Subcellular Fractionation Procedures and Metabolic Labeling Using [35S] Sulfate to Isolate Dense Core Secretory Granules from Neuroendocrine Cell Lines

I. INTRODUCTION
Subcellular fractionation techniques have been developed to allow isolation of a particular subcellular compartments. Typically subcellular fractionation is used as a starting point to characterize the composition of subcellular organelles, but has also been used frequently to provide membranes for cell-free assays that reconstitute a particular intracellular event or process.

Ideally, for both purposes (characterization and cellfree assays) the subcellular compartment should be purified to homogeneity. However, this is difficult to achieve, particularly when the starting material is from cell lines where amounts are limiting. Therefore, it is realistic to aim to achieve a highly enriched fraction (defined here as greater than 90% purity) with a good yield. In addition, it is important to carefully consider the purpose of the fractionation protocol as this will dictate how critical the purity is.

To study the formation of secretory granules in the PC12 cell line, a protocol was developed to identify nascent immature granules by optimizing their separation from the donor compartment, the Golgi complex (Tooze and Huttner, 1990). A combination of velocity-controlled differential and equilibrium density centrifugation achieved a separation of immature secretory granules and the trans-Golgi network (TGN). This protocol was also used to obtain a population highly enriched in immature secretory granules (ISG) and one highly enriched in mature secretory granules (MSG) (Tooze et al., 1991). MSGs are easier to obtain in a more purified form as MSGs are denser than other subcellular compartments. This protocol has been used to characterise the ISG and MSG (Tooze et al., 1991), study low pH-dependent prohormone processing (Urbé et al., 1997), clathrin coat recruitment to ISGs (Dittié et al., 1996), and the role of ADPribosylation factor (ARF) (Austin et al., 2000), sorting of proteins in the ISG (Dittié et al., 1999), cell-free homotypic fusion (Urbé et al., 1998; Wendler et al., 2001), and the recruitment of lipid kinases to ISGs (Panaretou and Tooze, 2002). This protocol has also been used by other researchers to investigate sorting sequences in regulated secretory proteins (Krömer et al., 1996) and with, for example, the AtT20 cell line (Eaton et al., 2000).

II. MATERIALS AND INSTRUMENTATION
PC12 cells are maintained in growth medium consisting of 10% horse serum, 5% fetal calf serum, and Dulbecco's modified Eagle's medium (DMEM) with 3.5 g/liter glucose, penicillin/streptomycin, and 4 mM glutamine (note this is twice the standard concentration) under 10% CO₂. Reagents for cell culture are obtained from Sigma Aldrich and GIBCO.

PC12 cells are passaged once a week at a dilution of 1:6. For a standard granule preparation, nine, 245× 245-mm confluent plates of PC12 cells are used. These plates are prepared from five, 175-cm² flasks harvested with trypsin, seeded, and grown for about 7 days or from nine, 175-cm² flasks harvested with trypsin, seeded, and grown for about 3 days.

All chemicals are available from Sigma-Aldrich, except sucrose (ultrapure grade Cat. 15503-02), which is from GIBCO.

A cell cracker (EMBL Workshop, EMBL Heidelberg, Germany) with a range of titanium balls. Assemble cell cracker with chosen ball, precool and wash cell cracker with 1 ml of homogenization buffer (HB)/protease inhibitor cocktail (pi) before use. For PC12 cells, use a 18-µm clearance.

Gradients are centrifuged in an SW40 rotor using Beckman ultraclear centrifuge tubes (Cat. No. 344060).

Velocity gradients are prepared using a BioComp gradient master (http://www.biocompinstruments. com), and equilibrium gradients are prepared using a Labconco Auto Densiflow gradient maker (http:// www.labconco.com). Both velocity and equilibrium gradients are collected using the Auto Densiflow.

A cell scraper is made from a silicon rubber bung by cutting the bung with a single-sided razor blade first horizontally (across the widest part) and then vertically in half. The tip of a plastic 10-ml pipette is then inserted into the center of the cut bung.

III. PROCEDURES
A. Preparation of a Postnuclear Supernatant
Solutions

1. Tris-buffered saline (TBS): 137 mM NaCl, 4.5 mM KCl, 0.7 mM Na₂HPO₄, 25 mM Tris-HCl, pH 7.4
2. Protease inhibitor (pi) cocktail: 0.5 mM phenylmethylsulfonyl fluoride and 5-µg/ml leupeptin
3. Homogenization buffer (HB): 0.25M sucrose, 10mM HEPES-KOH, pH 7.2, 1 mM EDTA, and 1 mM MgOAc

Steps
For preparation of sufficient postnuclear supernatant (PNS) to load six SW40 gradients, use nine 245 × 245-mm plates. All solutions must be at 4°C.

1. Place one 245 × 245-mm plate on ice.
2. Remove growth medium by decanting and wash gently twice with 40ml TBS.
3. Wash each plate again with 40ml TBS/pi.
4. Add 40ml TBS/pi and remove PC12 cells from the plates by scraping with a cell scraper. Collect the suspension of cells from each plate into a 50-ml Falcon tube. Repeat for all nine plates.
5. Spin each tube for 7min at 84g. Remove supernatant and resuspend each pellet in 1 ml of HB/pi. Pool in a 50-ml Falcon tube containing 24 ml of HB/pi and divide into six 15-ml Falcon tubes.
6. Spin for 7 min at 500g. Remove supernatant and resuspend each pellet in 700 µl HB/pi and pool all nine tubes of cell suspension into a 15-ml Falcon tube.
7. Using a 1-ml syringe with a 21- or 22-gauge needle, draw up 1 ml of cells and pass back and forth through the needle six to seven times. Check that cells are well dispersed by resuspending 5 µl of cells with 10 µl of trypan blue solution and monitoring by phasecontrast microscopy. Ensure that the cells are now dispersed uniformly and not still in clumps. A maximum of 20% of the cells should be permeable to trypan blue. Repeat until all the cells have been passed through the needle.
8. Using a 1-ml syringe and 21- or 22-gauge needle, draw up 1 ml of the cell suspension and place syringe on one port of the cell cracker. Place an empty syringe on the other port. Pass cells through chamber six to seven times. Check breakage with trypan blue as in step 7. Greater than 90% of the cells should be broken. Nuclei should be intact and spherical but free of subcellular membranes (see Tooze and Huttner, 1992).
9. Remove homogenate and repeat with remaining cell suspension. Rinse the cell cracker out with 1 ml of HB/pi. Pool all the homogenate into a 15-ml Falcon tube. Total volume will be between 9 and 10ml.
10. Spin for 10 min at 1700g to pellet nuclei. Remove postnuclear supernatant, avoiding the nuclear pellet. Because the interface between the supernatant and the pellet can be difficult to see, illuminate the tube from behind when removing the supernatant.
11. Respin the PNS for 5 min at 1700g to ensure that all nuclear material is removed and collect the supernatant. Adjust volume if necessary to 9.0 ml.

B. Velocity Gradient Centrifugation
This method achieves separation of organelles by size rather than density. Thus, larger subcellular compartments will sediment faster during this centrifugation than smaller ones. This is the basis for the separation of ISGs and MSGs from the trans-Golgi network (TGN). In particular, this step is essential for separation of ISGs from the TGN as these organelles have the same equilibrium density.

Solutions
Sucrose solutions: 0.3M sucrose, 10mM HEPESKOH, pH 7.2, and 1.2M sucrose, 10mM HEPES-KOH, pH 7.2

Steps

1. Prepare the velocity gradients. Use 5.5ml of 0.3M sucrose and 6ml of 1.2M sucrose per gradient. Using the Biocomp gradient maker, perform two steps: step 1 for 10min at a 50°C angle at 30rpm and step 2 for 1 min at a 80°C angle at 12rpm. If preparing the gradients manually, use a 15-ml gradient maker, gently mixing the heavy into the light. Using a narrow pipette, pump solution from the light chamber into the bottom of the tube and displace the solution upward with the heavier solutions.
2. Load 1.5 ml of PNS on top of each gradient. Spin at 25,000rpm in a SW40 rotor at 4°C. Using maximum acceleration, allow centrifuge to reach 25,000rpm and then spin for exactly 15 min with the brake applied at the end of the run.
3. Collect thirteen 1-ml fractions from the top. Fraction 1 will contain most of the soluble proteins from the PNS. Fractions 2-4 will contain the ISG and constitutive secretory vesicles (CSVs), fractions 5-7 will contain MSGs, and fractions 8-11 will contain TGN membranes. Note: There will be other subcellular membranes in these fractions so the ISGs, MSGs, and TGN fractions are only slightly enriched. Pool fractions 1-4 for preparation of ISGs and fractions 5-7 for preparation of MSGs.

C. Preparation of ISGs and MSGs by Equilibrium Gradient Centrifugation
This method achieves separation of organelles by density, allowing vesicles of the same size but different densities to be separated. ISG, which have an average diameter of 80nm, can be separated Golgiderived CSVs, which are reported to have a diameter of 50-200nm (Salamero et al., 1990).

Solutions
Sucrose solutions: 0.8M sucrose, 10 mM HEPESKOH, pH 7.2; 1.0M sucrose, 10 mM HEPES-KOH, pH 7.2; 1.2M sucrose, 10mM HEPES-KOH, pH 7.2; 1.4M sucrose, 10mM HEPES-KOH, pH 7.2; and 1.6M sucrose, 10 mM HEPES-KOH, pH 7.2

Steps

1. Prepare equilibrium gradients. Pipette 1 ml of 1.6M and then layer 2ml of 1.4, 1.2, 1.0, and then 1.0 ml of 0.8 M sucrose into a SW40 tube, either by hand or using the Auto Densiflow, with probe moving upward.
2. Adjust pooled fractions from the velocity gradient to a final volume of 4 ml per gradient with 10mM HEPES-KOH, pH 7.2. Load pooled fractions from velocity gradient onto prepared equilibrium gradients.
3. Spin gradients in a SW40 rotor at 25,000rpm overnight or for at least 5.5 h at 4°C.
4. Collect twelve 1-ml fractions from the top of the gradient. If preparing ISGs, the ISGs will be found in fractions 7-9. If preparing MSGs, the MSGs will be found in fractions 10-12. These respective fractions can be pooled and aliquoted for storage for up to 6 months in liquid nitrogen.

D. Other Procedures
The centrifugation procedure can be checked by assaying for TGN, secretory granule, or other compartment-specific markers in each fraction from the velocity and equilibrium gradients. This can be done by Western blotting using antibodies to the marker proteins or by metabolic labelling. For the technique described earlier using PC12 cells, the most accurate method is posttranslational labelling of proteins with [³⁵S]sulfate on tyrosine residues (for a review, see Moore, 2003).

Protein tyrosine sulfation is a posttranslational modification found in some secretory proteins, including secretogranin II (SgII) and an unidentified constitutively secreted heparan sulphated proteoglycan (HSPG). The enzyme tyrosylprotein sulfotransferase responsible for sulfation is a resident TGN protein (Lee and Huttner, 1985). Thus, by incubating or pulsing the cells for a short period of time by the addition of [³⁵S]sulfate, proteins that contain the sulfation motif can be labelled in the TGN and can be used to identify TGN membranes. In addition, exit of sulphated proteins from the TGN into CSVs, which contain HSPG, and ISGs and MSGs, which contain SgII, can be followed accurately during the chase. Identification of CSV and ISGs by sulfate label of the specific markers, followed by subcellular fractionation, can be used to confirm that the cultured cells are correctly sorting regulated and constitutively secreted proteins into separate vesicles (Tooze and Huttner, 1990). Metabolic labelling with sulfate can also be applied more generally (Tooze, 1999). It is advisable to use a scaled down version of the procedure described earlier to avoid having to use large amounts of radioactivity.

Solutions

1. Labelling medium (sulfate-free DMEM): DMEM is prepared by substituting MgCl₂·6H₂O for MgSO₄·7H₂O and reducing the normal concentration of cysteine and methionine to 1% of the original concentration; 0.1% dialysed horse serum; 0.05% dialysed fetal calf serum; and 2mM glutamine. It is important not to include any sulphated antibiotics.
2. [³⁵S]Sulfate (40-100mCi/ml): Amersham Biosciences (Cat. No. SJS.1).
3. Chase medium: growth medium supplemented with 1.6 mM NaS0₄.

Steps
To check the ability of PC12 cells to sort regulated proteins using the gradients described earlier, two 150- mm² dishes of PC12 cells should be prepared for each condition. Ideally the experiment should confirm the position of the TGN on the gradients, identified by a 5-min pulse of sulfate; the CSVs and ISGs, identified by a 5-min pulse followed by a 15-min chase; and the MSGs, identified by a 5-min pulse and 90-min chase. Alternatively, MSGs can be labelled using one-tenth the amount of [³⁵S]sulfate for 6 h and chased overnight.

1. Wash each 150-mm 2 dish once with labelling medium at 37°C. Replace with 20 ml labelling medium. Incubate for 20min at 37°C to deplete endogenous sulfate.
2. Remove medium and replace with fresh medium at 37~ containing lmCi/ml [³⁵S]sulfate, or 0.1mCi/ ml for long-term MSG labelling. Incubate at 37°C for precisely 5 min. Dishes can be rocked gently to reduce the amount of labelling medium required. Five to 7 ml is sufficient when dishes are being rocked.
3. Remove labelling medium. To identify the TGN, transfer the dishes immediately to 4~ and add 10ml of TBS at 4°C. To identify ISGs, add 20ml chase medium at 37°C and return cells to incubator for 15 min. To label MSGs, add 20ml chase medium at 37°C and return cells to incubator for 90min. Alternatively, use the longer label-and-chase protocol for MSGs.
4. At the end of the chase period, transfer remaining dishes to 4°C remove the chase medium, and add 10 ml TBS at 4°C.
5. Wash and harvest dishes as described previously. Pool the two dishes from each condition. Prepare a PNS as described earlier and load the PNS from both labelled 150-mm 2 dishes on one velocity gradient.
6. Collect thirteen 1-ml fractions from each gradient. Remove 30µl from each fraction and analyse the [³⁵S]sulfate-labelled proteins by SDS-PAGE. For optimal resolution of SgII and HSPG, use a 7.5% gel.
7. Save the remaining material for equilibrium gradient centrifugation. Pool fractions 1-4 from the ISG gradient, and pool fractions 5-7 from the MSG gradient. Load each pool onto on an equilibrium gradient and continue as described in Section III.C.

V. PITFALLS

1. The PC12 cells are not passaged correctly and thus do not achieve maximum confluency, resulting in a very low yield of cells. The cells must be plated as single cell suspensions. This is best achieved by tituration using a flamed narrowed disposable glass pipette.
2. The cells are not completely homogenized, resulting in fewer broken cells as judged by trypan blue. This will result in a reduced yield.
3. The homogenate is very viscous. This is most likely a result of nuclei lysis during homogenization. A clean postnuclear supernatant cannot be obtained from this homogenate, and the experiment should be stopped as the subcellular organelles will not be separated properly on the gradients.

References
Austin, C., Hinners, I., and Tooze, S. A. (2000). Direct and GTPdependent interaction of ADP-ribosylation factor 1 with clathrin adaptor protein AP-1 on immature secretory granules. J. Biol. Chem. 275, 21862-21869.

Ditti6, A. S., Hajibagheri, N., and Tooze, S. A. (1996). The AP-1 adaptor complex binds to immature secretory granules from PC12 cells, and is regulated by ADP-ribosylation factor. J. Cell Biol. 132, 523-536.

Ditti6, A. S., Klumperman, J., and Tooze, S. A. (1999). Differential distribution of mannose-6-phosphate receptors and furin in immature secretory granules. J. Cell Sci. 112, 3955-3966.

Eaton, B. A., Haugwitz, M., Lau, D., and Moore, H. P. (2000). Biogenesis of regulated exocytotic carriers in neuroendocrine cells. J. Neurosci. 20, 7334-7344.

Howell, K. E., Schmid, R., Ugelstad, J., and Gruenberg, J. (1989). "Immunoisolation using Magnetic Solid Supports: Subcellular Fractionation for Cell Free Functional Studies," pp. 265-292. Academic Press, New York.

Krömer, A., Samenfeld, P., Loef, I., Huttner, W. B., and Gerdes, H.- H. (1996). Essential role of the granin-loop for sorting to secretory granules is revealed by expression of chromogranin B in the absence of endogenous protein synthesis. J. Cell Biol. 140, 1059-1070.

Lee, R. W. H., and Huttner, W. B. (1985). (Glu62, Ala30, Tyr8)n serves as high-affinity substrate for tyrolsylprotein sulfotransferase; a Golgi enzyme. Proc. Natl. Acad. Sci. USA 82, 6143-6147.

Moore, K. L. (2003). The biology and enzymology of protein tyrosine O-sulfation. J. Biol. Chem. 278, 24243-24246.

Panaretou, C., and Tooze, S. A. (2002). Regulation and recruitment of phosphatidylinositol 4-kinase on immature secretory granules is independent of ADP-ribosylation Factor 1. Biochem. J. 363, 289-295.

Salamero, J., Sztul, E., and Howell, K. (1990). Exocytic transport vesicles generated in vitro from the trans-Golgi network carry secretory and plasma membrane proteins. Proc. Natl. Acad. Sci. USA 87, 7717-7721.

Tooze, S. A. (1999). Metabolic labeling with sulfate. In "Current Protocols in Cell Biology" (J. S. Bonifacino, M. Dasso, J. B. Harford, J. Lippincott-Schwartz, and K. M. Yamada, eds.), Vol. 1, pp. 7.3.1-7.3.7, Wiley, New York.

Tooze, S. A., Flatmark, T., Tooze, J., and Huttner, W. B. (1991). Characterization of the immature secretory granule, an intermediate in granule biogenesis. J. Cell Biol. 115, 1491-1503.

Tooze, S. A., and Huttner, W. B. (1990). Cell-free protein sorting to the regulated and constitutive secretory pathways. Cell 60, 837-847.

Tooze, S. A., and Huttner, W. B. (1992). Cell-free formation of immature secretory granules and constitutive secretory vesicles from trans-Golgi Network. Methods Enzymol. 219, 81-93.

Urbé, S., Ditti6, A., and Tooze, S. A. (1997). pH-dependent processing of secretogranin II by the endopeptidase PC2 in isolated immature secretory granules. Biochem. J. 321, 65-74.

Urbé, S., Page, L. J., and Tooze, S. A. (1998). Homotypic fusion of immature secretory granules during maturation in a cell-free assay. J. Cell Biol. 143, 1831-1844.

Wendler, E, Page, L., Urbé, S., and Tooze, S. A. (2001). Homotypic fusion of immature secretory granules during maturation requires syntaxin 6. Mol. Biol. Cell 12, 1699-1709.

Wu, C. C., MacCoss, M. J., Howell, K. E., and Yates, J. R. III. (2003). A method for the comprehensive proteomic analysis of membrane proteins. Nature Biotechnol. 21, 532-538.