Vol.
175,
March
No.
29,
BIOCHEMICAL
3, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
1991
THE 56 kDa ANDROGEN
Fred Pereira*,
“Department
BINDING
Eduardo Leonard
PROTEIN
IS AN ALDEHYDE
DEHYDROGENASE
Rosenmann”, Edward Nylen”, Morris Pinsky+ and Klaus Wrogemann*
of Biochemistry Winnipeg,
831-838
and Molecular Biology, University Manitoba, Canada, R3E OW3
Kaufman+,
of
Manitoba,
+Lady Davis Institute for Medical Research and Centre for Human Genetics, McGill University, Montreal, Quebec, Canada, H3A 1 Bl
Received
January
17,
1991
We have described a 56 kDa protein from genital skin fibroblasts that specifically binds androgen and that is generally not expressed in genital skin fibroblasts from patients with androgen insensitivity due to genetic defects of the androgen receptor. We have isolated a partial cDNA clone for the 56 kDa protein from an expression library of genital skin fibroblasts. In vitro translation of message selected with this clone faithfully produces the 56 kDa protein which can be immunepricipitated with an anti56 kDa antiserum. Northern blots probed with this clone show a 2.2 kb message, which parallels the expression of the 56 kDa protein. The sequence of this 998bp clone is identical to human liver aldehyde dehydrogenase 1, the cytoplasmic isoenzyme. On activity gels of genital skin fibroblast cytosol covalently labelled with androgen, aldehyde dehydrogenase activity comigrates with the single band labelled specifically with androgen. Thus, the 56 kDa androgen binding protein is an aldehyde dehydrogenase, which is prominently expressed in normal genital skin fibroblasts, but not in non-genital skin fibroblasts. B 1991 Academic Wess, inc.
In a comparison (GSF)
we have discovered
denaturing points patients
of protein maps from non-genital
conditions,
a protein
the protein
which
positions
androgen
insensitivity
genital skin fibroblasts
is only expressed as a doublet
of 6.5 and 6.7 (1,3,4). This relatively abundant with complete
versus
in GSF (1,2). Under
of 56 kDa with
protein is not detected
due to genetic
defects
isoelectric in most
of the androgen
Abbreviations: DHTBrAc, dihydrotestosterone bromoacetate; GSF, genital skin fibrobla sts; CHAPS, (3-([3-Chlolamidopropyl)-dimethylammoniol] I-propanesulfonate); PIPES, 1,4-piperazine diethane sulphonic acid; HEPES, (N-2-Hydroxyethylpiperazine-N’-2ethanesulphonic acid); Eagle’s MEM, Eagle’s minimal essential medium; PAGE, polyacrylamide gel electrophoresis; BSA, bovine serum albumin; PVP, polyvinylpyrrolidone.
831
0006-291X/91 $1.50 Copwight 0 1991 bv Academic Press, Im. All rights of reproduction in a+ form reserved.
Vol.
175,
No.
receptor
3, 1991
BIOCHEMICAL
(3-5). However,
several
AND
patients do express
that the protein is not androgen-induced (4-6)
albeit at a lower
discovered receptor were
this
and concluded
receptor
that it was
of it (57). It was clearly shown
with
androgen
receptor
defects
receptor
(6) and also not an unorthodox
androgen
receptor
gene,
receptor cDNA
indeed
coding
sequence
of this “androgen
aldehyde
dehydrogenase
as one patient still expresses binding
protein”
with
(6). Other
1 and that this enzyme
groups
the human
recent evidence
product
a deletion
specifically
androgen
from studies is not the
synthesized of the entire
100% identity with binds
also
investigators
that the protein
this protein (9). We show shows
suggesting
binds androgen
that these different
indicated
androgen
COMMUNICATIONS
the protein, strongly
indeed dealing with the same protein (8). However,
of patients
RESEARCH
(4,6). The protein specifically
affinity than to the androgen
protein
or a fragment
BIOPHYSICAL
from
the
androgen
here that a partial the cytoplasmic androgen.
METHODS Specific Labellina of the 56 kDa Protein: Human genital skin fibroblasts (GSF) were cultured under standard conditions in Human McCoy’s 5A medium (Bocknek) supplemented with 10% fetal calf serum. Confluent 60 mm dishes were rinsed with 3 ml of sterile buffer (15 mM HEPES, pH 7.4 in Eagle’s minimal essential medium). The dishes were incubated with 5 nM 3H-DHTBrAc synthesized according to (10) starting from [1,2,4,5,6,7,16,1 7-3H,] dihydrotestosterone (179.4 Ci/mmol, New England Nuclear) with or without 1 PM (200-fold excess) of radioinert DHTBrAc for 2 h in the incubator. After incubation, the used media were removed and the cells lysed directly with CHAPS lysis buffer for 2D-PAGE (3) or with SDS sample buffer for SDS-PAGE. One and two-dimension gel electrophoresis has been described by us and others previously (3,ll ,12). Gels were impregnated with Autofluor (National Diagnostics) prior to drying and fluorography by exposing to preflashed Kodak XAR-5 X-ray film at 80°C. Silver staining was performed as described (13). cDNA Librarv Buildino, lmmunoscreeninq and DNA Seauencinq: Poly A+ RNA was isolated from cell strains using oligo-dT cellulose supplied with a Fast Tract mRNA isolation kii (Invitrogen Corporation). A cDNA library was constructed in a lambda gtl 1 expression vector with poly A+ RNA from a human GSF strain MCHG, which expresses the 56 kDa protein (4) and was immunoscreened as described (14,15), using a previously described rabbit antiserum against the 56 kDa protein (6). Positive clones were plaque purified and subcloned into pUC19 for further amplification and insert isolation for probing blots or Ml3 for single strand dideoxy chain termination DNA sequencing (16). Northern Blot Analvsis: 10 pg of poly A+ RNA of each sample was denatured with 50% (v/v) formamide and 2.2 M formaldehyde and electrophoresed through a 1% (w/v) agarose gel containing 2.2 M formaldehyde and blotted onto a nitrocellulose filter. The filter was baked at 80°C for 2 h in a vacuum oven, prehybridized in hybridization buffer and hybridized to a 0.7 kb cDNA insert from PUMBl which was labelled with 32P by random priming (17) to a specific activity of 2 x lo8 cpm/pg. Hybridizations, usually for 16 h, were performed at 42°C in the presence of 50% (v/v) formamide, 5 x Denhardt’s (1 x solution was 0.02% (w/v) each of BSA, Ficoll and polyvinylpyrrolidone), 832
Vol.
175,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
5 x buffer (1 x solution was 1.15 M NaCI, 10 mM NaH,PO, and 1 mM EDTA), 250 pg/ml denatured salmon sperm DNA and 0.1% SDS. The filter was washed 2 x 20 min at room temperature and at 65°C in a buffer (0.1 x SSC and 0.1% SDS). Filters were exposed to X-ray film with a Lightning Plus (New England Nuclear) intensifying screen at -80°C. Hvbrid Select in Vitro Translation with Clone PUMB 1: 50 pg of the positive clone PUMB 1 or carrier plasmid pUC19 were denatured and bound to nitrocellulose filters (15). The filters were used to hybridize against 40 pg of poly A+ RNA from strain MCH6 in 300 ~1 of hybridization buffer (50% deionized formamide, 0.4 M NaCI, 0.2% SDS, 30 mM PIPES, pH 6.5, and 50 pg/ml yeast tRNA) at 50°C for 3 h. The filters were washed 10 times in a Tris buffer (10 mM Tris-Cl pH 7.6, 1 mM EDTA, 150 mM NaCl and 0.5% SDS) at 65°C and the messages released into 300 PI of water containing 2 ~1 of 10 mg/ml yeast tRNA by boiling and quick-freezing in a dry iceethanol bath. The solution was allowed to thaw at room temperature and the filters removed. The mRNA was phenol and chloroform extracted and then precipitated with two volumes of ethanol and l/50 vol. of NaCI. The precipitated messages were translated by a wheat germ in vitro translation system (Promega) for 1 h in the presence of L-%S-methionine (1070 Ci/mmol, New England Nuclear). The radioactive translation products were incubated with anti-56 kDa antiserum for 1 h in immuneprecipitation buffer (10 mM Tris-Cl pH 7.4, 2 mM EDTA, 150 mM NaCl and 10% Nonidet P-40) and precipitated with Sepharose protein-A (Pharmacia). The bound proteins and antibodies were released by boiling in the presence of SDS sample buffer. The supernatants were fractionated by SDS-PAGE and the gels treated for fluorography as above. Non-Denaturinq Isoelectric Focusinq of Radiolabelled Cvtosols: Confluent 150 mm dishes of human GSF were rinsed with 5 ml of sterile buffer (15 mM HEPES pH 7.4 in Eagle’s MEM). The dishes were incubated with 5 nM 3H-DHTBrAc with or without 1 PM (200-fold excess) of radioinert DHTBrAc for 2 h in the incubator. After incubation, the used media were removed and the cells lysed directly by scraping into a phosphate buffer (10 mM phosphate pH 7.5, 10% glycerol, 3 mM EDTA, 2 pg/ml of pepstatin A and leupeptin), triturated through a 10 ~1 Hamilton syringe and centrifuged for 2 minutes in a microcentrifuge at 4 “C. Non-denaturing isoelectric focusing of protein supernatants in phosphate buffer were carried out as follows: 20 pg of cytosol labelled with 3H-DHTBrAc + 200-fold excess cold ligand and 80 kg of unlabelled cytosolic proteins were applied to the same sample pad on a prefocused acrylamide slab gel (5% acrylamide, 3% bis, 10% glycerol, and 2% Pharmalyte (Pharmacia) pH 3-l 0). The gel was focused for 3 h, for a total of 1800 Vh, at 4°C using 1 M H,PO, and 1 M NaOH as the electrolytes. After activity staining (18) the proteins were fixed in 10% TCA and the gel rinsed in 50% methanol and 10% acetic acid. A photograph was taken and the appropriate lanes were rinsed in water and sliced into 3 mm fractions, digested in 0.5 ml NCS tissue solubilizer (Amersham) containing 10% water for 1 h and neutralized with 13 ~1 of glacial acetic acid before counting with 15 ml of 0.4% Omnifluor (New England Nuclear) in toluene. Two separate lanes were also sliced prior to activity staining and the pieces soaked in 0.5 ml of water and the pH of each slice determined to produce a pH profile of the focused slab gel.
RESULTS Specific
Labellina
Figure 1A shows
of the 56 kDa Protein with 3H-Dihvdrotestosterone the relative abundance
Bromoacetate:
of the 56 kDa protein in the cytosolic 833
fraction
Vol.
175,
o-acid
No.
3, 1991
-alkaline-,
BIOCHEMICAL
eacid-
AND
BIOPHYSICAL
-alkaline,
RESEARCH
+acid----
COMMUNICATIONS
-alkaline+
Fiaure 1. Specific labelling of the 56 kDa protein by 13H]-DHTBrAc in cytosol fractionated by 2D-PAGE and SDS-PAGE. A, Silver stained protein map of GSF cytosol from a normal individual, MCH 6. B and C, 2D-PAGE fluorogram of cells labelled with 5 nM [3H]-DHTBrAc + 200-fold excess cold ligand respectively. D and E, SDS-PAGE fluorogram of cells labelled with 5 nM 3H-DHTBrAc+ 200-fold excess cold ligand. The arrows point to the 56 kDa protein doublet or band.
of GSF. This protein can be efficiently and selectively labelled covalently with
5 nM
3H-DHTBrAc (Figure 1B). Labelling is saturable, as deduced from the suppression of labelling by the inclusion of 1 uM (200-fold excess) of radioinert ligand (Figure IC). Figures 1D and 1E show the specific labelling of the 56 kDa protein on onedimensional SDS-PAGE gels, confirming that the only major species of all cellular proteins labelled with androgen is the 56 kDa protein. Isolation of a cDNA Candidate Clone for the 56 kDa Protein: A cDNA library was built in the expression vector lambda gtll
and screened with a polyclonal rabbit antiserum
prepared against pl 6.7 56 kDa protein spots cut out from two-dimensional gels (6). One of the clones detected (PUMBl) has a 1.0 kb insert. When PUMBl is used as a probe on Northern blots of Poly A+ RNA of a variety of normal and mutant cell lines, a 2.2 kb message is detected, only in those cells which express the 56 kDa protein (Figure 2). When PUMBl is used for hybrid selection of mRNA for in vitro translation, the message for a 56 kDa protein is selected, and this 56 kDa protein is precipitable with the specific anti-56 kDa antiserum (Figure 3). Sequence of the PUMBl
Clone: The cDNA sequence of clone PUMB
1 was
determined after subcloning into Ml3 by the dideoxy chain termination method. When compared to the sequences in GenBank, 100% homology was detected with human aldehyde dehydrogenase
1 (EC 1.2.1.3) for the 998 bps of this clone (Figure 4). 834
Vol.
02
175,
4+
No.
l+
3, 1991
Cl
0
BIOCHEMICAL
0
4+
0
AND
BIOPHYSICAL
e-
18s
4--
Expression 56 kDa Protein
RESEARCH
COMMUNICATIONS
03
Fiaure 2. Northern Blot Analysis of PUMB 1 mRNA extracted from: I, GSF of a normal individual (MCHG); 2, GSF from an androgen insensitive patient (DB); 3, non-GSF from a normal individual (WPO9); 4, Lymph Node Carcinoma of the Prostate cell line (LNCaP) which over expresses the androgen receptor; 5, 6, and 7 are GSF from androgen insensitive patients 6779, NHL, and LEL respectively. The expression of the 2.2 kb message parallels the expression of the 56 kDa protein as illustrated by the scores from 0 (not detected) to 4+ (most strongly expressed) (3). Fiqure 3. In vitro translation of mRNA, total or hybrid selected with clone PUMB 1, from normal GSF. A, translation of total mRNA; B, hybrid selected with cDNA clone PUMB 1; C, hybrid selected with the vector pUCI9; D, immuneprecipitate of the translation products of B; E, molecular weight markers in kilodaltons.
Aldehyde dehydrogenase 1 is the cytoplasmic isoenzyme (ALDHl), which has been cloned and sequenced from human liver (19). Aldehvde Dehvdroaenase Activitv Associated with the 56 kDa Protein. To test whether normal genital skin fibroblasts indeed have aldehyde dehydrogenase activity, GSF cytosol was separated on non-denaturing isoelectric focusing gels and stained for aldehyde dehydrogenase activity with acetaldehyde and benzaldehyde as substrates. A single prominent band was detected with a pl of 5.1 (Figure 5). In the same experiment, the cells had been labelled in situ with 5 nM 3H-DHT8rAc 2 200-fold excess cold ligand. After staining and photography the gel was sliced and counted,
988 PUMB
bp
1 cDNA
SEQUENCING
STRATEGY
ALDH 1 cDNA SEQUENCE IDENTICAL TO PUMB 1
Fioure 4. Schematic to ALDHI sequence.
representation
“Lv I 748
I 1746
of cDNA sequence
835
of clone PUMBl
as compared
Vol. 175, No. 3, 1991
BIOCHEMICAL
a.
%_
-I.
--a... ‘0...._. ““0 . ..._ “CL,
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
8
=-..... ‘0~%..., ‘9
I a 6
in..,. 3H-DHTBrAc l#“....~lll.~ 3H-DHTBrAc --“-e-PH
““-.*. .. + Cold -f
40
60
Distance
...TPl =“....
5.1 -2.
\.
4
2
(mm)
Figure 5. Aldehyde dehydrogenase activity associatedwith 3H-DHTBrAc labelled GSF cytosol from the normal individual MCHG.Activity gel is shown as an insert. Lane A, cells labelled with 5 nM 3H-DHTBrAC;lane B, as A, but with the inclusion of 200-fold cold DHTBrAc. The pH profile and counts per minute are plotted against the length of the non-denaturing isoelectric focusing gel. The enzyme activity and radioactivity bands both focussed at pl 5.1 (arrow).
showing that the specific labelling of the only protein in GSF cytosol comigrates with the aldehyde dehydrogenase activity (Figure 5).
DISCUSSION Based on a number of criteria, the partial cDNA clone isolated here codes for the 56 kDa “androgen binding protein” described by us (l-4) and others (5,7): (i) the clone detects a message that reflects quantitatively the 56 kDa protein, in a variety of normal and mutant cell lines which do or do not make this protein; (ii) the PUMBl clone specifically selects a message which translates into a 56 kDa protein recognized by the purified antiserum (6), which is specific for the 56 kDa protein, and which was used to isolate the clone; (iii) the partial sequence shows 100% identity with human aldehyde dehydrogenase
1 (19); (iv) the subunit size, isoelectric point on non-
denaturing gels and message size for the 56 kDa protein are in close agreement with those for the enzyme (18,19,20); (v) GSF show aldehyde dehydrogenase activity, and this activity comigrates with the band that contains the protein specifically labelled with androgen; (vi) the data agree with the observation that ALDHl is not expressed in skin fibroblasts (21), which now has to be specified as non-genital skin fibroblasts. 836
Thus,
Vol.
175,
No.
3, 1991
BIOCHEMICAL
the 56 kDa protein aldehyde
is the subunit
dehydrogenase
cDNA
is necessary
ALDHl
of human
In spite of the specific our previous supergene
hypothesis family
RESEARCH
it may represent
binding
that this protein
identical
member
the data rule out
of the steroid receptor
that specific androgen
prove that a protein is a receptor
to the
isoform.
of the androgen
(4,5,7,8). The finding shows
likely
of this GSF
in all aspects
to this protein,
is another
most
and sequencing
another
of androgen
COMMUNICATIONS
dehydrogenase,
cloning
it is indeed
(4) or that it is a fragment
also been suggested not necessarily
complete
whether
liver or whether
BIOPHYSICAL
form of an aldehyde
1. However,
to determine
AND
or belongs
receptor
itself as has binding does
to the steroid
binding
protein family. The aldehyde of a broad reaction
dehydrogenases
range
of aldehydes
(18). Whether
enzyme’s
at other sites preferential defective
remains
of this protein
or other steroids
or esterase to be studied.
expression androgen
to the corresponding
androgens
dehydrogenase
(EC 1.2.1.3) catalyze the irreversible
in genital
receptor
in X-linked
activities,
acids
Of even greater
skin fibroblasts,
insensitivity
(22) for this
they bind to this protein
interest to us at this time is its
and the mechanism
gene often impairs that expression
androgen
in a NAD-dependant
are natural substrates
or whether
oxidation
by which
and the possible
a role
syndrome.
ACKNOWLEDGMENTS This work
was supported
Research
Council
of Canada
Hospital
of Winnipeg
Research
by the Manitoba Group
Grant
Health Research
in Medical
Genetics
Council, the Medical and the Children’s
Foundation.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
Rosenmann, E., Kreis, C., Thompson, R. G., Dobbs. M., Hamenon, J. L. & Wrogemann, K. (1982) Nature 298, 563-565. Thompson, R. G., Nickel, B., Finlayson, B., Meuser, R., Hamerton, J. L., & Wrogemann, K. (1983) Nature 394, 740-741. Nickel, B., Schwartz, A., Rosenmann, E., Kaufman, M., Pinsky, L. & Wrogemann, K. (1988) Clin. Invest. Med. 11, 22-33. Wrogemann, K., Pereira, F., Belsham, B., Kaufman, M., Pinsky, L. & Rosenmann, E. (1988) Biochem. Biophys. Res. Commun. 155, 907-913. Kovacs, W. J. & Turney, K. M. (1988) J. Clin. Invest. 81, 342-348. Pereira, F., Belsham, D., Duerksen, K., Rosenmann, E., Kaufman, M., Pinsky, L. & Wrogemann K. (1990) Molec. Cell. Endocrinol. 68, 195-204. Kovacs, W. J., Turney, M. K. & Skinner, M. K. (1989) Endocrinology 124, 1270-l 277. 837
Vol.
175, No. 3, 1991
8.
9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
Belsham, D. D., Rosenmann, E., Pereira, F. A., Williams, S. G., Turney, M. K., Kovacs, W. J., Faber, L. E. & Wrogemann, K. (1989) J. Steroid Biochem. 33, 389-394. Trifiro, M., Gottlieb, B., Pinsky, L., Kaufman, M., Belsham, D. D., Wrogemann, K., Brown, C. J., Willard, H. F., Trapman, J., Brinkmann, A. O., Chang, C., Liao, S., Sergovich, F. & Jung, J. (1990) Molec. Cell. Endocrinol., In press. Martin, K. 0. & Monder, C. (1978) J. Steroid Biochem. 9, 1233-1240. Laemmli, U. K. (1970) Nature 227, 680-685. O’Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021. Morrissey, J. H. (1981) Anal. Biochem. 117, 307-310. Young, R. A. and Davis, R. W. (1983) Science 222, 778-782. Huynh, T. V., Young, R. A. and Davis, R. W. In DNA Cloning: A Practical Approach, Vol. 2 (ed. D. M. Glover) 49-78 (IRL Press, Oxford 1985). Sanger, F., Nicklen, S. and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. Feinberg, A. B. and Vogelstein, B. (1983) Anal. Biochem. 132, 6-13. Kurys, G., Ambroziak, A. & Pietruszko, R. (1989) J. Biol. Chem., 264, 4715-4721. Hsu, L. C., Tani, K., Fujiyoshi, T., Kurachi, K. & Yoshida, A. (1985) Proc. Natl. Acad. Sci. USA 82, 3771-3775. Agarwal, D. P. & Goedde, H. W. (1989) Alcohol 6, 517-523. Anthony, C. T., Kovacs, W. J. & Skinner, M. K. (1989) Endocrinology 125. 2628-2635. Hsu, L. C., Yoshida, A. & Mohandas, T. (1986) Am. J. Hum. Genet. 38, 641-648.
838