Contribution of translocon peptide channels to the permeation of low molecular mass anions was investigated in rat liver microsomes. Puromycin, which purges translocon pores of nascent polypeptides, creating additional empty pores, raised the microsomal uptake of radiolabeled UDP-glucuronic acid, while it did not increase the uptake of glucose-6-phosphate or glutathione. The role of translocon pores in the transport of small anions was also investigated by measuring the effect of puromycin on the activity of microsomal enzymes with intraluminal active sites. The mannose-6-phosphatase activity of glucose-6-phosphatase and the activity of UDP-glucuronosyltransferase were elevated upon addition of puromycin, but glucose-6-phosphatase and β-glucuronidase activities were not changed. The increase in enzyme activities was due to a better access of the substrates to the luminal compartment rather than to activation of the enzymes. Antibody against Sec61 translocon component decreased the activity of UDP-glucuronosyltransferase and antagonized the effect of puromycin. Similarly, the addition of the puromycin antagonist anisomycin or treatments of microsomes, resulting in the release of attached ribosomes, prevented the puromycin-dependent increase in the activity. Mannose-6-phosphatase and UDP-glucuronosyltransferase activities of smooth microsomal vesicles showed higher basal latencies that were not affected by puromycin. In conclusion, translationally inactive, ribosome-bound translocons allow small anions to cross the endoplasmic reticulum membrane. This pathway can contribute to the nonspecific substrate supply of enzymes with intraluminal active centers.
several intraluminal enzymes of the endoplasmic reticulum (ER) gain their substrate supply from the cytosol. The substrates are often anions, which cannot cross the membrane by simple diffusion. Specific transporters may facilitate their movement across the membrane, although only a few ER transporters have been characterized at the molecular level (6, 20, 22). It has been repeatedly observed that microsomal vesicles derived from the ER exhibit a basal permeability toward various compounds, including xenogenous molecules, which presumably do not have strictly specific transporters. This nonspecific permeability is usually attributed to the damage or to the improper orientation of the membrane. However, recent observations indicate that translocon protein channels may play a role in the permeation of small molecules across the ER membrane.
Co-translational protein translocation and integration into the membrane of the ER occur at sites termed translocons (24, 28, 35–37). The translocon is composed of several ER membrane proteins that are associated to form an aqueous pore. This pore must maintain a permeability barrier because calcium ion is stored in the ER and the unregulated release of this potent second messenger would abolish a fundamental signaling mechanism of the cell. During co-translational translocation, the aqueous translocon pore is sealed at its cytoplasmic end by the growing polypeptide chain (35). The permeability of the translocon has been proposed to be regulated at its luminal side by the prominent ER chaperone BiP, a protein released from the translocon shortly after the completion of ribosome nascent chain targeting. In its empty state, when the translocon pore is ribosome-bound but unoccupied by polypeptide, the complex seems to allow the passage of small molecules (18, 33, 35).
Lomax et al. (27) studied the Ca2+ leak pathways in the ER of pancreatic acinar cells by directly measuring [Ca2+] in the ER. It was demonstrated that the leak was not blocked by either the IP3 receptor antagonist heparin or the ryanodine receptor antagonist ruthenium red, but they found a puromycin-induced Ca2+ efflux from the ER. Puromycin is an antibiotic that selectively terminates ribosomal translation by releasing the nascent polypeptide from the protein channel of the ribosome (32, 35, 38).
Wonderline and coworkers (18, 33) tested the permeation of a small neutral molecule, 4-methyl-umbelliferyl-α-d-glucopyranoside, by measuring its hydrolysis by an ER resident α-glucosidase, which is dependent on its entry into the ER. They found that translocons are permeable to this neutral molecule, as long as empty ribosomes remain bound to them.
It has been observed that low molecular weight anions, which presumably do not have specific ER transporters, e.g., mannose-6-phosphate, can cross the ER membrane, although with a moderate speed. Therefore, the aim of the present study was to investigate the permeation of small anions through translocon channels in rat liver microsomal vesicles. The transport was studied by two methods: 1) by comparing the activity of enzymes with intraluminal active sites in intact and permeabilized microsomes and 2) by directly examining the transport of substrates using the light scattering and rapid filtration techniques. The results confirmed that translocon channels can contribute to the permeation of small anions across the ER membrane.
MATERIALS AND METHODS
Preparation of rat liver microsomes.
Liver microsomes were prepared from fed male Sprague-Dawley rats (180–230 g body wt; Charles River Hungary) as reported earlier (3). The livers were homogenized in sucrose-HEPES buffer (0.3 M sucrose and 0.02 M HEPES, pH 7.2) with a glass-Teflon homogenizer. The microsomal fraction was then isolated using fractional centrifugation. After centrifugation of liver homogenates for 10 min at 1,000 g, the supernatant was spun for 10 min at 12,000 g. The 12,000 g supernatant was centrifuged for 60 min at 100,000 g; the sediment was resuspended in buffer A containing (in mM) 100 KCl, 20 NaCl, 1 MgCl2, and 20 4-morpholinepropanesulfonic acid, pH 7.2, and was centrifuged again for 60 min at 100,000 g. The final pellet (microsomes) was resuspended in buffer A to give ∼50 mg/ml protein concentration, then immediately frozen in liquid nitrogen and kept in liquid nitrogen until use (within 2 mo). The microsomal fraction was characterized by measuring marker enzyme activities (5). The activity of cytochrome c oxidase, glucose-6-phosphatase, and 5′-nucleotidase in the microsomal fraction was 1.2 ± 0.3, 30.1 ± 5.0, and 6.0 ± 1.1, respectively, expressed as percents of activities measured in total homogenate (means ± SD, n = 3).
Rough and smooth subfractions of liver microsomes were prepared according to (5). The purity of the subfractions was evaluated by immunoblot analysis of Sec61α. Compared with the total microsomal fraction, 1.5-fold enrichment in the rough and fivefold impoverishment in the smooth subfraction was observed by the densitometry of the corresponding bands.
Integrity of microsomal vesicles was assessed by measuring the latency of mannose-6-phosphatase activity (8). Protein concentration of microsomes was determined using Bio-Rad protein assay with bovine serum albumin as a standard, according to the manufacturer's instructions.
Transport measurements by light-scattering technique.
The permeability of the microsomal membranes to sucrose was also measured by continuous detection of the osmotically induced changes in size and shape of microsomal vesicles (3, 25, 29). Briefly, ER vesicles (50 μg/ml protein) were equilibrated for 2 h in a hypotonic medium (5 mM 1,4-piperazine diethanesulfonic acid potassium salt, pH 7.0). Light scattering of the microsomal suspensions was then monitored at 400 nm with the use of a fluorimeter (Hitachi F-4500) equipped with a temperature-controlled cuvette holder (22°C) and magnetic stirrer. The addition of a small volume (<5% of the total incubation volume) of a concentrated osmolyte solution results in a rapid increase in light scattering (due to the shrinkage of the vesicles). This peak is followed by a gradual decrease in light scattering (which reflects vesicle swelling due to progressive equilibration of the osmolyte concentration between the extra- and intravesicular spaces). The relative height of the peak and the slope of the curve depend on the permeability of the vesicular membrane. Light-scattering registrations were terminated by the addition of the pore-forming compound alamethicin. The channels formed by alamethicin have been estimated to be 10 Å in diameter with no real selectivity between univalent ions. Alamethicin channels allow the passage of low-molecular-mass compounds such as dicarboxylates and nicotinamide nucleotides, but exclude folded proteins (Ref. 23 and references therein). Upon alamethicin addition, the signal returned to the basal level, indicating that the osmotic effects did not alter the vesicular structure of the microsomes.
Rapid filtration experiments.
For the measurement of the uptake and the accumulation of radiolabeled compounds, liver microsomes (1 mg protein/ml) were incubated in buffer A containing 0.02 mM UDP-glucuronic acid, sucrose, glucose-6-phosphate, or 0.4 mM glutathione plus their radiolabeled analogues (1 μCi/ml [14C]UDP-glucuronic acid, 1 μCi/ml [14C]sucrose, 10 μCi/ml [3H]glutathione, 2 μCi/ml d-[14C]glucose-6-phosphate) at 22°C. At the indicated time points, samples (0.1 ml) were rapidly filtered through cellulose acetate/nitrate filter membranes (pore size: 0.22 μm) and were washed with ice-cold buffer A containing 1 mM 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid. The radioactivity associated with microsomes retained by filters was measured by liquid scintillation counting. In each experiment, alamethicin (0.05 mg/mg protein) was added to parallel incubates to distinguish the intravesicular and the bound radioactivity (3). The alamethicin-permeabilized microsomes were recovered on filters and washed as above. More than 95% of microsomal protein was retained by filters in cases of both native and alamethicin-permeabilized vesicles, indicating that the alamethicin treatment did not affect the vesicular structure of microsomes. The alamethicin-permeabilized microsomes retained amounts of radioactivity <20% of that associated with untreated microsomes. Intravesicular radioactive compounds were bona fide lost during the washing procedure because the alamethicin nonreleasable portion did not further decrease even after extensive washing of filters (and microsomes). This allowed us to regard the alamethicin-releasable portion of radioactivity as intravesicular.
All enzymatic assays were performed in buffer A containing 1 mg/ml microsomal protein at 37°C. Permeabilized microsomal vesicles were treated with the pore-forming alamethicin (0.05 mg/mg protein) immediately before the assay. UDP-glucuronosyltransferase activity was measured by using p-nitrophenol as substrate (0.5 mM) in the presence of 3 mM UDP-glucuronic acid. Absorbance of p-nitrophenol was measured spectrophotometrically at 400 nm, at time 0 and 10 min of incubations. Activity was calculated on the basis of p-nitrophenol disappearance. Glucose-6-phosphatase activity was measured in the presence of 1 mM glucose-6-phosphate or mannose-6-phosphate for 30 min. The phosphate released by the enzymes was measured with the malachite green reagent (31). Absorbance at 660 nm was measured and compared with sodium phosphate standards. β-Glucuronidase activity was assayed by using either p-nitrophenyl glucuronide or phenolphthalein glucuronide (1–1 mM) as substrate (10). Incubation time was 30 min; the corresponding deliberated aglycones were measured after alkalinization of samples at 400 and 540 nm, respectively.
Microsomal proteins were separated by 11% SDS-PAGE and transferred to PVDF filter membranes by electroblotting. Lowfat milk (5%) in PBS containing 0.5% (vol/vol) Tween 20 solution was used for blocking and for dissolving the primary antibody, which was applied for 1 h at room temperature. Anti-Sec61 (catalog no. PA3-014, Affinity Bioreagents) rabbit polyclonal antibody was detected using a fluorescein-conjugated anti-rabbit Ig-specific secondary antibody. To amplify the signal an anti-fluorescein alkaline phosphatase-conjugated tertiary antibody was also used (all antibodies were purchased from Amersham Biosciences). Blots were overlaid with the fluorescent ECF substrate (Amersham Biosciences). Fluorescence was detected using a fluorescence imaging system (Typhoon, Molecular Dynamics). The primary antibody labeled a band of the expected molecular weight (38 kDa).
Statistical comparisons were made by ANOVA, followed by a Dunnett's test (Table 1) or by a Tukey's test (Table 3). In the case of Tables 2 and 4 and Fig. 1, paired two-tailed Student's t-test was used.
Glucose-6-phosphate (monosodium salt), mannose-6-phosphate (disodium salt), glutathione, UDP-glucuronic acid, p-nitrophenyl glucuronide, phenolphthalein glucuronide, alamethicin, puromycin, anisomycin, and 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid were obtained from Sigma. d-[14C]Glucose-6-phosphate (300 mCi/mmol) and [14C]sucrose (500 mCi/mmol) were from American Radiolabeled Chemicals (St. Louis, MO). [14C]UDP-glucuronic acid (2.2 mCi/mmol) was purchased from UVVVR (Prague, Czech Republic). [3H]Glutathione was obtained from NEN Life Science Products (Boston, MA). Dithiothreitol was removed from the [3H]glutathione solution by extraction as described by Butler et al. (9). Cellulose acetate/nitrate filter membranes (pore size 0.22 μm) were from Millipore. All other chemicals were of analytical grade.
Puromycin increases permeability of rat liver microsomes to citrate and sucrose.
In the first set of experiments, the effect of puromycin was validated by using membrane nonpermeant compounds. Sucrose and citrate are not metabolized inside the ER; therefore, presumably they do not have specific membrane transporters. The permeation of these compounds was checked by the light scattering method. Addition of sucrose or citrate to microsomal suspensions resulted in a rapid increase in the light scattering signal reflecting a shrinkage of vesicles, which was followed by a very slow decrease in the signal reflecting a the swelling phase due to the influx (Fig. 1). The slope of the second phase was steeper in microsomes preincubated with puromycin (0.5 mM), indicating that puromycin increased the permeability of the vesicles towards both sucrose and citrate (Fig. 1). The calculated slope values upon addition of sucrose were −2.26 ± 0.17 (control) and −4.23 ± 0.30 (puromycin) upon addition of citrate were −0.41 ± 0.28 (control) and −0.78 ± 0.19 (puromycin). Values are expressed as arbitrary units per minute, means ± SD. In both cases, the slope in puromycin-treated samples differed significantly from that of controls (P < 0.01). Puromycin concentrations >0.5 mM did not result in a more pronounced effect (data not shown).
To control whether this increase in permeability can really be attributed to the translocon pore (and not to a nonspecific effect on the membrane), the effect of puromycin was evaluated either in total (nonfractionated) and smooth microsomes, in which translocons are underrepresented. The uptake was measured with rapid filtration technique, by using radiolabeled sucrose. The basal rate of sucrose uptake was higher in total compared with smooth microsomes, calculated either on protein (Table 1) or on phospholipid basis (data not shown). Puromycin (0.5 mM) caused an increase in the rate of sucrose uptake of total microsomes only (Table 1).
Puromycin increases uptake of UDP-glucuronic acid in rat liver microsomes.
The uptake of various anions by rat liver microsomes was detected by using radiolabeled compounds by rapid filtration. The control time course of UDP-glucuronate (0.02 mM) influx showed the characteristic initial overshoot, followed by a decline, as it has been reported by others (7). A completely different curve was gained in the presence of puromycin (0.5 mM); both the overshoot and the declination phases disappeared and an equilibrating uptake was detected (Fig. 2). Puromycin was ineffective in case of glucose-6-phosphate and glutathione transport (Table 1). In case of glutathione a higher (0.4 mM) concentration should had to be used; at comparable (0.02 mM) concentration uptake could not be properly detected due to a high background caused by binding.
Puromycin promotes substrate access to intraluminal enzymes of ER.
UDP-glucuronosyltransferases (13), glucose-6-phosphatase (41), and β-glucuronidase (42) are ER enzymes with active sites located in the lumen. Hence, the ER membrane represents a barrier for their substrates; consequently, the activity of these enzymes is reduced in native microsomal vesicles. Agents that disrupt the integrity or increase the permeability of the membrane cause an increase in the activity of these enzymes.
Microsomal UDP-glucuronosyltransferase activity was measured by using p-nitrophenol and UDP-glucuronic acid as substrates. The activity in native microsomal vesicles was less than 10% of the total activity measured in permeabilized vesicles (Table 2). The addition of puromycin resulted in a concentration-dependent increase in the activity (Fig. 3); 30% activation was observed at 0.5 mM puromycin concentration. Puromycin concentrations >0.5 mM did not cause a more pronounced effect (data not shown). However, puromycin was ineffective in alamethicin-permeabilized microsomes (Table 2), which indicates that the effect was likely due to the improved access of UDP-glucuronic acid the enzyme to the enzyme and not a direct activation of the enzyme itself.
Microsomal β-glucuronidase activity was measured with p-nitrophenyl glucuronide and phenolphthalein glucuronide as substrates. In accordance with previous results (10), the enzyme had a moderate latency in control vesicles. Puromycin did not increase β-glucuronidase activity (Table 2).
The catalytic subunit of glucose-6-phosphatase is able to hydrolyze hexose phosphates other than glucose-6-phosphate (41). The specificity of glucose-6-phosphatase activity is guaranteed by an ER transport protein, which is strictly specific to glucose-6-phosphate (14, 15, 19). Consequently, the latency of the enzyme is lower with glucose-6-phosphate as substrate than with other hexose phosphates, e.g., with mannose-6-phosphate (8). Therefore, changes in hexose-6-phosphatase activities upon puromycin addition were monitored by using both substrates.
The addition of puromycin resulted in a concentration-dependent increase in mannose-6-phosphatase activity (Fig. 3). A maximal activation of ∼15% was observed. Puromycin was ineffective in alamethicin-permeabilized microsomes (Table 2), which argues against a direct activation of the enzyme. Puromycin was unable to increase the activity when glucose-6-phosphate was used as substrate (Table 2 and Fig. 3).
Effect of puromycin on UDP-glucuronosyltransferase activity can be prevented by a translocon-binding antibody, by puromycin antagonist anisomycin, and by release of ribosomes.
It has been reported that an antibody (or its proteolytic fragment) binding to the Sec61 protein of the translocon occludes the pore of the channel (17). Therefore, the translocon-specific effect of puromycin was controlled by investigating the effect of this antibody on UDP-glucuronosyltransferase activity. Specific binding of the antibody was checked by immunoblot analysis of the microsomal proteins (data not shown). As expected, the addition of the antibody to microsomes counteracted the effect of puromycin. Moreover, it decreased the activity of UDP-glucuronosyltransferase even in the absence of puromycin (Table 3). The inhibition was specific, since neither the heat-denatured antibody, nor an anti-rabbit secondary antibody, was effective (Table 3).
Anisomycin is a known antagonist of puromycin (21), which is effective only if added beforehand. Anisomycin has been successfully used in microsomal experiments to prevent various translocon-dependent effects of puromycin (27, 33). In our experiments, puromycin failed to increase UDP-glucuronosyltransferase activity when added after anisomycin. If anisomycin was added after puromycin, it was ineffective (Table 3).
The release of ribosomes from the surface of microsomal vesicles also prevented the effect of puromycin. Pretreatment with EDTA (5 mM), which is known to detach ribosomes (12), decreased UDP-glucuronosyltransferase activity by 37%, and completely abolished the effect of puromycin. As shown in Table 3, puromycin was also ineffective in microsomes pretreated with a high-salt buffer, another maneuver that also removes ribosomes from the microsomal membrane (30). Treatment of native microsomes with high-salt buffer did not affect UDP-glucuronosyltransferase activity; in permeabilized microsomes the buffer slightly increased the activity (data not shown).
Different membrane permeability in smooth and rough microsomes.
Because recent studies (30, 34), have indicated that nontranslating empty ribosomes are present on the surface of the ER, this condition may therefore contribute to the permeability of the ER membrane. By using the above approach, we compared the latency of UDP-glucuronosyltransferase, glucose-, and mannose-6-phosphatase activities in rough (i.e., translocon rich) and smooth microsomal fractions. All of these enzyme activities were present in both microsomal fractions (Table 4). The latency of UDP-glucuronosyltransferase activity was much greater in smooth vesicles. The latency of glucose-6-phosphatase enzyme measured with glucose-6-phosphate was similar in the two fractions. However, when mannose-6-phosphate was added as substrate, a higher latency was found in smooth microsomes (Table 4).
The present findings show that translocon protein channels can mediate the permeation of small molecular weight anions (besides calcium ions and neutral polar molecules) across the ER membrane. As reported previously (35, 36), the translocon may permit the transmembrane flux of small anions, such as glutamate and HEPES. In the present study, the membrane permeation of six anions of different size and shape (glucose-6-phosphate, mannose-6-phosphate, glutathione, UDP-glucuronic acid, phenolphthalein glucuronide, and p-nitrophenyl glucuronide) was examined. All of these compounds are (or can be) metabolized by enzymes with an intraluminal activity, therefore, a possible contribution of the translocon channel to their membrane transport has a physiological relevance.
Puromycin was used to empty the translocon channel by releasing the incomplete nascent polypeptide from the ribosome. The effect of puromycin was specific because 1) it was absent in the smooth microsomal subfraction poor in translocons (Table 1.); 2) it did not cause a generalized permeability of the vesicles (Fig. 3 and Table 2); 3) it was ineffective in permeabilized vesicles (Table 2); and 4) it could be antagonized by anti-Sec61 antibody, by anisomycin, and by the release of ribosomes from the membrane (Table 3).
The puromycin concentrations used in our study were similar to those in previous studies, and the effective concentrations were in good agreement with other observations. In a recent study, Van Coppenolle et al. (40) applied 200 μM puromycin. Others have used similar concentrations (100–1,000 μM) to estimate the puromycin effect on ER permeability (18, 27, 33).
Although the pore of the translocon complex in the ER is large enough to be permeated by small molecules, it is generally believed that permeation is prevented by a barrier constituted by BiP at the luminal end of the pore (17). A recent study (18) rejected this assumption and concluded that a small, neutral molecule can permeate the empty pore of a translationally inactive, ribosome-bound translocon even in the presence of luminal proteins. Our findings support this observation and affirm that the translocon could provide a pathway for small neutral and anionic molecules to cross (although slowly) the ER membrane.
The effect of puromycin on anion transport is controversial at first sight. The permeation of mannose-6-phosphate and UDP-glucuronic acid was increased upon puromycin addition (Figs. 2 and 3 and Table 2), whereas in case of glucose-6-phosphate, glutathione, p-nitrophenyl glucuronide and phenolphthalein glucuronide, no effect could be demonstrated (Fig. 3, Tables 1 and 2). The discrepancy cannot be explained by the different size or charge of the indicated compounds; e.g., mannose-6-phosphate and glucose-6-phosphate are epimers. A more reasonable explanation is that the contribution of the translocon to the transmembrane flux is remarkable in cases when other possibilities for membrane transport are absent (or have a very limited capacity). Indeed, a high capacity ER transporter has been already characterized for glucose-6-phosphate (14, 15, 19). Glutathione (2) and glucuronide (1, 4, 11) transporters are also known, although they have been explored only functionally. On the other hand, no transporter for mannose-6-phosphate has been even hypothesized in the ER, and the existence and functioning of the ER UDP-glucuronic acid transporter is controversial (1, 7).
This explanation is supported by the finding that puromycin exerted only a slight increase even in cases when it was effective. Effects of similar magnitude have been found in previous studies on calcium (27, 40) and 4-methylumbelliferyl-glucoside permeation (18). However, it does not necessarily mean that translocon channel plays a minor role in the permeation across the ER membrane, because the basal rate of transport (i.e., in the absence of puromycin) observed in microsomal vesicles might well be due to translocons unoccupied by polypeptide chains. The difference in permeability between smooth and rough microsomal vesicles (Table 4) seems to favor this hypothesis.
The flux of low molecular weight compounds through translocon channels can be regarded as a structural imperfection which alters the barrier function of the ER membrane. Nevertheless, it might have a physiological function as a low-affinity, low-capacity transporter with low selectivity. A variety of structurally unrelated compounds are (or should be) able to cross the ER membrane, but only a few transporters have been reported so far (26). For example, various xenobiotics should reach the luminal compartment of the ER to be glucuronidated. Translocon channels may have a role in their transport. Moreover, nucleotides and nucleotide sugars are usually transported by antiport mechanisms in the ER (6, 22). The counter ion in these cases can be supplied by translocon. The role of the translocon has been also hypothesized in the transport of such a bulky anion like FAD (39) or in the permeation of exogenous biotin derivatives (26).
In summary, the translocon channel mediates the permeation of low molecular weight anions in the ER. The phenomenon may play a role in the substrate supply and/or product elimination of some intraluminal enzymatic activities in the ER.
This work was supported by the Italian Ministry of University and Research Grant RBAU014PJA, the Hungarian Ministry of Health Grants ETT 090/2003 and 613/2003, National Scientific Research Fund Grants OTKA F37484, F46740, T48939, T38312, and TS49851, and a grant from University of Siena (quota progetti; to A. Benedetti).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society