P-Glycoprotein Activity in Human Caucasian Male

Lymphocytes Does Not Follow its Increased

Expression During Aging

Vania Vilas-Boas,1* Renata Silva,1 A. Rita Gaio,2 Ana Margarida Martins,1 Sofia Costa Lima,3

Anabela Cordeiro-da-Silva,3,4 Maria de Lourdes Bastos,1 Fernando Remia~o1*

� AbstractP-glycoprotein (P-gp) is a transmembrane protein that mediates the efflux of innu-merous structurally unrelated compounds. It was initially found over-expressed in tu-mor cells, associated to a multidrug resistance phenotype (MDR). Then, P-gp wasfound constitutively expressed in excretory cells/tissues and in circulating cells, suchas lymphocytes. Considering the importance of this transporter in the establishmentof therapeutic protocols and the existence of contradictory results, this study aimedat evaluating the influence of aging in the expression and function of P-gp in humanlymphocytes, comparing two different methodologies to assess both parameters. P-gpactivity and expression were evaluated in lymphocytes isolated from whole bloodsamples of 65 healthy caucasian male donors, divided into two groups according toage (group 1: under 30-years old; group 2: above 60-years old). P-gp expression wasassessed using the anti-P-gp monoclonal antibody, UIC2, in the presence and in ab-sence of vinblastine (Vbl). P-gp activity was evaluated measuring the efflux rate of thefluorescent P-gp substrate rhodamine 123 (Rho 123) and also using UIC2 shift assay.Flow cytometric analysis was performed to assess all the proceedings. Furthermore, P-gp expression and each of the P-gp activity determination methods were compared,through correlation analysis and linear regression models. We observed a significantage-dependent increase in mean P-gp expression (p ¼ 0.029), which was not reflectedin the transporter’s activity (p > 0.050). Statistical analysis allowed selection of UIC2shift assay over Rho 123 efflux assay as a more selective method to assess P-gp activity.Despite the significant correlation between P-gp expression and P-gp activity foundin lymphocytes (Gp1(group 1)��r ¼ 0.609, p < 0.001; Gp2��r ¼ 0.461, p ¼ 0.012),using UIC2 shift assay, these data reinforce the need for P-gp activity assessment,rather than P-gp expression determination alone, when starting new therapeutic regi-mens with P-gp substrates, especially in men older than 60 years of age. ' 2011 Inter-

national Society for Advancement of Cytometry

� Key termsP-glycoprotein; aging; UIC2 shift assay; Rho 123 efflux; flow cytometry

P-GLYCOPROTEIN (P-gp), the product of multidrug-resistance 1 (MDR-1) gene in

humans, is one of the most important and interesting members of the numerous,

highly conserved, widespread membrane transporters of the ATP-binding cassette

superfamily. It was first detected as a surface phosphoglycoprotein over-expressed in

Chinese hamster ovary cultured cells selected for MDR, where it mediated resistance

to many amphipathic drugs (1). Then, P-gp was also found constitutively expressed

in epithelial cells from tissues with excretory function such as intestine, liver, kidneys,

lungs, blood–brain barrier, and placenta, and also in blood cells such as lymphocytes

(2), where P-gp expression may limit tissue penetration of its substrates. Thus, P-gp

plays an important role in pharmacokinetics and pharmacodynamics of many

1REQUIMTE��Laboratorio de Toxicologia,Departamento de Ciencias Biologicas,Faculdade de Farmacia, Universidade doPorto, Rua An�ıbal, Cunha, 164, 4099-030Porto, Portugal2Departamento de Matematica,Faculdade de Ciencias, Universidadedo Porto, Rua do Campo Alegre, 687,4169-007 Porto, Portugal3IBMC��Instituto de Biologia Molecular eCelular, Universidade do Porto, Rua doCampo Alegre, 823, 4150-180 PortoPortugal4Departamento de Ciencias Biologicas,Laboratorio de Bioqu�ımica, Faculdadede Farmacia, Universidade do Porto, RuaAn�ıbal Cunha, 164, 4099-030 Porto,Portugal

Received 25 March 2011; RevisionReceived 30 June 2011; Accepted 4August 2011

Additional Supporting Information may befound in the online version of this article.

Grant sponsor: Fundac~ao para a Cienciae Tecnologia (FCT); Grant number: PTDC/SAU-OSM/101437/2008; SFRH/BD/29559/2006; SFRH/BPD/37880/2007; Grant spon-sor: COMPETE - Operational Programmefor Competitiveness Factors.

*Correspondence to: Vania Vilas-Boas,REQUIMTE��Laboratorio deToxicologia, Departamento de CienciasBiologicas, Faculdade de Farmacia,Universidade do Porto, Rua An�ıbal,Cunha, 164, 4099-030 Porto, Portugal.E-mail: v_vilasboas@yahoo.com or

Original Article

Cytometry Part A � 79A: 912�919, 2011

compounds due to its widespread expression and its large list

of therapeutic substrates (3).

The P-gp substrates’ group includes components of natu-

rally occurring grape and citrus fruit juice (4), pesticides (5),

fluorescent dyes (e.g., rhodamine 123) and also many thera-

peutic agents such as HIV protease inhibitors (e.g., ritonavir),

antiarrhythmic drugs (e.g., digoxin), steroids (e.g., dexameth-

asone), antibiotics (e.g., rifampicin, erythromycin), immuno-

suppressants (e.g., cyclosporin A), anticancer drugs (e.g., vin-

blastine, doxorubicin, etoposide), among others (6). This ran-

dom efflux transport is, in fact, one of the most striking

features about this protein. Some compounds may act as P-gp

inhibitors and/or inducers, including many of its substrates.

P-gp inhibitors have been therapeutically used as a means to

enhance accumulation of P-gp substrates in cancer cells, thus

reverting the MDR phenotype associated to over-expression of

P-gp (7,8). On the other hand, P-gp induction was proposed

by our group as an antidotal pathway for P-gp substrates’

induced toxicity, protecting cells from the toxic side effects of

poisons (e.g., paraquat) (9,10).

Aging is a complex process associated with progressive

and significant immunological alterations in humans (11).

Previous studies of the evolution of P-gp expression and activ-

ity lead to contradictory conclusions. In fact, while Aggarwal

et al. observed an elevation both in P-gp expression and activ-

ity in aging T lymphocytes (12), Brenner and Klotz found no

age-related difference in P-gp activity (13) and other authors

described diminished P-gp function in aging lymphocytes

(14,15). Further studies are definitely required to fully answer

this question. As the elderly are often simultaneously sub-

mitted to more than one drug, many of them with capacity to

influence P-gp activity, it is important to undoubtedly estab-

lish whether aging also brings about differences in the expres-

sion and/or activity of such an influential transporter.

The aim of this work was to evaluate the influence of

aging in the expression and function of P-gp in human male

lymphocytes and also to compare two P-gp expression and ac-

tivity assessment techniques. For these purposes, blood sam-

ples were collected from 65 healthy Caucasian men, separated

into two groups according to their age: Group 1 (from 18- to

30-years old) and Group 2 (from 60- to 78-years old), and

then analyzed to assess P-gp expression and activity in lym-

phocytes. Flow cytometry was the selected technique for the

development of this study, as consensually recommended

(16). P-gp expression was evaluated using the monoclonal

antibody against the extracellular conformational epitope of

P-gp, UIC2, in the presence and absence of vinblastine (Vbl).

P-gp activity was assessed both by the widely used rhodamine

123 (Rho123) efflux assay and also by the UIC2 shift assay.

The former assay measures the efflux rate of P-gp fluorescent

substrate Rho 123 while the latter one is based on the UIC2

shift phenomenon (i.e., increased UIC2 reactivity in the pre-

sence of P-gp substrates under physiologic conditions) to

simultaneously evaluate P-gp expression and function (17,18).

MATERIALS AND METHODS

Materials

AccuGENE1 (13 PBS buffer) and Williams E cell culture

media were obtained from Lonza Laboratories (Verviers, Bel-

gium). Fetal bovine serum (FBS) was obtained from Gibco Lab-

oratories (Lenexa, KS). Methyl violet and glacial acetic acid (to

prepare Turk’s solution) were obtained from Merck KGaA,

(Darmstadt, Germany). Histopaque1-1077, rhodamine 123,

vinblastine (Vbl), cyclosporine A (CyA), and sodium azide were

obtained from Sigma-Aldrich (St. Louis, MO). Flow cytometry

reagents, Facs Flow (BD FacsFlowTM) and Facs Clean, were

obtained from Becton, Dickinson and Company, (San Jose,

CA). Propidium Iodide (Pi) was from BD Biosciences, Pharmin-

gen (San Jose, CA). UIC2 monoclonal antibody (mouse IgG2a)

conjugated with phycoerythrin (PE) was obtained from Beck-

man Coulter (Fullerton, USA). IgG2a (negative monoclonal

antibody control to UIC2) conjugated with PE was obtained

from ImmunoTools GmbH (Friesoythe, Germany).

All other reagents used were of analytical grade or of the

highest available grade.

Blood Samples

Sixty-five human blood samples were separated into two

groups according to age: Group 1 (34 caucasian healthy men

aged between 18- and 30-years old) and Group 2 (31 Cauca-

sian healthy men aged between 60- and 78-years old). All sam-

ples were collected into tubes containing heparin or EDTA as

anticoagulants. Informed consent was obtained from all the

volunteers in this study.

Isolation of Peripheral Blood Mononuclear Cells

(PBMC)

PBMC were isolated by density gradient using Histopa-

que-1077, according to manufacturer’s instructions.

Whole blood was carefully layered onto the same volume

of Histopaque-1077 previously kept at room temperature. Af-

ter centrifuging the tubes at 650g for 10 min, at room temper-

ature, the opaque interface containing mononuclear cells was

collected and washed twice with PBS. Cells were then diluted

in Turk’s solution (0.1% methyl violet and 1% glacial acetic

acid) and counted using a Neubauer-counting chamber.

Aliquots of 250,000 cells were made in PBS with 2% FBS

and kept overnight at 48C.

Fernando Remi~ao, REQUIMTE, Toxicology Department, Faculty ofPharmacy, University of Porto Rua Anibal Cunha 164, 4099-030Porto, Portugal.E-mail: remiao@ff.up.pt

Published online 8 September 2011 in Wiley Online Library(wileyonlinelibrary.com)DOI: 10.1002/cyto.a.21135

� 2011 International Society for Advancement of Cytometry

ORIGINAL ARTICLE

Cytometry Part A � 79A: 912�919, 2011 913

UIC2 Shift Assay

All analyzed parameters were previously tested and opti-

mized. Briefly, PBMC aliquots (250,000 cells/tube) were cen-

trifuged, then suspended in PBS plus 2% FBS, and allowed to

equilibrate for 10 min at 378C in a water bath. Cell aliquots

were submitted to a 10 min incubation period with Vbl (22

lM) or control diluent (PBS) at 378C and subsequently incu-

bated with UIC2 antibody for 30 min, at the same tempera-

ture (UIC2 staining in the presence of vinblastine vs. UIC2

conventional staining) (18). An anti-mouse antibody IgG2a,

the isotype control conjugate, was used in the place of UIC2 as

a negative control to trace unspecific binding. PBMC were

then washed twice with ice-cold PBS plus 2% FBS with 0.01%

sodium azide (Shift Stop Buffer, SSB). The samples were kept

on ice and in the dark until flow cytometric analysis.

Rhodamine 123 Efflux Assay

All parameters were previously tested and optimized

in preliminary titration experiments, including selection of

P-gp inhibitor and respective concentrations, Rho 123

loading concentration and accumulation/efflux period.

CyA was chosen as P-gp selective inhibitor on the basis of

previous results (19). Rho 123 and CyA cytotoxicity (evalu-

ated after a 30 min incubation period, the time necessary

for Rho 123 accumulation) were determined by the Pi

incorporation assay (data not shown). After optimization

of the assay conditions, PBMC were exposed to Rho 123, in

the presence of CyA, allowing its maximum accumulation

within the cells (Ai). Then, P-gp activity was assessed

comparing Rho 123 extrusion capacity in a Rho 123 free

medium (AiE, inhibited accumulation followed by efflux).

In more detail, to perform the CyA inhibited Rho 123

accumulation assay (Ai, inhibited accumulation), the fluores-

cent substrate was loaded into cells at a final concentration of

0.65 lM in PBS for 30 min at 378C, in the dark, in the pre-

sence of 10 lM CyA to prevent Rho 123 efflux at this stage. Af-

ter this accumulation period, cells were washed twice with

PBS plus 2% FBS and centrifuged at 650g for 10 min at 48C.Cells were then suspended in ice cold PBS and analyzed by

flow cytometry to determine the loading Rho 123 fluores-

cence.

To perform the Rho 123 efflux protocol (AiE), cells were

initially allowed to accumulate the fluorescent P-gp substrate

as described above (Ai). Subsequently, PBMC were resus-

pended in warmed Williams E growth medium (without pH

indicator) and enabled to efflux Rho 123 for 45 min at 378C,in the dark. After this efflux period, cells were washed with

PBS plus 2% FBS and centrifuged at 650g for 10 min at 48C.Flow cytometry analysis was promptly performed to assess the

fluorescence emission of remaining intracellular Rho 123.

Flow Cytometry

According to consensus recommendations presented in

Beck et al., P-gp expression values shall be established as Mean

Fluorescence Intensity (MFI) values or percentage of P-gp1

cells (16). Therefore, these parameters were determined for

each sample using a Becton Dickinson FACSCaliburTM flow

cytometer (Becton Dickinson, Mountain View, CA) equipped

with a 488 nm argon ion laser. Samples were injected at a flow

rate of 60 � 7 lL min21 and acquired as listmode files (FCS

files). To restrict the analysis to lymphocytes only, a polygon

gate was set according to their light scattering properties (in a

forward vs. side scatter plot) excluding cell debris and other

blood cells. Dead cells were also excluded from the analysis

according to their forward and side light scatters and the Pi (5

lg mL21) incorporation, based on the acquisition of data for

10 000 cells. Fluorescence signals were collected in logarithmic

mode and the cytofluorimetric data were analyzed by the

BDIS CellQuest Pro software (Becton Dickinson, NJ).

Fluorescence due to the R-phycoerythrin (PE) bound to

UIC2 antibody [fluorescent label recommended by consensus

guidelines (16)] was followed in channel 2 (FL2) and plotted

as a histogram of FL2 staining. Background correction was

performed in all cases subtracting the MFI values of the iso-

type control. For the evaluation of P-gp1 cells, a marker was

set considering the isotype control curve for each sample as P-

gp2 cells (see Fig. 1). P-gp activity in lymphocytes assessed by

UIC2 shift assay was defined as UIC2 shift calculated through

the difference in MFI of UIC2 binding in the presence and in

the absence of vinblastine (UIC2 shift), corrected for the re-

spective MFI values for isotype control, as follows [Eq. (1)]:

UIC2 shift ¼ ðMFI of P-gpþ cell UIC2þ VbIÞ�ðMFI of P-gpþ cells UIC2Þ ð1Þ

The green fluorescence due to Rho 123 was followed in channel

1 (FL1) and plotted as a histogram of FL1 staining. P-gp func-

tion assessed by Rho 123 efflux assay was expressed as ‘‘percent-

age of pumped Rho 123,’’ through the following Eq. (2):

% pumped Rho 123 ¼ MFI accumulated Rho 123 ðAiÞ �MFI remaining Rho 123 ðAiEÞMFI accumulated Rho 123 ðAiÞ 3100 ð2Þ

Statistical Analyses

Initial sample size consisted of 65 healthy Caucasian

men. Three individuals with negative results were

excluded from the analysis concerning P-gp activity for

UCI2 shift assay (Group 1: 33 individuals, Group 2: 29

individuals). One outlier individual was excluded from

the analysis concerning P-gp activity for Rho 123 efflux

assay (Group 1: 33 individuals, Group 2: 31 individuals).

All remaining analyses comprised the initially collected

samples.

Empirical distributions of data obtained through MFI

and through percentage of P-gp1 cells were relatively

ORIGINAL ARTICLE

914 Age-Related P-gp Expression and Activity

symmetric, therefore summarized graphically as mean

(1 S.E.M.) and compared or inferred alone through para-

metric statistical methodologies. In particular, correlation

and linear modeling techniques were used to evaluate the

magnitude of the association between the two assessment

methods.

Independent t tests were used to compare mean responses

across age groups for each treatment (with/without Vbl), and

paired t tests were used to compare mean responses across

treatments for each age group.

Significance level was fixed at 0.05. Statistical analyses

were performed by SPSS (PASW Statistics 18.0) and GraphPad

Prism Software v.5.01 (GraphPad Software, San Diego, CA).

RESULTS

UIC2 Conventional Staining

We observed an age-related significant increase in the

amount of detected P-gp, which is expressed by the respective

increase in MFI of UIC2 (p ¼ 0.014) (Fig. 2a). A significant

increase with age was also found in the percentage of P-gp

expressing cells (p ¼ 0.020; Fig. 2b).

UIC2 Staining in Presence of Vinblastine

In our study, an increase in MFI of UIC2 binding and in

the percentage of P-gp1 detected cells was observed, in the

presence of Vbl, for all tested samples, Figure 1a, as previously

suggested (17,18). The histogram overlays in Figure 1 are rep-

resentative of each age group and suggest a heterogeneous P-

gp expression, especially when the assay is performed in the

presence of Vbl. Binding of the IgG2a monoclonal antibody

(isotype control for UIC2 antibody) was not affected by the

presence of Vbl (Fig. 1b).

In the presence of Vbl, the MFI of UIC2 binding

remained significantly higher for Group 2 (p ¼ 0.029), as the

results from the conventional assay indicated (Fig. 2a). How-

ever, no significant differences in the percentage of P-gp1 cells

were observed across age groups in the same experimental

conditions (p ¼ 0.458; Fig. 2b).

The results from MFI of UIC2 binding in the presence of

Vbl were considered P-gp expression results.

UIC2 shift assay also enabled the assessment of P-gp ac-

tivity [UIC2 shift��Eq. (1)]. Using this methodology, no age-

related differences in P-gp activity were found (p ¼ 0.656), as

shown in Figure 3a.

Rhodamine 123 Efflux Assay

P-gp activity was measured as the ‘‘percentage of pumped

Rho 123,’’ calculated by the previously given Eq. (2). The

results are graphically summarized in Fig. 3b) and they indi-

cated the existence of no age-related significant differences in

P-gp activity between the groups (p ¼ 0.100). These results are

in accordance with those obtained using the UIC2 shift assay.

Comparative Analysis on Two P-gp Activity

Assessment Methods

Graphical analyses showed no marked asymmetries in the

empirical distribution of each of the two P-gp activity assess-

ment methods described above. The sample estimate for Pear-

son’s correlation coefficient between the two parameters was

low and not significant (available as Supporting Information;

r ¼ 0.05, p ¼ 0.709). No valid linear regression model was

Figure 1. Representative histograms of UIC2 binding in the absence or in the presence of vinblastine and isotype control. (a) Histogram

overlay for Groups 1 and 2 showing isotype control (black filled), UIC2_PE (dark grey) and UIC2_PE1Vbl (light grey) curves. A marker wasset to establish P-gp1 cells considering the isotype control curve as P-gp2 cells. (b) Isotype control binding (IgG2a_PE; black filled) was not

affected by the presence of Vbl (IgG2a_PE1Vbl; light grey).

ORIGINAL ARTICLE

Cytometry Part A � 79A: 912�919, 2011 915

obtained, even after transformation of variables and/or adjust-

ment for (and interaction with) age.

P-gp Expression/Activity Correlation

The latter results indicated the absence of a significant

linear association between the two P-gp activity determination

methods. To indirectly perform such evaluation, we studied

the association between each P-gp activity assessment method

and P-gp expression previously defined as MFI values of UIC2

binding in the presence of vinblastine.

Using the UIC2 shift assay, a significant correlation

between P-gp expression and P-gp activity was found for each

age group (Gp1��Pearson r ¼ 0.609, p < 0.001; Gp2��Pearson

r ¼ 0.461, p ¼ 0.012 – Fig. 5a) and a significant regression

model could be fitted (F ¼ 7,783, p ¼ 0.000), accounting for

an interaction between the expression and the age group. On

the other hand, when P-gp activity was assessed using the Rho

123 efflux assay, no significant correlation between the two pa-

rameters was identified in neither of the groups (Fig. 5b) and

the linear model applied to the previous situation was not

valid (F ¼ 2.607, p ¼ 0.060).

DISCUSSION

The establishment of P-gp expression/activity evolution

during senescence, which is usually associated with prolonged

therapeutic use of some P-gp substrates, may be a valuable

tool for the correct use of these agents, intensifying their bene-

fits and preventing adverse reactions. Previously published

conflicting results concerning P-gp expression/activity altera-

tions in lymphocytes associated with the aging process (12–

15) require further evaluation about this matter. For that pur-

pose, accurate techniques to assess both P-gp expression and

activity are required. In this study, two such different techni-

ques were used and compared. Lymphocytes were chosen

because they express significantly more P-gp than other leuko-

cytes (20). Within lymphocytes, P-gp expression seems to

change between different subtypes (20) and, thus, we decided

to study this population as a whole (instead of its subtypes

independently) to understand the global impact of such differ-

ences. To our knowledge, this is the first study that aimed at

Figure 3. P-gp activity in lymphocytes. The transporter’s functionality was assessed by (a) the UIC2 shift assay, which allowed evaluation

of P-gp activity levels as the shift detected in UIC2 staining in the presence and in the absence of Vbl [see Materials and methods section,

Eq. (1)], and (b) the Rho 123 efflux assay. The graphs represent mean (1S.E.M.) of the calculated values. Independent samples t test: ns,not significant for differences across age groups.

Figure 2. Detection of P-gp by UIC2 staining in lymphocytes

conventional staining vs. staining in the presence of vinblastine.

Results refer to mean (1S.E.M.) of (a) Mean Fluorescence Inten-sity (MFI) values corrected for isotype control signals, and (b) % of

P-gp1 cells. Mean MFI of UIC2 binding significantly differed

across age groups, independently of the presence or absence of

Vbl (p ¼ 0.029 and 0.014, respectively); average percentage of P-

gp1 cells significantly differed across groups only in the absence

of Vbl (p ¼ 0.020). Paired t-test *** p ‹ 0.001, for differences

between UIC2 and UIC21Vbl within the same age group; Inde-pendent t test x p ‹ 0.05 and ns, not significant, for differencesacross age groups in UIC2 conventional staining and in UIC21Vblstaining. White bars for Group 1; Black bars for Group 2.

ORIGINAL ARTICLE

916 Age-Related P-gp Expression and Activity

comparing UIC2 shift assay to Rho 123 efflux assay in healthy

Caucasian male lymphocytes.

UIC2 antibody specifically recognizes an extracellular

conformational epitope of human P-gp. This antibody has

been extensively used to evaluate P-gp expression in cells and

tissues (10,18,21–23). The results of the UIC2 conventional

staining showed a significant increase in MFI of UIC2 binding

and in the percentage of P-gp1 cells for group 2 (Fig. 2), sug-

gesting that, in the older group, more cells expressed more P-

gp. In the presence of Vbl, although the age-related difference

in the percentage of P-gp1 cells was dissipated, there remained

a significant increase in the MFI of UIC2 binding for group 2.

This age-related increase in P-gp expression has already been

found in T lymphocytes from normal adults (12). The pre-

sence of Vbl significantly improved the detection of both sig-

nals, for all tested samples (Figs. 1 and 2), inducing a confor-

mational change, which enabled the detection of some P-gp

molecules that were previously inaccessible to UIC2 antibody

(18). Although Vbl is known to induce P-gp (24), this possi-

bility was refused considering the short incubation period

with Vbl (40 min). On the basis of these observations, we

recommend P-gp expression assessment with UIC2 conforma-

tional sensitive antibody to be performed in the presence of

Vbl, to increase the technique’s sensitivity, tracking the other-

wise ‘‘hidden’’ P-gp molecules. Moreover, P-gp expression eva-

luation must be interpreted on the basis of MFI values, instead

of percentage of P-gp1 cells, because they unequivocally, in

the presence of Vbl, describe the rate of bound antibody and,

consequently, are proportional to the amount of P-gp mole-

cules. The existence of no significant differences in the per-

centage of P-gp1 cells associated to a significant increase in P-

gp expression is suggestive of differences in protein distribu-

tion between the groups, i.e., the elderly seem to express more

P-gp molecules per cell.

P-gp function was assessed both by UIC2 shift assay and

by the extensively used Rho 123 efflux assay. To our knowl-

edge, it is the first time a comparison is drawn between such

techniques using healthy Caucasian male lymphocytes. The

sensitivity of the Rho 123 efflux functional test has been pre-

viously demonstrated in clinical samples (25), it has been

extensively performed to evaluate P-gp functionality

(10,14,15,26–29) and Rho 123 is one of the consensually

recommended probes for P-gp activity determination (16).

These were the basis that supported our interest in using this

technique. The analysis of single parameter FL1 histograms

allowed the observation of two lymphocyte populations for all

tested samples, both prior to and after the efflux phase, sug-

gesting heterogeneous pumping activity, that is, different

capacities of accumulating and extruding Rho 123 (Fig. 4).

This suggestion is in accordance with the fact that P-gp

expression seems to be, itself, also heterogeneous, as shown in

Figure 1a. P-gp activity was calculated applying the previously

presented Eq. (2), considering both populations as a whole.

No significant differences were found between the groups in

the percentage of pumped dye (Fig. 3b). Additionally to Rho

123 efflux assay, P-gp activity was also assessed by UIC2 shift

assay, which was proposed as the most selective methodology

to quantify P-gp function (30). The UIC2 shift assay allows for

simultaneous and selective quantification of P-gp expression

and functionality from the background of other efflux trans-

porters that may be present. This assay uses an antibody

against a conformational-sensitive epitope of P-gp (UIC2) in

the presence of a conformation-change-inducer substrate, like

Vbl (18). P-gp activity was, therefore, defined as the difference

Figure 4. Representative histograms of Rhodamine 123 Efflux Assay. The figure joins representative sample histograms exhibiting two

cell populations, after Rho 123-loading and after Rho 123-efflux, for each analyzed group.

ORIGINAL ARTICLE

Cytometry Part A � 79A: 912�919, 2011 917

(shift) observed in MFI in the presence and in the absence of

Vbl [Eq. (1)]. Applying this methodology, we found no signif-

icant differences in P-gp activity with aging. The results for

both methodologies were, therefore, agreeing and indicated

the absence of age-related differences in P-gp activity (Fig. 3).

This observation, along with the age-related increase in P-gp

expression, is suggestive of an age-dependent accumulation of

nonfunctional P-gp. Therefore, this rise in the expression of

the transporter may be seen as a defense mechanism to pre-

vent the loss of P-gp activity associated to aging process.

We then proceeded with a comparative study on both P-

gp activity assessment methods, to verify their association and

to establish the best methodology to evaluate P-gp function in

the analyzed samples. The analysis suggested that this correla-

tion was poor and not significant (available as Supporting In-

formation r ¼ 0.045, p ¼ 0.728), leading to the study of the

correlation between the results of each P-gp activity determi-

nation method and the P-gp expression established results.

The UIC2 shift assay results for P-gp activity significantly cor-

related with P-gp expression results for both age groups. Fit-

ting the data with a linear model that accounted for an inter-

action between the age groups and the expression values,

resulted in two lines with distinct slopes (though not signifi-

cantly different among them); the increase of the P-gp activity

with P-gp expression was shown to be only slightly higher in

Group 1 than in Group 2 (Fig. 5a). On the other hand, the

Rho 123 efflux assay results did not significantly correlate with

P-gp expression results, in neither of the groups. Conse-

quently, the linear model analogous to the one previously

described was not statistically significant and no valid slopes

could be estimated. Intercepts for each group essentially

reflected mean values for P-gp activity suggesting that high P-

gp activity levels were being measured for P-gp expression

levels around zero (Fig. 5b), which indicated background

transport of the probe. It is possible that this additional efflux

may be mediated by the Multidrug Resistance Protein-1

(MRP-1), which has been reported to also transport this dye

(31). These observations support the fact that Rho 123 efflux

assay is a poorly selective method to assess low levels of P-gp

activity and shall be replaced by a more selective method like

UIC2 shift assay. Previous studies on the correlation between

P-gp expression and activity have reached conflicting results.

In fact, while some authors found no significant association

between P-gp expression and function, both in healthy volun-

teers (28) and leukemic patients (32), a significant correlation

between P-gp expression and function has been described in

recently diagnosed acute leukemia patients (33), in a blood

brain barrier cell line (29) and also in circulating cells of

healthy volunteers (21). These contradictory observations,

along with the lack of specific functional probes for P-gp func-

tionality assessment have led others to suggest the determina-

tion of P-gp expression as a predictor of a drug’s therapeutic

success (28). To the best of our knowledge, this correlation has

never been studied in aging individuals. Even though the

authors and others (21) found a significant correlation

between both parameters in healthy Caucasian men, inde-

pendently of age, we recommend the implementation of thera-

peutic regimens with P-gp substrates to be made on the basis

of P-gp activity (and expression) evaluation by means of more

selective methods, such as the UIC2 shift assay.

Figure 5. The interaction between P-gp expression and P-gp activity for each group. The values of P-gp activity for each sample deter-

mined (a) by UIC2 shift assay, and (b) by Rho 123 efflux assay, were plotted against respective P-gp expression values. The best-fitted lines

were represented in the figure, according to the equation activity ¼ group 1 group 3 expression. Correspondent model parameters and

Pearson correlation coefficient estimates are presented in the table.

ORIGINAL ARTICLE

918 Age-Related P-gp Expression and Activity

In conclusion, it is important to consider changes on

some actual procedures, which involve P-gp expression deter-

mination alone as a prognostic tool to implement therapeutic

treatment with P-gp substrates. In fact, though a significant

correlation was found between P-gp expression and activity,

the significant increase in P-gp expression with aging was not

followed by a modification in P-gp activity, which may indi-

cate lack of necessity to adjust drug dosage. Recently devel-

oped techniques, such as UIC2 shift assay, allow simple, rapid,

and sensitive simultaneous determination of P-gp expression

and P-gp activity in lymphocytes, making it easier to investi-

gate the clinical aspects of P-gp.

ACKNOWLEDGMENTS

The authors are grateful to Dra Barbara Duarte, Dra Laura

Pereira, and Dr Miguel Soares for the collection of blood samples

and to Dra Maria Luis for technical support with flow cytometry.

LITERATURE CITED

1. Juliano RL, Ling V. A surface glycoprotein modulating drug permeability in Chinesehamster ovary cell mutants. Biochim Biophys Acta 1976;455:152–162.

2. Thiebaut F, Tsuruo T, Hamada H, Gottesman MM, Pastan I, Willingham MC. Cellu-lar localization of the multidrug-resistance gene product P-glycoprotein in normalhuman tissues. Proc Natl Acad Sci USA 1987;84:7735–7738.

3. McDevitt CA, Callaghan R. How can we best use structural information on P-glyco-protein to design inhibitors? Pharmacol Ther 2007;113:429–441.

4. Varma MV, Ashokraj Y, Dey CS, Panchagnula R. P-glycoprotein inhibitors and theirscreening: A perspective from bioavailability enhancement. Pharmacol Res2003;48:347–359.

5. Schinkel AH, Smit JJ, van Tellingen O, Beijnen JH, Wagenaar E, van Deemter L, MolCA, van der Valk MA, Robanus-Maandag EC, te Riele HP, et al. Disruption of themouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrierand to increased sensitivity to drugs. Cell 1994;77:491–502.

6. Fromm MF. The influence of MDR1 polymorphisms on P-glycoprotein expressionand function in humans. Adv Drug Deliv Rev 2002;54:1295–1310.

7. Benson AB III, Trump DL, Koeller JM, Egorin MI, Olman EA, Witte RS, Davis TE,Tormey DC. Phase I study of vinblastine and verapamil given by concurrent IV infu-sion. Cancer Treat Rep 1985;69:795–799.

8. Gruber A, Bjorkholm M, Brinch L, Evensen S, Gustavsson B, Hedenus M, Juliusson G,Lofvenberg E, Nesthus I, Simonsson B, et al. A phase I/II study of the MDR modulatorValspodar (PSC 833) combined with daunorubicin and cytarabine in patients withrelapsed and primary refractory acute myeloid leukemia. Leuk Res 2003;27:323–328.

9. Dinis-Oliveira RJ, Remiao F, Duarte JA, Ferreira R, Sanchez Navarro A, Bastos ML,Carvalho F. P-glycoprotein induction: An antidotal pathway for paraquat-inducedlung toxicity. Free Radic Biol Med 2006;41:1213–1224.

10. Silva R, Carmo H, Dinis-Oliveira R, Cordeiro-da-Silva A, Lima SC, Carvalho F, deLourdes Bastos M, Remiao F. In vitro study of P-glycoprotein induction as an anti-dotal pathway to prevent cytotoxicity in Caco-2 cells. Arch Toxicol 2010;85:315–326.

11. Miller RA. The aging immune system: Primer and prospectus. Science 1996;273:70–74.

12. Aggarwal S, Tsuruo T, Gupta S. Altered expression and function of P-glycoprotein(170 kDa), encoded by the MDR 1 gene, in T cell subsets from aging humans. J ClinImmunol 1997;17:448–454.

13. Brenner SS, Klotz U. P-glycoprotein function in the elderly. Eur J Clin Pharmacol2004;60:97–102.

14. Machado CG, Calado RT, Garcia AB, Falcao RP. Age-related changes of the multidrugresistance P-glycoprotein function in normal human peripheral blood T lympho-cytes. Braz J Med Biol Res 2003;36:1653–1657.

15. Storch CH, Klimm HD, Heinrich T, Haefeli WE, Weiss J. Plasma LDL cholesterolhas no impact on P-glycoprotein (MDR1/ABCB1) activity in humanperipheral blood mononuclear cells. Naunyn Schmiedebergs Arch Pharmacol 2007;376:135–143.

16. Beck WT, Grogan TM, Willman CL, Cordon-Cardo C, Parham DM, Kuttesch JF,Andreeff M, Bates SE, Berard CW, Boyett JM, et al. Methods to detect P-glycopro-tein-associated multidrug resistance in patients’ tumors: Consensus recommenda-tions. Cancer Res 1996;56:3010–3020.

17. Mechetner EB, Schott B, Morse BS, Stein WD, Druley T, Davis KA, Tsuruo T,Roninson IB. P-glycoprotein function involves conformational transitions detecta-ble by differential immunoreactivity. Proc Natl Acad Sci USA 1997;94:12908–12913.

18. Park SW, Lomri N, Simeoni LA, Fruehauf JP, Mechetner E. Analysis of P-glycopro-tein-mediated membrane transport in human peripheral blood lymphocytes usingthe UIC2 shift assay. Cytometry A 2003;53:67–78.

19. Klimecki WT, Taylor CW, Dalton WS. Inhibition of cell-mediated cytolysis and P-gly-coprotein function in natural killer cells by verapamil isomers and cyclosporine Aanalogs. J Clin Immunol 1995;15:152–158.

20. Klimecki WT, Futscher BW, Grogan TM, Dalton WS. P-glycoprotein expressionand function in circulating blood cells from normal volunteers. Blood 1994;83:2451–2458.

21. Chaudhary PM, Mechetner EB, Roninson IB. Expression and activity of the multi-drug resistance P-glycoprotein in human peripheral blood lymphocytes. Blood1992;80:2735–2739.

22. Cianfriglia M, Dupuis ML, Molinari A, Verdoliva A, Costi R, Galluzzo CM, AndreottiM, Cara A, Di Santo R, Palmisano L. HIV-1 integrase inhibitors are substrates for themultidrug transporter MDR1-P-glycoprotein. Retrovirology 2007;4:17.

23. Taylor BJ, Olson DP, Ivy SP. Detection of P-glycoprotein in cell lines and leukemicblasts: Failure of select monoclonal antibodies to detect clinically significant Pgplevels in primary cells. Leuk Res 2001;25:1127–1135.

24. Shirasaka Y, Kawasaki M, Sakane T, Omatsu H, Moriya Y, Nakamura T, Sakaeda T,Okumura K, Langguth P, Yamashita S. Induction of human P-glycoprotein in Caco-2cells: Development of a highly sensitive assay system for P-glycoprotein-mediateddrug transport. Drug Metab Pharmacokinet 2006;21:414–423.

25. Webb M, Raphael CL, Asbahr H, Erber WN, Meyer BF. The detection of rhodamine123 efflux at low levels of drug resistance. Br J Haematol 1996;93:650–655.

26. Henmi K, Yoshida M, Yoshikawa N, Hirano T. P-glycoprotein functions in periph-eral-blood CD41 cells of patients with systemic lupus erythematosus. Biol PharmBull 2008;31:873–878.

27. Hitzl M, Drescher S, van der Kuip H, Schaffeler E, Fischer J, Schwab M, EichelbaumM, Fromm MF. The C3435T mutation in the human MDR1 gene is associated withaltered efflux of the P-glycoprotein substrate rhodamine 123 from CD561 naturalkiller cells. Pharmacogenetics 2001;11:293–298.

28. Meaden ER, Hoggard PG, Khoo SH, Back DJ. Determination of P-gp and MRP1expression and function in peripheral blood mononuclear cells in vivo. J ImmunolMethods 2002;262:159–165.

29. Perloff MD, von Moltke LL, Fahey JM, Greenblatt DJ. Induction of P-glycoproteinexpression and activity by ritonavir in bovine brain microvessel endothelial cells. JPharm Pharmacol 2007;59:947–953.

30. Mechetner E. Detection of the MDR1 P-glycoprotein expression and function. Meth-ods Mol Biol 2007;378:175–193.

31. Daoud R, Kast C, Gros P, Georges E. Rhodamine 123 binds to multiple sites in themultidrug resistance protein (MRP1). Biochemistry 2000;39:15344–15352.

32. De Moerloose B, Dhooge C, Philippe J. Discordance of P-glycoprotein expressionand function in acute leukemia. Adv Exp Med Biol 1999;457:107–118.

33. Fazlina N, Maha A, Zarina AL, Hamidah A, Zulkifli SZ, Cheong SK, Ainoon O, JamalR, Hamidah NH. Assessment of P-gp and MRP1 activities using MultiDrugQuantassay kit: A preliminary study of correlation between protein expressions and itsfunctional activities in newly diagnosed acute leukaemia patients. Malays J Pathol2008;30:87–93.

ORIGINAL ARTICLE

Cytometry Part A � 79A: 912�919, 2011 919