Neuropeptides (1992) 22,23 5-240 8 Longman Group UK Ltd 1992

Rapid, High-yield Isolation of Human Chromogranin A From Chromaffin Granules of Pheochromocytomas U. SWERSEN*,

H. L. WALDUM* and D. T. O’CONNORt

*Institute of Cancer Research, University of Trondheim, Trondheim, Norway. tDepartment of Medicine and Center for Molecular Genetics, University of California and tDepattment of Veterans Affairs Medical Center, San Diego, California 92161, USA (Reprint requests to US) Abstract - Chromogranin A (CgA) is a useful probe of human neuroendocrine neoplasia and exocytotic sympathoadrenal activity, but the application of CgA immunoassays has not been widespread because of limited availability of purified human CgA. Here we describe a rapid, high yield isolation of human CgA. After obtaining and lysing pheochromocytoma chromaffin granules, the soluble core proteins (chromogranins) were depleted of dopamine-betahydroxylase by passage over a concanavalin A-Sepharose affinity column, then lyophilized, resuspended in volatile buffer, and gel filtered on Sephacryl S-300. SDS-PAGE-analyzed column fractions contained homogeneous human CgA, which was verified structurally (N-terminal amino acid sequence) and immunologically (radioimmunoassay and immunoblot). The overall 22.6 mg yield of purified CgA represented 5.7% of the starting vesicle core protein. This preparation will be useful in evaluating the sympathoadrenal system and endocrine neoplasia in man.

Introduction Chromogranin A (CgA) is a member of the chromogranin/secretogranin family, a group of acidic proteins found in the secretory granules of a wide variety of endocrine cells and neurons (1). CgA was characterized by Helle et al in 1966 (2). More recently, chromogranins B and C have been indentified (3, 4), and three other acidic proteins which may be related to the chromogranins, have been described (5). Date received 2 1 April 1992 Date revised 18 May 1992 Date accepted 18 May 1992

Various functions of CgA have been suggested, and there are indications that it may function as a prohormone (6,7). Evaluated by immunoassay and immunohistology, CgA has an almost universal distribution in endocrine and neuroendocrine cells (1). In addition, CgA is released from a variety of neuroendocrine tumors, and measurement of CgA has emerged as a useful tool in the evaluation of such tumors (8, 9). Determination of circulating CgA in human serum or plasma is therefore of increasing interest, and a few groups have succeeded in establishing immunological methods to measure CgA (10, 11). However, implementation of immunoassays for

235

236

NEUROPEPTIDES

CgA has been hampered by scarcity of human CgA as a result of tedious, low-yield purification methods (10, 12, 13). Chromaflin tissue has a high content of CgA, and pheochromocytomas are therefore a suitable source for purification of the protein (1, 10). In this paper we describe a rapid isolation procedure with a high yield of intact, immunoreactive human CgA from chromaffm granules of pheochromocytomas.

Methods Chromafin grade

(vesicle) isolation

All steps in the preparation were carried out at 04°C. Chromaffi vesicles were isolated as described previously from fresh human pheochromocytomas (12). The average weight of pheochromocytomas was 90 + 22 gm (mean + one standard error of the mean), and generally l/4 to l/2 was obtained for granule preparation. In short, a highly purified chromafiin granule subcellular fraction was obtained by a modification (12) of the sucrose density step gradient method of Smith and Winkler (14). The tissue was minced fYlnely,homogenized in icecold 0.3 M sucrose solution (at 20% tissue wt/vol), filtered through cheesecloth, and then centrifuged at 1000 g for 10 min to sediment nuclei and cell debris. The supematant was centrifuged at 25 000 g for 10 min. The crude granule fraction was gently resuspended in 0.3 M sucrose and layered onto step-density gradients of 1.6 M sucrose. The gradients were centrifuged at 10 000 g overnight (about 9.6 x lo6 g per min). A small portion of the pink-colored vesicle pellet was saved for transmission electron microscopy (1,12). The remaining pellet was resuspended and lysed in 0.001 M sodium phosphate, pH 6.5, by freezing and thawing, followed by ultracentrifugation at 100 000 g for 1 h, to separate soluble granule core contents (lysate) from granule membranes. The chromafhn vesicle soluble core lysates (CVLs) were stored frozen (-7O’C). Pur@cation of human chromogranin A Before further purification, the protein content of pheochromocytoma CVLs was determined (15), and 100 c(g of protein from each CVL was analyzed by SDS-PAGE followed by Coomassie blue staining (12). Seven vesicle samples showing CgA bands

(corresponding to an M, of about 70 kDa) were pooled for further purification steps. Dopamine-beta-hydroxylase was removed (16, 17) from the pooled pheochromocytoma CVLs by one passage at 10 ml/b through a 10 x 0.9 cm affinity column of concanavalin A-Sepharose (Pharmacia, Piscataway, NJ, USA). The remaining chromogranins were dialyzed overnight against 6 mM ammonium acetate, pH 6.5, lyophilized, resuspended in a volume of 10 mL, and gel filtered on a 2.6 x 78 cm preparative scale column of Sephacryl S-300 (Pharmacia), equilibrated and eluted at 25 ml/h (= 5 ml/cm2 of column crosssectional area) with the high ionic strength, volatile buffer 0.3 M ammonium acetate, pH 6.5. A high ionic strength buffer was chosen, based on the results of Smith and Kirshner (18), while volatility enabled buffer removal by lyophilization after chromatography. 5 mL fractions were collected, and monitored for protein by absorbance at 280 nm (A&. Representative fractions (30 l.tglyophilized protein) were analyzed by SDS-PAGE. Fractions highly enriched in CgA (M, 70 kDa) were pooled, lyophilized and reconstituted. CVLs were frozen at -70°C during preparative steps, including the time period while SDS-PAGE was being run. Analytical methods The N-terminal amino acid sequence of the purified human CgA preparation was determined on an Applied Biosystems model 470A gas phase sequenator, as previously described ( 19,20). Immunoblots on purified fractions were performed by the ProtoBlotTM Immunoscreening system (21). One pg and 0.5 B of protein were applied to a 10% polyacrylamide gel (5 x 5 x 0.15 cm) in a Laemmli electrophoresis unit. Proteins separated by SDS-PAGE were electrophoretically transferred to a nitrocellulose membrane. The membrane was incubated for 1 h at room temperature in phosphate buffered saline (PBS), containing 10% non-fat milk (Gifco). Incubation with specific anti-human CgA antibodies, 1 polyclonal (10) and 1 monoclonal (LK2HlO [22]) at 1: 1000 (vol/vol) dilution, wasperformed for 2 h. After washing 3 times in PBS containing 1% BSA, the membrane was incubated for 1 h with alkaline phosphatase-conjugated anti-rabbit

CHROMOGlUNKN

IgG (H + L) or anti-mouse IgG (Fe) (Protoblot), at a dilution of 1:7500 (voYvo1). After extensive washing, immunoreactive bands were visualized by adding nitro blue tetrazolium as substrate and 5bromo-4-chloro-3-indolyl phosphate. 10 pg of the human CgA preparation was radioiodinated by the chloramine T method (23) and used in a radioimmunoassay of human CgA, as previously described (24). Protein concentration was determined either by the Coomassie blue dye binding method (15), or by Azso on highly purified CgA fractions, with the knowledge that a 1 mg/mL solution of purified CgA yields an Azso= 0.708 (18). The work was approved by the Human Subjects Committee of the UCSD School of Medicine. Since the project utilized tissue that would otherwise be discarded, written informed consent was not required. Results and Discussion Transmission electron micrographs of pheochromocytoma chromaf8n granules, obtained after

Fig. 1

237 sucrose gradient centrifugation of the tissue homogenates, revealed electron dense core chromaffi granules, without visible mitochondrial contamination (Fig. 1). Inspection of 8 pheochromocytoma CVLs by SDS-PAGE (Fig. 2) showed a visible CgA band (M, 70 kDa) in 7 samples (all except sample #2020-6), which were pooled for further analysis. CgB (M, 11O-l 20 kDa) was prominent in only 1 pheochromocytoma sample (#371-5). After removal of dopamine-beta-hydroxylase on concanavalin A-Sepharose, the remaining cbromogranins were size-fractionated by gel filtration (Fig. 3). When representative fractions were analyzed by SDS-PAGE (Fig. 4), fractions 23-31 were highly enriched in a M, 70 kilodalton @Da) band. When these fractions were pooled and sequenced (Figs 4 82 5), a single sequence emerged: LPVNSP. . ., matching that expected for the N-terminus of human CgA (6,18,25). Beyond fraction 3 1 (Fig. 4), lower mol wt peptides were increasingly prominent, as expected for a gel filtration size fractionation. Gut of a total of 397 mg of protein in the pooled original vesicle soluble core fractions, the yield of

electron micrograph of a chromaflin granule fraction isolated from a human pheochromocytoms Ihomogenate TImission by sucrose: density gradient homogenization. The electron dense core structures are chromaffin granules. There is no visible mitochondrial t:ontamination. The magnification is 13 500 diameters.

238

NEUROPEPTIDES

Pheochromocytoma chromaffin vesicle fractions

--CgB

llO-3 n s. r‘

a4--

--CgA

47--

01 0

i!i z g cn

33--

’

10

4

20

’

’

I

30

40

50

I

00 Fractbn Number

I

I

70

80

90

1 D

Fig. 3 Sephacryl S-300 gel filtration of human chromogranins atIer removal of dopamine-beta-hydroxylase, 5 mL fractions were collected, and protein was monitored by A2s0.V, indicates the total inter& volume of the cohmm.

24-16--

Fig. 2 SDS-PAGE of 8 pheochromocytoma chroma%n vesicle lysates. 100 pg of vesicle soluble protein (reduced with dithiothreitol) was loaded into each gel lane. The mobilities of chromogranins A (CgA) and B (COB) are shown. Numbers at the bottom are traction sample numbers.

purified human CgA was 22.6 mg, or 5.7% (Table). By contrast, previously published yields of human CgA were in the sub-mg range, even after multiple chromatographic or electrophoretic steps ( 12, 13). The purified protein was efficiently radioiodinated, and gave rise to a human CgA radioimmunoassay (24) standard curve (Fig. 6) with a lower limit of detection (at 80% B/B,,; see Fig. 6 legend) of 20 &ml. Using this tracer, the intra-assay coef-

Pooled fraction sequence: LPVNSP...

I+ 4n

25

% 21 23 25 26 29 31 35 40 43 46

Fraction ’ CgA

Y

SDS-PAGE and ammo acid sequencing results on gel filtration fractions. 30 pg of protein from relevant fractions (Fig. 3) Fig. 4 was lyophiliaed, reduced with dithiothreitof and subjected to SDS-PAGE. Fractions highfy enriched in human CgA (Mr 70 kDa) were pooled, lyophilized, resuspended, and subjected to gas phase sequenation (Fig. 5). LPVNSP. . . is the N-terminal sequence of human CgA.

239

CHROMOGRANIN

Table

Isolation of human CgA from pheochromocytoma chromaffin granules

Sample (stage) Initial CVLs (pooled from 7 pheochromocytomas) After dialysis and concanavalin A-Sepharose After lyophilization, before gel filtration After Sephacryl S-300 gel tiltration and lyophilization of pooled CgA-enriched fractions 23-3 1

P

N 100

P

s

~~ - . 0

i

i

j 4 Cycle wmber

;

4

207 210 10

397 239 196

100 60 49

;

Fig. 5 N-terminal amino acid sequenator yields. Protein from pooled, putative CgA peak fractions (#23-31; Fig. 4) was subjected to gas phase N-terminal amino acid sequenation. The yield (pmol) of a particular amino acid residue is presented as a fbnction of automated Edrnan degradation cycle number. The N-terminal sequence (LPVNSP. . .) matches that of human CgA.

ficient of variation for a low (23 q/ml) assay standard was 7.2%, while that for a high (65 @ml) assay standard was 0.84%. The results of anti CgA immunoblotting, using both a polyclonal antiserum and a monoclonal antiI

Yield W)

22.6

5.7

body, are shown in Figure 7. There is a dense, predominant immunoreactive band with an M, of about 70 kDa visualized by both antibodies, compatible with the SDS-PAGE Coomassie blue staining results (Fig. 4). Our preparation had two features that made the purification rapid. Only two chromatography steps

v

!

Protein tm@

10

l_ 200

Volume (ml)

Anti human chrcmogranin A antibody Rabbit polyclonal

Mouse monoclonal

106-

80-

B

.o

28-

0

0.6-v

19-

0.4-.

0.2-,

2

3

I 5

4

In [CgA,

6

ng/mll

Fig. 6 Human CgA radioimmunoassay standard curve. Ten pg of the purified human CgA product was radioiodinated by the chloramine T method and used as a tracer in a human chromogranin A radioimmunoaasay. The results are plotted as % B/B, versus log, (or ln) of added, unlabeled human CgA standard. B = cpm per antigen-antibody pellet in the absence of ad&d unlabeled standard or unknown sample, while B, = cpm per antigenantibody pellet in any assay tube.

0.5

1

0.5

1

Human chromogranin A dose, mcg Fig. 7 Immunoblotting of purified human CgA. One and 0.5 pg of the pooled fractions were separated by SDS-PAGE and incubated with either a polyclonal antiserum or a monoclonal anti- i body, each directed against human CgA, and each at 1: 1000 titer 8 (volhol). With both antibodies there is a main band with an M, of 70 kDa, corresponding to the SDS-PAGE Coomassie blue staining results (see Fig. 4).

240 (affinity

NEUROPEPTIDES

and gel filtration)

were required,

and

volatile buffers (ammonium acetate) were easily removed by lyophilization. Furthermore, the prepa-

ration gave a homogenous product (Figs 4 & 5) at high yield (Table). In previous reports, more extensive chromatographic and electrophoretic purifica-

tion steps have been necessary to obtain acceptable purity of human CgA (12,13). The immunoreactivity of the human CgA preparation (Figs 6 & 7) indicates that it will be useful in the detection and measurement of CgA in human tissues and serum.

Acknowledgements Supported by the Department of Veterans Affairs and the National Institutes of Health, and Krefifondet, Trondheim. Dr R. V. Lloyd (Department of Pathology, University of Michigan, USA) provided monoclonal antibody LK2HlO against human CgA. We acknowledge those physicians who allowed us to procure pheochromocytoma tissue at operation from their patients. MS Annie Chen and MI James Peirce assisted with the chromaffm granule preparations. Mr Matthew Williamson (Department of Biology, UCSD) assisted with the amino acid sequence determinationandprotein sequence database search. Dr M. R. Pandian (Nichols Institute, San Juan Capistrano, CA, USA) assisted with CgA radioiodination and radioimmunoassay.

References 1. O’Connor, D. T. (1983). Chromogranin: widespread immunoreactivity in polypeptide hormone producing tissues and in serum. Regul. Peptides 6: 263-280. 2. Helle, K. B. (1966). Some chemical and physical properties of the soluble protein fraction of bovine adrenal chromaf% granules. Molec. Pharmacol. 2: 298-310. 3. Fischer-Colbrie, R and Frischenschlager, I. (1985). Immunological characterization of secretory proteins of chromafEn granules: chromogranins A, chromogranins B, and enkephalin-containing peptides. J. Neurochem. 44: 1854-1861. 4. Rosa, P. and Zanini, A. (1981). Characterization of adenohypophyseal polypeptides by two-dimensional gel electroDhoresis. Mol. Cell Endocrinol. 24: 18 l-193. 5. H&ner,W.B.,Gerdes,H.-H.andRosa,P.(199l).Thegranin (chromogranin/secretogranin) family. TIBS 16: 27-30. 6. Konecki, D. S., Benedum, U. M., Gerdes, H.-H. and Huttner, W. B. (1987). The primarystructure ofhumanchromogranin A and pancreastatin. J. Biol. Chem. 262: 17 026-17 030. 7. Cohn, D. V., Fasciotto, B. H. and Gorr, S.-U., et al. (1990). The putative role of secretory protein-l/chromogranin A as a precursor for regulatory hormones. Mol. Cell. Regal. Calcium Phosphate Metab. Alan R. Liss, Inc., p. 51-66. 8. O’Connor, D. T. and Deftos, L. J. (1986). Secretion of chromogranin A by peptide-producing endocrine neoplasms. N. Engl. J. Med. 314/18: 1145-1151.

9. Eriksson, B., Amberg, H. and C)berg, K., et al. (1987).

Chromogranins - new sensitive markers for neuroendocrine tumors. Acta Oncologica 28: 325-9. 10. O’Connor, D. T. (1986). Radioimmunoassays of chromo-

granin. In: Krstulbvic (ed) Quantitative a&lysis of catecholamines and related compounds. John Wiley, London, p. 302-315. 11. Dillen, L., De Block, J., Van Lear, L. and De Potter, W. (1989). Enzyme-linked

immunosorbent assay for chromo-

graninA. Clin. Chem. 3519: 1934-1938. 13 O’Connor. D. T.. F&on. R. P. and Sokoloff. R. L. (1984). Human chromo&& Al Purification and characte&atioh from catecholamine storage vesicles of human pheochromocytoma. Hypertension 6: 2-12. 13. Wilson, B. S., Phan, S. H. and Lloyd, R. V. (1986). Chromogmnin from normal human adrenal glands: purification by monoclonal antibody affiity chromatography and partial N-terminal amino acid sequence. Regul. Peptides 13: 207-23. 14. Smith., A. D. and Winkler, H. (1967). A simple method for the isolation of adrenal chromafiin granules on a large scale. Biochem. J. 103: 480-2. 15. Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. J. Analyt. Biochem. 72: 248-54. 16. O’Connor, D. T., Frigon, R. P. and Stone, R. A. (1979). Human pheochromocytoma dopamine-beta-hydroxylase: purificationandmolecularparametersofthetetramer.Molec. Pharmacol. 16: 529-38. 17. Sokoloff, R. L., Frigon, R. P. and O’Connor, D. T. (1985). Dopamine-beta-hydroxylase: struc~l comparisons of membrane-bound versus soluble forms from adrenal medulla and pheochromocytoma. J. Neurochem. 44: 441-50. 18. Smith, W. J. and Kirshner, N. (1967). A specific soluble protein from the catecholamine storage vesicles of bovine adrenal medulla. I. Purification and chemical characterization. Molec. Pharmacol. 3: 52-62. 19. Kruggel, W., O’Connor, D. T. and Lewis, R. (1985). The amino terminal sequences of bovine and human chromogranin A and secretory protein I are identical. Biochem. Biophys. Res. Commun. 127: 380-3. 20. Barbosa. J. A.. Gill. B. M.. Takivvuddin M. A. and O’Connor. D. T. (1~91).‘Ch&mogr&in x post&nslational modifica: tions in secretory granules. Endocrinology 128: 174-90. 21. Huynh, T. V., Young, R. A. and Davies, R. W. (1985). In: Glover, D. M. (ed) DNA-cloning: a practical approach. IRL Press, Oxford, 1: 49-78. 22. Lloyd, R. V. and Wilson, B. S. (1983). Specific endocrine tissue marker defined by a monoclonal antibody. Science 222: 628-630. 23. Odell, W. D. and Franchiiont, P. (1983). Principles of competitive protein binding assays. John Wiley, New York, 2nd ed. 24. O’Connor, D. T., Pandian, M. R., Carlton, E., Cervenka, J. H. and Hsiao, R. J. (1989). Rapid radioimmunoassay of circulating chromogranin A: in vitro stability, exploration of the neuroendocrine character of neoplasia, and assessment of the effects of organ failure. Clin. Chem. 35: 163 1-1637. 25. Helman, L. J., Ahn, T. G. and Levine, M. A., et al. (1988). Molecular cloning and primaty structure of human chromogranin A (secretory protein I) cDNA. J. Biol. Chem. 263: 11 559-11 563.