemical and Analytical Techniques

El-Armouche A, Rau T, Zolk O, Ditz D, Pamminger T, Zimmermann WH, Jäckel E, Harding SE, Boknik P, Neumann J,
Eschenhagen T (2003) Evidence for protein phosphatase inhibitor-1 playing an amplifier role in β-adrenergic
signaling in cardiac myocytes. FASEB J 17:437–439
Fahimi-Vahid M, Gosau N, Michalek C, Han L, Jakobs KH, Schmidt M, Roberts N, Avkiran M, Wieland T (2002) Dis-
tinct signaling pathways mediate cardiomyocyte phospholipase D stimulation by endothelin-1 and thrombin. J
Mol Cell Cardiol 34:441–453
Feil R, Brocard J, Mascrez B, LeMeur M, Metzger D, Chambon P (1996) Ligand-activated site-specific recombination
in mice. Proc Natl Acad Sci USA 93:10887–10890
Fitzgerald SM, Gan L, Wickman A, Bergstrom G (2003) Cardiovascular and renal phenotyping of genetically modi-
fied mice: a challenge for traditional physiology. Clin Exp Pharmacol Physiol 30:207–216
Giraldo P, Montoliu L (2001) Size matters: use of YACs, BACs and PACs in transgenic animals. Transgenic Res 10:
83–103
Gosau N, Fahimi-Vahid M, Michalek C, Schmidt M, Wieland T (2002) Signalling components involved in the coupling
of α1-adrenoceptors to phospholipase D in neonatal rat cardiac myocytes. Naunyn-Schmiedeberg´s Arch
Pharmacol 365:468–476
Gossen M, Freundlieb S, Bender G, Muller G, Hillen W, Bujard H (1995) Transcriptional activation by tetracyclines in
mammalian cells. Science 268:1766–1769
He TC, Zhou S, Da Costa LT, Yu J, Kinzler KW, Vogelstein B (1998) A simplified system for generating recombinant
adenoviruses. Proc Natl Acad Sci USA 95:2509–2514
Heintz N (2001) BAC to the future: the use of bac transgenic mice for neuroscience research. Nat Rev Neurosci 2:
861–870
Hilal-Dandan R, Kanter JR, Brunton LL (2000) Characterization of G-protein signaling in ventricular myocytes from
the adult mouse heart: differences from the rat. J Mol Cell Cardiol 32:1211–1221
Janssen PM, Schillinger W, Donahue JK, Zeitz O, Emami S, Lehnart SE, Weil J, Eschenhagen T, Hasenfuss G, Prestle J
(2002) Intracellular β-blockade: overexpression of Gαi2 depresses the β-adrenergic response in intact myocardium.
In-Vitro Techniques

Cardiovasc Res 55:300–308

Ju H, Gros R, You X, Tsang S, Husain M, Rabinovitch M (2001) Conditional and targeted overexpression of vascular
chymase causes hypertension in transgenic mice. Proc Natl Acad Sci USA 98:7469–7474
Kellendonk C, Tronche F, Casanova E, Anlag K, Opherk C, Schutz G (1999) Inducible site-specific recombination in the
brain. J Mol Biol 285:175–182
Kim SJ, Iizuka K, Kelly RA, Geng YJ, Bishop SP, Yang G, Kudej A, McConnell BK, Seidman CE, Seidman JG, Vatner SF
(1999) An α-cardiac myosin heavy chain gene mutation impairs contraction and relaxation function of cardiac
myocytes. Am J Physiol 276:H1780–H1787
Korhonen J, Lahtinen I, Halmekyto M, Alhonen L, Janne J, Dumont D, Alitalo K (1995) Endothelial-specific gene ex-
pression directed by the tie gene promoter in vivo. Blood 86:1828–1835
Le Y, Sauer B (2000) Conditional gene knockout using cre recombinase. Methods Mol Biol 136:477–485
Li L, Miano JM, Mercer B, Olson EN (1996) Expression of the SM22α promoter in transgenic mice provides evidence
for distinct transcriptional regulatory programs in vascular and visceral smooth muscle cells. J Cell Biol 132:
849–859
Mittmann C, Chung CH, Höppner G, Michalek C, Nose M, Schuh A, Eschenhagen T, Weil J, Pieske B, Hirt S, Wieland T
(2002) Expression of 10 RGS proteins in human myocardium: functional characterization of an upregulation of
RGS4 in heart failure. Cardiovasc Res 55:778–786
Neyroud N, Nuss HB, Leppo MK, Marban E, Donahue JK (2002) Gene delivery to cardiac muscle. Methods Enzymol
346:323–334
Pawloski-Dahm CM, Song G, Kirkpatrick DL, Palermo J, Gulick J, Dorn GW 2nd, Robbins J, Walsh RA (1998) Effects
of total replacement of atrial myosin light chain-2 with the ventricular isoform in atrial myocytes of transgenic
mice. Circulation 97:1508–1513
Rajewsky K, Gu H, Kuhn R, Betz UA, Muller W, Roes J, Schwenk F (1996) Conditional gene targeting. J Clin Invest 98:600–
603
Regan CP, Manabe I, Owens GK (2000) Development of a smooth muscle-targeted cre recombinase mouse reveals novel
insights regarding smooth muscle myosin heavy chain promoter regulation. Circ Res 87:363–369
Sarao R, Dumont DJ (1998) Conditional transgene expression in endothelial cells. Transgenic Res 7:421–427
Schlaeger TM, Bartunkova S, Lawitts JA, Teichmann G, Risau W, Deutsch U, Sato TN (1997) Uniform vascular-endot-
helial-cell-specific gene expression in both embryonic and adult transgenic mice. Proc Natl Acad Sci USA 94:3058–
3063
Soonpaa MH, Kim KK, Pajak L, Franklin M, Field LJ (1996) Cardiomyocyte DNA synthesis and binucleation during
murine development. Am J Physiol 271:H2183–H2189
Subramaniam A, Jones WK, Gulick J, Wert S, Neumann J, Robbins J (1991) Tissue-specific regulation of the α-myosin
heavy chain gene promoter in transgenic mice. J Biol Chem 266:24613–24620
Valencik ML, McDonald JA (2001) Codon optimization markedly improves doxycycline regulated gene expression in
the mouse heart. Transgenic Res 10:269–275
Wettschureck N, Rutten H, Zywietz A, Gehring D, Wilkie TM, Chen J, Chien KR, Offermanns S (2001) Absence of pres-
sure overload induced myocardial hypertrophy after conditional inactivation of Gαq/Gα11 in cardiomyocytes. Nat
Med 7:1236–1240
 Receptor and Binding Studies 783 6.3
Wiechert S, El-Armouche A, Rau T, Zimmermann WH, Eschenhagen T (2003) 24-h Langendorff-perfused neonatal rat
heart to study the impact of adenoviral gene transfer. Am J Physiol 285: H907–914
Yu Z, Redfern CS, Fishman GI (1996) Conditional transgene expression in the heart. Circ Res 79:691–697
Zheng B, Ma YC, Ostrom RS, Lavoie C, Gill GN, Insel PA, Huang XY, Farquhar MG (2001) RGS-PX1, a GAP for Gαs and
sorting nexin in vesicular trafficking. Science 294:1939–1942