Cell Surface Biotinylation and Other Techniques for Determination of Surface Polarity of Epithelial Monolayers

I. INTRODUCTION
A fundamental property of epithelial cells is the polarized distribution of proteins and lipids in the apical and basolateral domains of the plasma membrane. These two domains are physically separated from each other by the tight junction. Many studies have been done over the past 20 years to understand the mechanisms that lead to the establishment and maintenance of the polarized distribution of proteins and lipids in the plasma membrane of epithelial cells (Rodriguez-Boulan et al., 2005; Yeaman et al., 1999; Keller and Simons, 1997).

A major advance in the study of epithelial cell polarity was achieved with the introduction of porous filter supports for the growth of epithelial cell cultures (reviewed in Rodriguez-Boulan et al., 2005). This method differs from classical cell culture in that it allows direct access to the basolateral surface of cultured cells. Epithelial cells and cell lines grown on such filter supports (either nitrocellulose or polycarbonate) attain a more differentiated appearance and become polarized after relatively short times in culture. While most studies on epithelial polarity and trafficking of plasma membrane proteins have been performed using a limited number of cell lines (MDCK, FRT, Caco-2), this culture technique has been gaining in popularity and has been used now for primary cultures as well. The confluency of cells grown on permeable supports can be determined by the measurement of transepithelial electrical resistance or transepithelial [³H]inulin flux (Hanzel et al., 1991).

One of the biggest advantages of this culture system is the accessibility of either the apical or the basolateral surface to any reagent added to the medium and the ability to add different reagents to contact either surface. This is the basis of the biotinylation techniques that have been developed to selectively label proteins present on the apical or basolateral domains of the plasma membrane of filter-grown epithelial cells.

The proteins present on the surface of filter-grown monolayers can be selectively modified by the watersoluble cell-impermeable biotin analog sulfo-NHSbiotin. Taking advantage of the access afforded by the filter support, the addition of sulfo-NHS-biotin to only one surface of the cell results in the selective labeling of only the apical or basolateral surface proteins. The biotinylated proteins can then be detected by blotting with [¹²⁵I]streptavidin or streptavidin conjugated to any number of enzymatic reporters. Furthermore, the cells can be metabolically pulse labeled and the proteins of interest can then be studied using biotinylation, immunoprecipitation and subsequent streptavidin- agarose precipitation. This technique is very versatile and is applicable to the study of diverse aspects of epithelial cell polarity, such as the steadystate distribution of specific antigens, or dynamic processes, such as targeting to the cell surface and transcytosis of membrane proteins. Several biotin analogs are available, including one that contains a disulfide bond. By differentially labeling the surfaces of epithelia in situ with the cleavable NHS-S-S-biotin and a noncleavable biotin, we have been able to study the polarity of a native epithelium in situ (Marmorstein et al., 1996).

The first edition of this article described a basic protocol for selective cell surface biotinylation, plus some modifications of the assay to study protein targeting and endocytosis. The second edition included a basic protocol for in situ domain-specific biotinylation. In this edition, we have added a protocol of surface immunolabeling as an alternative to biotin labeling for determination of polarity. This technique is critical in cell lines that may exhibit leakiness to biotin. Additionally, we have also included two sections determining the polarity of a protein by means of intranuclear microinjection of its cDNA and quantitative microscopic analysis to determine distribution of the protein relative to known polarized markers.

II. MATERIALS
Sulfo-NHS-biotin (sulfosuccinimidobiotin, Cat. No. 21217); NHS-LC-biotin (sulfosuccinimidyl-6-(biotinamido)- hexanoate, Cat. No. 21335); NHS-SS-biotin (sulfosuccinimidyl-2-(biotinamido)-ethyl-l,3-dithioproprionate, Cat. No. 21331); and immunopure-immobilized streptavidin (Cat. No. 20347) are from Pierce (Rockville, IL). Protein A-Sepharose Cl-4B (Cat. No. 17-0780-01) is from Pharmacia/LKB (Piscataway, NJ). Glutathione (Cat. No. G-6529) and cycloheximide (Cat. No. C-7698) are from Sigma Chemical Co. (St. Louis, MO). Staphylococcus aureus cells (Pansorbin, Cat. No. 507858) are from Calbiochem (La Jolla, CA). Cells are grown on polycarbonate filters (Transwell, 12mm diameter, Cat. No. 3401; 24mm diameter, Cat. No. 3412) from Corning-Costar (Cambridge, MA); MEMSelect- Amine kits (Cat. No. 19050-012) are from GIBCO BRL Life Technologies (Grand Island, NY); [³⁵S]EXPRE³⁵S ³⁵S (methionine/cysteine) and [³⁵S]cysteine are from Dupont NEN (Boston, MA) (Cat. No. NEG 072 for Express and Cat. No. NEG 022T for [³⁵S]cysteine); [¹²⁵I]streptavidin can be obtained from Amersham (Arlington Heights, IL) (Cat. No. IM236); streptavidin conjugated to horseradish peroxidase can be obtained from Sigma (Cat. No. S-5512).

III. PROCEDURES
A. Cell Surface Biotinylation
This procedure is used to determine the relative percentage of a plasma membrane protein(s) in the apical versus basolateral plasma membrane of epithelial cells grown on permeable filter supports (modified from Sargiacomo et al., 1989).

Solutions

1. PBS-CM: Phosphate-buffered saline containing 1.0 mM MgCl₂, and 1.3 mM CaCl₂
2. Sulfo-NHS-biotin or sulfo-NHS-LC-biotin: Stock solution is 200mg/ml in dimethyl sulfoxide (DMSO), which can be stored for up to 2 months at -20°C. Thaw just prior to use and dilute to a final concentration of 0.5 mg/ml in PBS-CM. Use immediately.
3. 50mM NH₄Cl in PBS-CM or Dulbecco's modified Eagle's medium (DMEM): Use to quench the excess biotin at the end of the labeling reaction

Steps
All steps are carried out on ice and with ice-cold reagents.

1. For all experiments, use confluent monolayers of cells plated at confluency (for most cell lines, 2.5-3.5 × 10⁵ cells/cm² of filter) 4-5 days prior to biotinylation. Measure transepithelial electrical resistance (TER) and discard monolayers that do not exhibit acceptable resistances (different cell lines exhibit different TER values ranging from tens to thousands of Ω · cm²; monolayers should be used that exhibit TER values in the normal range for your cell line. We have successfully performed this assay on cells with TERs as low as 50 Ω · cm²).
2. Wash filters on both sides three times with icecold PBS-CM.
3. Add a fresh solution of sulfo-NHS-biotin (0.5mg/ml in PBS-CM) to the apical or basolateral chamber. Add PBS-CM to the other chamber. We use 0.7ml apical and 1.4ml basolateral for 24-mmdiameter filters and 0.4 and 0.8 ml for 12-mm-diameter filters. Incubate with gentle shaking for 20min at 4°C and then repeat this step.
4. Quench the reaction by removing the solutions from both chambers and replacing with 1 ml of 50 mM NH₄Cl in PBS-CM. Incubate with gentle shaking for 10min at 4°C.
5. Rinse twice with PBS-CM.
6. Excise filters and either freeze at -80°C (the freeze thaw involved with storage at -80°C appears to inactivate some proteases) or immediately proceed with the extraction of biotinylated proteins as outlined in Section III,F.

Analysis of Results
The amount of protein present on the apical or basolateral surface is determined by a densitometric analysis of the autoradiographs. Multiple exposures are necessary if using the film to ensure that the values obtained are in the linear range of the film. Polarity is expressed as the percentage of total surface protein present on one surface of the monolayer.

B. Biotin Targeting Assay
This procedure is used to determine if proteins are delivered directly, indirectly (transcytotically), or nonpolarly to the apical and/or basolateral surface of an epithelial cell (modified from Le Bivic et al., 1990).

Solutions

1. Starvation medium: DMEM without methionine or cysteine. This solution is prepared using a MEM Select-Amine kit by not adding the methionine and cysteine.
2. [³⁵S]EXPRESS (methionine/cysteine:) or [³⁵S]cysteine: 1 mCi per multiwell plate (12 x 1.2-cm or 6 x 2.4-cm-diameter filters). In some cases, proteins are effectively labeled with [³⁵S]SO₄, an advantage for studies of post-Golgi sorting, because the addition of sulfate occurs in the trans-Golgi network (see chapter by Kreitzer et al.).
3. Chase medium: DMEM containing a 10x concentration of methionine and cysteine (made by addition of methionine and cysteine to starving medium) or the normal medium in which the cells grow.
4. HCO₃-free DMEM containing 20mM HEPES and 0.2% bovine serum albumin (BSA)
5. Sulfo-NHS-biotin (NHS-LC-biotin or NHS-SSbiotin): 0.5 mg/ml in PBS-CM + all of the reagents used in the cell surface biotinylation protocol.
6. Lysis buffer: 1% Triton X-100 in 20mM Tris, 150mM NaCl, 5mM EDTA, 0.2% BSA, pH 8.0, and protease inhibitors
7. Immunopure-immobilized streptavidin on agarose beads
8. 10% SDS: sodium dodecyl sulfate

Steps

1. Wash cells on filters three times with starvation medium and incubate for 20-40min in starvation medium. The starving period will depend on your cell type. MDCK cells work well with a 20-min starvation; RPE-J require longer times.
2. Pulse for 20-30min (again depends on cell line) in starving medium containing [³⁵S]EXPRESS of [³⁵S]cysteine at 37°C. The pulse solution is starving medium plus the ³⁵S label. Minimal volumes are recommended. For MDCK cells we pulse with 20-40µl from the basolateral surface. A drop of pulse medium is placed on a strip of Parafilm in a humidified chamber (we use a plastic box lined with wet towels), and the insert containing the filter and MDCK monolayer is removed from the multiwell and dropped on top of the pulse medium. Some cell lines (i.e., RPE-J) are better labeled from the apical surface. For apical pulse, starvation medium is removed from both chambers and pulse medium is applied only to the apical chamber. For 1.2-cm-diameter filters we use a 100-µl volume; for 2.4-cm-diameter filters we use a 350-µl volume of pulse medium.
3. The pulse is terminated by washing with chase medium three times.
4. At different chase times, aspirate chase medium and replace with ice-cold NaHCO₃-free DMEM containing 20mM HEPES and 0.2% BSA and store on ice until all chase points have been collected.
5. Proceed to apical or basolateral biotinylation following the protocol described earlier for cell surface biotinylation.
6. Excise filters and either freeze at -80°C or immediately lyse cells and immunoprecipitate specific proteins as described in the section extraction of biotinylated proteins.
7. Remove immunoprecipitated proteins from beads by adding 40µl of 10% SDS and heating for 5 min at 95°C. Immediately dilute with 460µl of lysis buffer and pellet for 1 min in a microfuge. Remove 450µl and place in a new tube. Dilute the remaining 50µl with 50µl of 2x Laemmli sample buffer. This sample is used to normalize for differential incorporation of radiolabel from filter to filter. The remaining 450µl is diluted with a further 1 ml of lysis buffer to which is added an additional 50µl of streptavidin agarose that has been preblocked for 1-12h with lysis buffer.
8. Streptavidin precipitation is allowed to proceed for 1 h to overnight at 4°C. Then the beads are washed successively in TPII, TPIII, and TPIV as described in Section III,E After the final wash the beads are resuspended in Laemmli sample buffer and heated to 95°C for 5 min.
9. Both the sample representing total and surface protein are resolved on SDS-PAGE gels. The gels are dried and exposed for autoradiography.

C. Targeting Assay by Surface Immunolabeling
Epithelial cells such as LLC-PK1 may not form a tight monolayer and hence could be leaky to biotin analogs (MW~ 400-600). To overcome this problem, we have used antibody labeling against the protein of interest (Gan et al., 2002). Antibodies are less likely to traverse through leaky monolayers. Leakiness of antibodies should be directly determined before using this method.

Solutions

1. PBS-CM: See Section III,A
2. DMEM containing 0.2% BSA

Steps

1. The initial steps are similar to Section III,B (steps 1-4). Label the ice-cold filters from different chase time points with antibody added to either apical or basolateral domains for 1 h on ice in a cold room kept at 4°C. Dilute the antibody in DMEM containing 0.2% BSA at an approximate concentration of 1 µg/ml. After 1 h, wash filters four times in ice-cold PBS-CM containing 0.2% BSA.
2. Excise filters and either freeze at -80°C or immediately lyse cells in lysis buffer.
3. Pull down the antigen-antibody complex with protein A or G beads from nine-tenths of the postnuclear supernatants. Subject one-tenth of the supernatant again to immunoprecipitation to measure total labeled protein.
4. Wash immunoprecipitates on the beads successively in TPII, TPIII, and TPIV as described in Section III,E
5. After the final wash, resuspend the beads in Laemmli sample buffer and heat to 95°C for 5 min.
6. Resolve both the sample representing total and surface proteins on SDS-PAGE gels. Dry and expose the gels for autoradiography.

Analysis of Results
The polarity of the protein is determined at each time point by densitometric analysis of the autoradiographic data. The values obtained for the surface protein should be normalized against the values obtained form the totals (including precursor forms). This controls for differences in the incorporation of label (specific activity) between monolayers. If the protein is highly polarized from the first time point at which it is detected on the cell surface, then it is delivered directly to that surface. If it is polarized on one surface early in the chase and then switches polarity later in the chase, then it is delivered indirectly. If the protein is nonpolar early in the chase and acquires polarity only after longer chase times, then it is not sorted in the TGN, but its final polarity is acquired by differential stability on the apical and basolateral surfaces.

D. Biotin Assay for Endocytosis
This assay examines the internalization of plasma membrane proteins (from Graeve et al., 1989)

Solutions

1. PBS-CM
2. Cleavable biotin reagent: NHS-SS-biotin
3. DMEM containing 0.2% BSA
4. Reducing solution: 310mg glutathione (free acid) dissolved in 17ml H₂O (50mM). Add 1 ml of 1.5M NaCl, 0.12ml of 50% NaOH, and 2ml of serum just before use.
5. Quenching solution: 5mg/ml iodoacetamide in PBS-CM containing 1% BSA

Steps

1. Wash cells on filters four times, 15 min each time with ice-cold PBS-CM.
2. Add 1 ml of NHS-SS-biotin (0.5mg/ml in icecold PBS-CM) to the chamber being labeled and PBSCM to the other chamber. Incubate for 20min at 4°C and repeat with fresh solutions.
3. Wash filters twice with DMEM/0.2% BSA. Keep two filters on ice (one of these will represent the total amount of proteins at the surface before internalization and the other will be treated with the reducing solution and represents your control of efficiency of reduction) and transfer the other filters to 37°C for various times to allow the biotinylated proteins to be internalized.
4. Stop incubation by transferring filters back to 4°C.
5. Wash twice in PBS-CM + 10% serum.
6. Incubate filters for 20min in reducing solution. Repeat. (Mock treat one filter.)
7. After washing, quench free SH groups in 5mg/ml iodoacetamide in PBS-CM + 1% BSA for 15 min.
8. Lyse cells and immunoprecipitate as described in the section extraction of biotinylated proteins.
9. Run the samples on SDS-PAGE gels, transfer to nitrocellulose or PVDF, blot with [¹²⁵I]streptavidin, and expose for autoradiography.

Analysis of Results
Endocytosis of the protein of interest is indicated by protection of the NHS-SS-biotin-labeled surface protein from reduction by glutathione. By chasing the cells for various lengths of time, a rate of endocytosis can be calculated by comparing the percentage of protein protected at each time point. Obviously if none of the protein is protected from reduction, it is 100% at the surface; conversely, if all of the protein is protected, 100% has been internalized.

E. In Situ Domain Selective Biotinylation of Retinal Pigment Epithelial Cells
The retinal pigment epithelium is uniquely suited for biochemical studies of polarity in situ. The RPE exists as a natural monolayer with a broad apical surface that is easily exposed after gentle enzymatic treatment to remove the adjacent neural retina. The apical surface of the tissue is labeled with a noncleavable biotin analog such as NHS-LC-biotin for the identification of apical proteins. For identification of basolateral proteins the biotinylatable sites on the apical surface are labeled with the cleavable NHS-SSbiotin. After isolation of RPE cells, the nonlabeled basolateral proteins are labeled in suspension with the noncleavable form. Removal of the cleavable NHS-SSbiotin by reduction with 2-mercaptoethanol results in the presence of biotin only in the population of proteins present on the basolateral surface of the RPE (Marmorstein et al., 1996)

Solutions

1. Sulfo-NHS-biotin, or NHS-LC-biotin, and NHS-SSbiotin stock solutions in DMSO
2. HBSS: 10mM HEPES buffered Hank's balanced salt solution
3. PBS-CM
4. DMEM containing 10 mM HEPES
5. Bovine testicular hyaluronidase
6. CMF-PBS: PBS calcium and magnesium free
7. PBS-EDTA: CMF-PBS + 1 mM EDTA

Steps
All steps are carried out on ice unless otherwise indicated.

1. Rats are euthanized by CO₂ asphyxiation, and the eyes are enucleated and stored for 3 h to overnight in the dark on ice in HBSS.
2. A circumferential incision is made above the ora serrata, and the cornea, iris, lens, and vitreous are removed.
3. The eyecups are incubated for 10-30 min at 37°C in HBSS containing 290 units/ml bovine testicular hyaluronidase.
4. The ora serrata is removed, and the neural retina is peeled carefully away from the RPE. The optic nerve head is severed and the neural retina is removed. The RPE is inspected under the dissecting microscope. Black spots on the outer surface of the retina or tracts of smooth reflective surface in the eyecup indicate damage. Damaged eyecups are discarded.
5. Soluble components of the interphotoreceptor matrix are removed by incubation in 2-ml microcentrifuge tubes (one eye per tube) on a rotator in ice-cold HBSS for 20min. This is repeated three times.
6. The apical surface of the RPE in one eyecup is biotinylated with 1 ml of PBS-CM containing 2mg of sulfo-NHS-biotin. The other eyecup is biotinylated with 1 ml of PBS-CM containing 2mg of NHS-SSbiotin. This procedure is repeated three times.
7. The reaction is quenched with 1 ml of 10mM HEPES buffered DMEM for 10min at 4°C.
8. The eyecups are rinsed once in ice-cold CMF-PBS and are then incubated in PBS-EDTA on ice for 30 min.
9. The RPE is gently teased from the inner surface of the eyecup using a 22-gauge needle. The RPE layer is collected in a 1.5-ml microcentrifuge tube and pelleted for 10 s in a microfuge. Cells labeled apically with noncleavable sulfo-NHS-biotin or NHS-LC-biotin at this stage are held on ice until the basolateral samples are ready.
10. For cells labeled with cleavable NHS-SS-biotin, the pellet is resuspended in 1 ml of PBS-CM containing 2mg/ml sulfo-NHS-biotin or NHS-LC biotin and incubated on a rotator at 4°C. After 20min the cells are pelleted for 10s in a microfuge and this step is repeated.
11. The reaction is quenched with 50 mM NH₄Cl in PBS-CM for 10min. The cells are then pelleted in the microfuge for 10 s.
12. At this point, both apical and basolaterally labeled pellets are frozen dry at -80°C or immediately lysed and specific proteins immunoprecipitated as described in the section extraction of biotinylated proteins.
13. Immunoprecipitated proteins are resuspended in Laemmli sample buffer containing 5% 2- mercaptoethanol or 50mM dithiothreitol to release the NHS-SS-biotin from the apical proteins in basolateral samples. After heating to 95°C for 5 min, samples are resolved by SDS-PAGE, transferred to nitrocellulose or PVDF membranes, and blotted with streptavidin.

Analysis of Results
Analysis of the results proceeds as in the section on cell surface biotinylation.

F. Extraction and Immunoprecipitation of Biotinylated Proteins

Solutions

1. TPI: 1% Triton X-100, 20mM Tris, 150mM NaCl, 5 mM EDTA, pH 8.0, containing 0.2% BSA, and protease inhibitors
2. TPII: 0.1% SDS, 20mM Tris, 150mM NaCl, 5 mM EDTA, pH 8.0, containing 0.2% BSA
3. TPIII: 20mM Tris, 500mM NaCl, 5 mM EDTA, pH 8.0, containing 0.2% BSA
4. TPIV: 50mM Tris, pH 8.0
5. Laemmli sample buffer
6. Protein A-Sepharose
7. Staphlycoccus A cells: Pansorbin

Steps

1. Excise filters from inserts using a #11 scalpel blade or razor blade. Lyse in 1 ml of TPI at 4°C. In some cases it may be necessary to use more stringent conditions for lysis (i.e., RIPA buffer, which contains 0.5% deoxycholate and 0.1% SDS in addition to 1% Triton X- 100).
2. Wash 100 µl/sample of Pansorbin three times with TPI. Do not omit protease inhibitors.
3. Centrifuge lysate in a microfuge at 4°C at 13,000g for 10min. Collect the supernatant and discard the pellet.
4. Add 100µl of washed Pansorbin to each supernatant. Preclear for 1h at 4°C.
5. Resuspend 5-10 mg of protein A-Sepharose/ lysate in 1 ml TPI /lysate. If you are immunoprecipitating with a mouse IgG, after 10min, when the Sepharose beads are swollen, add 2mg of rabbit antimouse IgG. After 1 h wash the beads three times with TPI. On the last wash, pellet the beads in microcentrifuge tubes.
6. Pellet the Pansorbin by centrifugation at 13,000g for 10min. Collect the supernatant, add an appropriate volume of antibody, and incubate at 4°C for 1 h to overnight.
7. Transfer the immunoprecipitates to the tubes containing the protein A-Sepharose and incubate for 1 h at 4°C.
8. Centrifuge the immunoprecipitates in a microfuge at 13,000g for 30s. Remove the supernatant and resuspend the beads in 1 ml of TPI. Repeat this step three times with TPII, three times with TPIII, and once with TPIV.
9. Resuspend with an appropriate volume of Laemmli sample buffer, heat to 95°C for 5min, and resolve by SDS-PAGE. For streptavidin blotting, transfer the gel to nitrocellulose or PVDF.

G. Streptavidin Blotting
Solutions

1. Blocking buffer: 5% Carnation instant milk, 0.3% BSA, in PBS-CM
2. Rinse buffer: 1% BSA, 0.2% Triton X-100 in PBSCM
3. [¹²⁵I]Streptavidin or streptavidin conjugated to horseradish peroxidase or alkaline phosphatase (streptavidin- HRP)
4. A phosphorimager, Kodak X-OMAT AR film and a cassette with an intensifying screen, or an enhanced chemiluminescence reagent kit (such as those supplied by Amersham) and appropriate film.

Steps

1. After transfer to nitrocellulose or PVDF (PVDF is superior for chemiluminescent detection systems and is used in our laboratory for most streptavidin blots), block the blot for 1 h in blocking buffer.
2. Rinse once with rinse buffer.
3. Incubate for 1 h in 40ml rinse buffer containing 1-2 × 10⁶ cpm/ml of [¹²⁵I] streptavidin or 0.5- 1.0 mg/ml streptavidin-HRP.
4. Wash three to five times for 5-10min each with rinse buffer.
5. Dry the blot and expose to a phosphorimager screen, autoradiograph it with Kodak X-OMAT AR film, or use any of the many enhanced chemiluminescent kits that are available.

H. Quantitation of Polarized Distribution by Confocal Microscopy Imaging Technique
Modern quantitative optical microscopy offers an alternative method to quantify the relative polarity of fluorescently labeled cell surface proteins in cells developing polarity in settings such as the calcium switch assay (Rajasekaran et al., 1996). This quantitation is based on the relative fluorescent pixel intensities in serial horizontal confocal sections of the target protein to a known polarized marker.

Requirements

1. Confocal laser-scanning microscope
2. Software to measure fluorescence intensity, e.g., Metamorph, Image Space Software.

Steps

1. Samples subjected to immunofluorescence technique (not discussed here) are imaged on a laser confocal microscopic system. Samples are subjected to optical sectioning (xy) of the entire thickness (z) of the monolayer. We select an interval between sections for optimized collection of fluorescence from a given plane without contribution from the neighboring z planes. Choice of interval depends on pinhole dimension, which in turn depends on characteristics of the excitation wavelength. [Refer to a handbook on confocal microscopy, e.g., Pawley (1995), or a confocal microscope manufacturer's manual for optimizing microscopic parameters.]
2. For each optical section, quantify the average per-pixel fluorescence intensity of the labeled proteins using the imaging software. Determine the ratio of intensity obtained for the protein of interest to that of a known marker.

Analysis of Results
The ratio of fluorescent intensity obtained represents the relative distribution of the target protein and is interpreted as follows.

1. A constant pixel intensity ratio for all optical sections of a given monolayer suggests overlapping distribution of the target protein with the known marker.
2. Decreasing pixel intensity ratio in from basolateral to apical domain of the target protein and the apical or tight junction marker would mean that the protein is localized in the basolateral domain.
3. Increasing intensity ratio with a basolateral marker would mean that the target protein is apically targeted (see Rajasekaran et al., 1996).

I. Determination of Plasma Membrane Protein Polarity after Intranuclear Microinjection of Its cDNA
This procedure is used for rapid qualitative determination of the steady-state polarity of a newly synthesized protein in polarized cells. Analysis is performed using either a fluorescence wide-field or a confocal microscope. We have utilized this technique to study the regulation of polarity of basolateral and apical membrane markers by GTPases and their downstream effectors. Normal polarization of the monolayer needs to be confirmed by studying the localization of known apical or basolateral markers.

Solutions

1. Microinjection buffer: H-KCl buffer containing 10 mM HEPES, pH 7.4, 140 mM KCl. Dissolve 1.04 g of KCl in 99ml deionized H₂O and then add 1.0ml HEPES from a 1M stock (pH 7.4). Sterilize the buffer by passing through a 0.22-µm filter and store at 4°C (Müsch et al., 2001).
2. PBS/CM: See Section III,A
3. H-DMEM: DMEM containing HEPES. Dissolve bicarbonate-free powdered DMEM in 900ml deionized H₂O. Add 20ml HEPES from a 1M stock and adjust the pH to 7.4. Sterilize the medium through a 0.22-µm filter and store at 4°C.
4. B-DMEM: DMEM containing sodium bicarbonate

Steps

1. Plate MDCK II cells on sterile glass coverslips at a concentration of 1.6 × 10⁶ cells/ml or at an approximate plating density of 2 × 10⁵ cells per cm² in BDMEM and 10% fetal bovine serum (FBS). Allow the cells to polarize for 4-5 days and change the medium only once on day 2 postplating. Growth conditions required to attain polarity for different cell lines vary and require optimization.
2. Dilute the stock of cDNA (stock prepared in deionized H₂O at a concentration of 0.5mg/ml) in microinjection buffer to a concentration of around 10µg/ml. It is highly recommended that cDNA constructs also contain a tag sequence, such as Myc, HA, GFP, and its variants, in-frame with the gene of interest so as to distinguish the newly synthesized proteins from endogenous proteins. For experiments involving more than one cDNA construct, cDNAs can be coinjected. However, the efficiency of expression of a construct may vary in the presence of another. Therefore, proper conditions should be established for good expression of each coinjected construct. A range of concentrations between 1 and 20 µg/ml of DNA should be tested to optimize their expression levels.
3. Prepare microinjection needles by pulling 1-mmdiameter and 6-in.-long borosilicate glass capillaries (1B100F-6, World Precision Instruments, Inc, Sarasota, FL) using a micropipette puller (e.g., Flaming/Brown Micropipette Puller Model P-97, Sutter Instrument Co., Novato, CA).
4. Load the cDNA diluted in microinjection buffer through the blunt end of the needle into the needle holder of the micromanipulator (Narishige Company, Ltd., SE-TAGAYA-KU, Tokyo, Japan) attached to the inverted microscope (Zeiss-Axiovert 25, Germany).
5. Transfer coverslip into 35-mm-diameter tissue culture dishes. Add 4ml of H-DMEM containing 5% FBS to each dish and place the dish on the dish holder of the micromanipulator-microscope described earlier. Microinject the nucleus of cells. Avoid microinjecting cells that are right next to each other. This simplifies the analysis of distribution of apical and basolateral markers. In order to avoid unsynchronized protein synthesis, preferably microinject within 10-15 min of transferring the dish to the microscope stage.
6. Incubate the microinjected cells with BDMEM- 10% FBS medium at 37°C. Most of the proteins accumulate in the ER within 60 x 90min at 37°C postmicroinjection. Different levels of expression should be tested to ensure that the sorting pathways are not saturated. For coinjections, it is necessary to standardize the conditions for sufficient expression of each protein. Adjustment of DNA concentrations (as described in step 2) and time of incubation at 37°C postmicroinjection are two steps that need to be tuned for expression of multiple constructs.
7. After appropriate incubation at 37°C, replace medium with B-DMEM-10% FBS containing 100 µg/ml cycloheximide (concentration may be lowered down to 20µg/ml if cells detach from coverslip) to inhibit new protein synthesis, the chase time for plasma membrane delivery of protein is initiated at 37°C for 3-4h.
8. After an appropriate chase period, fix cells with either -20°C chilled methanol for 10min or 2% paraformaldehyde at room temperature for 15min. Methanol fixation should be followed by a blocking step at room temperature with 1% BSA prepared in PBS-CM for 30min. Paraformaldehyde fixation of cells is followed by permeabilization at room temperature for 30min with either 0.2% Triton-X 100 or 0.075% saponin prepared in PBS-CM containing 1% BSA. Cells can now be processed for immunofluorescence with the appropriate primary and secondary antibodies.

Analysis of Results
Cells processed for immunofluorescence are imaged on either a confocal or a wide-field microscope. The correct orientation of the cells is determined by analyzing the staining of known polarized markers. In case of a wide-field microscope, the entire monolayer is subjected to z sectioning at least at 0.5-µm intervals with a 60x 1.4 NA objective, and standard deconvolution software is used to enhance the resolution. Alternatively, we use a confocal microscope in frame-scan mode and collect xyz stacks of the entire monolayer and display the xyz stack in the orthogonal plane. The cells can be displayed directly as a xz cross section by doing a line scan, i.e., scanning in the xzy mode. Localization of the protein is determined depending on the staining pattern, e.g., relative to a tight junctional marker such as ZO-1.

IV. COMMENTS
The methods described here represent examples of applications of the biotinylation technique; other examples of possible applications are (1) a transcytotic assay using a combination of the targeting and endocytosis protocols (Le Bivic et al., 1989; Zurzolo et al., 1992) and (2) detection of GPI-anchored proteins at the cell surface using Triton X-114 phase separation and PI-PLC digestion in place of the standard lysis procedure (Lisanti and Rodriguez-Boulan, 1990). Another analog of biotin, biotin hydrazide, can be used to label oligosaccharides of surface glycoproteins following periodate oxidation (Lisanti et al., 1989).

V. PITFALLS/RECOMMENDATIONS

1. It has been suggested that the use of pH 9.0 buffer to dilute sulfo-NHS-biotin would enhance the efficiency of labeling of surface proteins (Gottardi and Caplan, 1993). In our experience this is not always true and depends on different proteins and cell lines.
2. Always cut the filters out of the plastic holder before lysis. We have found that the cells can grow along the inside of the plastic ring supporting the filter. Lysis of these cells can result in erroneous results (Zurzolo and Rodriguez-Boulan, 1993).
3. Occasionally, in targeting experiments, intracellular nonbiotinylated forms are recovered on streptavidin beads. Using NHS-LC-biotin and keeping the SDS concentration at 0.4% helps reduce this. Another approach is to use NHS-SS-biotin and remove it from the streptavidin beads by incubation in 50mM dithiothreitol of 5-10% 2-mercaptoethanol in 62.5 mM Tris, pH 8.0. Then spin the beads out and dilute the supernatant 1:1 with 2x Laemmli sample buffer.
4. The in situ biotinylation assay works best for proteins that are restricted to the RPE cell (i.e., RET-PE2 antigen). Quantification of proteins present in adjacent tissues (particularly the choroid) can contaminate the basolaterally labeled material and yield an incorrectly high level of basolateral labeling.
5. Determination of polarity by confocal microscopy can provide visual validation of the biochemical assay. It is useful for measuring the polarity of steady-state protein and dynamic changes involved in tight junction assembly in the Ca²⁺ switch assay. A thorough analysis with known markers of different domains is recommended before determining polarity of the target protein.
6. For determination of polarity by means of intranuclear microinjection of cDNA, thorough optimization regarding the protein expression level and the incubation time is necessary. Moreover, it is critical to avoid saturating the sorting pathway. Hence, coinjecting and monitoring a known apical or basolateral marker are critical for final evaluation. Because the basolateral domain of the polarized monolayer on the coverslip is not accessible to the antibody without permeabilization, it is difficult to distinguish the pool of basolateral protein that is associated with submembrane structures closely juxtaposed to the cytoplasmic side of plasma membrane and the pool present in the external leaflet of the plasma membrane.

References
Gan, Y., McGraw, T.E., and Rodriguez-Boulan, E. (2002). The epithelial- specific adaptor APIB mediates post-endocytic recycling to the basolateral membrane. Nature Cell Biol. 4, 605-609.

Gottardi, C., and Caplan, M. (1993). Cell surface biotinylation in the determination of epithelial membrane polarity. J. Tissue Culture Methods 14, 173-180.

Graeve, L., Drickamer, K., and Rodriguez-Boulan, E. (1989). Functional expression of the chicken liver asialoglycoprotein receptor in the basolateral surface of MDCK cells. J. Cell Biol. 109, 2909-2816.

Hanzel, D., Nabi, I. R., Zurolo, C., Powell, S. K., and Rodriguez- Boulan, E. (1991) New techniques lead to advances in epithelial cell polarity. Semin. Cell Biol. 2, 341-353.

Keller R Simons K. (1997). Post-golgi biosynthetic trafficking. J. Cell Sci. 1103001-1103009.

Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685.

Le Bivic, A., Real, E X., and Rodriguez-Boulan, E. (1989). Vectorial targeting of apical and basolateral plasma membrane proteins in a human adenocarcinoma cell line. Proc. Natl. Acad. Sci. USA 86, 9313-9317.

Le Bivic, A., Sambuy, Y., Mostov, K., and Rodriguez-Boulan, E. (1990). Vectorial targeting of an endogenous apical membrane sialoglycoprotein and uvomorulin in MDCK cells. J. Cell Biol. 110, 1533-1539.

Le Gall, A., Yeaman, C., Muesch, A., and Rodriguez-Boulan, E. (1995). Epithelial cell polarity: New perspectives. Semin. Nephrol. 15(4), 272-284.

Lisanti, M., Le Bivic, A., Sargiacomo, M., and Rodriguez-Boulan, E. (1989). Steady state distribution and biogenesis of endogenous MDCK glycoproteins: Evidence for intracellular sorting and polarized surface delivery. J. Cell Biol. 109, 2117-2128.

Lisanti, M., and Rodriguez-Boulan, E. (1990). Glycosphingolipid membrane anchoring provides clues to the mecanism of protein sorting in polarized epiothelial cells. Trends Biochem. Sci. 113- 118.

Marmorstein, A. D., Bonilha, V. L., Chiflet, S., Neill, J. M., and Rodriguez-Boulan E. (1996). The polarity of the plasma membrane protein RET-PE2 in retinal pigment epithelium is developmentally regulated. J. Cell. Sci. 109, 3025-3034.

Mfisch A., Cohen D., Kreitzer G., and Rodriguez-Boulan E. (2001). cdc42 regulates the exit of apical and basolateral proteins from the trans-Golgi network. EMBO J. 20, 2171-2179.

Pawley J. B. (ed.) (1995). "Handbook of Biological Confocal Microscopy." Plenum Press, New York.

Rajasekaran, A.K., Hojo, M., Huima, T., and Rodriguez-Boulan, E. (1996). Catenins and zonula occudens-1 form a complex during early stages in the assembly of tight junctions. J. Cell Biol. 132, 451-463.

Rodriguez-Boulan, E., Kreitzer, G., and Muesch, A. (2005). Organization of vesicular trafficking in epithelia. Nature Rev. Mol. Cell Biol. 6, 233-247.

Rodriguez-Boulan, E., and Powell, S. K. (1992). Polarity of epithelial and neuronal cells. Annu. Rev. Cell Biol. 8, 395-427.

Sargiacomo, M., Lisanti, M., Graeve, L., Le Bivic, A., and Rodriguez- Boulan, E. (1989). Integral and peripheral protein compositions of the apical and basolateral plasma membrane domains of MDCK cells. J. Membr. Biol. 107, 277-286.

Yeaman, C., Grindstaff, K.K., and Nelson, W.J. (1999). New perspectives on mechanisms involved in generating epithelial cell polarity. Physiol. Rev. 79,73-98.

Zurzolo, C., Le Bivic, A., Quaroni, A., Nitsch, L., and Rodriguez- Boulan, E. (1992). Modulation of transcytotic and direct targeting pathways in a polarized thyroid cell line. EMBO J. 11, 2337- 2344.

Zurzolo, C., and Rodriguez-Boulan, E. (1993). Delivery of Na,KATPase in polarized epithelial cells. Science 260, 550-552.