In Situ Study of the Effect of Naringin, Talinolol and Protein–Energy Undernutrition on Intestinal Absorption of Saquinavir in Rats


  • Ana Catalán-Latorre,

    1. Department of Pharmacy and Pharmacy Technology, Faculty of Pharmacy, University of Valencia, Valencia, Spain
    2. Institute of Molecular Recognition and Technological Development (IDM). Polithecnic University-University of Valencia, Valencia, Spain
    Search for more papers by this author
  • Amparo Nácher,

    1. Department of Pharmacy and Pharmacy Technology, Faculty of Pharmacy, University of Valencia, Valencia, Spain
    2. Institute of Molecular Recognition and Technological Development (IDM). Polithecnic University-University of Valencia, Valencia, Spain
    Search for more papers by this author
  • Virginia Merino,

    1. Department of Pharmacy and Pharmacy Technology, Faculty of Pharmacy, University of Valencia, Valencia, Spain
    2. Institute of Molecular Recognition and Technological Development (IDM). Polithecnic University-University of Valencia, Valencia, Spain
    Search for more papers by this author
  • Nicolás Víctor Jiménez-Torres,

    1. Department of Pharmacy and Pharmacy Technology, Faculty of Pharmacy, University of Valencia, Valencia, Spain
    2. Pharmacy Service, Dr Peset University Hospital, Valencia, Spain
    Search for more papers by this author
  • Matilde Merino-Sanjuán

    1. Department of Pharmacy and Pharmacy Technology, Faculty of Pharmacy, University of Valencia, Valencia, Spain
    2. Institute of Molecular Recognition and Technological Development (IDM). Polithecnic University-University of Valencia, Valencia, Spain
    Search for more papers by this author

Author for correspondence: Matilde Merino-Sanjuán, Department of Pharmacy and Pharmaceutical Technology. Faculty of Pharmacy, University of Valencia, Vicente Andrés Estellés avenue, 46100 Burjassot, Valencia, Spain (fax +34963544911, e-mail


Abstract:  To study the potential interactions of naringin (NAR), talinolol (TAL) and protein–energy undernutrition (PEU) in the absorption process of saquinavir (SQV), perfusion experiments were performed in the small intestine of rats at different SQV concentrations. The results obtained demonstrated that SQV intestinal absorption was described by simultaneous passive diffusion (kdif = 3.44 hr) and saturable absorption (Vma = 127.31 μM/hr; Kma = 10.50 μM) together with a capacity-limited efflux (Vms = 270.53 μM/hr; Kms = 23.44 μM). The competitive inhibition constants of NAR on the SQV input and efflux processes were [IC50]a = 3.98 μM and [IC50]s = 5.00 μM, respectively. NAR significantly decreased (23–29%; p < 0.05) or kept unaltered the absorption rate constant (ka) of SQV in function of the concentration of both compounds administered. Finally, SQV ka significantly increased in PEU status (around 1.8 times) when the drug was perfused either in the presence (p < 0.05) or in the absence (p < 0.01) of NAR. The variations of SQV ka when the antiretroviral drug is co-administered with NAR and/or TAL reinforce their interaction in the absorptive process. Malnutrition may result in altered SQV absorption, and further studies are strongly recommended to analyse the impact of this finding on the pharmacokinetic drug profile.

For administration of drugs, the oral route is a non-invasive method, usually the safest and least expensive, and also it is the most common one used. One of the major limitations associated with this route is low and variable bioavailability that some xenobiotics undergo, so the possibility that absorption can be altered by several circumstances has to be considered. There is high evidence that factors such as interactions in intestinal transport, when absorption obeys a saturable mechanism, play an important role in the oral absorption. These interactions can have an effect on both influx and efflux transporters [1–4]. Malnutrition has dramatic effects on small intestinal mucosal structure and transport activity, and its influence on drug absorption remains controversial. Mucosal atrophy induced by malnutrition reduces the total intestinal absorption of nutrients, although nutrient absorption normalized to mucosal mass may actually be enhanced by a variety of mechanisms, including increased transporter gene expression, electrochemical gradients and changes in the ratio of mature to immature cells [5,6]. In fact, poor prognosis observed in some patients could be related to the detrimental impact of malnutrition on the efficacy, toxicity or pharmacokinetics of the drugs [7–9], so study of drug effects in animal models that reproduce the undernourishment state could help to accurately simulate what happens in ill patients [10,11]. Taking into account that polymedication, food insecurity and also undernutrition are common among HIV (human immunodeficiency virus)-infected patients, this study was carried out to measure and understand SQV intestinal absorption in undernutrition status and in the presence of inhibitors of intestinal influx and efflux carriers, because it constitutes a suitable example because of its pharmacokinetic characteristics [12].

Saquinavir is a protease inhibitor that is used in the management of different stages of HIV [13]. Its oral bioavailability is variable and limited by different factors. On the one hand, efflux transporters located in the intestine such as the export pump P-glycoprotein (P-gp) or multidrug resistance protein 2 (MRP2) and influx transporters like the organic anion-transporting polypeptide (OATP) are implicated in its intestinal absorption [14–18]. On the other hand, SQV suffers of an important first-pass metabolism [3,19]. Additionally, there appears to be an overlap in the substrate specificity between the efflux transporter P-gp and the influx transporter OATP [20], which could lead to opposing influences on the net absorption of a shared substrate [21,22].

Inhibition and induction of P-gp efflux function is a well-established mechanism of drug–drug interactions [15,23], and it is possible to alter that function with dietary constituents, such as citrus fruit juices [24]. Alteration of transporter-mediated drug uptake by concomitantly administered drugs or food components may also result in a change in drug pharmacokinetics [25–27]. Naringin (NAR), a flavonoid glycoside compound of grapefruit, has demonstrated to inhibit the OATP and the P-gp [28–30]. The simultaneous oral administration of NAR with drugs like fexofenadine or SQV that are absorbed by means of OATP can produce an alteration in the pharmacokinetics of the drugs [31]. For example, the intestinal permeability of talinolol (TAL), a drug substrate of both transporters, decreases with moderate doses (200 μM) of the flavonoid, but it increases if greater doses of NAR (2000 μM) are co-administered with TAL [32]. This observed effect indicated that the inhibitory effect of grapefruit juice on OATP and P-gp-mediated TAL absorption in rats depends on the NAR concentration ingested.

This work was performed to determine the impact of energy–protein undernutrition on the SQV intestinal absorption and to study the interactions when NAR and TAL, as inhibitors of saturable intestinal absorption, are co-administered.

Materials and Methods

Administrative/legal prerequisites for the performance of the animal study.  The animal protocol, duration of treatments and doses, anaesthesia, surgical and perfusion procedures as well as the statistical methods were justified in detail and were approved by the Pharmacy Faculty Ethics Commission of Valencia, Spain. The study complied with European requirements for animal studies. The protocol needed to be optimized in such a way that it permitted us to obtain maximum information using a small number of animals.

Protocol, animals and experimental procedures.  The animals included in the study were selected according to biochemical and biometric criteria. A total of 60 male Wistar rats of 8–9 weeks of age, weighing 226–240 g, with a serum albumin ranging from 2.2 to 2.45 g/dL and serum cholesterol ranging from 57 to 77 mg/dL, were placed in individual polyethylene cages in a controlled room (22–23°C, 50–60% humidity) under a 12-hr light/dark cycle and were randomly divided into two groups: WN group (well nourished with regular nutrition diet) and PEU group (under-nourished with protein–energy restricted diet). All animals had free access to tap water but controlled food intake. WN group was subjected to a standard pellet diet (2014 from Harlan laboratory), according to the normal daily requirements of a rat (proteins 14%, dose of diet intake per day: 20 g/60.2 kcal) for 23–25 days of adaptation period. The PEU group received a reduced protein, carbohydrate and fat pellet diet (TD 99168 from Harlan laboratory; proteins 5%, dose of diet intake per day: 10g/38 kcal) during the same period of time.

In both groups, all rats were weighed daily, and serum albumin and total cholesterol were quantified, at least once a week, using standard commercial kits, according to the manufacturer’s instructions (QCA Laboratory, Tarragona, Spain).

Animals from the PEU group were considered malnourished if, at the end of the adaptation period, the weight was lower than 80% of the weight reached in the WN animal group and serum albumin lower than 2.2 g/dL [33,34]. Moreover, qualitative parameters were considered, such as piloerection, nail status, decreased locomotor activity and constant fearful state.

Absorption assays: surgical procedures.  The in situ single-pass intestinal perfusion method was performed at the end of the period of adaptation. The surgical procedure described by Doluisio [35] and modified as previously reported [36] was performed. On the study day, prior to surgery, the WN and PEU rats were fasted for 10 hr but had unlimited access to tap water. After anaesthesia via intraperitoneal administration of 30 mg/kg of a thiopental sodium solution 20% (w/v) (Dolethal®; Vétoquinol SA, Lure, France), the intestine of the animal was exposed by a midline abdominal incision, and the whole small intestine was canulated and the biliary duct was ligated. The rats were placed on a 37°C heating pad to maintain their body temperature. After rinsing with physiological saline solution, to eliminate faecal residues and debris, 10 mL of the SQV isotonized and thermostated solution were perfused. The SQV remaining concentrations in the intestinal lumen were measured every 5 min., for a total time of 30 min., using 300 μL samples. Samples were centrifuged (1500 × g for 10 min.) and frozen (−80°C) until quantification. In these conditions, no degradation of the drug was detected. The remaining drug concentrations were properly corrected by the calculation of water reabsorption that was evaluated for each animal taking into account that this process obeys apparent zero-order kinetics [37,38].

Perfusion solutions.  The perfusate consisted of 1% dimethyl sulfoxide, DMSO, in a saline-buffered solution (pH 6.4) in which SQV was dissolved. Four different SQV concentrations were assayed in the absence of additives, and some of them were also perfused in the presence of NAR and/or TAL. The combinations are shown in table 1. The study included ten experimental groups, with six rats in each treatment group.

Table 1. 
Perfusates composition and concentration (μM) in the different groups assayed.
  1. WN, well-nourished groups; PEU, protein–energy undernutrition groups; SQV, saquinavir; NAR, naringin; TAL, talinolol; N, numbers of rats.

(N = 48)
(N = 12)

Analytical procedures.  Saquinavir concentrations in perfusate sample aliquots were determined with a high-performance liquid chromatography (HPLC) method with UV detection. Samples (50 μL) were injected into a Waters HPLC system equipped with a 200-μL loop and Breeze chromatographic software. The wavelength was set to 235 nm and the flow rate to 1 mL/min. The analytical column was a Symmetry analytical column (C-18, 150 × 4.6 mm), preceded by a pre-column (Tecnokroma-C18). SQV was eluted using a mobile phase consisting of a mixture of acetonitrile and 99% bidistilled water together with 1% phosphate buffer 1/15 M (pH 6.9) solution (61/39, V/V). Excellent linearity between peak area and concentration was observed for the compound over the entire range of concentrations assayed (r > 0.999). Accuracy was evaluated by calculating the relative error, which was always <14%. Precision was evaluated by calculating the coefficient of variation, which was always <3.46%. The limit of SQV quantification was 0.54 μM. These results were considered satisfactory [39].

Fitting of models to data; exploratory analysis.  To detect non-linearities in the intestinal absorption of SQV, the apparent first-order rate constant was calculated according to the classical expression by fitting the Equation (1) to the experimental data by the non-linear least-squares procedure.


where A values represent the remaining drug concentrations in the luminal content, already corrected for water reabsorption, at each sampling time, t; ka is the apparent first-order rate constant, and A0 is the calculated intercept at zero time, which is always lower than the initial concentration perfused because of membrane adsorption. This equation was used because it gives an accurate description of the profiles within a limited time interval (i.e. from 0 to 30 min., with good correlation coefficients). Moreover, the apparent rate constant, ka, constitutes a good basis for the interpretation of non-linearities and inhibition phenomena. Only samples obtained between 5 and 30 min. were used for calculations, i.e. the zero-time samples were not used for regression [35,38]. Both parameters (A0 and ka) were then calculated for each animal.

Data are expressed as the mean of at least six rats with standard error (S.E.). To assess non-linearity phenomena in SQV intestinal absorption, the ka values found for each initial concentration of the drug were statistically compared using the Dunnett’s T3 test (Levene test p < 0.05), using the SPSS 15.0 program. Statistical significance level was fixed at 0.05.

Kinetic analysis.  Previous analyses have demonstrated that SQV undergoes efflux process from enterocyte to intestinal lumen mediated by transporters of the ATP-binding cassette family and, in particular, P-gp, and also, it has influx carriers like Oatp [14,17,18,21]. SQV absorption process was evaluated using six structural pharmacokinetic models that are described as follows:

Model (1). Passive absorption.


Model (2). Active absorption.


Model (3). Combined (passive and active) absorption.


Model (4). Passive absorption and active secretion.


Model (5). Active absorption and active secretion.


Model (6). Combined (passive and active) absorption and active secretion.


where dA/dt is the absorption rate (concentration/time), A the remaining drug concentration in luminal fluid at each sampling time, kdif the passive absorption rate constant, Vma the maximal saturable absorption rate, Kma the concentration of the drug at which the saturable absorption rate is half maximal, Vms the maximal efflux rate and Kms the concentration at which the secretion rate is half maximal. These fits were performed using the Winnonlin® 5.0.1 program (Pharsight Corporation, Mountain View, CA, USA) (Nelder–Mead algorithm, weighing factor = 1/Ateor2). The sum squared residuals (SS) and Snedecor F test (α = 0.05) were calculated to assess the goodness of fits, and the Akaike information criterion (AIC) was also calculated to select the most representative model.

Inhibitory study.  Inhibitory effects on SQV absorption process resulting from the addition of NAR were detected by statistical comparison between the apparent first-order rate constant values found in the presence and absence of the selected inhibitor.

A set of differential equations were fitted to the whole set of concentration versus time data obtained in the presence of NAR. The parameters corresponding to the absorption process of SQV were fixed to the values obtained in the previous section. NAR, majoritary component of the grapefruit juice, has been demonstrated to affect both the carrier system of absorption and the efflux system. Thus, three possibilities have been explored:

Model (7). Competitive inhibition of the active absorption.


Model (8). Competitive inhibition of the efflux.


Model (9). Competitive inhibition of the active absorption and the efflux.


where [IC50]a and [IC50]s are the concentration of a NAR needed to reduce the active influx and efflux rate by 50%. These fits were performed and evaluated as described above.


Detection of non-linearity in SQV absorption process.

The four WN rat groups (n = 6, in each group) were administered SQV in free solution (5, 50, 100 or 500 μM). ka was different depending on the concentration of SQV administered: ka5 = 3.24 ± 0.35 hr (R2 > 0.992); ka50 = 3.41 ± 0.36 hr (R2 > 0.995); ka100 = 2.49 ± 0.21 hr (R2 > 0.996) and ka500 = 2.67 ± 0.33 hr (R2 > 0.992). Dunnet’s T3 test demonstrated that the ka of SQV 100 μM was significantly different when compared with SQV 5 and 50 μM (p < 0.05). AIC and SS values for the models assayed are shown in table 2.

Table 2. 
AIC and SS values for the different models assayed.
 Number of model assayedKinetic and processAICSS
  1. Akaike information criterion (AIC) and Sum of squares (SS) obtained for SQV intestinal absorption models assayed. SQV, saquinavir.

Saquinavir intestinal absorption1Passive absorption319644,788
2Active absorption279430,796
3Passive and active absorption266263,535
4Passive absorption and efflux259288,666
5Active absorption and efflux224199,924
6Passive and active absorption and efflux207134,910
Effect of Narangin on SQV intestinal absorption7Inhibition on absorption process483141,941
8Inhibition on efflux process473151,906
9Inhibition on absorption and efflux processes42030,199

Interaction studies: NAR, TAL and PEU.

Influence of NAR and TAL on SQV absorption.  The NAR and TAL inhibitory effects on mediated uptake of SQV (5 or 100 μM) are shown in fig. 1. As can be seen, the ka value of SQV 5 μM (ka5 = 3.24 ± 0.35 hr) significantly decreased by 23.15% (p < 0.05) and 29.32% (p < 0.05) when the drug was co-administered with NAR 28 μM (ka5N28 = 2.49 ± 0.22 hr) or 1400 μM (ka5N1400 = 2.29 ± 0.25 hr), respectively. In contrast, when SQV and NAR were co-administered at 100 and 1400 μM, respectively, absorption rate constant of SQV (ka100 = 2.49 ± 0.21 hr) tended to increase although no statistical differences were obtained (ka100N1400 = 2.70 ± 0.68 hr; p > 0.05).

Figure 1.

 Effect of NAR concentration (28 or 1400 μM) and talinolol (TAL) 50 μM on saquinavir (SQV) absorption rate pseudoconstant (ka) perfused at 5 or 100 μM. *p < 0.05 significantly different from SQV 5 μM. Data are shown as mean ± Standard error. **p < 0.01 Significant differences with ka of SQV 5 μM with 28 μM Naringin (NAR) and SQV 5 μM added to 28 μM NAR and 50 μM TAL.

Presence of TAL 50 μM in a solution with SQV 5 and NAR 28 μM significantly increased SQV ka by 110.04% (ka5N28T50 = 5.23 ± 0.94; p < 0.01) when compared with SQV ka when the drug was administered in the presence of NAR alone (ka5N28 = 2.49 ± 0.22 hr).

Influence of undernourishment on SQV absorption. ka of SQV in PEU groups was significantly increased when compared with ka of WN groups (fig. 2). SQV ka in PEU animals (kaPEU5 = 5.49 ± 0.39 hr) was 1.69 times greater (p < 0.01) than ka in WN animals (ka5 = 3.24 ± 0.35 hr), and the presence of NAR 28 μM increased the absorption rate constant by 1.85 times (kaPEU5N28 = 4.60 ± 0.51 hr) compared with the WN control group (ka5N28 = 2.49 ± 0.22 hr; p < 0.05). No statistical differences in ka were found between the two PEU groups (p > 0.05).

Figure 2.

ka of saquinavir (SQV) 5 in the absence or presence of 28 μM Naringin in WN and protein–energy undernutrition groups (PEU) animals. *p < 0.05; **p < 0.01 when comparing SQV ka in PEU groups with control WN animals. Data are shown as mean ± S.E.: Standard error.

Kinetic analysis.

The absorption rate constants obtained for SQV in rat small intestine showed statistically significant differences (p < 0.05) regarding the involvement of a carrier-mediated process. The hypothesis of a saturable mediated transport was tested by mathematical fitting of six different absorption models, considering a possible saturable input and a saturable secretion that ran simultaneously with the passive diffusion. The AIC and SS parameters obtained in the different models assayed are listed in table 2. As can be observed, the better fitting corresponds to the more complicated model, which combines two different input processes (saturable and passive) simultaneously with an efflux process (fig. 3).

Figure 3.

 Diagram of the absorptive process of saquinavir. kdif: passive absorption rate constant (diffusion); Vma: maximal absorption rate; Kma: concentration of the drug at which the absorption rate is half maximal; Vms: maximal secretion rate; Kms: concentration at which the secretion rate is half maximal.

After fixing the constant values obtained in the SQV kinetics analysis, the final model to describe the effect of NAR on SQV intestinal absorption is summarized by model 9. NAR has an influence on both influx and efflux of SQV. The parameters obtained are shown in table 3. Fig. 4 shows the correlation between population-predicted and observed SQV intestinal concentration values obtained in the absence or presence of NAR, and the weighted residuals and time, showing the goodness of fit of the selected model to the experimental data.

Table 3. 
Parameters defining the SQV absorption kinetics (model 15) and the statistical parameters calculated.
ParametersSQV absorption kinetics
Estimated valueS.E.
  1. S.E., standard error; SQV, saquinavir.

Kdif (hr)3.440.898
Vma (μM/hr)12723.10
Kma (μM)10.4956.77
Vms (μM/hr)27021.44
Kms (μM)23.441.29
[IC50]a (μM)3.982.90
[IC50]s (μM)5.005.39
Figure 4.

 Goodness of fits. Population-predicted versus observed saquinavir (SQV) intestinal concentrations (left) and weighted residuals versus time (right) for SQV in free solution (•, model 6) and interaction SQV–NAR (bsl00066, model 9). NAR, Naringin.


Drug–drug and drug–food interactions have been described frequently. For example, pharmacokinetic drug interactions involving fruit-derived beverages have received a great deal of scientific and public media attention during the past 15–20 years. The incidence of these processes increases when patients are polymedicated by the oral route as occurs in HIV-infected patients, because the current antiretroviral therapy always contains several drugs. SQV is used in the management of different stages of HIV infection, and its pharmacokinetic profile is characterized by low and variable oral bioavailability as a consequence of an extensive pre-systemic loss attributed to influx- and efflux-mediated transport as well as intestinal and hepatic first-pass metabolism [14,19].

This study was undertaken to investigate the relative contribution of several factors as dietary compounds (NAR) and drugs (TAL) that could share the same carrier system. The effect of nutritional status on the intestinal absorption of SQV was also investigated because this compound is frequently administered in these circumstances.

Evidence of specialized transport mechanisms.

To detect the non-linearity in the SQV absorption process, four concentrations (5, 50, 100 and 500 μM) were perfused. The study was limited to this drug concentration range because of its poor hydrosolubility. For this reason, 1% DMSO was added to the SQV solutions, because it has previously been demonstrated that at this concentration, the cosolvent does not modify the intestinal membrane permeability [36].

Statistical analysis of the ka as a function of the SQV concentration in the perfusion fluid shows that there is non-linearity. ka of SQV administered at 5 or 50 μM was significantly greater (p < 0.05) than SQV ka when administered at the concentration of 100 μM. Also, a slight increase in ka value can be seen when the concentrations of SQV raise from 100 to 500 μM, although statistical differences between both ka were not found (p > 0.05). The present findings could be explained whether SQV absorption is mediated by at least two opposite transport systems. The hypothesis of a saturable mediated transport was tested by mathematical fitting, and the better fitting corresponds to the more complex model, which combines two different input processes (saturable and passive) with simultaneous efflux. It should be kept in mind that non-linear regression methods calculate the standard error of parameters from linear approximations and give asymptotic values, which are overestimated. Nevertheless, the predicted and experimental values are well correlated, demonstrating the functionality of the estimated parameters despite their low precision. The absorption mechanism developed and supported by our experimental data agrees with data published by other authors, indicating that SQV is a substrate of an absorptive transporter, OATP and of secretory carriers such as P-gp and/or MRP2 [14,17,18,21,31]. Kinetic parameters indicate that the absorptive transporter has double affinity and half capacity than the secretory carrier system, so active input at SQV 100 and 500 μM would be saturated, resulting in a constant influx rate. In contrast, the efflux process at SQV 100 μM would not be saturated, and saturation would begin when SQV is administered at 500 μM (model 6, table 3), producing a slight increase in ka when SQV is perfused at the highest concentration assayed.

Interaction studies: effect of NAR, TAL and PEU on SQV absorption.

Influence of NAR and TAL.  Naringin, one of the major constituents of grapefruit juice, interacts with intestinal transporters as Oatp/OATP or P-gp [40,41] and also with metabolism enzymes (CYP3A4) [19,42,43]. Based on the literature reports, the variety of drugs affected by grapefruit juice includes SQV, TAL, cyclosporine, simvastatin, midazolam, colchicine or montelukast [31,40,44]. On these bases, we suspected that the concomitant administration of NAR, SQV and TAL [32,45,46] could restrict the absorption of any of them to various extents because of interferences on absorption or secretion transporters. To understand the potential interaction among SQV, TAL and NAR, a competitive inhibition model was checked.

Concentrations of NAR were chosen according to the following criteria: a commercial grapefruit juice concentration, 1400 μM, and a NAR concentration 50 times lower than the commercial concentration assayed, NAR 28 μM [29,41]. According to our findings, the flavonoid acts as an inhibitor of SQV influx and has a slight effect on the intestinal efflux of the drug (model 9, table 3), leading to a statistically significant decrease in SQV ka (p < 0.05) when the drug is administered at the 5 μM concentration with NAR (fig. 1).

Results reported indicate that the inhibitory effect of NAR on influx and efflux carriers involved on SQV absorption is concentration dependent of the drug and concentration independent of NAR. The net result when SQV is perfused at 5 μM in the presence of NAR at 28 or 1400 μM is a similar reduction of SQV absorption, which seems to be independent of the concentration of NAR used in this study. This SQV ka reduction was not observed at 100 μM concentration of SQV co-administered with NAR 1400 μM (p > 0.05). On these bases, NAR seems to affect predominantly the influx carrier of SQV at lower doses of SQV.

Only when TAL, a potent inhibitor of mdr1, is co-administered with NAR and SQV, the ka of the protease inhibitor significantly increases 2.07 times (p < 0.01) (fig. 1), reflecting that TAL has a profound impact on efflux of SQV.

Taking into account that NAR preferentially inhibits OATP over P-gp [30,32] and TAL is considered a candidate for the determination of intestinal P-glycoprotein function [47], data reported in this study suggest that OATP/Oatp and MDR1/Mdr1 play roles in the intestinal absorption of SQV as influx and efflux transporters, respectively. In this way, the reported km value of SQV for OATP1A2-mediated transport is 36.4 μM, and OATP1A2-mediated NAR transport had an IC50 value of 3.6 μM [48].

Pharmacokinetic model tools used in this in situ study do not provide direct evidence of the involvement of Oatp or gp-P on the SQV absorption process, but the results obtained are consistent with the role of these carriers in SQV uptake described by other authors [14,17,18,21]. This is why these tools can be considered to be of great utility in this type of studies, because they allow us to avoid more complex techniques and reduce the number of animals used. In the light of this result, we can conclude that SQV absorption in the rat small intestine is mediated by carriers shared by NAR and TAL.

Influence of undernourishment on SQV absorption.

Protein–energy malnutrition has an elevated incidence in hospitalized patients, and it is associated with many pathological states, such as infections, cancer or metabolic diseases like diabetes [49].

Various studies have demonstrated that undernourishment has dramatic effects on small intestinal mucosal structure and transport activity [7,50]. The consequent effect of malnutrition on nutrients and drugs absorption is highly variable and contradictory [51], being responsible for either increase or decrease in absorption, whether measured in terms of rate, as absorption rate constant, and in terms of extent, as area under the curve (AUC) [5].

ka of SQV significantly increases in the PEU group when compared with the WN groups, even if SQV is administered in the absence (p < 0.01) or in the presence of NAR 28 μM (p < 0.05) (fig. 2).

The higher ka obtained in PEU animals compared with WN animals may be a consequence of changes in the absorptive membrane, like altered mucosal protein concentrations (absorptive and secretory carriers, and also metabolic enzymes), reduced villus heights and crypt depths, and also disrupted intestinal membrane integrity or damage in tight junctions [7,50,52–55]. In this nutritional status, the number of protein carriers would be decreased, and the repercussion of inhibitors of influx or efflux transport would have a minor effect.

Nutritional status has a profound impact on drug disposition and metabolism. Metabolic pathways may be impaired in malnourished subjects, and they may result in decreased clearance and increased elimination half-life of some drugs [56]. These alterations in drug metabolism associated with the changes in the SQV absorption produced by NAR administration and nutritional status observed in this study may lead to greater drug exposure and, according to the FDA criteria, would be clinically significant if they produce at least a 30% modification in the drug AUC values [57]. In accordance with this notion, the findings obtained in this study at the SQV 5 μM concentration suggest that co-administration of TAL and malnutrition status are risk factors predisposing to SQV overdosing and toxicity, whereas co-administration of the drug with NAR could lead to SQV plasma levels lower than the minimum effective concentration.

Further in vivo studies are recommended to elucidate whether raises in SQV rate absorption imply an increase in SQV plasma levels.


Saquinavir is absorbed through passive absorption and Michaelis–Menten kinetics and excreted to the luminal fluid by a saturable secretion pathway. The ka obtained at the SQV concentrations assayed support that saturation of the absorptive carrier occurs at lower concentrations than the saturation of the active efflux process. Modification of ka values of SQV when NAR and/or TAL is co-administered with the drug reinforces their competitive interaction in the absorptive process; this may be because the compounds assayed share at least one intestinal carrier system. NAR decreases SQV intestinal absorption, whereas TAL increases the drug uptake at the concentrations assayed. Malnutrition may result in altered SQV absorption, and further studies are strongly recommended to analyse the impact of this finding on drug response and toxicity. The results of these studies will benefit both developing and developed countries patients that need SQV in their pharmacotherapy. Potential effects of undernutrition and drug–food interactions should be tested in drugs with similar pharmacokinetic profile as SQV.


We thank Pharsight Corporation for the academic licence of Winnonlin 5.0.1

Conflict of Interest Disclosure

The authors declare no conflicts of interest.