UNIT 7.

5 Determination of Compound Binding to

Plasma Proteins
The pharmacokinetic and pharmacodynamic properties of a compound are profoundly
affected by the extent of its binding to plasma proteins. Consequently, the determination
of the plasma protein binding properties of a compound is essential during drug develop-
ment and is increasingly required during lead prioritization.
The protocols described in this unit detail the techniques of ultrafiltration and equilibrium
dialysis for determination of plasma protein binding. Equilibrium dialysis is the more
robust of the two techniques, whereas ultrafiltration (see Basic Protocol 1) is the more
rapid. Two versions of the equilibrium dialysis technique are detailed in these protocols,
the traditional approach (see Basic Protocol 2) and a higher throughput 96-well plate
version (see Alternate Protocol).
CAUTION: Human and animal plasma should be handled with appropriate care and
disposed of responsibly. Specifically, gloves, labcoats, and safety glasses should be worn,
spills cleaned and sterilized promptly, and waste disposed of as appropriate for a
biohazard.
CAUTION: Radioactive materials require special handling. Refer to the institutional
Radiation Safety Office for guidelines concerning proper handling and disposal.
BASIC
PROTOCOL 1
DETERMINATION OF PLASMA PROTEIN BINDING BY
ULTRAFILTRATION
The technique of ultrafiltration relies on centrifugal force to drive the unbound test article
(compound of interest) through a size-selective membrane. A representation of the device
is shown in Figure 7.5.1. Ideally, the degree of nonspecific binding of the test article to
the apparatus should be determined prior to investigating plasma protein binding. If the
nonspecific binding exceeds 5%, ultrafiltration is not recommended for the determination
of plasma protein binding. An alternative technique such as equilibrium dialysis should
be investigated. It is not advisable to attempt to subtract nonspecific binding when
Supplement 18
Contributed by Natasha Dow
Current Protocols in Pharmacology (2002) 7.5.1-7.5.14
Copyright 2002 by John Wiley & Sons, Inc.
stopper
membrane
filter
upper portion
lower portion
Figure 7.5.1 Micropartition device for determination of plasma protein binding by ultrafiltration.
7.5.1
Pharmacokinetics
determining plasma protein binding as the presence of plasma may alter its extent. Plasma
from at least three donors should be pooled by species to ensure true representation of the
species under investigation. Plasma may be stored frozen at 20C prior to use.
Materials
Test article(s)
Solvent (e.g., acetonitrile, methanol, DMSO)
0.01 M phosphate-buffered saline (PBS), pH 7.4 (see recipe)
Plasma samples from all species to be investigated, fresh or frozen
Sodium phosphate monobasic (NaH
2
PO
4
), solid
Sodium phosphate dibasic (Na
2
HPO
4
), solid
Fixed-angle rotor
Micropartition devices containing membrane filters with a 10 to 30 kDa MWCO
(e.g., Centrifree; Millipore)
Additional reagents and equipment for liquid scintillation counting (LSC;
radiolabeled test articles only), determining protein concentration (APPENDIX 3B),
and HPLC or LC/MS analysis
Determine nonspecific binding (NSB)
1. Set centrifuge and fixed-angle rotor to warm to 37C.
This may require centrifugation of the rotor at a high centrifugal force.
2. Prepare a stock solution or solutions of the test article or articles in a suitable solvent
(e.g., acetonitrile, methanol, DMSO) at concentrations 100 times those required for
investigation. If a radiolabeled test article is used, dilute with unlabeled test article
to achieve a maximum radioactive concentration of 0.5 Ci/ml, thus conserving
expensive radiolabelled material and reducing disposal costs.
The preparation of the concentrated stock ensures that the maximum final solvent concen-
tration is only 1% (v/v).
Typical concentrations for investigation are the anticipated maximum plasma concentra-
tion (C
max
) and 10 C
max
, or 10 and 100 M when C
max
is unknown. Preferentially the
lowest concentration should be investigated to determine the maximal percentage nonspe-
cific binding.
3. Aliquot 990 l of 0.01 M PBS, pH 7.4, into two 2-ml microcentrifuge tubes (i.e.,
duplicate) per concentration of each test article. Add 10 l of 100 test article stock
solution to each. Vortex mix and incubate 20 min in an 37C water bath.
Preparation of individual samples instead of a pool is preferred for loading of the Centrifree
micropartition units, as a Pasteur pipet can be used for transfer into the centrifuge tubes
instead of the wider pipet tip required for delivering a specific volume.
4. Following 20 min of incubation, remove the samples from the water bath. If the test
article is radiolabeled, remove appropriate duplicate aliquots for determination of the
actual dosed concentration by liquid scintillation counting (LSC). Using a long glass
Pasteur pipet, transfer remaining sample directly to the upper portion of a micropar-
tition device containing a membrane filter with a 10 to 30 kDa MWCO (e.g.,
Centrifree).
The 20 min incubation time allows the sample to reach physiological temperature and is
convenient when performing multiple batches, as a new set of incubations can be initiated
once centrifugation of the previous set is started.
Take care not to introduce air bubbles while adding the sample to the unit. If an air bubble
is noted, gently tap the entire unit on the bench until the bubble is removed.
Supplement 18 Current Protocols in Pharmacology
7.5.2
Determination of
Compound
Binding to
Plasma Proteins
5. Centrifuge the samples 20 min at 1000 g, 37C.
6. Remove and appropriately discard the upper portion of the unit.
7. For radiolabelled test articles, remove duplicate aliquots of the ultrafiltrate (lower
portion) and determine the radioactive concentration by LSC. For nonlabeled test
articles, determine the test article concentration of the ultrafiltrate and of a 1:100 PBS
dilution of the original 100 stock solution by an appropriate analytical technique.
Quantification against a standard curve is not strictly necessary as the percentage recovery
can be determined by comparison of the responses instead of absolute concentrations.
Ultrafiltrate portions may be capped and stored frozen prior to analysis. The temperature
and duration of storage depends on the stability of the test article. If the stability of the test
article is not characterized, storage should be up to 3 months at 70C.
8. Determine the percentage nonspecific binding (NSB) according to the equation
below:
Prepare plasma for determination of plasma protein binding
9. Defrost frozen plasma from all species to be investigated at room temperature until
thawed.
10. Centrifuge defrosted plasma 5 min at 2000 g, room temperature. Decant and retain
the supernatant.
Freeze-thaw of plasma can produce a fibrous precipitate.
11. Pool plasma by species and measure pH.
12. Adjust the pH to 7.4 by adding small amounts (i.e., a few grains) of either solid
NaH
2
PO
4
(decrease pH) or Na
2
HPO
4
(increase pH).
Variation in plasma pH can dramatically affect the extent of plasma protein binding of a
test article, as exemplified by propranolol (Paxton and Calder, 1983).
13. Retain an aliquot of each plasma sample for determination of total protein concen-
tration (APPENDIX 3B) and if required, specific plasma protein concentrations.
This step is to confirm that the plasma used contains protein concentrations representative
of the species and strain. This can be achieved using a clinical chemistry analyzer such as
the Hitachi 912 and a suitable diagnostic kit (e.g., JAS Diagnostics). Note that the
methodology is not described in this unit.
Determine plasma protein binding
14. Determine plasma protein binding in the same manner as described for determination
of nonspecific binding, replacing PBS with pooled plasma (see steps 1 to 6).
15. Calculate the percentage plasma protein binding according to the equation:
concentration in ultrafiltrate
NSB(%) =100 100
concentration added




concentrationin ultrafiltrate
plasma protein binding(%) 100 100
concentrationadded

=


Current Protocols in Pharmacology Supplement 18
7.5.3
Pharmacokinetics
16. If required, subject a portion of the ultrafiltrate to analysis by HPLC or liquid
chromatography/mass spectroscopy (LC/MS) to confirm the identity of the free
fraction. This step is particularly important if the test article contains impurities, is
prone to degradation (or if the plasma stability is unknown), and/or is highly protein
bound.
The volume of ultrafiltrate should not exceed 40% of the initial volume to avoid dissociation
of bound compound due to removal of the free compound (Whitlam and Brown, 1981).
BASIC
PROTOCOL 2
DETERMINATION OF PLASMA PROTEIN BINDING BY TRADITIONAL
EQUILIBRIUM DIALYSIS
The technique of equilibrium dialysis is conducted with two Teflon cells, one containing
plasma and the other an equal volume of buffer, separated by a size-selective membrane.
The unbound test article is allowed to reach equilibrium between the two compartments.
Thus, measurement of the test article concentration in each compartment allows the
degree of plasma protein binding to be calculated.
Equilibrium dialysis is theoretically the most accurate method for determination of
plasma protein binding because the equilibrium is not shifted when aliquots are taken
from either side of the membrane (Sebille et al., 1990). Some advantages of equilibrium
dialysis are that reproducible results can be obtained even for low affinity compounds,
and that meaningful results can be determined even for compounds bound nonspecifically
to the apparatus. Disadvantages of equilibrium dialysis include degradation of the
compound due to the duration of exposure at 37C, and volume shifts due to osmotic
equilibrium. It is recommended that the equilibration time be determined in the absence
of plasma to ease analysis, preserve plasma, allow determination of nonspecific binding,
and to simply measure the time required for equilibration of free drug between compart-
ments, avoiding any complication presented by protein interactions.
Materials
PBS, pH 7.4 (see recipe)
Test article or articles
Solvent (e.g., acetonitrile, methanol, DMSO)
Plasma samples from all species to be investigated, either fresh or frozen
Sodium phosphate monobasic (NaH
2
PO
4
), solid
Sodium phosphate dibasic (Na
2
HPO
4
), solid
47-mm-diameter precut Spectra/Por 4 membrane discs with 12 to 14 kDa MWCO
(Spectrum Laboratories)
Spectra/Por equilibrium dialyzer with stopper plugs and semi-micro Teflon cells
(Spectrum Laboratories) labeled 1 to 20 in indelible ink
37C acrylic water bath (e.g., Spectrum Laboratories or equivalent)
Blunt-nose 22-G needles and syringes
Determine equilibration time and NSB
1. For each concentration of test article, soak ten 47-mm-diameter precut Spectra/Por
4 membrane discs with 12 to 14 kDa MWCO in a dish of room-temperature deionized
water for at least 30 min. Replace the water with PBS, pH 7.4, and incubate a further
30 min.
This step is required to remove preservatives and impurities in the membrane discs and to
equilibrate them to dialysis conditions. The dish used for soaking should be sufficiently
large to allow the membranes to soak as a single layer.
Supplement 18 Current Protocols in Pharmacology
7.5.4
Determination of
Compound
Binding to
Plasma Proteins
2. Prepare a 100 stock solution or solutions of test article or articles in suitable solvent
as described for ultrafiltration investigations (see Basic Protocol 1, step 2).
3. Add 120 l each test article stock solution to 11.88 ml PBS and vortex mix to prepare
dosing solutions.
4. Assemble the first set of five semi-micro Teflon cells of the Spectra/Por equilibrium
dialyzer according to the manufacturers instructions, placing the larger half of each
cell on the bottom of the guide rod and aligning the filling ports. Plug the single port
on each cell half with the supplied stopper. Maintain temperature with a 37C acrylic
water bath.
5. Using a blunt-nose 22-G needle and syringe, fill the larger half of each cell with 1 ml
PBS through one of the double filling ports.
This cell half will be referred to as the buffer compartment.
6. Similarly, fill the other cell half with the appropriate PBS solution spiked with test
article (step 3).
This cell half will be referred to as the dosing compartment.
7. Stopper all filling ports and place the complete cell set in the drive unit maintained
in the temperature controlled environment (37C). Set the drive unit to 20 rpm and
observe to ensure the stoppers remain in place during rotation. Record the start time
as when the rotation commences.
8. Repeat steps 4 to 7 as necessary, such that 10 cells are prepared for each concentration
of each test article.
9. Following 1, 2, 4, 6, and 20 hr of dialysis, completely remove the liquid from both
compartments of two cells per test article concentration, using blunt-nosed 22-G
needles and syringes through the single port. Use a different filling port to remove
the sample than the one used to add the sample, thus avoiding contamination. Remove
a stopper from both cell halves to allow air flow and successful sample removal.
10. Determine the concentration of test article in both compartments and in the PBS
dosing solution by an appropriate method, such as liquid scintillation counting (LSC)
for radiolabeled test articles or LC/MS for nonlabeled test articles.
11. Plot the concentration of test article in each compartment against duration of dialysis
and determine the time at which equilibrium is achieved.
12. Calculate the percentage nonspecific binding (NSB) at equilibration time according
to the equation:
where dose concentration is the test article concentration (e.g., disintegrations per
minute, nanomoles, micrograms) expressed per milliliter and amount recovered
equals the concentration in the dosing compartment plus the concentration in buffer
compartment.
This equation is valid since the compartment volumes are 1 ml.
dose concentration amount recovered
NSB() 100 100
dose concentration

=


Current Protocols in Pharmacology Supplement 22
7.5.5
Pharmacokinetics
Determination of plasma protein binding
13. For each concentration of each test article, soak two Spectra/Por 4 membrane discs
in a dish of deionized water followed by PBS buffer, as described for the nonspecific
binding determination (step 1).
14. Prepare test article stock solutions as described for ultrafiltration investigations (see
Basic Protocol 1, step 2).
15. As described for ultrafiltration (see Basic Protocol 1, steps 9 to 13), thaw (if
necessary), centrifuge, and pool plasma samples from all species to be tested. Adjust
the pH of the plasma for investigation and retain a portion for determination of total
protein content.
16. Add 25 l of each 100 test article stock solution to 2.475 ml plasma and vortex to
mix.
17. Set up the equilibrium dialysis equipment as described in steps 4 to 8, such that two
cells are prepared for each test article concentration per species. Note the identity of
each cell.
18. Following the duration of dialysis previously determined as necessary to attain
equilibrium (step 11), remove both the dosing and buffer compartments. Note the
volume recovered from each compartment and record any variation >15%.
If a >15% variation in volume recovered from each compartment is noted, the volume shift
may affect the final results. The experiment should be repeated, measuring the final
compartment volumes and accounting for any differences when calculating percentage
bound (Tozer et al., 1983).
19. Determine the concentration of test article in both compartments and in the plasma
dosing solution by an appropriate method.
If the test article is radiolabeled, aliquots of plasma and buffer samples should be subjected
to LSC to determine the test article concentration.
If the test article is not radiolabeled, any plasma samples should be extracted with a suitable
solvent (such as 1 vol acetonitrile) to precipitate and remove the protein prior to analysis
by HPLC with UV, fluorescence, or MS detection. The efficiency of this extraction must be
confirmed, by comparison of the response of an extracted plasma curve to that of a
protein-free buffer curve treated in the same manner.
20a. For reactions not requiring compensation for volume change: Calculate the percent-
age recovery and plasma protein binding not accounting for volume changes
according to the following expressions:
Where Cpe is the test article concentration in the plasma compartment after dialysis,
Cbe is its concentration in the buffer compartment after dialysis, and Cpi is its
concentration in plasma compartment prior to dialysis.
Cpe Cbe
protein binding () 100
Cpe

Cpe Cbe
recovery () 100
Cpi

Supplement 22 Current Protocols in Pharmacology
7.5.6
Determination of
Compound
Binding to
Plasma Proteins
20b. For reactions in which volume change must be accounted: Calculate the percentage
recovery and plasma protein binding, this time accounting for volume changes
according to the following expressions:
Where Vpe is the volume in the plasma compartment after dialysis, Vbe is the volume
of the buffer compartment after dialysis, and Vpi is the volume of plasma loaded into
cell (1 ml). All volumes are expressed in milliliters and all concentrations are
expressed in units per milliliter (e.g., disintegrations per minute, nanomoles, micro-
grams per milliliter).
If the concentration of test article in the buffer compartment is below the limit of detection,
the minimum percentage protein binding may be estimated by substituting the limit of
detection in place of the buffer compartment concentration.
22. If required, analyze a portion of the buffer compartment to determine the nature of
the unbound material.
ALTERNATE
PROTOCOL
DETERMINATION OF PLASMA PROTEIN BINDING BY 96-WELL PLATE
EQUILIBRIUM DIALYSYS
The Dialyzer-96 is a 96-well plate with each well bisected horizontally by a size-selective
membrane. The plate allows multiple equilibrium dialysis investigations to be performed
simultaneously and utilizes less plasma. The determination of equilibration time when
multiple compounds are investigated on the plate is impractical, thus this step is omitted
and the dialysis performed over-night to ensure equilibration is achieved. This method-
ology is not recommended for precise determination of plasma protein binding of
compounds in development, but is useful during compound screening and selection. Note
that plate membranes are compromised upon exposure to liquid and thus cannot be
re-used.
Additional Materials (also see Basic Protocol 2)
Equilibrium Dialyzer-96 plates with 10-kDa MWCO membrane (Harvard
Apparatus)
1. Prepare the test article stock solutions and pooled human plasma as described (see
Basic Protocol 1, step 2 and steps 9 to 13, respectively).
2. Spike 990 l pooled plasma, pH adjusted as described (see Basic Protocol 1, step 14),
with 10 l each 100 test article stock solution. Prepare spiked PBS in the same
fashion.
3. Aliquot 0.25-ml portions of the spiked plasma and PBS into duplicate wells of an
Equilibrium Dialyzer-96 plate with 10-kDa MWCO membrane. Cap the plasma
compartments.
(Cpe Cbe) (Vpe/ Vpi)
protein binding (%) = 100
[(Cpe Cbe) (Vpe/ Vpi)] + Cbe

(Cpe Vpe) + (Cbe Vbe)
recovery (%) = 100
(Cpi Vpi)

Current Protocols in Pharmacology Supplement 18

7.5.7
Pharmacokinetics
4. Invert the plate and aliquot 0.25 ml PBS into the underside of all wells to be
investigated. Cap the buffer compartments.
5. Incubate at 37C with shaking, as per manufacturers instructions.
6. Following 20 hr of incubation, remove the plasma and buffer compartments to clean
96-well plates with a multichannel pipet.
7. Determine the concentration of test article in both compartments as previously
described.
8. Calculate the percentage recovery and percentage protein binding according to the
equations detailed previously (see Basic Protocol 2, step 20a).
REAGENTS AND SOLUTIONS
Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see
APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.
Phosphate-buffered saline (PBS), 0.01 M, pH 7.4
Dissolve the required number of phosphate-buffered-saline tablets (Sigma) in
HPLC-grade water to achieve 0.01 M phosphate buffer, pH 7.4/0.0027 M potassium
chloride/0.137 M sodium chloride. Store up to 3 months at 4C.
COMMENTARY
Background Information
The determination of the plasma protein
binding of a compound is critical for interpre-
tation of the associated pharmacodynamic,
pharmacokinetic, and toxicity profiles. It is
generally accepted that the pharmacological
effects of a drug are regulated by the unbound
(free) concentration in plasma. This view fol-
lows the assumption that the unbound com-
pound in plasma is in equilibrium with that in
the extravascular regions where the drug target
is situated (Sansom and Evans, 1995). The
effect of the fraction of unbound drug on the
pharmacokinetic profile of a compound is in-
fluenced by many parameters, such as the
physicochemical properties of the compound
and the extraction ratios of the metabolizing
organs. Typical effects are summarized as fol-
lows (McEnlay and DArcy, 1983):
1. Absorption is favored if a compound is
highly bound due to the large concentration
gradient between the free compound at the
absorption site and in the plasma.
2. Drug distribution is decreased if a com-
pound is highly protein bound, since bound
drug is not available for excretion into the
tissues, leading to a low apparent volume of
distribution.
3. In the absence of extensive first pass meta-
bolism or active renal clearance, clearance is
decreased if a compound is highly protein
bound.
4. Saturation of binding sites at increased
doses will increase the concentration of un-
bound drug and may result in nonlinear phar-
macokinetics.
The plasma proteins typically involved in
the binding of drug molecules are albumin and
-1 acid glycoprotein. Albumin is the most
abundant plasma protein (concentration in nor-
mal human plasma is 0.6 mM) with several
binding sites and accounts for binding of both
acidic and basic compounds. -1 acid glyco-
protein binds primarily basic compounds and
is present at a concentration of approximately
between 10 and 25 M in normal human
plasma (Sansom and Evans, 1995). Both iso-
lated human proteins are commercially avail-
able, allowing binding parameters to be deter-
mined.
The concentration, structure, and function
of human plasma proteins may be altered ac-
cording to age and disease, resulting in a change
in the degree of binding of compounds (Lindup
and Orme, 1981). Binding can also be affected
by the presence of competing endogenous com-
pounds (Sansom and Evans, 1995). In addition
to these inherent variations, the presence of
another drug may affect plasma protein bind-
ing. While the clinical implications of protein
binding interactions are rarely significant, it is
prudent to investigate the potential for interac-
tion in vitro and analyze the findings in the
context of clearance and therapeutic index (the
Supplement 18 Current Protocols in Pharmacology
7.5.8
Determination of
Compound
Binding to
Plasma Proteins
ratio of the effective dose compared to the toxic
dose). An algorithm for determining the clinical
significance of potential protein binding dis-
placement interactions is presented in Figure
7.5.2 (Rolan, 1994).
The protocols described in this unit provide
simple methods for determining the binding of
a compound to plasma or serum proteins from
the full range of toxicological species and man,
as well as to isolated proteins. Any of the
described techniques can also be used to deter-
mine binding parameters and to investigate the
potential for binding interactions with co-ad-
ministered compounds. A useful explanation of
the theory and limitations of equilibrium dialy-
sis and ultrafiltration is presented in Sebille et
al. (1990).
Critical Parameters
Ultrafiltration
The technique of ultrafiltration (see Basic
Protocol 1) determines only the unbound test
article concentration. The remaining test article
is assumed to be bound to plasma proteins, not
the apparatus. Thus, the presence of nonspecific
binding will lead to an overestimation of the
percent bound to plasma proteins. It is not
advisable to simply subtract the nonspecific
binding determined in the absence of plasma
from the percentage plasma protein bound as
the presence of the plasma may affect the affin-
ity of the test article for the apparatus. If the
nonspecific binding is found to be >5%, an
alternative technique should be used, such as
equilibrium dialysis.
The aim of in vitro measurement is to predict
the situation in the human body following a
dose of the test article. As such, the experiment
should mimic physiological conditions as
closely as possible. Ensure that the centrifuge
and rotor are at 37C prior to spinning ultrafil-
tration units and adjust the pH of the plasma to
pH 7.4. Variations from these physiological
conditions may lead to variable and unpre-
dictive results, as exemplified by propranolol
(Paxton and Calder, 1983).
Ultrafiltration relies on smooth transition of
the plasma water through the membrane. The
existence of an air pocket caused by air bubbles
Table 7.5.1 Troubleshooting Guide to Plasma Protein Binding Experiments
Problem Possible cause Solution
Low percentage recovery Excessive nonspecific binding
to the equipment
Investigate alternative method
of determination. Equilibrium
dialysis is the method of
preference for compounds
which tend to bind
nonspecifically.
Degradation of test article
during experiment
Minimize incubation times. If
using equilibrium dialysis
Poor extraction from plasma
(equilibrium dialysis)
Investigate alternative
extraction procedures
Deviation from literature
values
Alteration in plasma pH Carefully adjust plasma pH to
7.4 prior to spiking
Atypical protein concentration
in plasma
Repeat experiment with a
different batch of plasma
Volume shift during
equilibrium dialysis
Account for volume change in
calculations
Compromised membrane or
filter
Inspect membrane or filter
and replace if necessary
High variation in results Problems with analytical
technique
Optimize analytical technique
Poor mixing of dosing stock Repeat
Equilibrium not achieved
(equilibrium dialysis)
Confirm equilibration at
dialysis time selected
Compromised membrane or
filter
Inspect membrane or filter
and repeat if necessary
Current Protocols in Pharmacology Supplement 18
7.5.9
Pharmacokinetics
Table 7.5.2 Percentage Binding of a Test Article to Plasma Proteins from Various Species
Replicate
Mean protein binding at
10 M (%)
Mean protein binding at
100 M (%)
Buffer
a
1 5.0 4.0
2 4.0 1.0
Mean 4.5 2.5
Mouse
1 69 63
2 88 68
Mean 88 66
Dog
1 87 74
2 84 75
Mean 86 75
Monkey
1 83 67
2 85 67
Mean 84 67
Human
1 81 77
2 82 77
Mean 82 77
a
Percentage nonspecific binding.
Table 7.5.3 Percentage Binding of a Test Article to Plasma Proteins from Various Species
a
Replicate
Protein bound at
10 M (%)
Protein bound at
50 M (%)
Recovery at
10 M (%)
Recovery at
50 M (%)
Buffer
b
1 NA NA 77 84
2 NA NA 79 85
Mean NA NA 78 85
Rat
1 99.9 99.2 100 93
2 99.6 98.7 100 90
Mean 99.7 98.9 100 92
Dog
1 98.5 97.6 99 100
2 98.4 97.1 100 100
Mean 98.5 97.4 100 100
Human
1 99.6 98.9 83 100
2 99.8 98.2 86 100
Mean 99.7 98.5 85 100
a
Abbreviations: NA, not applicable.
b
Percentage nonspecific binding.
Supplement 18 Current Protocols in Pharmacology
7.5.10
Determination of
Compound
Binding to
Plasma Proteins
Is drug interest >90%
protein bound
Does the drug have a
narrow therapeutic index
What is the hepatic
extraction ratio of the drug
Would a transient increase
in free drug concentration
be clinically relevant
Clinically significant
interaction not likely
Is the drug given i.v.
Clinically significant
interaction likely.
Perform a clinical study
to quantify effects.
yes
yes
yes
yes
high
low no
no
no
no
Figure 7.5.2 Algorithm for determining the clinical significance of potential protein binding dis-
placement interactions.
P
r
o
t
e
i
n

b
i
n
d
i
n
g

(
%
)
0
10
20
30
40
50
60
70
80
90
100
Buffer Mouse Rat Dog Monkey Human
Figure 7.5.3 Mean percentage binding of propranolol to plasma proteins from various species
and nonspecific binding (NSB) to the ultrafiltration apparatus (buffer). Doses are 1 (light gray) and
10 g/l (dark gray).
Current Protocols in Pharmacology Supplement 18
7.5.11
Pharmacokinetics
in the unit will hinder this transition. Remove
all air bubbles from the ultrafiltration units prior
to centrifugation to ensure accuracy.
The production of an ultrafiltrate volume
>40% of the plasma volume may cause the
bound compound to dissociate due to a shift in
the equilibrium, leading to an underestimation
of the percent protein bound (Whitlam and
Brown, 1981). If such a large ultrafiltrate vol-
ume is observed, the procedure should be re-
peated, adjusting the centrifugal force as appro-
priate.
The use of radiolabeled test articles eases
the analysis of the ultrafiltrate considerably;
however, any test article-related impurities,
metabolites, or degradants containing the radi-
olabel would be quantified as the test article,
leading to an underestimation of the percent-
age bound. Thus the purity of the free com-
pound should be confirmed by LC/UV or
LC/MS to confirm that it is indeed the test
article that is failing to bind to the plasma
proteins.
Equilibrium dialysis
The dialysis membranes contain preserv-
atives that may interfere with the binding or
analysis of the compounds. Thus, they should
be soaked as described in the protocol to equili-
brate them to dialysis conditions.
The aim of the in vitro measurement is to
predict the situation in the human body follow-
ing a dose of the test article, as such, conditions
during the experiment should mimic physi-
ological conditions as closely as possible. Thus,
adjust the pH of the plasma to pH 7.4 and ensure
that the dialysis is performed at 73C. Vari-
ations from these physiological conditions may
lead to variable and unpredictive results, as
exemplified by propranolol (Paxton and Cal-
der, 1983).
The aim of the equilibrium dialysis is to
reach conditions at which the bound and free
test article are in equilibrium. If the compart-
ments are removed prior to reaching equilibra-
tion, the percent protein bound will be overes-
timated. Thus, the dialysis should be performed
for at least the equilibration time determined in
the initial experiment.
Analysis of the plasma compartment by
LC/UV or LC/MS requires extraction from the
protein-rich matrix. If the extraction of the test
article is variable or incomplete, the calculated
test article concentration and thus the percent
protein binding will be incorrect. The efficiency
of any plasma extraction should be >90% with
a variation of <5% to be acceptable.
A >15% variation in volume recovered
from each compartment may affect the final
results. The experiment should be repeated,
measuring the final compartment volumes and
accounting for any differences when calculat-
ing percent bound (Tozer et al., 1983).
The use of radiolabeled test articles eases
the analysis of the buffer compartment consid-
erably; however, any test article-related impu-
rities, metabolites, or degradants containing the
radiolabel would be quantified as the test arti-
cle. This can lead to an underestimation of the
percent bound. Thus, the purity of the free
compound should be confirmed by LC/UV or
LC/MS to confirm that it indeed the test article
that is failing to bind to the plasma proteins.
Table 7.5.1 presents some common prob-
lems with the assays detailed in this unit, along
with their possible causes and solutions.
Anticipated Results
The degree of binding of compounds to
plasma proteins varies greatlyi.e., anywhere
between <1% and >99%. Representative re-
sults are presented in this section.
Ultrafiltration
The degree of binding of the representative
compound propranolol to mouse, rat, dog,
monkey, and human plasma proteins, and to the
ultrafiltration equipment is presented in Table
7.5.2 and depicted graphically in Figure 7.5.3.
The low degree of nonspecific binding to the
ultrafiltration units (listed as buffer in the
table and figure) confirms that the technique of
ultrafiltration is suitable for this compound.
Binding affinity of propranolol to proteins in
mouse, rat, dog, monkey, and human plasma is
moderately high. The decreased percentage
protein binding at the higher propranolol con-
centrations (in all but mouse plasma) suggests
saturation of the binding sites.
Equilibrium dialysis
The equilibration profile of a proprietary test
article during dialysis is depicted in Figure
7.5.4. The dialysis time subsequently selected
for determination of plasma protein binding of
this test article was 4 hr to guarantee that equili-
bration was achieved.
Example plasma protein binding results
generated by equilibrium dialysis are presented
in Table 7.5.3 and depicted in Figure 7.5.5.
These results illustrate the advantages of both
compartments being readily available for
analysis following equilibrium dialysis. The
compound recovery in the absence of plasma
Supplement 18 Current Protocols in Pharmacology
7.5.12
Determination of
Compound
Binding to
Plasma Proteins
is only 78% at 10 M and 85% at 50 M,
suggesting that nonspecific binding would be
an issue when determining plasma protein
binding. However, in the presence of plasma
the compound recovery is much improved, the
exception being in the presence of human
plasma at 10 M (Table 7.5.3). The percentage
protein binding determined with the reduced
recovery in human plasma is still meaningful,
although the actual concentration of test article
represented is 15% of nominal.
Time Considerations
Ultrafiltration
The number of samples that can be investi-
gated in one day depends on the capacity of the
centrifuge rotor (note that only a single concen-
tric circle should be used in order to standardize
the centrifugal force on each sample). If the
number of samples required to fill the rotor is
considered a batch, then a single batch requires
1 hr to process to the analytical stage. If
incubation of additional batches is initiated
once centrifugation of the previous batch has
0
10
20
30
40
50
60
70
80
90
100
Rat Dog Human
P
r
o
t
e
i
n

b
i
n
d
i
n
g

(
%
)
Figure 7.5.5 Mean percentage binding of a test article to plasma proteins from various species
as determined by equilibrium analysis. Doses are 10 (light gray) and 50 M (dark gray).
0
0 1 2 3 4 5 6 7
20
40
60
80
100
120
140
160
T
e
s
t

a
r
t
i
c
l
e

c
o
n
c
e
n
t
r
a
t
i
o
n

(
n
g
/
m
l
)
Duration of dialysis (hr)
Figure 7.5.4 Equilibration time of a test article during dialysis of fortified PBS. Concentration of
article is measured in both the dosing (circles) and buffer (triangles) compartments.
Current Protocols in Pharmacology Supplement 18
7.5.13
Pharmacokinetics
commenced, the time required per batch is
reduced by 30 min. Nonspecific binding of the
test article to the equipment should be deter-
mined prior to protein binding determinations.
Traditional equilibrium dialysis
The initial nonspecific binding and equili-
bration time determination requires 1 hr for
set-up and 20 min at each sampling point. Each
dialyzer has a capacity of 20 cells, thus two test
articles can be investigated utilizing a single
dialyzer (each test article investigated at five
time points in duplicate).
The total time required for the plasma pro-
tein binding determination is dependent on the
required equilibration time. The set up, includ-
ing test article formulation and plasma spiking,
requires 2 hr, while terminal sampling re-
quires 1 hr.
96-well-plate equilibrium dialysis
Set-up of the 96-well plate requires 2 hr
including spiking of the plasma. The dialysis
lasts 20 hr and terminal sampling requires only
20 min.
Literature Cited
Lindup, W.E. and Orme, M.C. 1981. Plasma protein
binding of drugs. Br. Med. J. 282:212-214.
McEnlay, J.C. and DArcy, P.F. 1983. Protein bind-
ing displacement interactions and their clinical
importance. Drugs 25:495-513.
Paxton, J.W and Calder, R.L. 1983. Propranolol
binding in serum: Comparison of methods and
investigation of effects of drug concentration,
pH, and temperature. J. Pharmacol. Meth. 10:1-
11.
Rolan, P.E. 1994. Plasma protein binding displace-
ment interactions why are they still regarded as
clinically important? Br. J. Clin. Pharmacol.
37:125-128.
Sansom, L.N. and Evans, A.M. 1995. What is the
true clinical significance of plasma protein bind-
ing displacement interactions? Drug Safety
12:227-233.
Sebille, B., Zini, R., Madjar, C.V., Thaud, N., and
Tillement, J.P. 1990. Separation procedures used
to reveal and follow drug-protein binding. J.
Chromatogr. 531:51-77.
Tozer, T.N., Gambertoglio, J.G., Furst, D.E., Avery,
D.S., and Holford, N.H.G. 1983. Volume shifts
and protein binding estimates using equilibrium
dialysis: Application to prednisolone binding in
humans. J. Pharm. Sci. 72:1442-1446.
Whitlam, J.B. and Brown, K.F. 1981. Ultrafiltration
in serum protein binding determinations. J.
Pharm. Sci. 70:146-150.
Contributed by Natasha Dow
Ricerca LLC
Concord, Ohio
Supplement 18 Current Protocols in Pharmacology
7.5.14
Determination of
Compound
Binding to
Plasma Proteins