Inhibitory properties of cystatin F and its localization
in U937 promonocyte cells
Tomaz
ˇ
Langerholc
1
, Valentina Zavas
ˇ
nik-Bergant
1
, Boris Turk
1
, Vito Turk
1
, Magnus Abrahamson
2
and Janko Kos
3
1 Department of Biochemistry and Molecular Biology, Joz
ˇ
ef Stefan Institute, Ljubljana, Slovenia
2 Department of Clinical Chemistry, Institute of Laboratory Medicine, University of Lund, Sweden
3 Faculty of Pharmacy, Department of Pharmaceutical Biology, University of Ljubljana, Slovenia
Human papain-like cathepsins were long believed to be
responsible for terminal protein degradation in the
lysosomes. This view changed dramatically when they
were found to be involved in a number of important
cellular processes, such as antigen presentation [1],
bone resorption [2], apoptosis [3] and protein process-
ing [4], as well as several pathologies such as cancer
progression [5], inflammation [6] and neurodegenera-

Tel: +386 14773573
E-mail:
(Received 9 November 2004, revised 31
January 2005, accepted 2 February 2005)
doi:10.1111/j.1742-4658.2005.04594.x
Cystatin F is a recently discovered type II cystatin expressed almost exclu-
sively in immune cells. It is present intracellularly in lysosome-like vesicles,
which suggests a potential role in regulating papain-like cathepsins involved
in antigen presentation. Therefore, interactions of cystatin F with several
of its potential targets, cathepsins F, K, V, S, H, X and C, were studied
in vitro. Cystatin F tightly inhibited cathepsins F, K and V with K
i
values
ranging from 0.17 nm to 0.35 nm, whereas cathepsins S and H were inhib-
ited with 100-fold lower affinities (K
i
% 30 nm). The exopeptidases, cathep-
sins C and X were not inhibited by cystatin F. In order to investigate the
biological significance of the inhibition data, the intracellular localization
of cystatin F and its potential targets, cathepsins B, H, L, S, C and K,
were studied by confocal microscopy in U937 promonocyte cells. Although
vesicular staining was observed for all the enzymes, only cathepsins H and
X were found to be colocalized with the inhibitor. This suggests that cysta-
tin F in U937 cells may function as a regulatory inhibitor of proteolytic
activity of cathepsin H or, more likely, as a protection against cathepsins
misdirected to specific cystatin F containing endosomal ⁄ lysosomal vesicles.
The finding that cystatin F was not colocalized with cystatin C suggests
distinct functions for these two cysteine protease inhibitors in U937 cells.
Abbreviations
mAb, monoclonal antibody; pAb, polyclonal antibody; M6P, mannose-6-phosphate.

[17]. Immunocytochemical staining of cystatin F in
human promonocyte U937 cells displays a vesicular
pattern [18]. In subcellular fractionation experiments
cystatin F coeluted with the peak of b-hexosaminidase
activity, an enzyme typically located in lysosome-like
organelles. Independently, Journet et al. [19] detected
cystatin F as a soluble protein after affinity puri-
fication of mannose-6-phosphate (M6P) containing
proteins. This means that M6P was present in the
N-linked carbohydrate moiety in cystatin F or, alter-
natively, that cystatin F was in complex with another
M6P containing protein. Nevertheless, despite secre-
tion of cystatin F from U937 cells, a high proportion
seems to reside intracellularly in lysosomes or lyso-
some-like organelles [18].
The aim of our study was to identify potential tar-
gets of cystatin F among endogenous lysosomal cys-
teine proteases. First we found that dimers of cystatin
F are inactive as inhibitors of cysteine proteases and
that the monomeric form has to be restored for the
inhibitory potential. After activation of cystatin F we
have studied the in vitro kinetics of the interaction
between cystatin F and several cathepsins, as well
as their intracellular localization in promonocyte
U937 cells, using specific antibodies and confocal
microscopy.
Results
Activation of cystatin F
Recombinant cystatin F showed one band on
SDS ⁄ PAGE (Fig. 1A) at 17 kDa under reducing con-

F were observed at dithiotreitol concentrations above
30 mm (Fig. 1B). In addition, the effect of increasing the
concentration of reductant on cystatin F was followed
by electrophoresis under native conditions, where the
transition between 20 and 30 mm dithiotreitol was
accompanied by a shift to a smaller molecular mass
(Fig. 1C). These results suggest that dimerization of
cystatin F is linked to disulfide bond formation, which is
responsible for the loss of inhibitory activity of the pro-
tein. We also noticed that monomerization of cystatin F
was enhanced in acidic environment, especially below
pH 5 (T. Langerholc, unpublished data).
Cystatin F was remarkably stable under reducing
conditions, as no loss of inhibitory activity was
observed even after prolonged incubation at dithiotrei-
tol concentrations as high as 100 mm, and the K
i
value
for the inhibition of papain (K
i
¼ 1.4 nm) was similar
to that reported for inhibition at low dithiotreitol con-
centration (K
i
¼ 1.1 nm) [13].
Based on these results, 100 mm dithiotreitol was
used in the cystatin F activation buffer in order to
ensure total conversion of cystatin F to the active
monomeric state prior to kinetic studies. It should be
noted that, after dilution of cystatin F solution to the

i
values, which were
corrected for substrate competition, are listed in
Table 1. Cystatin F was observed to be a tight binding
inhibitor of cathepsins F, K, L, V, with K
i
values ran-
ging from 0.17 to 0.35 nm. Surprisingly cathepsin S,
despite being an endopeptidase, was inhibited by cysta-
tin F substantially more weakly, with K
i
¼ 33 nm,
comparable to the inhibition of the aminopeptidase
cathepsin H (K
i
¼ 30 nm ). In comparison with other
cystatins, cystatin F is a rather slow binding inhibitor
of the cathepsins, characterized by k
ass
values in the
range of 10
6
)10
7
m
)1
, and high k
diss
values in the
range of 10

Cathepsin K 0.35 ± 0.15 11 ± 3 3.2 ± 0.6 Z-FR-AMC
Cathepsin V 0.30 ± 0.15 4.8 ± 1.4 1.6 ± 0.3 Z-FR-AMC
Cathepsin S 33 ± 13 3.7 ± 0.7 0.011 ± 0.002 Z-FR-AMC
Cathepsin H 36 ± 15 0.57 ± 0.2 0.0016 ± 0.00024 H-R-AMC
Cathepsin C > 100 H-SY-bNA
Cathepsin X > 100 Dnp-GFFW
Papain 1.4 ± 0.4 3.5 ± 0.6 0.25 ± 0.03 Z-FR-AMC
Cathepsin L
a
0.31 Z-FR-AMC
Legumain
b
10 Z-AAN-AMC
Cathepsin B
a
>1000 Z-FR-AMC
Papain
a
1.1 Z-FR-AMC
a
[13].
b
[15].
T. Langerholc et al. Cystatin F in U937 cells
FEBS Journal 272 (2005) 1535–1545 ª 2005 FEBS 1537
lower inhibition constants for cathepsins S and H was
due mainly to the low k
ass
values.
Colocalization of cystatin F and potential target

Discussion
Cystatin F has been known for some years, but its
activity and functional properties have not been com-
pletely determined. However, initial studies revealed an
inhibitory profile that was not typical of other type II
cystatins [13]. Cystatin F, isolated from a baculovirus
expression system, can form disulfide-bonded dimers,
as shown for the inhibitor expressed in Escherichia coli
[12]. This type of dimerization mechanism is different
from general domain-swapping in the cystatin family
[20]. Although both additional cysteines in cystatin F
at positions 1 and 37 (cystatin C numbering) can form
a disulfide bond [13], the cysteine at position 1 has
been suggested to be involved in dimerization of cysta-
tin F [12], similar to cysteine 3 in stefin B [21]. Higher
dithiotreitol concentrations than previously reported
[12] were needed to restore monomers and inhibi-
tory activity under nondenaturing conditions. Loss of
inhibitory potential of dimerized cystatin F can be
explained by blocking of the N-terminal part, disabling
protease access, or by a conformational change result-
ing from disruption of an intramolecular disulfide
bond between cysteines 1 and 37. As dimers have been
observed in U937 cells under physiological conditions
[22], dimerization of cystatin F could be a process
regulating its inhibitory properties.
Screening of proteases for their inhibition showed
that cystatin F is different from other cystatins, both
in terms of specificity and strength of binding to the
target enzyme. Cystatins are generally rather non-

affinity for cathepsins B, H, L and S [24] and is
replaced by an unfavorable proline in cystatin F, thus
partially explaining the overall lowered affinity of cyst-
atin F for these enzymes. Proline in the S
2
site is a fea-
ture of human stefin A and cystatins F, S and SN, all
of which are significantly less potent inhibitors of cath-
epsin B than cystatin C [9,26]. Unlike Val10, Leu9
which occupies the S
3
pocket in cystatin C is the most
discriminating residue for binding to cathepsins B, H,
L, S [24]. No L9K mutants of cystatin C have been
prepared yet to study the effect of incorporating the
Cystatin F in U937 cells T. Langerholc et al.
1538 FEBS Journal 272 (2005) 1535–1545 ª 2005 FEBS
Fig. 2. Immunolabeling of cystatin F in U937 cells, where colocalization was found. Specific monoclonal (mAb) and polyclonal (pAb) antibod-
ies were applied. In all pictures, cystatin F was labeled with primary rabbit anti-(cystatin F) pAb and goat anti-rabbit Alexa Fluorä 488-labeled
secondary antibody (Ab) (green). Red color originates from labeling with: (A) mouse anti-(LAMP-2) mAb and goat anti-mouse Alexa Fluorä
546-labeled secondary Ab; (B) mouse anti-CD68 mAb and goat anti-mouse Alexa Fluorä 546-labeled secondary Ab; (C) mouse anti-(cathepsin
X) 1F12 mAb and goat anti-mouse Alexa Fluorä 546-labeled secondary Ab; (D) sheep anti-(cathepsin H) pAb and donkey anti-sheep Alexa
Fluorä 546-labeled secondary Ab. Before merging the images, signals for red and green fluorescence were adjusted to comparable levels.
The sites of colocalization are shown in yellow.
T. Langerholc et al. Cystatin F in U937 cells
FEBS Journal 272 (2005) 1535–1545 ª 2005 FEBS 1539
bulky, charged lysine, which is unique for cystatin F at
this position.
Cystatins are considered to be typical representatives
of the so-called emergency inhibitors. They are present

1540 FEBS Journal 272 (2005) 1535–1545 ª 2005 FEBS
fact that cystatin F does not colocalize with cystatin C
in U937 cells suggests different intracellular functions
for these cystatins. Although cystatin C has been sug-
gested to control antigen presentation by regulating
the activity of cathepsin S [16], this view has recently
been challenged [29]. Cystatin F is present predomin-
antly in cells of the immune system, and it would
therefore be expected to be a better candidate to con-
trol activity of cathepsins. Although it is partially
secreted from promonocyte U937 cells, a large propor-
tion resides in the lysosome-like vesicles [18,19]. The
colocalization of cystatin F with LAMP-2 and CD68
shown here confirms these observations, in contrast to
a recent report that cystatin F is not targeted to endo-
somes and lysosomes [22]. A role for cystatin F, invol-
ving a function other than its inhibition of cysteine
proteases, cannot be excluded, as shown for chicken
cystatin [30,31] and cystatin C [32].
If the same criteria that are valid for emergency type
inhibitors were met in lysosomes, proteases would be
inactivated and there would be no proteolyis. Hence
the concept that inhibitors can modulate protease
activity, and not only abolish it [33]. These modulatory
inhibitors are often colocalized with their targets. Cyst-
atin F could be a candidate for modulating the activity
of cathepsin H; in vitro inhibition is tight enough to
impair its activity at concentrations, which can be
found inside the lysosomes [34]. Additionally, such an
inhibitor would have a protective role against misdirec-

trans-Epoxysuccinyl-l-leucylamido-(4-guanidino)butane (E-
64) was obtained from the Peptide Research Institute
(Osaka, Japan). Fluorogenic substrates benzyloxycarbonyl-
FR-7-amido-4-methylcoumarin (Z-FR-AMC), R-AMC and
SY-b-naphthylamide (SY-bNA) were purchased from
Bachem (Bubendorf, Switzerland). The specific cathepsin X
substrate 2,4-dinitrophenyl (Dnp)-GFFW-OH [38] was a
gift of L. Juliano (University of Sao Paolo, Brazil). Stock
solutions of substrates were made in dimethysulfoxide
(Merck, Darmstadt, Germany).
Enzymes and inhibitors
Cystatin F was produced in a baculovirus expression system
and purified to homogeneity as described [13]. Papain (2·
crystallized; Sigma, St. Louis, MO, USA) was further puri-
fied by affinity chromatography as described [39]. Human
cathepsins were expressed in E. coli (cathepsin K [40]), in
Pichia pastoris (cathepsin F (M. Fonovic
ˇ
, Jozˇ ef Stefan Insti-
tute, Ljubljana, Slovenia, unpublished data), cathepsin S
(M. Mihelic
ˇ
, Jozˇ ef Stefan Institute, Ljubljana, Slovenia, un-
published data) and cathepsin V [41]) or isolated from spleen
(cathepsin C [42]) or liver (cathepsin X [38]). Cathepsin H
was isolated from porcine spleen [43]. All enzymes were 10%
(cathepsin K) to 100% active (papain) as determined by act-
ive site titration with E-64 or chicken egg white cystatin [44].
Activation of cystatin F
Twenty microliters of 0.1 m phosphate buffer, pH 6.0, con-

After this and all subsequent steps membranes were washed
with NaCl⁄ P
i
, pH 7.2, containing 0.05% Tween. Mem-
branes were incubated with primary anti-(human cystatin
F) polyclonal antibody [13], followed by goat anti-(rabbit
IgG) secondary antibody (Jackson Immunoresearch Labor-
atory, West Grove, PA, USA). Bands were detected using
0.05% 3,3¢-Diaminobenzidine (Sigma-Aldrich, Steinheim,
Germany) and 0.01% H
2
O
2
in 0.05 m Tris ⁄ HCl buffer,
pH 7.5.
Kinetic measurements
All measurements were performed at 37 °C under pseudo-
first order conditions with at least 10-fold molar excess of
the inhibitor. The following assay buffers were used: 0.1 m
phosphate buffer, pH 5.5 (for cathepsin F), pH 6.0
(papain), pH 6.5 (cathepsin S) or pH 6.8 (cathepsins C and
H), or 0.1 m acetate buffer, pH 5.5, for cathepsins K, V
and X. All buffers contained 2.5 mm EDTA and 0.1%
(w ⁄ v) polyethyleneglycol. In addition, the assay buffer for
cathepsin C contained 0.02 m NaCl. Activating buffers for
all the enzymes consisted of 5 mm dithiotreitol in assay
buffer. The fluorogenic substrates used to measure the
activity of each cathepsin are listed in Table 1. In all experi-
ments the dimethylsulfoxide concentration was less than
2% and the final dithiotreitol concentration was 2.5 mm.

z
and v
s
are the initial and
steady-state velocities, respectively, and k is the pseudo-first
order rate constant describing the presteady state of the
reaction. Based on previous results (reviewed in [8]) we
assumed a competitive mechanism of inhibition without a
pre-equilibrium step for the interaction between cystatin F
and its potential target proteases. In this mechanism k is
given by the following equation [45]:
k ¼ k
ass
Á½I
0
=ð1 þ½S
0
=K
m
Þþk
diss
ð2Þ
where I
o
is the inhibitor concentration, K
m
is the Michaelis–
Menten constant for the enzyme-substrate pair, S
o
is the

(Hyclone, Logan, USA), in a humidified atmosphere con-
taining 5% (v ⁄ v) CO
2
at 37 °C. For immunolabeling
experiments, 4 · 10
5
cellsÆmL
)1
were grown in fresh culture
medium for 24 h.
Immunofluorescence
For colocalization studies, cystatin F, paired successively
with the cathepsins, cystatin C, LAMP-2 and CD68, was
double immunolabeled. Cells (10
5
) were cytocentrifuged
onto poly(l-lysine) coated slides. Cells were fixed in NaCl ⁄ P
i
,
pH 7.4, containing 4% (v ⁄ v) paraformaldehyde for 30 min
and permeabilized for an additional 10 min in 0.1% (v ⁄ v)
Triton X-100. Non-specific staining was blocked with 3%
BSA (Sigma-Aldrich) and 10% normal serum (Sigma,
St. Louis, MO, USA). Either polyclonal or monoclonal high
affinity primary antibodies were used, and were all tested
for cross-reactivity to other cathepsins or cystatins. The poly-
clonal antibodies were rabbit anti-(human cystatin F) [13],
rabbit anti-(human cathepsin F) (H-110, Santa Cruz Bio-
technology, Santa Cruz, CA, USA) and sheep anti-(human
cathepsin H). The mouse monoclonal antibodies 2A2–1F5,

488 and Alexa Fluor
TM
546
were excited with an argon (488 nm) and He ⁄ Ne (543 nm)
laser, respectively, and emission was filtered using a narrow
band LP 505–530 nm (green fluorescence) and LP 560 nm
(red fluorescence) filter, respectively.
Molecular modeling
Cystatin F was modeled on chicken cystatin (1CEW) with
the program modeller [48]. Models of cystatin F-cathepsin
S and cystatin F-cathepsin L were made with the program
main [49] by fitting modeled cystatin F, human cathepsin S
(1GLO) and human cathepsin L (1ICF) to the structure of
stefin B–papain complex (1STF).
Acknowledgements
The authors thank Marko Fonovic
ˇ
, Mojca Trstenjak-
Prebanda, Sas
ˇ
a Jenko Kokalj, Olga Vasiljeva, Dieter
Bro
¨
mme and Ivica Klemenc
ˇ
ic
ˇ
for kindly providing us
with cathepsins F, K, H, S, V and X. We thank Sas
ˇ

7 Felbor U, Kessler B, Mothes W, Goebel HH, Ploegh HL,
Bronson RT & Olsen BR (2002) Neuronal loss and brain
atrophy in mice lacking cathepsins B and L. Proc Natl
Acad Sci USA 99, 7883–7888.
8 Turk B, Turk D & Salvesen GS (2002) Regulating
cysteine protease activity: essential role of protease inhi-
bitors as guardians and regulators. Curr Pharm Des 8,
1623–1637.
9 Abrahamson M (1994) Cystatins. Methods Enzymol 244,
685–700.
10 Rawlings ND & Barrett AJ (1990) Evolution of proteins
of the cystatin superfamily. J Mol Evol 30, 60–71.
11 Turk B, Turk V & Turk D (1997) Structural and func-
tional aspects of papain-like cysteine proteinases and
their protein inhibitors. Biol Chem 378, 141–150.
12 Halfon S, Ford J, Foster J, Dowling L, Lucian L, Sterling
M, Xu Y, Weiss M, Ikeda M, Liggett D, Helms A, Caux
C, Lebecque S, Hannum C, Menon S, McClanahan T,
Gorman D & Zurawski G (1998) Leukocystatin, a new
class II cystatin expressed selectively by hematopoietic
cells. J Biol Chem 273, 16400–16408.
13 Ni J, Fernandez MA, Danielsson L, Chillakuru RA,
Zhang J, Grubb A, Su J, Gentz R & Abrahamson M
(1998) Cystatin F is a glycosylated human low mole-
cular weight cysteine proteinase inhibitor. J Biol Chem
273, 24797–24804.
14 Morita M, Hara Y, Tamai Y, Arakawa H & Nishimura
S (2000) Genomic construct and mapping of the gene
for CMAP (leukocystatin ⁄ cystatin F, CST7) and identi-
fication of a proximal novel gene, BSCv (C20orf3).

21 Lenarcic B, Ritonja A, Sali M, Kotnik M, Turk V &
Machleidt W (1986) Cysteine Proteinases and Their Inhi-
bitors (Turk V, ed), pp. 473–487. Walter de Gruyter,
Berlin.
22 Cappello F, Gatti E, Camossetto V, David A, Lelouard
H & Pierre P (2004) Cystatin F is secreted, but artificial
modification of its C-terminus can induce its endocytic
targeting. Exp Cell Res 297, 607–618.
23 McGrath ME, Palmer JT, Bromme D & Somoza JR
(1998) Crystal structure of human cathepsin S. Protein
Sci 7, 1294–1302.
24 Hall A, Hakansson K, Mason RW, Grubb A & Abra-
hamson M (1995) Structural basis for the biological spe-
cificity of cystatin C. Identification of leucine 9 in the
N-terminal binding region as a selectivity-conferring
residue in the inhibition of mammalian cysteine pepti-
dases. J Biol Chem 270, 5115–5121.
25 Lindahl P, Abrahamson M & Bjork I (1992) Interaction
of recombinant human cystatin C with the cysteine pro-
teinases papain and actinidin. Biochem J 281, 49–55.
26 Saitoh E & Isemura S (1994) Production of human sali-
vary type cysteine proteinase inhibitors (cystatins) by an
Escherichia coli system and partial characterization of
recombinant cystatin S and its mutant (117 arginine fi
tryptophan). J Biochem (Tokyo) 116, 399–405.
27 Werle B, Sauckel K, Nathanson CM, Bjarnadottir M,
Spiess E, Ebert W & Abrahamson M (2003) Cystatins
C, E ⁄ M and F in human pleural fluids of patients with
neoplastic and inflammatory lung disorders. Biol Chem
384, 281–287.

Haley KJ, Riese R, Ploegh HL & Chapman HA (1999)
Cathepsin S required for normal MHC class II peptide
loading and germinal center development. Immunity 10,
197–206.
36 Claus V, Jahraus A, Tjelle T, Berg T, Kirschke H,
Faulstich H & Griffiths G (1998) Lysosomal enzyme
trafficking between phagosomes, endosomes, and lyso-
somes in J774 macrophages: enrichment of cathepsin H
in early endosomes. J Biol Chem 273, 9842–9851.
37 Lautwein A, Kraus M, Reich M, Burster T, Brandenburg
J, Overkleeft HS, Schwarz G, Kammer W, Weber E,
Kalbacher H, Nordheim A & Driessen C (2004) Human
B lymphoblastoid cells contain distinct patterns of
cathepsin activity in endocytic compartments and regu-
late MHC class II transport in a cathepsin S-independent
manner. J Leukoc Biol 75, 844–855.
38 Klemencic I, Carmona AK, Cezari MH, Juliano MA,
Juliano L, Guncar G, Turk D, Krizaj I, Turk V & Turk
B (2000) Biochemical characterization of human cathe-
psin X revealed that the enzyme is an exopeptidase, act-
ing as carboxymonopeptidase or carboxydipeptidase.
Eur J Biochem 267, 5404–5412.
39 Blumberg S, Schechter I & Berger A (1970) The purifi-
cation of papain by affinity chromatography. Eur J Bio-
chem 15, 97–102.
40 D’Alessio KJ, McQueney MS, Brun KA, Orsini MJ &
Debouck CM (1999) Expression in Escherichia coli,
refolding, and purification of human procathepsin K, an
osteoclast-specific protease. Protein Expr Purif 15, 213–
220.

patients with nonsmall cell lung carcinoma. Cancer 89,
2282–2291.
48 Sali A & Blundell TL (1993) Comparative protein mod-
elling by satisfaction of spatial restraints. J Mol Biol
234, 779–815.
49 Turk D (1992) Weiterentwicklung eines Programms fu
¨
r
Moleku
¨
lgraphik und Elektrondichte-Manipulation und
seine Anwendung auf Verschiedene Protein-Stru-
kturaufkla
¨
rungen. PhD Thesis, Technische Universita
¨
t
Mu
¨
nchen, Munich, Germany.
T. Langerholc et al. Cystatin F in U937 cells
FEBS Journal 272 (2005) 1535–1545 ª 2005 FEBS 1545