The Structure of the P-Glycoprotein

Figure 8: The overall structure of the P-Glycoprotein (Ambudkar et al, 2003)

The human P-glycoprotein is a transmembrane protein responsible for actively pumping toxic hydrophobic substances across the cell's plasma membrane. Therapeutic drugs are transported out of the cell by the P-glycoprotein by a process that requires two ATP-binding domains (circled in Figure 8). These two ATP binding domains are a defining characteristic of a family of transporters known as the ABC (ATP Binding Cassette) family. In addition to the two ATP-binding domains, members of the ABC family are comprised of two homologous subunits, each of which has six transmembrane helices.

The exact mechanism of drug efflux from cells due to P-glycoprotein transport is not well understood, but it might involve either direct transport out of the cytoplasm or redistribution to the basal-lateral side of cells of the drug as it transverses the plasma membrane. The MDR1 gene that encodes for the human P-glycoprotein was one of the first human ABC transporters to have been isolated and described (Soldner et al, 1990).

Function and Localization in the Body

The P-glycoprotein is localized on the luminal surface of epithelial cells in the small intestine (Figure 9), the bile canalicular membrane of the liver (Figure 10), and the proximal tubule of the kidney (Figure 11).

Figure 9: The intestinal epithelial lumen cells

Figure 10: Antibody staining of P-glycoprotein in the bile canalicular membrane of the liver showing numerous localization of P-glycoprotein (Buschman et al, 1992)

Figure 11: A schematic representation of a nephron, the functional unit of a kidney. Arrow 5 points to the proximal tubule of the kidney where P-glycoprotein is greatly expressed

In addition to these surfaces, the P-glycoprotein is also found in the endothelial cells that comprise the blood-brain, blood-testes, and maternal-fetal placental barriers. All of these cells types, where the protein is normally found, readily encounter environmental substances that may be toxic. These cells are all part of tissues with barrier functions. As a result, the P-glycoprotein functions as a "hydrophobic vacuum cleaner," capable of preventing the entry of potentially damaging agents into the body (Dresser et al, 2003).

Specifically, the P-glycoprotein acts on neutral or cationic hydrophobic compounds. In the plasma membrane, it serves to translocate these substrates from the inside to the outside of cells by acting as a flippase, consuming two ATPs for every one molecule of substrate pumped through. Interestingly, many of the substances that are substrates for the P-glycoprotein are also substrates for CYP3A4. Additionally, the tissue localization of the two proteins are overlapping (Fromm, 2003).

The Catalytic Cycle: How Does It Work?

While the P-glycoprotein plays an active role in pumping toxic substances out of cells, it is unable to differentiate between toxic environmental pollutants and xenobiotics. Thus, the P-glycoprotein is as apt to pump out life-saving drugs as it is to remove cells of harmful pesticides. Once pumped out of the cell and back into the lumen, P-glycoprotein substrate is then cleared from the body through excretion.

The current understanding of how drugs are normally transported through the P-glycoprotein is illustrated in Figure 12:

Figure 12: The catalytic cycle model of the P-glycoprotein (Ambudkar et al, 2003)

Unlike other ATP binding proteins such as myosin or the F1F0 ATP synthase, the P-glycoprotein has relatively low affinity for ATP. Additionally, no covalent phosphorylated intermediate has been shown during substrate transport. These observations led to the postulate that the ATP hydrolysis during the catalytic cycle generates a high state of chemical potential energy. It is the relaxation of this chemical potential state that is ultimately responsible for driving the extrusion of substrate from the P-glycoprotein.

Figure 10 shows the most recent proposed scheme for the P-glycoprotein catalytic cycle. In the first step, the drug (substrate) and ATP first bind to the protein. In this proposed model, there is no energetic requirement for the drug to bind. The drug binds to the high affinity site labeled "On." The ATP can bind to either of the two ATP binding pockets. In step 2, the ATP is hydrolyzed, and as a result, the drug is moved from the "On" site to the lower affinity "Off" site. Following hydrolysis of ATP, inorganic phosphate and the drug are both released because of the subsequent low affinity for the drug (step 3). Step 4 involves the spontaneous release of ADP from the complex. The model proposes that the dissociation of ADP is accompanied by a conformation change that allows nucleotide binding, while continuing to reduce substrate-binding affinity.

Step 5 occurs when the P-glycoprotein binds another molecule of ATP in either binding site. Just as in step 2, this ATP is hydrolyzed in step 6. The subsequent release of ADP and Pi constitute steps 7 and 8. This second binding and hydrolyzing of ATP is what resets the P-glycoprotein to its initial state. At this point, the protein is capable of binding another drug molecule and transporting it out of the cell.

The Multi-Drug Resistance Phenomenon

The P-glycoprotein was first identified as the potential cause of a problem in drug therapy known as multi-drug resistance. In all the epithelial cells in which it is localized (kidneys, liver, gastrointestinal tract, capillaries of the brain), the P-glycoprotein acts as a barrier to uptake of xenobiotics. Two of the most problematic clinical examples of MDR occur with chemotherapy agents and HIV protease inhibitors.

Use of chemotherapy to treat cancer has been limited by the development of resistant cancer cells which express the P-glycoprotein. 85-90% of all cancers are carcinomas; in other words, most human cancers are caused by the proliferation of epithelial cells. These are the very same cells where the drug transporter normally localizes. Thus, these cancerous cells have inherent anti-chemotherapy pumps that prevent drug localization intracellularly. Drugs that inhibit processes such as DNA replication in rapidly dividing cancerous cells are of no use if they cannot get beyond the plasma membrane. Because of the problem of MDR, substances that have the potential to block the P-glycoprotein activity are being investigated to mitigate its pumping effects.

The development of inhibitors of HIV-1 protease was also a major advancement in the treatment of HIV infection. Their use has dramatically reduced the HIV viral load in both plasma and tissue of many patients. However, success of protease therapy is greatly impaired by the actions of the P-glycoprotein. The poor and or variable transport across biological membranes into cells has presented a large obstacle to treatment of HIV. Especially frustrating is the apparent low absorption of the drug into the brain which could potentially help alleviate the neurological manifestations of late stage HIV-1 infection. When researchers cloned knock-out mice for the P-glycoprotein, they found a significantly higher level of drug absorption into the body leading to hopes that future therapies can target the protein. Inhibition of P-glycoprotein may lead to improvement of many current therapies for the treatment of cancer and HIV.

The P-Glycoprotein and its Relationship to CYP3A4

Figure 13: Functional interaction between the P-glycoprotein and CYP3A4 (Fromm, 2003)

The first step in Figure 13 shows the passive diffusion of a drug into an enterocyte. Step 2 shows the metabolism of the drug by CYP3A4. Step 3 shows the role of the P-glycoprotein in transporting either metabolized or unmetabolized drug from the enterocyte (intestinal cell). Those substances that make it through the apical to the basal lateral side of the cell are then translocated across the basal lateral membrane into other tissues.

This is a complex process with different implications for different drugs. In some drugs, the drug metabolite is the active form that has an effect on the body. For other drugs, it is the unmetabolized form that is the active therapeutic form of the agent. Thus, modification by CYP3A4 has different implications for different drugs. Transport, however, of the drug by the P-glycoprotein, greatly reduces any form of active drug from entering into tissues.

What about Grapefruit Juice and the P-Glycoprotein?

Mixed Results

Current research into the effects of grapefruit juice consumption on the P-glycoprotein have yielded conflicting results. Most studies have focused on the effects of grapefruit juice on CYP3A4. However, many researchers have postulated that because of the substrate commonalities of both CYP3A4 and the P-glycoprotein, the two proteins should act in a coordinated manner to determine oral drug bioavailability and should both play a role in the grapefruit drug interactions observed in patients. However, preliminary studies show mixed results. Some research indicates that grapefruit juice causes an inhibition of the P-glycoprotein; other studies show increased activation of the P-glycoprotein.

Inhibition Studies

Results of a study conducted by Japanese researchers in 2002 led to the conclusion that grapefruit juice components inhibit the action of the P-glycoprotein. The researchers chose the drug fexofenadine, a non-sedating histamine receptor antagonist that is also a substrate of the P-glycoprotein. In cell lines expressing the P-glycoprotein, such a drug has been seen to be transported from the basal to the apical side. The drug was radiolabeled with 14C and incubated with rat intestinal cells along with grapefruit juice and other P-glycoprotein inhibitors.

First, the researchers determined that P-glycoprotein was indeed expressed in the epithelial intestinal cells of the rat membrane fractions.

Figure 14: Western blot for P-glycoprotein in the small intestine (Tian et al, 2002)

In the Western blot analysis, it was determined that the P-glycoprotein was present in both the ileum and jejunum of the small intestine.

Figure 15: Results of experiments of cells incubated with drug with and without grapefruit juice (Tian et al, 2002)

Researchers added grapefruit juice mixed with orange juice in a 50%:50% ratio and incubated cells with it on the apical side. The white dots indicate no grapefruit juice. The black dots represent the presence of grapefruit juice. Panels A-D represent drug transport in the ileum while panels E-H represent transport in the jejunum. Overall, each time researchers added grapefruit juice with varying concentrations of drugs, the concentration of Fexofenadine fell on the apical side. This indicated that the P-glycoprotein was being inhibited by some substance in the juice that reduced the amount of drug being transported out to the apical side.

This result led researchers to conclude that the P-glycoprotein is inhibited by substances in grapefruit juice, and that this can account for the increased bioavailability of drugs as seen in patients who consumer grapefruit juice while taking their medicine.

Activation Studies

In a study done by researchers from UCSF in 1999, the drug vinblastine was added to intestinal cells that expressed the MDR1 and consequently the P-glycoprotein. The experiments were run in vitro with 0.05% - 5% Minute Maid Grapefruit juice. A gradient was established that allowed researchers to measure the amount of radiolabeled vinblastine on both the apical and basal lateral sides of the cell monolayers. The higher the amount of radiolabeled drug on the apical side, the more transport of the drug had occurred through the P-glycoprotein. Some of the results are shown below:

Figure 16: Transport of drug to apical side of cells over time with increasing concentrations of grapefruit juice (Soldner et al, 1999)

The net secretion or net transport of drug was measured as the ratio of basal lateral to apical amounts of the drug compared to apical to basal lateral amounts of the drug. Figure 16 shows that the higher the grapefruit juice concentrations in the media, the higher the vinblastine transport seen in the intestinal cells. With time, the transport markedly increased.

Then 10uM of cyclosporine (a potent P-glycoprotein inhibitor) was added. The increased net secretion seen with the addition of grapefruit juice was then completely abolished. The resulting drug translocation of radiolabeled vinblastine no longer exhibited any directional transport. Instead, the transport resembled a purely passive process with an average net secretion calculated to be 1, meaning neither apical nor basal-lateral secretion was favored.

Figure 17: Transport of drug to apical side of cells over time when a P-glycoprotein inhibitor is added (Soldner et al, 1999)

CsA represents the inhibitor of P-glycoprotein. The drop in transport of the drug with the addition of the inhibitor suggests that the P-glycoprotein is indeed activated by the addition of grapefruit juice and not some other enzyme.

These results indicate that grapefruit juice significantly activate the P-glycoprotein to increase drug efflux (basal to apical) out of cells. These researchers have proposed that there are perhaps two distinct and positively cooperative drug-binding sites of the P-glycoprotein. They postulate that grapefruit juice components may bind directly to such allosteric sites to enhance drug transport.

What Research Could be Done to Further Understanding?

The two studies-inhibition and activation-shown above find opposing results for the effect of grapefruit juice on the P-glycoprotein. While the inhibition of the P-glycoprotein is consistent with the observed increased drug bioavailability results seen in patients upon grapefruit juice consumption, the results remain inconclusive.

Using standard doses of grapefruit juice in future in vivo and in vitro studies may help to make results more comparable. For example, the inhibition study (Tian et al, 2002) used a 50%/50% solution of orange juice and grapefruit juice while the researchers that concluded that the P-glycoprotein is activated by grapefruit juice used .05% to 5% grapefruit juice (Soldner et al, 1999).

Experimental results were all obtained either in vitro or ex vivo in animal models. Thus, they may not be applicable to the in vivo results seen in humans upon grapefruit juice consumption. Furthermore, the complexity of drug metabolism could mean that different drugs are influenced in different ways by grapefruit juice action upon the P-glycoprotein. In the end, it may be that some activation of the P-glycoprotein could be seen with some drugs while in others, inhibition of the P-glycoprotein will always be the case. Future research may elucidate the answers to these questions.

Xuemei Cai · caix@email.arizona.edu

Biochemistry 462b Honors Project · The University of Arizona

Instructor Dr. Don Bourque

Last Revised May 2004