Studying Exit and Surface Delivery of Post-Golgi Transport Intermediates Using in vitro and Live-Cell Microscopy-Based Approaches

I. INTRODUCTION
Study of the mechanisms of polarized protein sorting in epithelial cells has been facilitated greatly by the use of enveloped RNA viruses, such as vesicular stomatitis virus (VSV) and influenza virus, which bud from the basolateral and apical plasma membranes, respectively (Rodriguez-Boulan and Sabatini, 1978). Following infection, a rapid onset of viral protein synthesis occurs, leading to the vectorial transport of envelope glycoproteins to either the apical or the basolateral surface. This model continues to provide information on the mechanisms of protein sorting (Musch et al., 1996) and the basic protocols are included here. However, because cells cannot be coinfected efficiently with both types of viruses due to reciprocal inhibition of protein synthesis, a major drawback of this paradigm is the inability to study the segregation of apical and basolateral proteins from one another in the same cell.

Two alternative approaches have been developed that improve upon and greatly facilitate studying the molecular effectors of protein sorting in the trans-Golgi network (TGN) and polarized transport routes to the plasma membrane in epithelial cells. These approaches utilize either recombinant adenovirus vectors or intranuclear microinjection of cDNAs to introduce exogenous biosynthetic markers into cells. Both methodologies advance previous techniques in numerous ways: (i) they allow for high-level, simultaneous expression of two markers (Marmorstein et al., 2000); (ii) they are amenable to the use of temperature blocks, which allow for accumulation in and synchronous release of newly synthesized proteins from the TGN; (iii) neither method interferes with the ability of cells to synthesize and transport endogenous proteins, permitting the study of marker proteins in a normal cellular environment; (iv) adenoviral infection generally results in transduction of all cells in the culture and is thus ideal for metabolic labeling studies and biochemical analysis of biosynthetic events; (v) microinjection results in the rapid expression of cDNAs, providing a means by which to study anterograde membrane trafficking events selectively and dynamically in individual cells; and (vi) cDNAs or adenoviral particles can be introduced easily into a wide variety of cultured cells, making it relatively simple to compare secretory sorting pathways in different multiple cell types.

The most important advance provided by cDNA microinjection and adenoviral-mediated gene transfer in studying protein-sorting events is the ability to cointroduce two or more genes into cells and express simultaneously multiple secretory cargoes that follow divergent routes out of the TGN. This allows one to evaluate the role(s) of potential molecular effectors of protein sorting and targeting to different cellular domains. Furthermore, the level of expression of exogenous genes can be manipulated [by changing the amount of DNA introduced and the expression time allowed in microinjection-based assays or by varying either the multiplicity of infection (moi) or the time in culture following adenoviral infection] to allow study of the ability to saturate the various sorting pathways available to the cell (Marmorstein et al., 2000). Adenoviral infection also results in high-level expression of reporter proteins in large cell populations, a factor essential in obtaining sufficient incorporation of radioactive amino acids for pulse-chase studies, as well as for immunoisolation of transport vesicles. (Once the adenovirus has been titrated and the infection conditions optimized, 100% of the cells express the desired proteins and remarkably consistent pools of cells are produced from experiment to experiment.) Procedures for adenoviral infection can be modified and adapted easily to a variety of different cell lines.

While cDNA microinjection is limited with respect to the number of cells that can be evaluated, it is exquisitely suited to live cell imaging studies aimed at evaluating highly dynamic membrane trafficking events occurring at specific points throughout the biosynthetic pathway (e.g., ER-to-Golgi events and Golgi-to-plasma membrane events, such as budding of Golgi membranes, transport of post-Golgi carriers, and exocytosis of post-Golgi carriers). Momentum in this area is due primarily to the advent of fluorescent tags, such as green fluorescent protein [GFP(Chalfie, 1994)] from the jellyfish Aequorea and dsRed from the coral Discosoma (Matz et al., 1999; Baird et al., 2000), which can be genetically appended to any DNA of choice. These tags serve as vital fluorescent indicators that facilitate direct observation and analysis of membranetrafficking events in living cells.

This article describes an assay that monitors post- Golgi vesicle budding from semi-intact MDCK cells following infection either with enveloped RNA viruses or with recombinant adenovirus vectors. The adenoviruses we have found most useful for these applications encode receptors for neurotrophins (p75 NTR) and low-density lipoprotein (LDLR), which were shown previously to be sorted to the apical and basolateral surfaces of polarized MDCK cells, respectively (Le Bivic et al., 1991; Hunziker et al., 1991; Gridstaff et al., 1998). Each of these proteins, when expressed in MDCK cells, incorporates radiosulfate into carbohydrate moieties during posttranslational processing late in the secretory pathway, providing a convenient method to label markers in the sorting compartment.

In addition, this article describes methods we have developed pertaining to the use of time-lapse fluorescence imaging to study the transport of plasma membrane proteins through the biosynthetic pathway in living cells. While we only discuss this technique specifically with respect to MDCK epithelial cells, with slight modifications we have also successfully employed this approach for similar studies in several different cell lines. Additionally, we have found that these studies can be used with numerous reporter proteins of interest. Thus while we focus on reporters marking the apical or basolateral plasma membrane, the system is not limited to the study of membraneassociated proteins. Finally, this article discusses some quantitative measurements that can be made from this type of data related to protein-trafficking events occurring in vivo.

II. MATERIALS AND INSTRUMENTS
Dulbecco's modified Eagle's medium (DMEM, Cat. No. 10-013-CV), MEM nonessential amino acids (Cat. No. 25-025-CI), Hank's balanced salt solution (HBSS, Cat. No.21-023-CV), and L-glutamine (Cat. No. 25-005- CI) are from Cellgro. MEM SelectAmine kits (Cat. No. 19050), MEM vitamins (Cat. No. 11120), penicillinstreptomycin (Cat. No. 15140), 7.5% bovine serum albumin (BSA, Cat. No. 15260), and donor horse serum (Cat. No. 16050) can be purchased from Gibco-BRL (Grand Island, NY). Heat-inactivated fetal bovine serum (FBS, Cat. No. 100-106) is from Gemini Bioproducts. HEPES (Cat. No. H-4034), D-glucose (Cat. No. G-8270), and cycloheximide (Cat. No. C-7698) are from Sigma Chemicals. Polycarbonate filters (Transwells; 0.4mm pore size, Cat. No. 3412 for 24-mm filters) can be purchased from Corning Costar Corp. (Cambridge, MA). Tissue culture grade plasticware is from Corning Plasticware. Tran³⁵S-label (Cat. No. 51006), [³⁵S]cysteine (Cat. No. 51002), and H₂³⁵SO₄ (Cat. No. 64040) can be purchased from ICN (Costa Mesa, CA). Reagents for the production of viruses are described elsewhere (Rodriguez-Boulan and Sabatini, 1978; see article by Hitt et al.) All other reagents are standard reagent grade and available from several sources. Sterile solutions are autoclaved or sterilized by ultrafiltration (0.2 mm).

There are numerous manufacturers of microscope heating and cooling chambers. We use the Harvard apparatus recording chamber (PDMI-2) and temperature controller (TC-202A). A variety of microinjection apparatuses are available from Narishige Inc. or from Eppendorf. Cooled charged-couple device (CCDs) cameras are also available from several manufacturers. For high-resolution time-lapse imaging, we recommend a camera capable of 12- to 16-bit digitization (4095-65,000 gray levels) with a 5- to 10-MHz controller for rapid acquisition rates. The Flaming Brown Micropipet puller (Model P-97) is from Sutter Instruments. Glass capillaries for pulling microinjection needles are available from World Precision Instruments (Cat. No. 1B100F-6). The Sykes-Moore culture chamber, consisting of a 32 × 7-mm chamber (Cat. No. 1943-11111), silicone gaskets (Cat. No. 1943-33315), and the wrench assembly (Cat. No. 1943-44444) can be purchased from Bellco. Number 1 thickness, 25-mm glass coverslips for the chamber (Cat. No. 483800-80) are from VWR.

III. PROCEDURES
1. Coinfection of MDCK Cells with Recombinant Adenovirus Vectors
Solutions

1. Dulbecco's modified Eagle's medium (DMEM): Dissolve DMEM powder in 1 liter H₂O. Sterilize by filtering through a 0.2-µm pore filter.
2. Complete DMEM (cDMEM): Add 50ml heatinactivated fetal bovine serum (FBS), 10ml 0.2M L-glutamine, 5 ml penicillin-streptomycin (10,000 U / ml and 10mg/ml, respectively), and 5 ml MEM nonessential amino acids to 430ml DMEM. Store at 4°C for up to 2 weeks.

Steps

1. For infection with replication-defective viruses, grow cells on semipermeable polycarbonate filter supports. Seed cells at confluency on day 1 and culture for 4 days with medium changed daily. MDCK strain II cells are confluent at a density of ca. 7 × 10⁵ cells/cm², so approximately 3.3 × 10⁶ cells should be seeded on each 24-mm filter.
2. Before applying adenovirus vectors, rinse cultures twice with serum-free DMEM (2 ml for the apical chamber and 2.5ml for the basolateral chamber). Removal of serum proteins from the culture medium results in enhanced adsorption of adenovirus particles to the cell surface and improves infection efficiency. MDCK cells do not appear to be affected adversely by serum deprivation during the infection period, but this should be checked with other cell types before attempting infection.
3. Add recombinant adenoviruses, diluted in serum-free DMEM, at moi ranging from I to 1000pfu/ cell. Incubate at 37° for 1 h, gently tilting the plates every 15 min to mix. Two or more adenovirus vectors can be mixed together and applied simultaneously to cells. It is recommended that serial three-fold dilutions of viruses be tested in order to determine the optimal moi required for the quantitative expression of the marker proteins.
4. Vectors must be applied to the apical domain of epithelial cells, as infection is markedly more efficient from this surface. The reason for this is unclear, but a preference for adenovirus entry through the apical surface has been observed in every polarized epithelial cell type we have examined, including canine (MDCK), bovine (MDBK), and porcine (LLC-PK₁) kidney; rat thyroid (FRT) and retinal pigment epithelia (RPE-J); and human intestine (Caco-2). Serum-free DMEM, without virus, should be applied to the basolateral chamber. Infection should be performed in a minimum volume of DMEM required to keep the filter submerged. A recommended volume is 0.25 ml apical/0.5 ml basolateral for a 24-mm filter.
5. Following the infection period, add 2 ml cDMEM to both apical and basolateral chambers. It is not necessary to aspirate the virus because the addition of serum effectively terminates the infection. Culture the cells at 37°C for the desired incubation time. Expression of adenovirus-encoded proteins should be detectable by 4-6 h following infection, will rise gradually over the next 18 h and should reach a plateau by 24 h. This level of expression will be maintained for at least 1 week, provided the cells are fed daily. We routinely use cultures of polarized MDCK cells between 20 and 24 h postinfection.
6. Monitor infection after 24 h as follows:
   1. Use indirect immunofluorescent staining to determine (i) the percentage of cells in the culture expressing the transfected gene products and (ii) the intracellular distribution of the gene products. Under optimum infection conditions, at least 95% of the cells will express each adenovirus-encoded protein. If the ultimate goal of the experiment is the study of the molecular trafficking of two or more proteins in the same cell, it is essential that all of the cells that express one marker also express the second marker. Methods for fixation, permeabilization, and staining of epithelial cells grown on polycarbonate filters are described elsewhere in this manual.
   2. Use immunoprecipitation or Western blotting to determine (i) the molecular weight of the adenovirus-encoded proteins and (ii) the level of expression of the proteins in the culture. Ideally, both adenovirus-encoded proteins will be expressed at comparable levels.

Pitfalls
Each batch of adenovirus virus vector must be tested for optimum transduction. We find a batch-to-batch variability in the correspondence of pfu obtained from plaque assays to the moi needed.

B. Infection of MDCK Cells with Enveloped RNA Viruses
Solutions

1. DEAE-dextran (100X stock): Dissolve 100 mg DEAE-dextran (Sigma Cat. No. Dl162) in 10 ml H₂O. Filter, sterilize, and store 1-ml aliquots at -20°C.
2. Infection medium: Add 13ml 7.5% bovine serum albumin and 5ml 1M HEPES, pH 7.4, to 482ml DMEM. Filter, sterilize, and store at 4°C. Immediately before use, add 0.1ml DEAE-dextran stock to 10ml medium. DEAE-dextran is only necessary for infection with VSV.
3. Virus stocks: Vesicular stomatitis virus, Indiana strain (VSV), and influenza virus A (WSN strain) are grown in MDCK strain II cells, harvested, and plaque assayed as described by Rodriguez-Boulan and Sabatini (1978).

Steps

1. Set up 10-cm dishes of MDCK strain II cells, passages 6-20, and allow them to reach confluency. Cultures are infected with VSV or influenza virus 3 days after becoming confluent.
2. Before infecting cells, rinse cultures twice with infection medium. For viral infection, inoculate MDCK cells with 50pfu/cell VSV or influenza WSN in 3.5 ml infection medium containing 0.1 mg/ml DEAE-dextran.
3. Incubate cultures for 1 h at 37°C.
4. Aspirate viral medium and rinse cultures twice with fresh infection medium.
5. Return VSV-infected cultures to 37°C and incubate a further 3.5 h before metabolic radiolabeling (see later). Incubate influenza WSN-infected cultures for 4.5 h at 37°C before labeling. Cultures should be examined hourly to monitor cytopathic effects.

C. Metabolic Radiolabeling and Accumulation of Marker Proteins in the Trans-Golgi Network
1. Radiosulfate Labeling of Glycoproteins at 20°C
Both p75^NTR and LDLR are sulfated when expressed in MDCK cells. Sulfation occurs largely on asparaginelinked carbohydrate moieties on both proteins. Because this posttranslational modification occurs late in the secretory pathway, likely in the trans-Golgi or TGN, it provides a convenient method to label markers in the sorting compartment. When labeling is performed at the reduced temperature of 20°C, the labeled markers accumulate in the TGN because post-Golgi vesicular transport is inhibited.

Solutions

1. Sulfate-free labeling medium: Essentially, labeling medium is DMEM in which the MgSO₄ is replaced by MgCl₂. Combine 100 ml 10× DME salts (Ca²⁺, Mg²⁺- free), 10 ml 100× Ca²⁺, Mg²⁺ stock, 10 ml MEM Vitamins, 10ml of 100× stock MEM amino acid solutions (arginine, glutamine, histidine, isoleucine, leucine, lysine, phenylalanine, threonine, tryptophan, tyrosine, glycine, serine, and valine), 1 ml of 100× stock MEM solutions of methionine and cysteine, 10ml MEM nonessential amino acid solution, 20ml 1M HEPES, pH 7.4, 27ml 7.5% BSA stock, and H₂O to a final volume of 1000ml. Filter, sterilize, and store at 4°C.
2. 10× DME salts (Ca²⁺ Mg²⁺-free): Combine 50ml 100× Fe(NO3)3, 50ml 100× NaH₂PO₄, 50 ml 100 x KCl, 22.5 g dextrose, 32 g NaCl, and 15 ml phenol red solution. Adjust volume to 500ml with H₂O. Filter, sterilize, and store at 4°C.
3. Fe(N0₃)₃ stocks: Add 0.05 g Fe(NO₃)₃ to 50ml H₂O to prepare a 100,000× stock solution. Dilute 100 µl into 100 ml H₂O to prepare 100× stock solution. Filter, sterilize, and store at 4°C.
4. 100× Na H₂PO₄: Add 1.25g NaH₂PO₄ to 100ml H₂O. Filter, sterilize, and store at 4°C.
5. 100× KCl: Add 4.0 g KCl to 100 ml H₂O. Filter, sterilize, and store at 4°C.
6. 100× Ca²⁺, Mg²⁺: Add 2.96g CaCl₂·2H₂O and 3.02 g MgCl₂·6H₂O to 100ml H₂O. Filter, sterilize, and store at 4°C.

Steps

1. Twenty-four to 48 h following adenovirus infection, aspirate culture medium and rinse filters three times with sulfate-free labeling medium. Sulfate starve cells for 30min at 37°C in this medium.
2. Label cells for 1 h at 20°C in sulfate-free labeling medium containing H₂³⁵SO₄ . To label cells on one 75mm Transwell filter, we use 0.5mCi H₂³⁵SO₄ in 500µl sulfate-free labeling medium. Place medium, preequilibrated at 20°C, on a sheet of Parafilm in a humid chamber and place the Transwell filter upon this so that label is exposed to the basolateral surface. Apply 2.5ml sulfate-free labeling medium, without label, to the apical chamber to prevent drying.

2. Pulse-Chase Labeling with [³⁵S]Methionine/ Cysteine
Solutions

1. Methionine/cysteine-free labeling medium: Prepare 1000 ml of medium following the product specification insert (Gibco SelectAmine kit, Cat. No. 19050), excluding the methionine and cysteine in the kit. Add 10ml 1M HEPES and 27ml 7.5% BSA stock solution. Filter, sterilize, and store at 4°C.
   For PC12 cells, replace the BSA will 20ml dialyzed serum. To prepare the dialyzed serum, under sterile conditions, dialyze a mixture of two-thirds fetal bovine serum and one-third horse serum against PBS in 12,000 molecular weight cutoff dialysis tubing for 12-20h. Dialyze serum to remove small molecules, which may be used to scavenge sulfate.
2. Chase medium: Add 5 ml of 100× MEM methionine and cysteine (left over from the Selectamine kit) solutions to 40ml complete DMEM (or PC12 growth medium for PC12 cells). Immediately before use, add cyclohexamide to a concentration of 20µg/ml.

Steps

1. Rinse MDCK cultures three times in methionine/ cysteine-free labeling medium before incubation at 37°C for the final 30min of incubation following viral infection. For PC12 cells, extensive rinsing may not be possible, so rinse once before the 30-min incubation.
2. Label VSV-infected MDCK cells with [³⁵S]methionine/ cysteine (Tran³⁵S-label, ICN Cat. No. 51006). Label influenza WSN-infected cells with [³⁵S]cysteine (ICN Cat. No. 51002). Use 0.5 mCi, in a total volume of 3.5ml labeling medium, to label each 10-cm plate. Pulse-label cells for 10min at 37°C. Medium can be recycled twice if multiple dishes are to be labeled. Label LDLR- and p75-infected PC12 cells with [35S]cysteine, due to the fact that these proteins are not sulfated as they are in MDCK cells. Use 0.5 mCi, in a total volume of 1 ml labeling medium, to label each 10-cm plate. Pulse-label cells for 15 min at 37°C, with gentle rocking every 5 min.
3. For MDCK cells, aspirate labeling medium and rinse plates three times with chase medium. For PC12 cultures, just aspirate medium.
4. Chase cultures at 20°C for 2h in chase medium. During the chase at 20°C, roughly 60% of the labeled VSV G/influenza HA proteins are accumulated in the TGN (Muesch et al., 1996).

D. Vesicle Budding from the TGN in Semi-intact Cells
Semi-intact MDCK cells are prepared after accumulating marker proteins in the TGN (Muesch et al., 1996). Cells are first swollen in a low salt buffer and are subsequently scraped from the substratum, which produces large tears in the plasma membrane. Endogenous cytosol and peripheral membrane proteins are removed by washing with a high salt buffer. Addition of an exogenous source of cytosol, an energyregenerating system, and incubation at 37°C typically result in the release of 25-65% of the total marker accumulated in the TGN into sealed vesicles. Budded vesicles are separated from the material that remains by a brief, low-speed centrifugation step.

Solutions

1. Swelling buffer: Add 7.5 ml 1M HEPES/KOH, pH 7.2, and 7.5 ml 1M KCl to 485 ml H₂O. Store at 4°C.
2. 10x transport buffer: Add 20ml 1M HEPES/KOH, pH 7.2, 2 ml 1M Mg(OAc)₂, and 18 ml 5 M KOAc to 60 ml H₂O. Store at 4°C.
3. 1× transport buffer: Add 1 ml 10× transport buffer to 9ml H₂O. Immediately before use, bring to 1 mM DTT and add protease inhibitors to 1× concentration.
4. High salt buffer: Add 10ml 1M HEPES/KOH, pH 7.2, 50ml 5M KOAc, and 1 ml 1M Mg(OAc)2 to 4391 ml H₂O. Store at 4°C. Immediately before use, bring to 1 mM DTT and add protease inhibitors to 1X concentration.
5. 1M dithiothreitol stock
6. 500× protease inhibitor stock: Dissolve 5mg of each of the following inhibitors individually in 330µl dimethyl sulfoxide (DMSO) and combine the three solutions.
   1. Pepstatin A: (Sigma Cat. No. P-4265)
   2. Leupeptin: (Sigma Cat. No. L-8511)
   3. Antipain: (Sigma Cat. No. A-6191)
7. 100 mM PMSF stock
8. Energy mix: Pipette in the following order: 3µl ATP, 2 µl GTP, 4 µl creatine phosphate, and 3 µl creatine kinase.
   1. 0.1M ATP (Boehringer Mannheim Cat. No. 519 987)
   2. 0.2M GTP (Boehringer Mannheim Cat. No. 106 399)
   3. 0.6M creatine phosphate (Boehringer Mannheim Cat. No. 621 722)
   4. 8 mg/ml creatine kinase (Boehringer Mannheim Cat. No. 127 566)
9. Bovine brain cytosol: Prepare gel-filtered bovine brain cytosol in batches exactly as described previously (Malhotra et al., 1989). The protein concentration should be 10-20mg/ml. Snap freeze 50-µl aliquots in liquid nitrogen and store at -80°C.

Steps
All steps are performed on ice, unless otherwise specified.

1. Following the 20°C incubation, wash the monolayer twice briefly with ice-cold swelling buffer and incubate in the same for 15 min.
2. Scrape cells from the filter (or plastic dish) with a rubber policeman into 2.5 ml transport buffer. DiSPo scrapers (Baxter Scientific, McGaw Park, IL) work well for this purpose. The best way to scrape cells from the Transwell filter is to place the filter inside the lid of the dish so that the bottom lies flat against the plastic. This prevents the scraper from poking through the filter, but care must be exercised to prevent tearing the filter. Scraping does not have to be vigorous, but should be done with long, gentle strokes. Transfer cells to 1.5-ml microfuge tubes and rinse the filter (or dish) with 2.5 ml fresh transport buffer. Combine with cells from first scraping. Discard the filter as radioactive waste.
3. Pellet cells by centrifugation at 800g for 5 min in a refrigerated microfuge.
4. Pool semi-intact cells into one tube and wash with 1.5 ml high salt buffer on ice for 10 min. Pellet cells by centrifugation at 800g for 5min in a refrigerated microfuge.
5. Resuspend cells in transport buffer. A volume of 250µl is used to resuspend cells from one filter (or dish).
6. Set up vesicle budding assay.
   1. In standard assays, suspend semi-intact cells (ca. 10µl = 20-25µg protein) in an assay volume of 50µl transport buffer supplemented with 50µg gel-filtered bovine brain cytosol and an energy-regenerating system (1 mM ATP, 1 mM GTP, 5mM creatine phosphate, and 0.2IU creatine kinase). Combine components in the following order: mix 26µl H₂O, 1 µl 100 mM PMSF, 4 µl 10× transport buffer, 4 µl energy mix, 5 µl cytosol (final concentration = 1 mg/ml), and 10µl semi-intact cells.
   2. Important controls include assays in which either the cytosol or the energy mix or both components are omitted. For the complete depletion of energy from the system, cytosol and semi-intact cells must be preincubated for 10 min on ice with 0.6 U/ml apyrase before assembling the assay. In the absence of either cytosol or energy, vesicular release from semiintact cells should be negligible. We suggest that serial dilutions of cytosol (i.e., 0.1-10 mg/ml) be tested in order to determine the optimal range of cytosol-dependent vesicle budding in the assay.
7. Incubate assays at 37°C for desired time. In standard assays, we incubate for 30-45 min. However, we suggest that a time course of vesicle budding be performed to optimize the assay for different marker proteins. As additional controls, two complete assays should be assembled and incubated at 0 and 20°C. At these reduced temperatures, vesicular release from semi-intact cells should be insignificant.
8. Pellet semi-intact cells by centrifugation at 800g for 5 min in a refrigerated microfuge. Transfer supernatant fractions to clean microfuge tubes. The pellets (containing nonbudded material remaining in the TGN) and supernatants (containing the vesicles released during the 37°C incubation) can be analyzed further.
   1. To quantify the efficiency of vesicular release of each marker under different conditions, samples can be lysed in SDS-PAGE sample buffer and analyzed directly by PAGE. In cells infected with VSV or influenza WSN, the viral proteins should be the only labeled proteins in the lysates. Alternatively, following adenovirus- mediated transfer of cDNAs encoding p75^NTR and LDLR into MDCK cells, these proteins are by far the most heavily labeled proteins when radiosulfate is used as a precursor.
   2. To confirm that markers are present inside sealed vesicles, the supernatant fraction should be treated with either proteinase K or trypsin. In the absence of Triton X-100, only the cytoplasmic domains of the proteins will be cleaved and this can be detected as a relatively small mobility increase during SDS-PAGE. In contrast, protease treatment in the presence of 1% Triton X-100 will result in complete digestion of markers. Use three 50- µl assay samples for this analysis. To tube 1, add nothing. To tube 2, add 2.5µl protease (10mg/ml stock → 0.5mg/ml final). To tube 3, add 2.5µl protease and 2.5µl 20% Triton X-100. Incubate on ice for 30min. Inactivate protease with 1 mM PMSF or 1mg/ml soybean trypsin inhibitor before lysing samples in SDS-PAGE sample buffer. Analyze products by SDS-PAGE.
   3. Immunoisolation of specific classes of transport vesicles is performed using antibodies against the cytoplasmic portions of cargo proteins as well as appropriate negative controls. (i) Use 5 mg protein A-Sepharose (Pharmacia, Piscataway, NJ; Cat. No. 17-0780-01) for each immunoisolation. Swell in transport buffer for 10min. (ii) If using a murine monoclonal primary antibody, use a bridge. Incubate 5 mg protein A-Sepharose with 50µg rabbit antimouse IgG (Rockland Labs, supplied by VWR, New York, Cat. No. 610-4102) in 1 ml transport buffer for 60min at room temperature. Wash twice with transport buffer. (iii) Block nonspecific binding sites in 1 ml transport buffer containing 0.2% BSA for 60min at room temperature. (iv) Couple primary antibody for 2 h at room temperature in transport buffer. Wash twice with transport buffer. Titrate each antibody to determine amount needed for quantitative recovery of vesicles. (v) Incubate immunoadsorbant with the supernatant fraction from the vesicle budding assay in a total volume of 1 ml transport buffer for 2-18 h at 4°C with end-over-end rotation. In some cases, vesicle coat proteins may mask epitopes on the cytoplasmic tails of cargo. Therefore, it may be necessary to wash the vesicles in high salt buffer prior to immunoisolation to strip coat proteins. Add 33µl of 1M KOAc to 50-µl vesicles. Incubate on ice for 10min. Add 333µl salt-free transport buffer [20 mM HEPES/KOH, 2 mM Mg(OAc)₂]. (vi) Wash immunoprecipitates six times with transport buffer. Elute bound markers by boiling 5min in SDS-PAGE sample buffer. Analyze by SDS-PAGE.

E. Vesicle Budding from TGN-Enriched Membranes
A TGN-enriched membrane fraction is prepared from metabolically labeled PC12 cells after marker proteins have been accumulated in the TGN. Initially, a postnuclear supernatant is prepared following the method of Tooze and Huttner (1992). From the postnuclear supernatant a TGN-enriched membrane fraction is prepared (Xu et al., 1995). As with semi-intact MDCK cells, addition of an exogenous source of cytosol and an energy-regenerating system leads to the release of accumulated marker protein from the TGN in sealed vesicles. All steps for the vesicle budding assay are identical to those for intact MDCK cells except that 50-µg aliquots of TGN-enriched membranes are used for each individual assay condition.

F. Expression of cDNA Using Microinjection
Solutions

1. Preparation of cDNA stocks for microinjection: Prepare DNA using either a midi or maxi preparation, making sure to suspend the DNA pellet (final concentration of at least 0.2mg/ml) in sterile water rather than Tris-EDTA. DNA stocks can be stored at either 4 or -20°C.
2. HEPEs:KCl microinjection buffer: 10 mM HEPES, 140 mM KCl, pH 7.4
3. cDNA for microinjection: Dilute cDNA stock in HKCl to a final concentration of 5-20µg/ml (see later for how to choose a concentration)
4. MDCK cell culture medium: DMEM prepared as per manufacturer's instructions. Add 50ml FBS (10% FBS final concentration) and 10ml of 1M HEPES, pH 7.4 (20mM final concentration), and 5ml of MEM nonessential amino acids to 435 ml DMEM.

Steps
For microinjection, cells must be cultured on sterilized glass coverslips.

1. Sterilize coverslips using either of the following methods.
   1. Autoclaving: Place coverslips in a glass petri dish and autoclave on the dry cycle for 20 min. If cells adhere well to the glass, this method is most convenient, as many coverslips can be sterilized at once.
   2. Acid washing: Place coverslips in histology staining racks. Wash coverslips in a beaker of 2N hydrochloric acid (2 × 5-min washes), rinse in distilled water (2 × 5-min rinses), and wash in 100% ethanol (3 × 2min). After the final ethanol wash, place the staining racks containing the cleaned coverslips into a dry beaker, cover with heavy-duty foil, and bake at 250°C for 1 h. Acid treatment "etches" the glass, creating a somewhat rough surface that is useful if cells are not sufficiently adherent on the coverslips.
2. Place the sterilized coverslips into a 10-cm tissue culture dish. Seven 25-mm coverslips can be placed in one 10-cm dish.
3. Trypsinize cells and seed onto coverslips placed previously in the 10-cm culture dish(es).
   1. For experiments in nonpolarized cells: Seed 5 × 10⁵ cells onto coverslips. Use a cell stock that is actively dividing (i.e., sparse cells). Cells should be used between 36 and 48 h after they are seeded onto the coverslips. Do not use the cells prior to 36 h postplating.
   2. For experiments in polarized cells: Seed cells at confluency onto coverslips. Change the medium 24 h after plating and culture the cells another 2-4 days prior to use. Do not change the culture medium again as this can result in changes in cell morphology.
4. On the day of the experiment, transfer coverslips to 3.5-cm culture dishes. Add fresh medium and microinject the cDNA into cell nuclei. As one goal of these experiments is to achieve synchronized exogenous protein expression in a population of cells it is important that you only inject cells on the same coverslip for -5 min. If a sufficient number of cells was not injected in that time, take another coverslip from the incubator and repeat.
5. After injection, place cells into the incubator and wait for protein to be expressed.
6. Monitor and identify the minimum expression time.¹ We routinely find that 1 h is sufficient for the expression of many different cDNAs in both nonpolarized and polarized MDCK cells. However, we have also found that there is heterogeneity in expression time depending on the cDNA being injected, as well as on the cell types being used. Minimum expression time can be determined by merely looking at the injected cells at a series of time points (e.g., 1-h intervals) after microinjection. Fluorescently tagged reporter proteins can be visualized directly without fixation. If your reporter is not fluorescently tagged, then fix the cells at 1-h intervals after microinjection and immunostain with antibodies against the exogenous protein.
   
   
   ¹ The minimum expression time is the time at which exogenous, newly expressed protein is present in the endoplasmic reticulum, but is not yet found at the plasma membrane. It is critical that no exogenous protein is at the cell surface for studies of Golgi to plasma membrane trafficking.

G. Synchronizing Transport through the Biosynthetic Pathway
Solutions

1. Bicarbonate-free MDCK cell culture medium: DMEM without bicarbonate prepared as per manufacturer instructions. Add 25 ml FBS (5% FBS final concentration), 10ml of 1M HEPES, pH 7.4 (20mM final concentration), and cycloheximide (final concentration 100µg/ml) to 475ml DMEM.
2. Recording medium: Hank's balanced salt solution with calcium and magnesium (HBSS-CM). Add 5ml FBS (1% FBS final concentration; serum does autofluoresce so it is important to keep the concentration low), 25 ml of 9% D-glucose (4.5 g/liter glucose final concentration), 5ml HEPES, pH 7.4 (10mM HEPES final concentration), and cycloheximide (final concentration 100µg/ml) to 465 ml HBSS-CM.

Steps
Synchronization of protein trafficking can be achieved through the use of a series of temperature shifts. Cycloheximide added during the temperature shifts will effectively create a pulse of newly synthesized protein that can be chased synchronously from ER to Golgi and from the Golgi to the plasma membrane. Newly synthesized protein can be accumulated in the ER or the Golgi when cells are incubated at 15 or 20°C, respectively. These proteins will leave the Golgi when cells are shifted to the permissive temperature for secretion, (30-37°C). Because most laboratories do not maintain a tissue culture incubator set to 15 or 20°C (our laboratory uses a small refrigerator set to the desired temperature), you will need to use a bicarbonate-free medium during these incubation periods. Check that this medium remains between pH 7.2 and 7.4 during the course of the temperature blocks.

One hour after microinjection (or the minimum expression time for your protein of interest, see earlier discussion), place cells into bicarbonate-free medium with cycloheximide.

1. To study ER to Golgi events, place cells into recording medium and incubate at 15°C in the thermally regulated recording chamber mounted on a microscope for time-lapse imaging. (For a more extensive discussion on equipping an imaging work station, see Mikhailov).
   1. Monitor the total cellular fluorescence by acquiring images of your cells at 10-min intervals. When total cellular fluorescence stabilizes for ~10min, acquire and save both transmitted light and fluorescent images of the cells you will be studying.
   2. Increase the temperature of the recording chamber to 20°C. Wait 5 min for the temperature and focus to stabilize.
   3. Acquire time-lapse images to evaluate ER to Golgi transport events. ER to Golgi transport can be studied at either low or high spatial and temporal resolution depending on the questions being addressed.
2. To study Golgi to plasma membrane events, place cells into bicarbonate-free medium and incubate at 20°C.
   1. Determine the time required to accumulate newly synthesized protein in the Golgi. We routinely find that 1-3 h is sufficient, but this time varies from protein to protein. The extent to which new protein has accumulated in the Golgi can be determined by assessing the degree of colocalization with Golgi markers at a series of time points (e.g., 30-min intervals) after shifting to 20°C.
   2. After accumulating protein in the Golgi in a 20°C incubator, place cells into recording medium and incubate at 20°C in the recording chamber mounted on a microscope for time-lapse imaging. Acquire and save both transmitted light and fluorescent images of the cells you will be studying.
   3. Increase the temperature of the recording chamber to 32°C. Wait 5 min for the temperature and focus to stabilize.
   4. Acquire time-lapse images to evaluate post- Golgi transport events. Post-Golgi transport can be studied at either low or high spatial and temporal resolution depending on the questions being addressed.
   5. At the end of every time-lapse recording, save a transmitted light image of the cells from which you recorded data.

H. Kinetics of Protein Transport through the Secretory Pathway
Steps
To determine the rates at which a fluorescent protein moves from the ER to the Golgi and then to the plasma membrane using time-lapse fluorescence microscopy, it is necessary to introduce a pulse of fluorescence into individual cells. We find that microinjection of cDNA is best for this purpose, as the expression of injected cDNA is generally rapid and can be controlled temporally. The kinetics of ER-to-Golgi and Golgi-to-plasma membrane cargo transport can be determined by measuring the ratio of Golgi-associated fluorescence/total fluorescence over time. It is imperative to be able to distinguish Golgi-associated from non-Golgi-associated fluorescent signals. This can be done in two ways. First, you can coexpress a fluorescently tagged Golgi resident protein to use as a reference during time-lapse recordings. Alternatively, you can make an educated deduction as to whether your reporter is in the Golgi based on its localization and the intensity of its fluorescent signal (supplemental data in Kreitzer et al., 2000). Using this differential in position and intensity, you can define a threshold above, which includes Golgi-associated fluorescence, and below, which includes all other cellular fluorescence. Given that you are trying to account for fluorescence present in the entire cell, these assays are best executed using low-magnification objectives capable of imaging an entire cell in a single focal plane.

One hour after microinjection (or the minimum expression time for your protein of interest, see earlier discussion), place cells into bicarbonate-free medium with cycloheximide and incubate at 20°C to accumulate newly synthesized protein in the Golgi.

1. To evaluate ER to Golgi transport kinetics, mount the coverslip into the precooled (20°C) recording chamber on the microscope in recording medium with cycloheximide.
   1. Acquire time-lapse images at 5- to 30-min intervals until the fluorescently tagged reporter has accumulated in the Golgi.
   2. Determine the minimum time required for maximal transport of protein from the ER to the Golgi by measuring the ratio of Golgiassociated fluorescence/total fluorescence over time.
2. To evaluate kinetics of Golgi emptying, place cells into bicarbonate-free DMEM with cycloheximide and incubate at 20°C. The duration of the 20°C temperature block depends on the time it takes to accumulate maximally your protein of interest in the Golgi (see earlier discussion).
   1. Acquire time-lapse images at 5- to 30-min intervals.
   2. Determine the rate at which fluorescently tagged cargo empties from the Golgi by measuring the ratio of Golgi-associated fluorescence/ total fluorescence over time in individual cells.

I. Measuring Delivery of Post-Golgi Carriers to the Plasma Membrane
Solutions and Materials

1. Phosphate-buffered saline with calcium and magnesium (PBS-CM)
2. 2% paraformaldehyde prepared freshly in PBS-CM
3. Antibodies reactive with an extracellular epitope contained in the plasma membrane reporter protein. Surface immunolabeling is dependent on having an antibody that recognizes an extracellular epitope on the plasma membrane protein being studied. If this is not available, it may be desirable to create a cDNA probe that contains an epitope tag (e.g., HA, myc or FLAG tag) that can be immunostained.

Steps
Delivery of newly synthesized proteins to the plasma membrane can be evaluated using single cell assays or by biochemical methods. Evaluation of protein delivery to the plasma membrane in single cells can be analyzed from either fixed or living samples. Analysis in fixed samples involves cell surface selective immunolabeling of the expressed reporter protein. Analysis in living samples requires a relatively robust expression of the GFP-tagged reporter protein and acquisition of time-lapse images at relatively high frame rates using either total internal reflection fluorescence microscopy or spinning disk confocal microscopy. In TIR-FM, membranebound transport intermediates containing GFP-tagged fusion proteins are detected only when they move into an evanescent field, which in our experiments was within ~120nm of the plasma membrane domain in contact with the substratum, i.e., the basal membrane. Schmoranzer and colleagues (2000) have established a quantitative method for detecting bona fide fusion of post-Golgi transport intermediates with the plasma membrane. Briefly, exocytic events are defined by a simultaneous rise in both the carrier's total fluorescence intensity and the area occupied by carrier fluorescence as it flattens into the plasma membrane and the cargo diffuses laterally. (For a more extensive description of TIR-FM in studying exocytosis, see Mikhailov). Exocytic events occurring in the lateral membrane of polarized epithelial cells can be identified using similar criteria when time-lapse images are acquired by high-speed confocal microscopy (for a complete description of lateral membrane fusion analysis, see Kreitzer et al., 2003). Biochemical methods for studying delivery to the plasma membrane in a large population of cells, such as pulse-chase, cell surface biotinylation assays, have been described in detail previously (see article by Rodriguez-Boulan et al.).

1. Measuring the rate of protein delivery to the cell surface in fixed cell, "time-lapse" experiments. If cells are grown on glass (as would be the case if exogenous proteins are expressed by cDNA microinjection), this method is useful in evaluating the delivery of proteins to the apical membrane only. For surface-labeling analysis of delivery to the basolateral membrane, cells must be grown on semipermeable filter supports and exogenous proteins must be introduced by transfection or viral infection methods.
   1. Microinject GFP-tagged cDNA into the cell nuclei.
   2. Accumulate newly synthesized protein in the Golgi at 20°C.
   3. Shift to the permissive temperature for transport out of the Golgi (37°C).
   4. At 15- to 30-min intervals after releasing the Golgi block, fix cells in paraformaldehyde for 5 min at room temperature. Do not permeabilize with detergent. Fixation in a nonpermeabilizing fixative, such as paraformaldehyde, enables selective immunolabeling of surfaceassociated proteins.
   5. Label, by indirect immunofluorescence, the surface-associated reporter protein. Make sure to use a fluorescently conjugated secondary antibody other than fluorescein (or any dye excitable at 488 nm).
   6. Acquire images of both the surface-associated (immunostained) and the total (GFP) protein expressed at each time point after release of the Golgi block. It is imperative to use identical acquisition settings for the individual fluorophores in each time-lapse sample as this is all that allows you to quantitatively (ratiometrically) evaluate the relative amount of reporter protein delivered to the cell surface.
   7. Calculate the integrated fluorescence intensity of both surface-associated and total fluorescence in each cell expressing the reporter protein. The ratio of surface fluorescence (immunostained) to total fluorescence (GFP) of your reporter reflects the relative amount of protein that has been delivered to the plasma membrane at each time point. Over time, this ratio should increase and will directly reflect the rate of protein delivery from the Golgi to the plasma membrane.
2. Analysis of exocytosis using time-lapse total internal reflection fluorescence microscopy (TIR-FM).
   1. Microinject GFP-tagged cDNA into the cell nuclei.
   2. Accumulate newly synthesized protein in the Golgi at 20°C.
   3. Mount coverslip in the recording chamber on a microscope equipped for TIR-FM. Acquire and save both bright-field and epifluorescent images of the cells you will be studying.
   4. Shift to the permissive temperature for transport out of the Golgi and wait 5 min for the temperature and focus to stabilize.
   5. Acquire time-lapse images (aim for at least four to five frames per second) to visualize exocytic events occurring in the basal plasma membrane. The typical duration of our recordings is 1-2 min.
   6. Analyze images for exocytic events as described in Schmoranzer et al. (2000).
3. Analysis of exocytosis using time-lapse, spinningdisk confocal microscopy.
   1. Microinject GFP-tagged cDNA into the cell nuclei.
   2. Accumulate newly synthesized protein in the Golgi at 20°C.
   3. Mount coverslip in the recording chamber on a microscope equipped with a spinning disk confocal head. Acquire and save both brightfield and epifluorescent images of the cells you will be studying.
   4. Shift to the permissive temperature for transport out of the Golgi and wait 5min for the temperature and focus to stabilize.
   5. Acquire time-lapse images (aim for four to five frames per second) to visualize exocytic events occurring along the lateral membrane. Photobleaching that occurs during confocal image acquisition typically limits the duration of time-lapse sequences to ~1-2min.
   6. To evaluate the spatial positioning of cargo delivery events, acquire time-lapse sequences as described in step e at multiple Z-axis positions throughout individual cells.
   7. Analyze images for exocytic events as described in Kreitzer et al. (2003).

References
Baird, G. S., Zacharias, D. A., et al. (2000). Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97(22), 11984-11989.

Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263, 802-805.

Grindstaff, K., Yeaman, C., Anandasabapathy, N, Hsu, S.-C., Rodriguez-Boulan, E., Scheller, R., and Nelson, W. J. (1998). Sec 6/8 complex is recruited to cadherin-medicated cell-cell contacts and specifies transport vesicle delivery to the basal-lateral membrane in polarized epithelial cells. Cell 93, 71-74.

Hunziker, W., Harter, C., et al. (1991). Basolateral sorting in MDCK cells requires a distinct cytoplasmic domain determinant. Cell 66, 907-920.

Kreitzer, G., Marmorstein, A., et al. (2000). Kinesin and dynamin are required for post-Golgi transport of a plasma-membrane protein. Nature Cell Biol. 2(2), 125-127.

Kreitzer, G., Schmoranzer, J., et al. (2003). Three-dimensional analysis of post-Golgi carrier exocytosis in epithelial cells. Nature Cell Biol. 5(2), 126-136.

Le Bivic, A., Sambuy, Y., et al. (1991). An internal deletion in the cytoplasmic tail reverses the apical localization of human NGF receptor in transfected MDCK cells. J. Cell Biol. 115, 607-618.

Marmorstein, A. D., Csaky, K. G., Baffi, J., Lam, L., Rahaal, E, and Rodriguez-Boulan, E. (2000). Saturation of, and competition for entry into, the apical secretory pathway. Proc. Natl. Acad. Sci. USA 97, 3248-3253

Matz, M. V., Fradkov, A. E, et al. (1999). Fluorescent proteins from nonbioluminescent Anthozoa species. Nature Biotechnol. 17(10), 969-973.

Mikhailov, A. (ed.) "Practical Fluorescence Microscopy: Protein Localization and Function in Mammalian Cells." Humana Press, Clifton, NJ.

Musch, A., Xu, H., et al. (1996). Transport of vesicular stomatitis virus G protein to the cell surface is signal mediated in polarized and nonpolarized cells. J. Cell Biol. 133(3), 543-558.

Rodriguez-Boulan, E., and Sabatini, D. D. (1978). Asymmetric budding of viruses in epithelial monlayers: A model system for study of epithelial polarity. Proc. Natl. Acad. Sci. USA 75, 5071-5075.

Schmoranzer, J., Goulian, M., et al. (2000). Imaging constitutive exocytosis with total internal reflection fluorescence microscopy. J. Cell Biol. 149(1), 23-32.