Correspondence to: Lana Grauer, Hybritech Incorporated, P.O. Box

269006, San Diego, California 92196-9006.

Received for publication December 5, 1995; accepted for publication

February 29, 1996.

353

Journal of Anthology. Vol. 17, No. 4, July/August 1996

Copyright © American Society of Andrology

Identification of Human Glandular Kallikrein hK2From LNCaP Cells

LANA S. GRAUER, M. CRISTINE CHARLESWORTH, MOHAMMAD S. SAEDI, JUDITH A. FINLAY,

RU-SHYA LIU, KRISTINE KUUS-REICHEL, CHARLES Y.-F YOUNG, AND DONALD J. TINDALL

From Hybritech Incorporated, San Diego, California and Department of Urology, Mayo Clinic/Foundation,

Rochester, Minnesota.

ABSTRACT: Based on studies indicating that human glandular kal-likrein (hK2) mANA is present in the prostate, we prepared a mono-clonal antibody to a synthetic peptide corresponding to the 41-56region of hK2 to try to identify the hK2 protein. Although prostate-specific antigen (PSA) and hK2 share 80% homology, the 41-56

amino acid sequence of hK2 is only 50% homologous with PSA. Amonoclonal antibody, HK1A523, was identified that demonstrateshigh specificity for hK2. In western blot analysis, the antibody has a1 .000-fold greater sensitivity for the detection of hK2 than for PSA.

The antibody was used to probe spent media from the prostate car-

cinoma cell line, LNCaP. An irnmunoreactive species was N-termin-ally sequenced and identified as mature hK2. HK1 A523 was alsoutilized to probe prostate tumor cytosols and seminal fluid whereputative forms of hK2 were also identified. The hK2 protein thereforeis expressed and secreted from prostate carcinoma cells.

Key words: Prostate carcinoma, rnonoclonal antibody, seminalfluid.

J Androl 1996;17:353-359

In 1987 a new human glandular kallikrein gene was iso-

lated (Schedlich et al, 1987). The predicted proteinproduct of this gene was named human glandular kalli-

krein or hGK-1 and later renamed hK2 (Berg et al, 1992).

The glandular kallikreins are serine proteases believed to

have a role in activation of growth factors and peptidehormones (Murray et al, 1990). Members of the kallikrein

family include prostate-specific antigen, PSA (hK3)

(Wang et al, 1979), and pancreatic/renal tissue kallikrein

(hKl) (Clements, 1989). The hK2 transcript was identi-

fied in prostate epithelial cells by Chapdelaine et al (1988)

and Young et al (1992) by in situ hybridization. The ex-

pression of hK2 mRNA was reported to be restricted to

prostate tissue (Morris, 1989; Henttu et al, 1990). Al-

though the mRNA for hK2 has been isolated, the protein

has yet to be identified. Early efforts to identify the pro-

tein proved difficult due to its similarity to PSA (Paradis

et al, 1989). The protein sequence for hK2 was deduced

from the cDNA sequence and shown to share 80% se-

quence homology with PSA (Schedlich et al, 1987). One

region of sequence diversity occurred at amino acids 41-

56. In this stretch of 16 residues there were 8 amino acid

differences between the two proteins. In addition, PSA

contained a putative glycosylation site in this region atAsn-45 (Lundwall and Lilja, 1987) that was not present

in hK2.

We have developed a monoclonal antibody (mab),

HK 1A523, against the 41-56 region of hK2 that reactswith recombinant forms of hK2 and shows minimal cross-

reactivity to PSA. The mab was used as a probe to iden-

tify hK2 in LNCaP spent media. The immunoreactive

band was sequenced and identified as hK2. Prostate car-

cinoma cytosol and seminal fluid were also probed with

the hK2-specific antibody. Both of these sources con-

tained HK1A523 immunoreactive protein.

Materials and Methods

Peptide Synthesis

The peptide corresponding to amino acids 41-56 of the hK2

protein was synthesized using standard F-moc solid-phase chem-

istry on an Applied Biosystems 431A peptide synthesizer. Pep-tides were purified by reverse-phase high-performance liquid

chromatography (HPLC) and analyzed by mass spectrometryand analytical HPLC prior to immunization and assay develop-

ment.

Cloning and Expression of Recombinant Constructs

PreprohK2 (pphK2) was prepared in bacteria as explained in

Saedi et al (1995). The hK2 cDNA fragment encoding the entiresequence of pphK2 was cloned into a T7 expression vector underthe control of the T7 promoter. The plasmid (pBpphK2) was

transformed into Escherichia coli BL21, and pphK2 expressionwas induced by the addition of isopropyl-3-D-thiogalactopyra-

noside. Bacterial pellets were lysed and the inclusion bodieswere subjected to SDS-PAGE. The pphK2 was purified by gel

354 Journal of Andrology . July/August 1996

elution. The preproPSA (ppPSA) was generated in a similar

fashion.ProhK2 was purified from the AV12-664 hamster tumor cell

transfected with pphK2 cDNA, and mature hK2 was obtained

from the recombinant prohk2 preparation following mild tryp-sinization (Mikolajczyk, unpublished data).

Antibodies

Six-week-old A/J mice obtained from Jackson Laboratories (BarHarbor, Maine) were immunized intraperitoneally with 50 pg of

hK2 41 -56-KLH. After two injections of immunogen, positive

titers were obtained and mice were final boosted intravenously

with 25 ig of the same immunogen. Hybridomas were producedby fusion of mouse splenocytes with P3.653 myeloma cells

(Kohler, 1975) obtained from American Type Culture Collection

in Bethesda, Maryland. Monoclonal antibodies were selectedbased on reactivity with peptide as well as binding to recombi-

nant pphK2 and recombinant prohK2, and failure to react with

recombinant ppPSA and native PSA purified from seminal fluid.Monoclonal antibodies were screened by indirect enzyme-linked

immunosorbent assay (ELISA) on biotinylated protein captured

on avidin-coated microtiter plates and detected by horse radishperoxidase (HRP)-conjugated goat anti-mouse immunoglobulin,

followed by color development with o-phenyl-aminediamine

(Sigma, St. Louis, Missouri). Ascites was produced in nude mice

and the mabs were purified by protein A chromatography.

Affinity-purified goat anti-protein C inhibitor (PCI) was pur-

chased from Enzyme Research Labs (South Bend, Indiana). The

anti-PSA mab P5M773 was previously prepared from a mouse

immunized with purified PSA. The polyclonal rabbit anti-PSA

was prepared previously by immunizing a rabbit with purified

PSA. All horseradish-conjugated secondary antibodies were ob-

tained from Jackson ImmunoLabs (West Grove, Pennsylvania).

LNCaP Cell Line and Mibolerone Induction

The prostate carcinoma cell line, LNCaP.FGC was obtained fromthe American Type Culture Collection (Bethesda, Maryland).The cells were maintained in RPMI media with 10% horse serum

at 37#{176}Cand 5% CO2. After cells had reached 80% confluency,

they were incubated in serum-free media containing 10 nM mi-

bolerone (Amersham, Arlington Heights, Illinois) for 48 or 100

hours under the same growth conditions. Supernatants were har-vested, concentrated by Amicon ultrafiltration to 50X concen-

trates, and stored at - 20#{176}C.

Western Blot Analysis of hK2

Reduced and denatured proteins were electrophoresed on 4-20%

polyacrylamide gels and blotted onto nitrocellulose (Towbin et

al, 1979). Successful transfer was checked by staining blots with

Ponceau S protein stain. Blots were blocked in 5% nonfat dry

milk in phosphate-buffered saline (PBS) and then incubated withantibodies at 4#{176}Covernight. Following washes, secondary anti-

bodies conjugated to HRP were incubated with the blots for 1

hour and again washed with PBS. Antibody reactivity was de-

tected by the addition of 10 mg/mI diaminobenzidine, 0.1% hy-

drogen peroxide in PBS, or using the enhanced chemilumines-

cence (ECL) reagent (Amersham, Arlington Heights, Illinois) ac-

cording to manufacturer’s instructions. In experiments where

41 58

hK2 HCLKKNSQVWLGRHNL

PSA HCIRNK SVI LLGRHS L

FIG. 1. The amino acid sequences of hK2 and PSA at residues 41-56.

ECL detection was used, blots were incubated with antibodies

for 1 hour at room temperature and exposed to X-ray film, typ-

ically for 30 seconds.

Partial Purification of hK2 from LNCaP Supernatants

Concentrated spent media from mibolerone-stimulated LNCaP

cells was chromatographed by anion exchange and hydrophobic

interaction to yield a partially purified preparation of hK2 and

PSA. These proteins were separated and identified by sodium

dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE)

and western blot analysis with HK 1 A523 and PSA-specific rab-

bit polyclonal antisera (Towbin et al, 1979). The proteins were

electrophoresed and blotted onto high-density polyvinylidene di-

fluoride (PVDF) membranes. The hK2 and PSA bands were cut

out and sequenced on an Applied Biosystems (ABI) 492 Procise

Automated Sequencer using pulse liquid mode. Data analysis

was performed using the ABI 610A sequence analysis software.

Cytosol PreparationSpecimens of prostate carcinoma obtained at surgical resection

were immediately frozen and stored at - 80#{176}C.Frozen sections

of the specimens were examined microscopically and those spec-imens containing greater than 90% tumor cells were selected for

homogenization. The tissue was cut into small pieces and then

homogenized in 4 volumes of 0.1 M tris-HCI pH 7.5, 100 mM

NaCI, 2 mM CaCI, 1 mM Phenylmethyl sulfonyl fluoride using

a Dounce homogenizer. Nuclei and intact cells were removed by

centrifugation at 500 x g for 5 minutes. The supernatant was

centrifuged at 39,000 X g for 30 minutes at 4#{176}C.The supernatant

was removed and frozen and used as the cytosol source in further

experiments.

PSA Quantitation MethodPSA levels were determined by an immunoradiometric assay,

Tandem#{174}-R PSA (Hybritech, San Diego, California).

Results

The 4 1-56 amino acid sequences of hK2 and PSA contain

significant amino acid diversity (50%) (Fig. l).We syn-thesized this region of hK2 as an immunogen for the pro-

duction of hK2-specific antibodies. HK1A523 (IgG2a)was selected based on its strong reactivity with recom-

binant pphK2 and limited binding to recombinant ppPSA

in ELISA. HK1A523 was examined more extensively bywestern blot analysis for potential cross-reactivity to PSA

purified from seminal fluid and recombinant hK2. Incu-

bation of HK1A523 with increasing amounts of purified

recombinant hK2 demonstrated that the antibody can de-

tect as little as 15 ng of the recombinant (Fig. 2).

Grauer et al - Human Glandular Kallikrein 355

A. hK2

B. PSA

19 ng 78 ng 312 ng 1.25 .tg 5 g 20 p.g

33 kDa

20 kDa

15 kDa

33 kDa

20 kDa

19.5 kDa

FIG. 2. Reactivity of HK1A523 with hK2 and PSA on western blot. Increasing amounts of hK2 or PSA were electrophoresed under reducingconditions on SDS-PAGE on a 4-20% gel and blotted to nitrocellulose. The blots were reacted with 10 /rnl of HK1A523 followed by goat anti-mouse lgG-HRP and detection with diaminobenzidine. (A) hK2; (B) PSA.

1 2 3 4

356 Journal of Andrology . July/August 1996

98kDa

64

50

36

30

16

6

FIG. 3. Induction of hK2 with mibolerone in LNCaP cells. Media from LNCaP cells grown in the presence or absence of 10 nM mibolerone wasconcentrated 50x and electrophoresed under reducing conditions on SDS-PAGE 4-20% gels (30 il/lane). Following transfer to nitrocellulose, theblot was probed with HK1A523 (10 jig/mI), followed by goat anti-mouse lgG-HRP and detection with diaminobenzidine. Lane 1, molecular weightmarkers; lane 2, 50x LNCaP spent media; lane 3, 50X LNCaP media + 10 nM mibolerone, 48 hours; lane 4, 50x LNCaP media + 10 nM mibolerone,100 hours.

HK1A523 detects the predominant form of hK2 at 33

kDa. At higher levels of hK2, in addition to the 33-kDa

form, two smaller species at approximately 20 kDa and

15 kDa are also detected with the mab. In contrast, in-

cubation of HK 1 A523 with increasing amounts of PSA

demonstrated that the antibody detects PSA only at the

highest level of (20 rag). An anti-PSA mab PSM773 stains

all levels of PSA (data not shown). Therefore, HK1A523

has a 1,000-fold greater sensitivity for the detection of

hK2 than for PSA in this format. The predominant cross-

reactivity seen with HK1A523 at the highest level of PSA

(20 p.g) is not with intact PSA at 33 kDa but rather with

two smaller bands at approximately 20 kDa and 19.5 kDa.

These results suggest that HK 1A523 can be used to detect

hK2 specifically by western blot with minimal cross-reac-

tivity to PSA.

In order to determine if the hK2 protein is expressed

in LNCaP cells and regulated by the androgen, mibole-

rone, concentrated media from LNCaP cells grown in the

presence or absence of androgen was probed with

HK1A523 in western blots. HK1A523 reacts with a band

at 33 kDa in LNCaP spent growth media (Fig. 3). The

addition of 10 nM mibolerone to the growth media for

48 hours increases the expression of the protein approx-

imately twofold (lane 3) and incubation for 100 hours

increases expression levels approximately fivefold (lane

4). The control LNCaP media (50X concentrate) contains

7 ag/ml PSA or 140 hg PSA/lane and the 100-hour mi-

bolerone-induced LNCaP media (50X concentrate) con-

tains 33 p.g/ml PSA or 660 ng PSA/lane. Because the

amount of PSA loaded per lane is significantly below the

detection level for HK1A523 (i.e., 20 pjg), the 33-kDa

immunoreactive bands are deduced to be hK2.In order to confirm that the antibody-reactive band is

hK2, the protein was partially purified and sequenced.

Mibolerone-induced LNCaP spent media was concentrat-

ed and fractionated by anion exchange and hydrophobicinteraction chromatography. The immunoreactive en-

riched material was separated by gel electrophoresis and

blotted to PVDF membranes. The blot was stained with

Coomassie blue and probed with the anti-hK2 mab. The

Coomassie blue-stained band that comigrated with the

HK1A523 immunoreactive band was N-terminal se-

quenced through 20 residues and shown to be mature

hK2. A second band of slightly higher molecular weight

was N-terminal sequenced through 20 residues and iden-

tified as PSA.

Prostate carcinoma cytosols prepared from surgicalspecimens were examined for the presence of hK2.

HK 1 A523 reacts with three protein bands migrating at 62

1 2 3 4 5

Grauer et al - Human Glandular Kallikrein 357

98 kDa

64

50

36

30

16

6

FIG. 4. Reactivity of HK1A523 with prostate carcinoma cytosol. Prostate carcinoma cytosol (50 jig/lane) and PSA (1 jig/lane) were electrophoresedunder reducing conditions on an SDS-PAGE 4-20% gel. Following transfer to nitrocellulose, the blot was probed with HK1A523 or anti-PSA mabfollowed by goat anti-mouse lgG-HRP and ECL detection. Lane 1, molecular weight markers; lane 2, prostate carcinoma cytosol specimen #1, probedwith HK1A523 (10 jig/mI); lane 3, prostate carcinoma cytosol specimen #2, probed with HK1A523 (10 jig/mI); lane 4, 2 jig of PSA probed withHK1A523 (10 jig/mI); lane 5, 2 jig of PSA probed with anti-PSA mab (0.2 jig/mI).

kDa, 33 kDa, and approximately 20 kDa in both prostate

carcinoma cytosol preparations (Fig. 4, lanes 2 and 3). In

lane 2 the cytosol loaded onto the gel contains 880 ng of

PSA and lane 3 contains 280 ng of PSA. Lanes 4 and 5

are controls showing that 2 g of purified PSA cannot be

detected by HK1A523 (lane 4) while the anti-PSA mab,

PSM773 binds the 2 p.g of PSA (lane 5). These data sug-

gest that western blot detection is specific for hK2 and

that hK2 exists as multiple forms in prostate tissue.

Seminal fluid was also evaluated by western blot anal-

ysis with the anti-hK2 mab. HK1A523 reacts with four

bands in seminal fluid at 85 kDa, 33 kDa, and approxi-

mately 20 kDa and 15 kDa (Fig. 5, lane 2). An 85-kDa

band is also seen when seminal fluid is probed with an

anti-protein C inhibitor (PCI) mab (lane 3). Because both

antibodies bound an 85-kDa protein, this protein appears

to represent an hK2-PCI complex. The lower band (55

kDa) reacting with anti-PCI antibody in lane 3 most likely

represents free PCI in the seminal fluid. The 33-kDa band

in seminal fluid comigrates with recombinant mature hk2

(lane 4). Each lane of seminal fluid (50 p.g protein) con-

tains 2 g of PSA determined by the PSA-specific im-

munoassay. This level of PSA was previously shown to

be too low to be detected with the anti-hK2 mab (Fig. 4).

Discussion

This report describes the first identification of the hK2

protein in prostate cancer cells. The 33-kDa protein was

identified as hK2 by its reactivity with the anti-hK2-spe-

cific mab, HK1A523, followed by N-terminal sequencing

of the immunoreactive band blotted to PVDF membranes.Amino acids at positions 17 and 19 in hK2 are different

from corresponding residues in PSA and serve as a means

to distinguish hK2 from structurally similar PSA. The ex-

pression of the protein in LNCaP cells was up-regulated

by the synthetic androgen, mibolerone. This amplification

of the hK2 protein in LNCaP cells is consistent with pre-

viously reported increases in hK2 mRNA in LNCaP cells

following androgen treatment (Young et al, 1992).A high-affinity mab, HK1A523, was identified, which

exhibited a 1,000-fold increased sensitivity to hK2 than

to PSA by western blot analysis. The cross-reactivity seen

at high levels of PSA (20 p.g) may represent an hK2 con-

taminant in the PSA preparation, which copurified with

PSA. When recombinant ppPSA was examined under

identical conditions even at the highest levels of ppPSA

(20 rig), no cross-reactivity was detected (data not

shown). The high specificity of the antibody for hK2 and

2

i., -

358 Journal of Andrology . July/August 1996

98 kDa

64

50

36

30

16

6

1 2 3 4

FIG. 5. Reactivity of HK1 A523 with seminal fluid. Pooled seminal fluid(50 jig/lane) was electrophoresed under reducing conditions on an SDS-PAGE 4-20% gel. Following transfer to nitrocellulose, the blot wasprobed with HK1 A523 and anti-PCI antibody followed by appropriate sec-ondary antibodies labeled with HAP and ECL detection. Lane 1, molec-ular weight markers; lane 2, 50 jig seminal fluid, probed with HK1A523(10 jig/mI); lane 3, 50 jig seminal fluid, probed with anti-PCI (0.5 jig/mI);lane 4, 20 ng recombinant hK2, probed with HK1A523 (10 jig/mI).

minimal cross-reactivity with PSA made this antibody a

useful tool to probe prostate carcinoma tissue and seminal

fluid for the presence of hK2.

Multiple forms of hK2 were detected on western blots

of prostate carcinoma cytosols and seminal fluid with

HK1A523. In seminal fluid, a band at 85 kDa appeared

to be a complex of hK2 and protein C inhibitor (PCI)because antibodies to both proteins reacted with the same

molecular weight band on western blots. PCI is known to

exist in seminal fluid at high concentrations (220 mgfL)

(Laurell et al, 1992) and PSA has been shown to bind to

PCI and form a complex with a molecular weight of 90

kDa (Espana et al, 1991). Identification of hK2 in seminal

fluid by sequence analysis was recently reported (Deper-

thes et al, 1995). This group also described multiple

forms, including a 55-kDa complex of hK2 and a frag-

ment of PCI.

A 62-kDa immunoreactive form was detected in pros-

tate carcinoma cytosols, which may also represent a com-

plex of hK2 with an inhibitor. The exact nature of this

complex will require additional experimentation. Com-

plex formation between proteases and inhibitors often re-sults in loss of proteolytic activity and thereby serves as

a means of regulation. These high molecular weight forms

may serve to regulate the proteolytic activity of hK2.

Low molecular weight immunoreactive forms were

also detected in prostate carcinoma cytosols and seminal

fluid, which may represent breakdown products of the

mature protein. Immunoreactive bands were seen at ap-

proximately 20 kDa in prostate carcinoma cytosols and

at approximately 20 kDa and 14 kDa in seminal fluid.

The recombinant form of hK2 shown in Figure 2 also

contained two smaller molecular weight species at ap-

proximately 20 and 14 kDa. The exact nature of these

smaller forms will be determined by sequence analysis.

The physiological role of hK2 is currently unknown,

Due to its high sequence homology with PSA it may have

similar functions. It has been accepted that PSA facilitates

male fertility by aiding in the proteolytic liquefaction of

seminal coagulum (Lilja et al, 1987). Lilja (1985) and

Watt et al (1986) suggested that hK2 may in fact be the

protease responsible for the clipping and inactivation of

PSA and thereby serve as a regulator of PSA activity.

Final elucidation of the role of hK2 awaits additional in

vitro and in vivo studies.

The hK2-specific mab HK1A523 described in this re-

port has been used to identify hK2 in the spent media

from the prostate carcinoma cell line, LNCaP. N-terminal

sequence analysis of the immunoreactive band from

LNCaP cells confirmed its identity as mature hk2. Human

glandular kallikrein represents a potential new marker for

prostate cancer that may be useful in augmenting the clin-

ical information derived from PSA assays.

ReferencesBerg T, Bradshaw RA. Carretero OA, Chao J, Chao L, Clements JA,

Fahnestock M, Fritz H, Gauthier RJ, MacDonald RI, Margolius HS,

Morris BJ, Richards RI, Scicli AG. A common nomenclature for

members of the tissue (glandular) kallikrein gene families. Agents

Actions 1992;38(suppl l):19-25.

Chapdelaine P. Paradis G, Tremblay RR, Dube JY. High level of expres-

sion in the prostate of a human glandular kallikrein mRNA related to

prostate-specific antigen. FEBS Left 1988;236:205-208.

Clements J. The glandular kallikrein family of enzymes: tissue-specific

expression and hormonal regulation. Endocrine Rev 1989; 10:393-

419.

Deperthes D, Chapdelaine P. Tremblay R, Brunet C, Berton J, Hebert J,

Lazure C, Dube J. Isolation of prostatic kallikrein hK2, also known

as hGK-l, in human seminal plasma. Biochim Biophys Ada

l995;1245:31 1-3 16.

Espana F, Gilabert J, Estelles A, Romeu A, Aznar J, Cabo A. Functionally

active protein C inhibitor/plasminogen activator inhibitor-3

(PCI/PAI-3) is secreted in seminal vesicles, occurs at high concentra-

tions in human seminal plasma and complexes with prostate-specific

antigen. Thromb Res 199 l;64:309-320.

Henttu P. Lukkarinen 0, Vihko P. Expression of the gene coding for

human prostate specific antigen and related hGK-I in benign and

malignant tumors of the human prostate. tnt J Cancer 1990:45:654-

660.Kohler G, Milstein C. Continuous cultures of fused cells secreting anti-

body of predefined specificity. Nature 1975:256:495-497.

Laurell M, Christensson A, Abrahamsson PA, Stenflo J, Lilja H. Protein

C inhibitor in human body fluids. Seminal plasma is rich in inhibitor

Grauer et al . Human Glandular Kallikrein 359

antigen deriving from cells throughout the male reproductive system.

J Clin Invest l992;89:1094-l 101.

Lilja H. A kallikrein-like serine protease in prostatic fluid cleaves the

predominant seminal vesicle protein. J Clin Invest l985;76:l899-

1903.Lilja H, Oldbring J, Rannevik G. Laurell CB. Seminal vesicle-secreted

proteins and their reactions during gelation and liquefaction of human

semen. J Clin Invest 1987;80:281-285.

Lundwall A, Lilja H. Molecular cloning of human prostate specific an-

tigen cDNA. FEBS Left 1987;214:3l7-322.

Morris BJ. hGK-1: A kallikrein gene expressed in human prostate. Clin

Exp Pharm Phys 1989;16:345-351.

Murray SR, Chao J, Lin FK, Chao L. Kallikrein multigene families and

the regulation of their expression. J Cardiovasc Pharmacol 1990;15:

S7-S 16.

Paradis G, Tremblay RR, Dube JY. Looking for human glandular kalli-

krein-1 in the prostate. Prostate 1989;15:343-353.

Saedi M, Cass MJ, Goel AS, Grauer L. Okaneya T, Young C, Tindall D.

Overexpression of a human prostate-specific glandular kallikrein, hK2

and generation of antibody. Mo! Cell Endocrinol 1995;109:237-241.

Schedlich Li, Bennetts B, Morris B. Primary structure of a human glan-

dular kallikrein gene. DNA I987;6:429-437.

Towbin H, Staehelin 1, Gordon J. Electrophoretic transfer of proteins

from polyacrylamide gels to nitrocellulose sheets: procedure and

some applications. Proc Nat! Acad Sci USA 1979;76:4350-4354.

Wang MC, Valenzuela LA, Murphy GP, Chu TM. Purification of a human

prostate specific antigen. Invest Urol 1979;l7:159-163.

Watt KWK, Lee PJL. Timkulu TM, Chan WP, Loor R. Human prostate-

specific antigen: structural and functional similarity with serine pro-

teases. Proc Nat! Acad Sci USA l986;83:3 166-3170.

Young CY-F, Andrews PE. Montgomery BT, Tindall DJ. Tissue-specific

and hormonal regulation of human prostate-specific glandular kalli.

krein. Biochemistry 1992:31:818-824.