BIOLOGY OF PROSTATE-SPECIFIC ANTIGEN

HANS LILJA

ABSTRACT
The human kallikrein (hk) family, located on chromosome 19, encodes prostate-specific antigen (PSA [or
hK3]), hK2, hK4, and hK15 (prostin), as well as other serine proteases. Although PSA has been used in the
detection of prostate cancer for several years, much remains unknown about its function and forms. The
regulatory mechanisms of PSA are vital to its understanding. A particular mechanism by which PSA forms
complexes with either ␣1-antichymotrypsin or ␣2-macroglobulin may provide important information for
disease detection and progression. Data are emerging that show that active hK2, hK4, and hK15 may be
important to convert pro-PSA to the active PSA enzyme. This information, along with insights into the precise
mechanisms of PSA expression, may be used to suggest that PSA and, perhaps, other members of the hK
family contribute critical control mechanisms to tumor invasion or progression. Although much remains to be
revealed on the role of these gene products in the detection and progression of prostate cancer, findings
from studies that show sensitive signaling of the disease ⱖ20 years before the diagnosis of clinically
significant prostate cancer may alter screening procedures and improve treatment options. UROLOGY 62
(Suppl 5A): 27–33, 2003. © 2003 Elsevier Inc.

R ecent investigation has expanded the human

kallikrein (hK) gene family, located on chro-
mosome 19, to include 15 genes that encode serine
proteinases, several of which are expressed at high
levels in the prostate (Figure 1).1,2 Levels of pros-
tate-specific antigen (PSA) released into the glan-
dular lumen from the epithelial cells, where it is
mainly functionally active, can range as high as 10
to 50 ␮mol/L. In contrast, the extracellular level of
latent or inactive PSA forms is significantly lower,
⬍0.1 nmol/L, which corresponds to only 10⫺7 of
those in seminal fluid. Both PSA and hK2 have
been the subjects of recent and ongoing investiga-
tions, and both have been shown to be released at
increased levels in blood early in the development
of prostate cancer or parallel disease progression.
To date, there is no evidence to suggest that PSA is
a mitogen. A better understanding of the latent,
active, and inactive forms of PSA and related pro-
From the Department of Laboratory Medicine, Division of Clini-
cal Chemistry, Lund University, University Hospital (UMAS),
Malmö, Sweden; and Departments of Clinical Laboratories, Urol-
ogy, and Medicine, Memorial Sloan-Kettering Cancer Center,
New York, New York, USA
Hans Lilja holds patents on free prostate-specific antigen test-
ing and human kallikrein–2 analysis
Reprint requests: Hans Lilja, MD, PhD, Department of Labo-
ratory Medicine, Division of Clinical Chemistry, Lund Univer-
sity, University Hospital (UMAS), S-205 02, Malmö, Sweden.
E-mail: Hans.lilja@klkemi.mas.lu.se
teases and their significance in the detection and
progression of prostate cancer could provide im-
portant insights into the prognosis and possible
treatment targets for patients with prostate cancer.
REGULATORY MECHANISMS OF
PROSTATE-SPECIFIC ANTIGEN
A key element in understanding PSA is regula-
tion of the protease activity. There are 4 different
regulatory mechanisms for serine proteinases: zy-
mogen activation, allosteric regulation, proteolytic
internal cleavage, and proteinase inhibition. Zy-
mogen activation, which is common to the serine
proteinases, is characterized by a conformational
change. The irreversible conversion from inactive
to active PSA occurs when the protein reaches sub-
cellular compartments or an extracellular environ-
ment.3 The process may be autocatalytic or may
require enzymatic or nonenzymatic molecules. On
release of the activation peptide, which for PSA and
glandular hKs is relatively short, enzymatic action
is gained by a conformational change induced in
the mature PSA.
There are further elements involved in the regu-
lation of active PSA and hK2, including divalent
ions, particularly zinc.4,5 Inhibition models show
that each hK2/PSA molecule is inhibited by bind-
ing to 2 zinc-binding sites with inhibitory con-
stants at micromolar levels, although the precise

© 2003 ELSEVIER INC. 0090-4295/03/$30.00

ALL RIGHTS RESERVED doi:10.1016/S0090-4295(03)00775-1 27
FIGURE 1. Human kallikrein (hK) gene locus chromosome 19q 13.3 to 13.4 ⬇300 kilobases (kb). The kallikrein
family, located on chromosome 19, includes ⱖ15 genes. Of these, prostate-specific antigen (PSA), hK2, hK4
(prostase), and hK15 (prostin) are expressed with abundance in prostatic tissue. Although their precise roles in the
detection and progression of prostate cancer remain incompletely defined, they have enhanced diagnostic effects
and may well provide information leading to more effective therapies. (Adapted from Trends Endocrinol Metab.1)

contribution of this mechanism in vivo remains
unknown.
In the third regulation pathway, PSA loses activ-
ity because of conversion of the mature, active,
single-chain PSA to a multichain or nicked enzyme
through internal cleavage. Sites where the cleavage
occurs include proteolysis occurring in the car-
boxy-terminal of amino acid residues lysine
(Lys)145 and/or Lys182. This suggests some type
of plasmin-like enzymatic activity to inactivate the
functional PSA enzyme, although, thus far, the re-
sponsible enzymes have not been identified.
The fourth regulatory mechanism, which was
first reported on in the early 1990s, is the interac-
tion between protease inhibitors, most of which
are produced in the liver and are present in great
excess (eg, 1 to 10 ␮mol/L, which corresponds to
about 10,000- to 100,000-fold molar excess) com-
pared with any PSA molecule released extracellu-
larly. The interactions between 2 different classes
of inhibitors and PSA are of particular impor-
tance.6 – 8 The first class is the serpin (serine pro-
tease inhibitor) type, of which ␣1-antitrypsin (also
called ␣1-protease inhibitor) is the archetype. This
class also includes ␣1-antichymotrypsin (ACT),
protein C inhibitor (or plasminogen activator in-
hibitor [PAI]–3), PAI-1, and antithrombin. Inhib-
itory activity of serpin-type antiproteases is repre-
sented by recent modifications of a mousetrap
model.9 In vitro kinetics are slow for inhibition of
active PSA by ACT by formation of a stable acyl-
enzyme, where the PSA molecule is predicted to
move 180° over to 1 side, while it permits insertion
of an antiparallel ␤-pleated sheet structure released
by the proteolysis manifested by PSA.4,8,10
The second class of protease inhibitors includes
␣2-macroglobulin (AMG) and the analogous preg-
nancy-zone protein that has demonstrated a high
capacity for interaction with PSA.8,11 In vitro kinet-
ics of PSA reacting with AMG is about 20-fold
more rapid than that of the formation of PSA–ACT
complexes. Further, compared with the serpin-
type ACT, AMG interacts with PSA by a different
mode of action, whereby PSA is encapsulated. This
makes it difficult to access the antigenic epitopes of
PSA to reliably measure levels of the PSA–AMG
complex, although 2 different approaches have
been used to develop experimental in-house re-
search assays: (1) distorting the conformational
shape of this complex, and (2) releasing PSA from
its complex ligand by hydrolysis at alkaline pH.
However, currently available experimental data
suggest that levels of PSA–AMG may remain low in
vivo, close to or below the limits of detection in
blood, at least at clinically localized stages of pros-
tate cancer. This is likely to be largely because of
rapid elimination by efficient clearance mecha-
nisms, despite the fact that this may be an impor-
tant pathway for elimination of active PSA from
extracellular fluids.

28 UROLOGY 62 (Supplement 5A), November 2003

 EFFECTS OF VARIOUS FACTORS ON
PROSTATE-SPECIFIC ANTIGEN
EXPRESSION
PSA expression is mainly dependent on ligand-
dependent (or ligand-independent) activation and
signaling by the androgen receptor.12 Elevation of
extracellular PSA levels is generally because of
damage in the relation between PSA expression
and the prostatic glandular ductal system, which is
aimed to protect extracellular environments from
exposure to the action of PSA and related proteases
(eg, hK2) and to maintain extracellular PSA levels
at ⬍0.1 nmol/L, which roughly corresponds to
about ⬍10⫺7 lower than the intraductal levels
(⬍0.1 mmol/L). Further, extracellular levels of
each form of PSA (and hK2) are also dependent on
metabolic pathways and clearance mechanisms:
free noncomplex PSA (referred to as free PSA
[fPSA]) forms are cleared from blood primarily by
glomerular filtration in the kidneys with an ap-
proximate half-life of 12 hours,13,14 and severe im-
pairment of renal function significantly elevates
percent fPSA, whereas PSA complexes are too large
to be subject to renal clearance.15 This may become
clinically relevant (eg, in the evaluation of individ-
uals considered for renal transplantation) because
men on dialysis have significantly elevated percent
fPSA.15
The clearance pathway for PSA–ACT remains
unclear, but the clearance rate for PSA–ACT levels
in men with clinically localized prostate cancer is
suggested to be very slow, corresponding to a de-
crease of PSA–ACT levels by ⬍1 ng/mL per day.13
This may be reflective of the capacity limitation
and inefficiency of the elimination pathway(s).
THE VALUE OF PROSTATE-SPECIFIC
ANTIGEN FORMS IN BLOOD AS
INDICATORS OF PROSTATE CANCER
DETECTION AND PROGRESSION
In the healthy prostate, there is a normal arrange-
ment of cell types, including epithelial cells, basal
cells, and basal lamina. Normal prostatic function
and production of PSA and other hK genes are
largely reliant on androgen-dependent activation
of the androgen receptor. Extracellular release of
PSA in advanced stages of prostate cancer can in-
crease the levels up to 10,000-fold of those found
in adult, prostate disease–free, healthy men. Fur-
thermore, the proportion of fPSA to complexed
PSA forms varies according to disease type: typi-
cally, significantly higher percent fPSA are found
in serum/plasma from men with benign gland en-
largements, whereas much more pronounced ele-
vations of PSA–ACT levels in serum/plasma from
men with prostate adenocarcinomas contribute to
UROLOGY 62 (Supplement 5A), November 2003
typical findings of significantly lower percent fPSA
in these samples.
It has now been ⬎10 years since the original
findings on the design and usefulness of measuring
fPSA in blood were reported,8,11 and although
novel at that time, this testing modality has under-
gone a large number of single- and multi-institu-
tional evaluations. In general, these studies have
confirmed the ability of percent fPSA to increase
the diagnostic specificity in the “diagnostic gray
zone,” in particular, at total PSA levels ranging
from 4 to 10 ng/mL.16 More recently reported data
suggest that there is significant heterogeneity of
fPSA comprising 3 major components: latent pro-
PSA, mature single-chain PSA (which, by un-
known mechanisms, remains unreactive with a
large excess of active protease inhibitors in the ex-
tracellular compartments), and multichain, inacti-
vated forms of PSA.17–21 Furthermore, latent pro-
PSA may be present in many different variants
because of more or less explicit truncation of the
activation peptide of pro-PSA.
Reagents have been developed by several differ-
ent research groups to selectively measure the dif-
ferent fPSA subfractions and to evaluate whether
they may further enhance the ability to discrimi-
nate men with only benign causes of moderate total
PSA elevations from those harboring clinically sig-
nificant prostate cancer. Differences in composi-
tion of fPSA subfractions were first reported by
characterization of PSA isolated from benign tissue
and from androgen-dependent cancer cells cul-
tured in vitro (LNCaP). About 50% of PSA from
LNCaP cells was secreted as latent pro-PSA; ⬎40%
was secreted as mature, single-chain enzyme; and
⬍10% was secreted as the N-terminal truncated,
single-chain form. In particular, PSA from LNCaP
cells contained no internally cleaved multichain
forms.22 By contrast, PSA from benign tissue con-
tains no latent pro-PSA, whereas internally cleaved
multichain forms contribute ⬎30% of the isolated
protein.
Using assays designed to discriminate single-
chain from multichain forms of fPSA (ie, nicked
PSA), Steuber et al.23 recently found (1) that there
are significantly larger proportions of nicked PSA
in serum from men with benign biopsies compared
with men who had prostate cancer, and (2) that
compared with other means of PSA testing, this
approach could be used to significantly enhance
discrimination of the 2 groups of men with mod-
erately elevated PSA levels. Future studies will
strive to position the role of nicked and intact PSA
in the diagnosis of prostate cancer.
There is abundant expression of both nicked and
intact PSA by the prostate epithelium. This is con-
sistent with what is found in each LNCaP cell,
where there might be about 103 copies per cell of
29
FIGURE 2. Hypothetical proposal on protease interactions resulting in a cascade of protease activation(s). The
activation of various members of the human kallikrein (hK) gene family could influence the function and activity of
other members. PAI ⫽ plasminogen activator inhibitor; PSA ⫽ prostate-specific antigen; uPA ⫽ urokinase plasmin-
ogen activator; UPAR ⫽ uPA receptor.
PSA messenger RNA (mRNA). Levels of hK2
mRNA in LNCaP cells are 25% to 30% of those of
PSA mRNA. This is in contrast to the total levels of
hK2 in seminal fluid and blood, which correspond
to only about 1% of those of PSA, a fact on which
there is unanimous agreement. This suggests a pos-
sible discrepancy in levels of mRNA to protein for
PSA versus that for hK2, which has not yet been
fully resolved. However, there is extensive struc-
tural similarity of the 2 proteins (about 80% iden-
tity), which makes it important to ensure careful
control of immunologic crossreactivity. Therefore,
it has been difficult, time consuming, and resource
demanding to develop specific hK2 assays with suf-
ficiently low functional detection limits (⬍10 pg/
mL) and insignificant immunologic crossreactivity
to PSA (⬍0.01%).24
In clinical studies, hK2 has also contributed sig-
nificant power to discriminate organ-confined
prostate cancer from extracapsular disease stages
in men with total PSA values ⬍10 ng/mL.25 Com-
pared with testing for total PSA, hK2 can, there-
fore, be used to significantly improve prediction of
extraprostatic growth of the disease. In vitro stud-
ies performed with physiologic ratios of hK2 to
30
PSA have demonstrated that active hK2 has the
capacity to convert latent pro-PSA to active PSA.3
When functional enzyme forms of 2 other genes,
hK4 and hK15—which, similar to PSA and hK2,
are also expressed at highly abundant levels in
prostate tissue—were examined, they were found
to also have capacity to convert latent pro-PSA to
active PSA in vitro.2,26 Moreover, the conversion of
latent pro-hK2 to active hK2 may not require an-
other enzyme, as it is strongly implicated to be the
result of an autocatalytic loop.27 When all these
factors are taken into account, it may be justifiable
to propose that these gene products interact in
stepwise activation pathway(s) that could affect
the functional stages, as well as net effects, of each
other (Figure 2).
According to previously reported data, the ac-
tions of these abundantly prostate-expressed hK
gene products may also contribute critical regula-
tory control of angiogenesis, growth factor release,
and extracellular matrix modification. For exam-
ple, active PSA and hK2 have the capacity (1) to act
on extracellular matrix proteins, such as fibronec-
tin and laminin; (2) to modify function of insulin-
like growth factor– binding protein 3, resulting in
UROLOGY 62 (Supplement 5A), November 2003
the release of insulinlike growth factor–1; (3) to
inactivate parathyroid hormone–related protein;
(4) to inhibit angiogenesis; and (5) to interact with
antithrombin, which is locally produced in the
prostate and has strong antiangiogenic activity.28
Therefore, disease-linked extracellular release of
latent and active forms of these proteases may con-
tribute important characteristics to phenotypic
growth patterns of prostate cancer, although so far,
there is no conclusive evidence reported that may
be used to fully substantiate any proposals on these
matters.
A distinctly different issue relates to several cur-
rently unresolved matters on the putative func-
tional significance of the proteolytic action of PSA
and/or hK2 after their release into various extracel-
lular compartments. Research has also shown that
LNCaP cells secrete adequate levels of active PSA
to convert an inactive prodrug component into an
active apoptosis inducer.29,30 To optimize this ef-
fect, it was proved critically important to identify
specific peptide sequences used to conjugate the
cytotoxic drug, which are selectively split by PSA
to release the drug at high concentrations in the
area of tumor cells.29,30 In addition, this is further
illustrated by reports on the limitations of using
another prodrug system design, which was found
to become prematurely activated in the blood-
stream.31
The biologic function of hK2 could be important
in controlling the functional stages of its own la-
tent form, as well as to control those of latent pro-
PSA. In addition, hK2 has also been shown to in-
teract with and convert latent prourokinase-type
plasminogen activator to active urokinase-type
plasminogen activator (uPA), to interact with sev-
eral PAIs (eg, PAI-1 and PAI-3) or protein C inhib-
itor, and to interact with extracellular matrix pro-
teins (eg, fibronectin). So far, there are no data
available on whether hK2 and/or PSA may also in-
teract with the urokinase receptor. This receptor is
the principal site of activation of uPA and plasmin-
ogen that may link to subsequent activations (eg,
of metalloproteinases). Lastly, a quite large num-
ber of novel genes located in a 678-kilobase locus,
which surrounds the genes coding for gel pro-
teins—major substrate proteins for PSA (and pos-
sibly also for hK2) in the ejaculate— have recently
been identified on the long arm of chromosome 20
(20q13.1). Most of these genes were found to be
expressed in reproductive tract tissues, and they
encode relatively small-sized gene products pre-
dicted to contain whey acidic protein or Kunitz-
type protease inhibitory domains.32,33 A few of
these gene products have been extensively investi-
gated, characterized, and found to manifest pro-
tease inhibitory function (eg, the secretory leuko-
cyte protease inhibitor that originally was
UROLOGY 62 (Supplement 5A), November 2003
identified and characterized in seminal fluid, but
which has no significant protease inhibitory capac-
ity against PSA).34
During the first experience in 1994 to 1995 with
detection of circulating prostate cells, which may
or may not be tumor cells, by use of a technique to
measure PSA mRNA in blood samples, my group
initially tried to perform nested reverse transcrip-
tion-polymerase chain reaction (RT-PCR) and col-
lected samples from the routine hematology lab,
where, typically, 50% of the samples from female
subjects gave a positive signal. In a collaborative
project between our laboratory and researchers su-
pervised by Dr. Timo Lövgren35 at the Department
of Biotechnology, University of Turku, Finland,
exogenously added internal standards were intro-
duced at the time when the sample was first
worked up to ensure the ability to keep track of
efficacy and recovery of mRNA copy number dur-
ing RT-PCR amplification of selected mRNA tar-
gets. Using this enhanced technology, detection of
circulating prostate cells by means of carefully con-
trolled, quantitative measurements of mRNA copy
numbers for hK2 and PSA in blood have recently
been reported to most efficiently discriminate be-
tween samples collected from men with cancer and
from those without cancer.35
In this recently completed study,35 which built
on novel techniques reported in several previous
studies designed to optimize reliability of these
mRNA measurements,36 –38 the investigators mea-
sured circulating cells in 70 subjects, 20 of whom
had organ-confined or clinically localized prostate
cancer, 19 of whom had benign biopsy, and the
remainder of whom had advanced stages of pros-
tate cancer.35 Blood from 18 of the 19 men with
benign biopsy and no prostate cancer did not con-
tain PSA or hK2 mRNA copy numbers above the
threshold of detection (ⱖ100 mRNA copies). By
contrast, samples from most men with localized
prostate cancer contained detectable levels of PSA
mRNA and/or hK2 mRNA in blood. In total, PSA
mRNA/hK2 mRNA levels were above the limits of
detection in the samples from 70% of those with
organ-confined disease and/or clinically localized
disease.35 It is noteworthy that the 3 patients with
organ-confined disease from whom the blood did
not contain any detectable levels of PSA mRNA/
hK2 mRNA were all found to have low pathologic
Gleason scores in the resected prostatectomy spec-
imens. Further, PSA mRNA/hK2 mRNA levels
were above the limits of detection and indicated
occurrence of circulating prostate cells in the blood
samples from all the men with advanced stages of
prostate cancer. Moreover, subjects with advanced
hormone-refractory disease were included among
the advanced cancer cases in this study. Recent
preliminary reports of data would strongly suggest
31
that during very slow, continuous progression of
initially small tumor lesions, the lesions are likely
to become surrounded by significantly elevated
levels of latent/active protease forms (eg, of PSA
and hK2) during very early stages of progression of
the malignant disease. Therefore, it is possible that
at some stage these lesions establish a link to the
systemic circulation from which prostate cells may
frequently shed into the blood (eg, at disease stages
when tumor growth is confined to the prostate
gland). If so, circulating hK2 and PSA levels could
become very important signs of clinically signifi-
cant cancer that may require curative treatment.
CONCLUSION
Measurement of several gene products from the
hK gene family, located on chromosome 19q13.4,
can show very early signs of prostate cancer pro-
gression. Further, some of these gene products
may contribute significantly to the control of crit-
ical mechanisms involved in tumor invasion or
progression. Elevated levels of latent, active, and
inactive hK gene-product forms in extracellular
fluids (eg, blood) are very early signs of prostate
cancer, and elevated levels parallel disease progres-
sion, which may be used to suggest that they are
actively involved in control of tumor progression.
However, a widely recognized and critical limita-
tion of the diagnostic efficacy of these measure-
ments is the common finding of PSA elevations in
blood because of various benign prostatic disease
conditions. In particular, this becomes problem-
atic in populations of elderly men in whom PSA
elevations in blood because of benign prostate dis-
ease is much more common than that caused by
malignant disease. Therefore, various multivariate
approach models might be needed to enhance the
positive predictive value of several forms of PSA
measurements. In addition, baseline testing of 45-
to 50-year-old men (an age range when the fre-
quency of the common confounding benign pros-
tate disease conditions is still low) to differentiate
very low or no current risk in men may prove use-
ful in combination with repeated follow-up testing
every 2 to 4 years for those at higher risk. However,
this should first be analyzed for both cost efficacy
and the ability to provide critically important in-
formation to selectively identify subjects at highest
risk for developing clinically significant prostate
cancer long before it reaches advanced disease
stages. Validated technology, however, remains ur-
gently needed to discriminate insignificant cancer
lesions that may only require chemopreventive
regimens from clinically significant malignant dis-
ease that would obtain the greatest benefit from
curative therapeutic interventions. In particular,
this technology could provide a means of early
32
identification of malignant disease conditions as-
sociated with significant risk for progression, inva-
sion, and the capability to form metastases. In this
regard, circulating cell technology may contribute
significant information, and the identification of
additional methods for identifying patients with
locally advanced and/or early metastatic stage of
this disease would also be valuable.
REFERENCES
1. Diamandis EP, Yousef GM, Luo L-Y, et al: The new
human kallikrein gene family: implications in carcinogenesis.
Trends Endocrinol Metab 11: 54 –60, 2000.
2. Takayama TK, Carter CA, and Deng T: Activation of
prostate-specific antigen precursor (pro-PSA) by prostin, a
novel human prostatic serine protease identified by degener-
ate PCR. Biochemistry 40: 1679 –1687, 2001.
3. Lövgren J, Rajakoski K, Karp M, et al: Activation of the
zymogen form of prostate-specific antigen by human glandu-
lar kallikrein 2. Biochem Biophys Res Commun 238: 549 –
555, 1997.
4. Malm J, Hellman J, Hogg P, et al: Enzymatic action of
prostate-specific antigen (PSA or hK3): substrate specificity
and regulation by Zn2⫹, a tight-binding inhibitor. Prostate 45:
123–139, 2000.
5. Lövgren J, Airas K, and Lilja H: Enzymatic action of
human glandular kallikrein 2 (hK2): substrate specificity and
regulation by Zn2⫹ and extracellular protease inhibitors. Eur
J Biochem 262: 781–789, 1999.
6. Lilja H: Role of the hK2, free PSA, and complexed PSA
measurements in the very early detection of prostate cancer.
Eur Urol 39(suppl 4): 47–48, 2001.
7. Lilja H: Prostate-specific antigen: molecular forms and
the human kallikrein gene family. Br J Urol 79(suppl 1): 44 –
48, 1997.
8. Christensson A, Laurell C-B, and Lilja H: Enzymatic
activity of prostate-specific antigen and its reactions with ex-
tracellular serine proteinase inhibitors. Eur J Biochem 194:
755–763, 1990.
9. Huntington JA, Read RJ, and Carrell RW: Structure of a
serpin-protease complex shows inhibition by deformation
[letter]. Nature 407: 923–926, 2000.
10. Piironen T, Villoutreix BO, Becker C, et al: Determina-
tion and analysis of antigenic epitopes of prostate specific an-
tigen (PSA) and human glandular kallikrein 2 (hK2) using
synthetic peptides and computer modeling. Protein Sci 7:
259 –269, 1998.
11. Lilja H, Christensson A, Dahlén U, et al: Prostate-spe-
cific antigen in serum occurs predominantly in complex with
␣1-antichymotrypsin. Clin Chem 37: 1618 –1625, 1991.
12. Young CY-F, Andrews PE, Montgomery BT, et al: Tis-
sue-specific and hormonal regulation of human prostate-spe-
cific glandular kallikrein. Biochemistry 31: 818 –824, 1992.
13. Björk T, Ljungberg B, Piironen T, et al: Rapid exponen-
tial elimination of free prostate-specific antigen (PSA) con-
trasts the slow, capacity-limited elimination of PSA com-
plexed to alpha1-antichymotrypsin from serum. Urology 51:
57–62, 1998.
14. Lilja H, Haese A, Björk T, et al: Significance and metab-
olism of complexed and noncomplexed prostate specific anti-
gen forms and human glandular kallikrein 2 in clinically lo-
calized prostate cancer before and after radical prostatectomy.
J Urol 162: 2029 –2035, 1999.
15. Bruun L, Björk T, Lilja H, et al: Percent free prostate
specific antigen is elevated in men on haemodialysis or peri-
toneal dialysis treatment. Nephrol Dial Transplant 18: 598 –
603, 2003.
UROLOGY 62 (Supplement 5A), November 2003
 16. Catalona WJ, Beiser JA, and Smith DS: Serum free pros-
tate specific antigen and prosate specific antigen density mea-
surements for predicting cancer in men with prior negative
prostatic biopsies. J Urol 158: 2180 –2181, 1997.
17. Mikolajczyk SD, Grauer LS, Millar LS, et al: A precursor
form of PSA (pPSA) is a component of the free PSA in prostate
cancer serum. Urology 50: 710 –714, 1997.
18. Noldus J, Chen Z, and Stamey TA: Isolation and char-
acterization of free form prostate specific antigen (f-PSA) in
sera of men with prostate cancer. J Urol 158: 1606 –1609,
1997.
19. Charrier J-P, Tournel C, Michel S, et al: Two-dimen-
sional electrophoresis of prostate-specific antigen in sera of
men with prostate cancer or benign prostate hyperplasia. Elec-
trophoresis 20: 1075–1081, 1999.
20. Nurmikko P, Pettersson K, Piironen T, et al: Discrimi-
nation of prostate cancer from benign disease by plasma mea-
surement of intact, free prostate-specific antigen lacking an
internal cleavage site at Lys145-Lys146. Clin Chem 47: 1415–
1423, 2001.
21. Niemela P, Lövgren J, Karp M, et al: Sensitive and spe-
cific enzymatic assay for the determination of precursor forms
of prostate-specific antigen after an activation step. Clin Chem
48: 1257–1264, 2002.
22. Väisänen V, Lövgren J, Hellman J, et al: Characteriza-
tion and processing of prostate-specific antigen (hK3) and
human glandular kallikrein 2 (hK2) secreted by LNCaP cells.
Prostate Cancer Prostatic Dis 2: 91–97, 1999.
23. Steuber T, Nurmikko P, Haese A, et al: Discrimination
of benign from malignant prostatic disease by selective mea-
surements of single chain, intact free prostate specific antigen.
J Urol 168: 1917–1922, 2002.
24. Haese A, Väisänen V, Finlay JA, et al: Standardization of
two immunoassays for human glandular kallikrein 2. Clin
Chem 49: 602–610, 2003.
25. Haese A, Graefen M, Steuber T, et al: Human glandular
kallikrein 2 levels in serum for discrimination of pathologi-
cally organ-confined from locally- advanced prostate cancer in
total PSA-levels below 10 ng/ml. Prostate 49: 101–109, 2001.
26. Takayama TK, McMullen BA, Nelson PS, et al: Charac-
terization of hK4 (prostase), a prostate-specific serine pro-
tease: activation of the precursor of prostate specific antigen
(pro-PSA) and single-chain urokinase-type plasminogen acti-
vator and degradation of prostatic acid phosphatase. Biochem-
istry 40: 15341–15348, 2001.
27. Denmeade SR, Lövgren J, Khan SR, et al: Activation of
UROLOGY 62 (Supplement 5A), November 2003
latent protease function of pro-hK2, but not pro-PSA, involves
autoprocessing. Prostate 48: 122–126, 2001.
28. Cao Y, Lundwall A, Gadaleanu V, et al: Anti-thrombin
is expressed in the benign prostatic epithelium and in prostate
cancer and is capable of forming complexes with prostate-
specific antigen and human glandular kallikrein 2. Am J
Pathol 161: 2053–2063, 2002.
29. Denmeade SR, Lou W, Lövgren J, et al: Specific and
efficient peptide substrates for assaying the proteolytic activity
of prostate-specific antigen. Cancer Res 57: 4924 –4930, 1997.
30. Denmeade SR, Nagy A, Gao J, et al: Enzymatic activa-
tion of a doxorubicin-peptide prodrug by prostate-specific an-
tigen. Cancer Res 58: 2537–2540, 1998.
31. DiPaola RS, Rinehart J, Nemunaitis J, et al: Character-
ization of a novel prostate-specific antigen-activated peptide-
doxorubicin conjugate in patients with prostate cancer. J Clin
Oncol 20: 1874 –1879, 2002.
32. Lundwall Å, and Clauss A: Identification of a novel
protease inhibitor gene that is highly expressed in the prostate.
Biochem Biophys Res Commun 290: 452–456, 2002.
33. Clauss A, Lilja H, and Lundwall A: A locus on human
chromosome 20 contains several genes expressing protease
inhibitor domains with homology to whey acidic protein. Bio-
chem J 368: 233–242, 2002.
34. Ohlsson K, Bjartell A, and Lilja H: Secretory leukocyte
protease inhibitor in the male genital tract: PSA-induced pro-
teolytic processing in human semen and tissue localization. J
Androl 16: 64 –74, 1995.
35. Ylikoski A, Pettersson K, Nurmi J, et al: Simultaneous
quantification of prostate-specific antigen and human kal-
likrein 2 mRNA in blood samples from patients with prostate
cancer and benign disease. Clin Chem 48: 1265–1271, 2002.
36. Ylikoski A, Sjöroos M, Lundwall Å, et al: Quantitative
reverse transcription-PCR assay with an internal standard for
the detection of prostate-specific antigen mRNA. Clin Chem
45: 1397–1407, 1999.
37. Ylikoski A, Karp M, Lilja H, et al: Dual-label detection
of amplified products in quantitative RT-PCR assay using lan-
thanide-labeled probes. Biotechniques 30: 832–836, 838, 840
passim, 2001.
38. Ylikoski A, Karp M, Pettersson K, et al: Simultaneous
quantification of human glandular kallikrein 2 and prostate-
specific antigen mRNAs in peripheral blood from prostate can-
cer patients. J Mol Diagn 3: 111–122, 2001.
33