Identification of Type IV Collagenase in Rat Testicular Cell Culture ...

BIOLOGY

OF

(1990)

43, 956-964

REPRODUCTION

Identification

of Type IV Collagenase Influence of Peritubular-Sertoli

QING-XIANG

SANG,3’4

WILLIAM

G.

and

Department

of Anatomy

Laboratory

of Pathology,5

in Rat Testicular Cell Cell Interactions1

STETLER-STEVENSON,5

STEPHEN

W.

LANCE

Cancer

Bethesda,

A. LIOTI’A,5

BYERS2’4

and Cell Biology,4 Georgetown Washington, District of Columbia National

Culture:

University 20007

Institute,

Maryland

National

Medical

Center

Institutes

of Health

20892

ABSTRACT In the In this be

of the

study

medium

we

and

However,

predominantly

confluent

rat testicular other

type

cells

activated

In contrast,

in culture

metalloproteases

the

produce

is regulated

different

by the

of type

(from

cell

activator

secrete

type

ratio

of mesenchymal

blast

migration

proteases pathological and

plays an processes.

invasion

of the

uterine

wall

and

metastasis In the

has

[3, 5-7]. reproductive

system,

it

been

reported

that

tivator

(t-PA).

FSH

and

LH stimulate

u-PA activity in mice [8,91. PM volved in ovulation in mammals are

secreted

erous tubule stage-dependent

by isolated segments manner

Sertoli

t-PA activity

cells

as well

in a hormone and [15-20]. Recently,

in rats

Accepted July 10, 1990. Received March 1, 1990. ‘This research was supported 2Correspondence

Stephen

that the proteases

spermatogenic a small matrix

functions.

ulated

by NIH lID 23744. Department

in culture.

secretion cells

and

and

reports

time

have

of Anatomy

and

Cell

of type much

in cultures

zymography

more

type

immunothan

results of this

show

that

enzyme

and

in culture.

also

implicated

mesenchymal-epi-

collagenase epithelium

involved in the production in the Anuran tadpole (carcinoma

IV coilaless

consisting and

These

to

72-kDa

metalloproteinases

activation

and

secretion. For and mesen-

of collagenase [221. In addition,

melanoma

as type

IV pro-collagenase.

cells)

amount present by time

The enzyme cells.

can

Cell

and tu-

stimulate

Georgetown University Medical Center, 3900 Reservoir Rd., Washington, DC 20007. ‘Current address: Department of Chemistry and Institute of Molecular Biophysics, Florida State University, Tallahassee, Fl. 32301-3015.

Isolation

AND

phys-

by the we show

that of other metallobut not PM, is reg-

and 20-day-old enzyme

METHODS

Primary

Cell Culture

Sprague-Dawley digestions according

ley et al. [31], with some modifications. tions, the cells were washed thoroughly 956

and

in culture.

Sertoli cells from isolated with several

Biology,

secretion

are influenced In addition,

of this enzyme and in culture medium,

MATERIALS

Testicular W. Byers,

both

iological activation of this ratio of Sertoli to peritubular

as by seminif-

ulation

cells

zymes

inPM

rat uterus (MMP-7) has been pu[21]. Thus, both serine proteases may be involved in the cellular reg-

by

by time

levels produce

and

secrete

the

cently demonstrated that cultured Sertoli cells secrete several metalloproteases in addition to PM [30]. Some of these are regulated by hormones and others by coculture with peritubular cells. In this report we identify one of these en-

and

are

metalloproteinase from rified and characterized and metalloproteinases of reproductive

mor

plasac-

and collagenolysis 110-14]. Furthermore,

detected

that

fibroblast collagenase production [23-26]. Similar interactions are also involved in bone resorption and corneal fibroblast collagenase activity in vitro [27-29]. We have re-

cul-

tured ovarian granulosa cells produce urokinase-type minogen activator (u-PA) and tissue-type plasminogen

was

to epithelial

chyme are hyaluronidase

involves

metalloproteases [1, 2]. are required for angiogenand plasminogen acwith tumor invasion and

IV pro-collagenase,

the

basal

coculture

it is likely

that

show

alone

type

that

indicating

produce cultured

proteolytic

cell

Sertoli-peritubular form,

IV collagenase

alone cells

with

spermatogenesis.

and

thelial interactions in regulating example, interactions between

in both tropho-

extracellular matrix-degrading Membrane and matrix proteases esis [3,4] and both metalloproteases tivators (PM) are clearly associated

of type

unaffected

Several important role For instance,

and

mass

to confluence)

are

cells

cell

cultured

LV collagenase

IV collagenase,

morphogenesis,

Sertoli

Peritubular

attachment

INTRODUCTiON Extracellular normal and

cells

together

(epithelium),

molecular

do produce

levels

cells

domains

cells.

in coculture

plasminogen

and

Sertoli

of cultured form

in rat

to a lower

to three

phase

Sertoli

development,

present

(p-APMA)

cytoplasm

transitional

the

of 72 kDa

cells

and

in testicular

IV pro-collagenases.

peritubular

and

cells,

phase.

role

antibodies

in the

during

growing

(mesenchyme)

acetate

using are

Sertoli

of peritubular

Cells the

protease

cells a central

a metalloproteinase

Immunoblots

a 76-kDa

IV collagenase.

during

that

play

by p-aminophenylmercuric

can be immunolocalized

that

blotting.

may

matrix,

demonstrate

collagenase.

protease

peritubular

between

extracellular

is activated

a latent

genase

interactions

testis,

remodeling

rats were to Had-

After enzyme digeswith at least five

TESTICUlAR

changes

of

medium

(Dulbecco’s

minimum

dium, DMEM; Irvine Scientific cells were cultured in serum-free for composition, tains cytosine were 1 x

on plastic

i05 cells/cm2.

dishes

pernatant

of

the

and

2 mg/ml

2 days

in

digestion

(2

Biochemical Co., II, Sigma Chemical [Type

culture media

cells

cell ratios from the

mg/mI

of su-

collagenase

Chemical

Co.]),

by Skinner and Fritz [32]. After several washes, cells from this supernatant were grown for 10%

fetal

bovine

serum

(FBS)

(HyClone

Labo-

ratories Inc., Logan, UT) in DMEM and subsequently, after several washes with DMEM to remove serum, in SFDM. Conditioned media were collected and prepared as described

elsewhere

[30].

[30}. then

The gels incubated

was carried modifications

out

using standard procedures as described by Sang et al.

were rinsed three in 0.1 M Tris-HC1

gels containing plasminogen, contains 50 mM Tris, 0.15 7.6) [35] in a 37#{176}C shaking After incubation, the gels

water bath at 44 RPM were fixed and stained

Brilliant Blue R-250 for 1 h and destained

acetic

for

Effect

of EDTA

The was

10-40

effect

of EDTA

gels)

was

used.

gels

were

overnight. in 0.125%

the

For (7-20 Aliquots

of secreted

two

gels

or

Co.)

used

synthesized

on

a Biosearch

activities of the of the proteolytic with

or without

EDTA

the

was evaluated (white bands)

in

10 mM gelatinby comon these

treatment.

H2-amino sequence residues

in the acetate

methodology

procedures

pared,

Ah45-amino sequence rabbit

purified,

[38]. The zyme-linked

and

were

(p-APMA,

activated Sigma

of Stetler-Stevenson

were Pep-

N pro-collagenase MBR-26-amino acid sequence VAA-

mercial

ELISA MD).

polyclonal

(Kirkegard

kit

acid residues 75-94 of type NPDVANYNFFPRKPKWDKNQ. antibodies

characterized

antibody preparations immunosorbent

ersburg,

(8%) method

were

as described

pre-

previously

were characterized assay (ELISA) using and

Perry

Samples

gels for of Laemmli

were

boiled

us-

Green

by ena com-

Laboratories,

with

peroxidase 000

in the

Gaith-

was premodifi-

some

double-strength

sample

of proteins

sheets

was

from

carried

monospecific

polyout

ac-

anti-synthetic

some

modifications.

The

of 1:100 in 10% in PBS with 0.05%

first

(Bio-Rad serum

in PBS

Co.) with

and

anti-

calf serum Tween-20

CA). The second antibody goat anti-rabbit IgG (H+L) calf

for 10 loaded

was

IV collagenase antibodies Hi, H2, for immunoblotting. The antibody blots was performed according to

con]ugate 10%

in

transfer

used at a dilution Laboratories Inc.)

(Bio-Rad Co., Richmond, grade affinity-purified

electrophoresis [39], with

of 5% 23-mercaptoethanol and 30 pg protein

Electrophoretic

et al. [41],

body was (HyClone

was

0.05%

was EIA horsediluted Tween-

Immunocytochemistry

Chemet al.

lane.

peptides of human type and MBR-26 were used staining of nitrocellulose

1:2 20. media

peptide

acid residues 472-490 of type DKPMGPLLVATFWPELPEK [37]. 371-385 of type IV pro-collagenase

Affinity-purified

radish

Pro-Collagenase

pro-collagenases and

in-

the same amount of distilled for 2 h. After electrophoresis,

proteases activities

Peptides synthesizer.

usual slab

shaking

EDTA were sample

were incubated zymography.

9600

sensitive assay, half the for 1.5-mm miniature

with

1 mM and the occasional vor-

tide antigens are as follows: Hi-amino acid residues 1-17 of type N pro-collagenase sequence APSPIIKFPGDVAPKTD.

Polyacrylamide pared by the

were

This cell

immunization

in each

medium

A 10-

in 50 mM NaOH. to the concentrated

to Synthetic

in the

actylamide gels to nitrocellulose cording to Towbin et al. [40]. Affinity-purified, polyclonal,

Chen

at 37#{176}C overnight

ingp-aminophenylmercuric ical

of Antibodies

The

with

of concentrated

incubated

Activation of Putative by OrganomerauriaL Putative

Preparation peptides

served

kDa).

reaction was stopped by adding of 20 mM, and the samples treatment with double-strength

[36],

and

the collagenase buffer or the same buffer containing EDTA but without 5 mM CaCl2. Inhibition of the degrading parison

texing. The p-APMA to a final concentration stored at 4#{176}C before

proteases

by Chen

a more ig/lane

was dissolved was added

buffer in the presence mm before electrophoresis,

activity

medium (70

Immunoblotting

methanol/i0% methanol/7.5%

Proteases

on

cubated in 10 mM EDTA water at room temperature the

dye/50% in 25%

culture

N collagenase

media to a final concentration of were incubated at 37#{176}C for 2 h with

cations.

as described

some modifications. amount of protein

water, [34] for

mm.

on Secreted

determined

in distilled (pH 8.1)

or collagenase buffer, which M NaCl, and 5 mM CaCl2 (pH

Coomassie acetic acid acid

times buffer

(A2058)

of type

HEFGHAMGLEHS. N pro-collagenase

Zymograpby Zymography [33], with some

cell

control

buffer. After electrophoresis, the samples with collagenase buffer as described for

Freehold, NJ], 20 tg/ Co., St. Louis, MO],

III, Sigma

melanoma

mM stock of p-APMA p-APMA stock solution

conCells

peritubular

Human

as a positive

of approximately

experiments,

hyaluronidase

as described the peritubular

medium division.

cocultured at original cells were collected

second

[CLS II, Worthington ml DNase 1 [Type

This cell

957

IV COLLAGENASE

[35].

me-

Ana, CA). Sertoli medium (SFDM;

at a density

In some

Sertoli cells were and 7:3. Peritubular

essential

Santa defined

see Hadley et al. [31]). arabinoside, which prevents

grown

and 3:7

Co.,

TYPE

use

Type N collagenase of affinity-purified

was detected rabbit antisera

in cultured cells by H1 and Ab45. Both

958

SANG

antibodies lated

produced

Sertoli

a similar

cells

were

staining

cultured

pattern.

on glass

Freshly

coverslips

ET

AL.

iso-

1234

in SIDM

for 4 days. Peritubular cells and Sertoli-peritubular cells were cultured in 10% FBS for 2 days and subsequently, after washes to remove serum, in SFDM for another 2 days. Immunocytochemical staining Herron et al. [42], with some per

milliliter

was carried modifications.

of nonimmune

out

or immune

according Six micrograms

rabbit

was

IgG

185-

to used

110-I

as the first antibody. The affinity-purified nonimmune rabbit IgG was kindly provided by Dr. V. Papadopoulos and Ms. C. Davis (Georgetown University Medical Center, Washington, DC). The second antibody used was 2.5 g/ml fluorescein-labeled ris, Inc.). The 80% with

anti-rabbit IgG (H&L; coverslips were mounted

HyClone on glass

Laboratoslides with

glycerol, 20% 0.04 M Tris-HCI (pH 8.0), the addition of one chip of para-phenylene

to prevent labeled X 400,

quenching

[43].

antibodies Kodak film

Cells

tive

light (Carl

and Zeiss,

with

rated

at 200

ASA. A Planapo

Thornwood,

NY)

objec-

used.

RESULTS p-APMA We

Activation report

of Putative

elsewhere

plasminogen-independent proteases of 185 72 kDa,

and

50 kDa.

testicular

122

proteases kDa, 110

Figure

1 shows

at 72 kDa and to a lesser extent and that a band of degradation

of

in the the

autocatalytic

clipping

collagenase

molecule,

cells

that

covered

melanoma

from

N collagenase

of the

portion

N-terminal

suggesting

dicates genase.

that

this

metalloproteinase

Fig.

known

the

indicating probably pattern control to contain

1, lanes

3 and

the

have re-

72-kDa

intensity of the activated that activated

both by

was similar (medium rehuman 4).

This

type in-

72 kDa may be rat type N pro-collaexperiment, the 50-kDa protease from

that protease In the same

Sertoli-penitubular indicating that

cells

activity

was diminished molecular mass and

p-APMA activation 2) to the positive

of 70 kDa;

include 76 kDa,

Organomercurials collagenases [35]

band ran as quite a discrete doublet, the 72 kDa and 76 kDa forms were The i and

several

protease

protease is a latent collagenase. Although the 76 kDa band was only slightly decreased,

this treatment. (Fig. 1, Lanes

secrete

[30]. These kDa, 83 kDa,

at 76 kDa at a lesser

appeared after p-APMA treatment. been used previously to activate sult

cell coculture was molecule is either or

that

it

was

not activated (Lane not a collagenase-like

already

completely

2), acti-

vated.

The 76-kDa and 72-kDa with Antibodies to Human type

Immunoblots N collagenase

FIG. 1. The activation of putative pro-collagenases with p-APMA. Lanes 1 and 2 show proteases from Sertoli-peritubular cell (7:3) coculture in the presence of FSH. Lanes 3 and 4 show proteases from human melanoma cell culture medium as a positive control. The band at about 70 kDa is the type IV collagenase. Lanes 1 and 3, samples without, and Lanes 2 and 4, samples with p.APMA treatment for 1 h at 37#{176}C. Protease of 72 kDa was activated by p-APMA. Molecular mass x iO.

Pro-Collagenases

that

kDa,

-65

on TnNY), rated

63/1.4

was

72-

50-

NaN3, diamine

on a Zeiss Photomicroscope were photographed with trans-

Inc.,

-76

fluorescein-

on coverslips were photographed (Eastman Kodak Co., Rochester,

at 400 ASA by epi-illumination III. Phase-contrast images mitted

stained

0.02%

83-

Gelatinases Type/V

Cross-react Collagenase

using antibodies to three peptides showed that

different human media from Ser-

toli cell culture tures contained

alone and Sertoli-peritubular cell a doublet of type N pro-collagenases

76/72 kDa (Fig. 2, Lanes N collagenase-containing positive antibodies

control and is shown specific for all three

lagenase munoblots,

molecule were we concluded

gelatmnases

were

Sertoli-penitubular 5) more type

rat

490). region nized

recognized

type

in Figure different

human served

of type as a

2, Lane 1. Since regions of the col-

positive on three different imthat both the 76-kDa and 72-kDa

N pro-collagenases.

Media

from

cell cocukures generally contained (lane N collagenase detected by all three antibod-

ies than did media other immunoreactive also

2, 3, 4, and 5). The melanoma medium

cocul-

from Sertoli cell cultures species at 62 kDa and

in lane

5 by antibody

H2 (amino

alone. 42 kDa

Two were

acids

472-

Antibody MBR-26, directed against the metal-binding common to most metalloproteinases, also recogtwo bands at 42 kDa and 18 kDa in medium from

Sertoli-peritubular cell cocultures (3:7). It is possible that these smaller immunoreactive peptides were proteolytic degradation products of type N collagenase. Alternatively, they could gelatinolytic under any

Type

be related, collagenase-like molecules, although activity at 42 kDa and 18 kDa was not detected of the conditions that we tested.

IV Collagenase Activation Cell to Peritubular

by Sertoli Sertoli of 76/72

Is Influenced Cell Ratio

cells secreted basal levels of type N collagenases kDa (Fig. 2, Lane 4; Fig. 3, Lane 1). At a 7:3 ratio

TESTICULAR

1YPE

IV

959

COLIAGENASE

1234 123

4

5 ui140 1

22

‘86

FIG. 3. The effect of Sertoli-peritubular cell ratio on type nase secretion. Lane 1 is medium from Sertoli cell monoculture the presence of FSH. Lane 2 is medium from Sertoli-peritubular coculture coculture.

with FSH. Lane 3 is the medium from Sertoli-peritubular cell (3:7) Lane 4 is the medium from peritubular cell monoculture. Note

the changes different

of the 76/72-kDa

culture

systems.

of Sertoli

cells amounts 2, Lanes 2 and activity (Fig. discrepancy

type

IV collagenases

Molecular

mass

and other

3, Lane 2) could have

as detected by arisen because

in

detected

changes

zymography. This the very sensitive

zymographic

analysis

els that were cells cultured

not resolved by immunoblotting. alone did not produce detectable

in collagenase

lev-

Peritubular type IV col-

activity (Fig. 3, lane 4). However, at a 3:7 ratio of cells to peritubular cells, the amount of immuno-

reactive type IV collagenase (Fig. the predominant form was the little of the the activated in molecular peritubular

proteases

x 10.

to peritubular cells, the medium contained of immunoreactive type IV collagenase (Fig. 3), but somewhat more type IV collagenase

similar

lagenase Sertoli

IV collagegrown in cell (7:3)

2, lane 72-kDa

5) was enzyme

greater with

and very

76 kDa enzyme. In addition, the presence of form of type N collagenase, slightly smaller mass (62 kDa), was detectable in Sertolicell

coculture

(3:7)

medium

by both

immuno-

FIG. 2. Immunoblotting reveals that proteases of 76 kDa and 72 kDa are type IV pro-collagenases. (A) antibody H1 directed against amino terminal sequence 1-17 of human latent type IV collagenase; (B) antibody H2 directed against internal residue 472-490; (C) antibody MBR 26 directed against the metal-binding region (371-386) of human type IV collagenase. Lane 1 is the positive control of type IV collagenase from human melanoma cell culture medium. Lanes 2 and 3 are Sertoli-peritubular cell (7:3) coculture media in the presence (Lane 2) and absence (Lane 3) of FSH. Lane 4 is the medium from Sertoli cell monoculture. Lane 5 is the medium from Sertoli-peritubular cell (3:7) coculture. Note collagenase is influenced by Sertoli-peritubular

io.

that

the secretion of type IV cell ratio. Molecular mass

960

SANG

ET

AL.

FIG. 4. Immunolocalization of type IV collagenase in testicular cell cultures. Panels A, C, E, and G are phase-contrast H are indirect immunofluoresence images. Panels A and B serve as a negative control (affinity-purified rabbit nonimmuno cells in culture. Panels E and F are Sertoli-peritubular cells in coculture. Panels G and H are peritubular cells in culture.

images and Panels B, D, F, and lgG). Panels C and D are Sertoli Cells were stained with Ab45.

TESTICULAR

1YPE

blotting (Fig. (Fig. 3, Lane

2, Panel 3; Fig.

reactivity

this

for

terminal epitope which can only

1851. 14O-“110

‘76 “62

54,’-

“50

5) and zymographic 4). The pattern of form

was

identical

for human melanoma was consistent with

upon enzyme recognize the

analysis immunoto

that

cell-activated the loss of the

re-

type amino

activation [37]. Antibody Hi, latent amino terminal amino

acid residues 1-17, cannot recognize the activated 62-kDa enzyme (Fig. 2, Panel A, lane 5). However, antibody H2, which recognized an internal sequence near the carboxy terminal,

72”

B, Lane 5, lane

activated

ported previously N collagenase and

A

83”-

961

IV COUAGENASE

recognized

the enzyme. metal-binding zyme (Fig.

both

activation,

structural

epitope

the

latent

and

activated

Antibody MBR-26, directed against region, did not recognize this 2, Panel C). It is possible that after changes

unavailable

for

of the

antibody

enzyme

forms

of

the putative 62-kDa enpro-enzyme

may

render

this

binding.

Type/V Collagenase is Localized in Cultured Testicular Cells Paired collagenase

phase-contrast immunostaining

tibodies

HI

and

is illustrated

one

and

and produced

Ab45

here

images of type N in Figure 4. An-

fluorescence are presented the

(Ab45).

same

staining

pattern,

Positive

staining

for type

N collagenase

was detected in small cytoplasmic vesicles cells (Panel D) and in both cell types in Sertolicell cocultures (Panel F). Faint but detectable of peritubular cells in monoculture was also ob-

in Sertoli peritubular staining served

(Panel

H). The

peninuclear

of type N collagenase that observed for many ent in the that other

control (Panel immunoreactive

fragments

of type

der

some

conditions

the due

positive solely

Effects

B).

Since peptides,

N collagenase,

immunoreactivity to intact type

pattern of pres-

Figure 2 demonstrates perhaps proteolytic

could

by these

staining

cells was characteristic proteins and was not

antibodies,

be

recognized

un-

it is possible

that

here

not

demonstrated

was

N collagenase.

of EDTA

Figure noma

vesicular

in these secreted

5 shows culture

that

medium

all

the

(Lane

proteases 1),

Sertoli

present cell

in melaculture

me-

dium (Lane 2), and Sertoli-peritubular cell (7:3) and (3:7) coculture media (Lanes 3, 4) were completely or partially inhibited by EDTA. In order to visualize the effects of EDTA, less type FIG. 5. The inhibitory effect of EDTA on testicular cell-secreted proteases. Concentrated cell culture media were treated without (Panel A) or with (Panel B) 10 mM EDTA. Lane 1 is medium from human melanoma cell culture. Lane 2 is medium from Sertoli cell culture alone. Media from Sertoli-peritubular cell coculture were loaded on Lanes 3 (Sertoli-peritubular cell ratio is 7:3) and 4 (the ratio is 3:7). Note that the proteolytic activities of melanoma cell culture medium (Lane Bi) and Sertoli-peritubular cell (ratiO 3:7; Lane B4) are completely blocked by EDTA. Molecular mass x iO.

protein than IV collagenase

usual was loaded band was very

tures. Figure 5A shows a Sertoli to peritubular a new 140-kDa gelatinase.

on these gels and the faint in the monocul-

that medium from cells cultured at cell ratio of 3:7 (lane 4) contained A protease of this molecular mass

was also present in melanoma kDa gelatinase present in the lar cells cultured at a 7:3 ratio blocked by EDTA. Although tained a 56/50-kDa gelatinase 5, Lane 1) this was completely

medium (Lane 1). The 50medium of Sertoli-peritubu(Lane 3) was not completely melanoma (interstitial blocked

medium also collagenase; by EDTA.

conFig.

962

SANG

Effects

of Time

Figure (Lane genase

in Culture

6 shows

ond 3-day proteolytic

ing the

during

the

cells secreted metalloproteinases

first

much

period

from

and other transitional

the

time

3 days

neither (Fig.

N collathe

at plating,

metalloproteinases remodeling phase

of plating

sec-

t-PA nor u-PA 6B) suggesting

to the

time

may (dur-

paper,

of initial

we

have

demonstrated

rat testicular

type A

protease by three

of

human

N pro-collagenase.

type

olytic activity was by organomercunial all anti-type N that, in addition testicular collagenase difference

identified treatment

A second that

exhibited and also

collagenase antibodies. to a 72-kDa type N

76-kDa

be

due

to

the

presence

enced over, eling secreted

suggest rat

type results other influ-

by the ratio of penitubular cells to Sertoli cells. Morewe demonstrated that during the transitional remodphase at early times in cell culture [44] testicular cells more

type

N collagenase

and

other

cells

activated

even

were

the

(62-kDa)

though

t-PA and

predominant

form

cell

of type

cell

interaction,

may

state of this enzyme. may occur in a fashion similar

metallopro-

However,

interstitial hal cell

in our

At this might

it is possible

collagenase ratio may be

type

IV collagen-

influence

the

This physiological to that determined change and is also dracells to messtromal pro-

time,

we

know

nothing

metalloprotease

testicular There are

the

about activity

mechanism

the in

cells, or the molecular several levels at which

exerted at the

upon proteolysis. transcriptional and

these main

molecules are made latent, form enzyme

activated.

that

Proteolytic

Protease perhaps

the

that

medium

from

basis of this regregulation can be

genes can translational

and secreted they inhibitor complexes,

activity

can

then

be

be regulated levels; once can

either reor become

exerted

upon

that

type

IV collagenase

penitubular tors. Proteolytic

cell

interactions

activities

regulation

is mediated

can

also

and inhibitors. It is known and inhibitors are produced inhibitor of metalloproteinases [47-50]. Recently, TIMP-2,

cifically

with

type

N collagenase,

and Most

characterized cells produce

be unlikely Therefore,

that Sertoli net enzyme

between

enzyme

the conformation itor complexes, zymography munologically.

it

Sertolifac-

by enzyme

ac-

that both plasminogen by testicular cells [46]. (TIMP-I) has been which associates spehas

been

isolated

from

medium [35]. In the reproductive inhibitor activity has been idenin human some form or penitubular activity may and

inhibitor.

of the pro-enzymes proteolytic activity and

by

via paracrine

be regulated

tivators activators A tissue described

melanoma cell culture system, metahloproteinase

in-

plasma work by ep-

for the production of interstitial cells in coculture [45]. Therefore

is possible

ance

colla-

molecules

teraction with the matrix or after binding to the membrane [3]. Of particular interest to the present are studies demonstrating that a cytokine produced

tified [51].

of

by mesenchymal-epithefrom that of type N

system.

influence

cultured ulation.

regulation different

ithelial cells is essential collagenase by stromal

FIG. 6. Time in culture influences type IV collagenase secretion. Gelatin zymographies without (A) or with (B) copolymerized 20-p.g/mI pIesminogen. Lane 1 is the concentrated medium from the first 3-day Sertoli cell culture. Lane 2 is the medium from the second 3-day culture. Note that t-PA (at 58 kOal and u-PA (at 39 kDa) activities did not change, but that the metalloproteinase activities decreased dramatically during the second 3-day period. Molecular mass x iO.

u-

duction of interstitial collagenase was absolutely dependent upon the presence of small numbers of epithelial cells, an increase in the number of epithelial cells actually inhibited

genase

of a glycosylated

form of the 72-kDa enzyme. In addition, we localized N pro-collagenase in testicular cells in culture. Our also showed that type IV collagenase and metalloproteinases in testicular cell culture were

confluence,

ase was detected in the media collected after 3 days. To our knowledge, this shows for the first time that physiological stimuli, such as an alteration of some aspect of mes-

collagenase.

of type N procell lines. This

cells may produce a second form that is not produced by human may

gelatin-

some activation cross-reacted with These data pro-collagenase,

cell

for in vitro activation, namely conformational autocatalysis [35]. Corneal collagenase activity matically influenced by the ratio of epithelial enchymal cells [28]. In this instance, although

N collagenase

in primary cultures of rat testicular cells. of 72 kDa was activated byp-APMA and recognized different antibodies specific for various regions

the

activation activation

DISCUSSION In this

after

unaffected. penitubular

enchymal-epithelial

[441.

confluence

than

PA were When

of culture

more type than during

period (Lane 2). However, activities were decreased

that type N collagenase be required during the

AL.

teinases

that

1), Sertoli and other

ET

specific sequences Consequently, our

ovarian follicular of TIMP and be

cells are the result Since

SDS

it

fluid would

exceptions. of the balcan

change

or the enzyme-inhibcan still be detected

by

can be recognized imresults do not provide

1YPE IV COUAGENASE

TESTICULAR

information on the presence or absence of metalloproteinase inhibitors. Type N collagenase secreted by migrating endothelial

local degradation to detach from the

the

adluminal

of the basal lamina, basement membrane

compartment.

our

results

indicate

and other metalloproteiin culture and at low den-

sity, conditions which are most likely velopment rather than in the adult.

to prevail

during

During testicular crucial and definitive

development event in the

and differentiation, formation of a testis

a sexually

gonad

organization

indifferent

testicular cords sential to cord basement 15 days be

membrane of gestation. the

This

of the

event

basement

the cords by 14in thickness and

membrane

1981;

memN col-

17. Vihko

the activities of other matrix in seminiferous tubule for-

ogen

sex

together with are involved cord

formation,

tubular cells or by Matrigel, this time, type N collagenase in the testis in vivo. In conclusion, brane action

since

between of locally

suggest proteases, peritubular

induced

play

a role

Vihko

of the

basement

and epithelium matrix-degrading

mem-

requires proteases,

the we

development

and

cells

1965;

12:408-418. C Tumor

Biophys 24.

Res

Hemandea

outgrowths:

RA, Werb a model

z.

Degradation for

of extracellular

implantation.

J Cell

DM,

Zhang

L, Hartman

properties

L, Grahi

of human

K, Zhang placental

G-Y,

Tarpey

cytotrophoblast

J, Damsky cells

in vitro. J Cell Biol 1989; 109:891-902. 3. Moscatelli I), Rificin DB. Membrane and matrix localization of proteinases: a common theme in tumor cell invasion and anglogenesis. Biochim Biophys Acts 1988; 948:67-85. 4. Mignaui P, Tsuboi R, Robbins E, RifIdn DB. In vitro angiogenesis on the human amniotic membrane: requirement for basic fibroblast growth factor-induced proteinases. J Cell Biol 1989; 108:671-682.

1984; MS.

1988;

by fibroblasts.

Biochem

in cocultures

of human

and human

fibroblasts

24:201-207.

to proliferate

Khidair

and

release

of basal

cell carci-

factor(s)

J Invest

collagenase.

which Dermatol

CW, Rees RC. Modulation of type IV collagenase in a hamster fibrosarcoma by basement membrane fibroblasts. Br J. Cancer 1988; 57:475-480.

IA, Potter

activator

epithelial-mesenchymal 1987;

B, Gross cell

J.

B. Regulation stimulation

Proc

of stromal and

Hadley

and

resorption

collagenase

of corned NatI Acad

collagenase Sci USA

cell collagenase by epithelial

inhibition

77:5331-5335. 30. Sang Q-X, Dym M, Byers SW. Secreted ture. Biol Reprod 1990; 43:946-955. Sci USA

M. Bone

interaction

Regulation

interactions.

cornea:

in vitro

CC. Abramson cell

in cholestea.

production.

Lar-

production:

epi.

97:854-859.

29. Johnson.Wint

31.

production

Postlethwaite AE. Establishment for a basal cell carcinoma-derived

and lung H, Honda Y, Huang

Moriyama

Acad

96:1108-1116. T-Y,

secreted

109:1026-1034.

stimulation Len evidence

plasminogen

components

Biol

of collagenase

1982;

fibroblasts

thelial-stromal A, Pedersen

stimulation

Hibbs

culture:

28. Johnson-Muller

REFERENCES

and degradative

of levels of plasminogen activator activity in a two-chamber assembly. Endocrinology

85:470-475.

and 27

cell

Commun

AD,

in

yngoscope

SJ, Cui

M. Stage-specific regulation of plasminogen actiepithelium. Endocrinology 1987; 120:142-

lB. Control

Biswas C. Collagenase tumor cells. Cancer

1985;

authors

CH. Adhesive

P. Nielsen La, Skriver L Adv Cancer Res 1985;

cancer.

rat seminiferous

maintained

23. Biswas

toma:

2. Fisher

and

21. Woessner JF Jr. Taplin Ci. PurifIcation and properties of a small latent matrix metalloproteinase of the rat uterus. J Biol them 1988; 263:16918-16925. 22. Eisen AZ, Gross J. The role of epithelium and mesenchyme in the production of a collagenolytic enzyme and a hyaluronidase in the Moran tadpole. Dev Biol

mor-

are grateful to NIADDK for the gift of FSH. We thank Dr. Vassilios Papadopoulos and Ms. Cynthia Davis for kindly providing purifIed nonimmune rabbit IgG and Dr. Charles Underhill for helping with immunoblotting procedures.

1983;

J, Kristensen

degradation,

J, Parvinen

in the

M, Fritz

Seitoli

26. Teale

Spindle

tissue

122:2613-2618.

in testicular

trophoblast

Can-

JJO,

KK, Toppari

noma

by mouse

matrix.

during

secretion

145. 20. Ailenberg

ACKNOWLEDGMENTS

matrix

extracellular

112:431-435.

25.

1. Glass RH, AggelerJ,

Lab

matrix.

Parvinen M. Cellular regulation of plasminogen actispermatogenesis. Biol Reprod 1984; 31:383-389. KK, Toppari J, Saksela 0, Suominen j(O, Parvinen M. Testicular plasminactivators during postnatal development in the rat. Acts Endocrinol 1986; Kl#{231}, Suominen

stimulates

The

extracellular

of the

PA, Grondahl-Hansen

activators,

secretion

18. Vthko

by

that in the testis, type N collagenases and other which can be influenced by interactions between cells (mesenchyme) and Sertoli cells (epithe-

hum), may phogenesis.

metastases-role

the

25:143-146.

vator

at

and

and

46:1-7.

K, Andreasen

vator

19.

by pen-

in vitro [31, 55]. However, has not been demonstrated

remodeling

mesenchyme secreted

either

invasion

44:139-266.

to

male

Tumor

8. Beers WH, Strickland 5, Reich E. Ovarian plasminogen activator: relationship to ovulation and hormonal regulation. Cell 1975; 6:387-394. 9. Canipari R, Strickland S. Studies on the hormonal regulation of plasminogen activator production in the rat ovary. Endocrinology 1986; 118:1652-1659. 10. Reich R, Miskin R, Tsafriri A. Follicular plasminogen activator: involvement in ovulation. Endocrinology 1985; 116:516-521, 11. Reich R, Tsafriri A, Mechanic GL. The involvement of collagenolysis in ovulation in the rat. Endocrinology 1985; 116:522-527. 12. Curry TE Jr, Dean DO, Woessner JF Jr, LeMaire WJ. The extraction of a tissue collagenase associated with ovulation in the rat. Biol Reprod 1985; 33:981991. 13. Curry TE Jr. Clark MR. Dean DO, Woessner JF Jr. LeMaire WJ. The preovulation increase in ovarian collagenase activity in the rat is independent of prostaglandin production. Endocrinology 1986; 118:1823-1828. 14. Morioka N, Brannstrom M, Woessner JF, LeMaire WJ. Mechanism of mammalian ovulation. In: Yashinaga K, Mon T (eds.), Development of Preimplantation Embryos and Their Environment New York: Alan It. Liss, Inc.; 1989: 65-85. 15. Lacroix M, Smith M, Fritz lB. Secretion of plasminogen activator by Sertoli cell enriched cultures. Mol Cell Endocrinol 1977; 9:227-236. 16. Lacroix M, Parvinen M, Fritz lB. Localization of testicular plasminogen activator in discrete portions (stages VII and VIII) of the seminiferous tubule. Biol Reprod

ap-

the

invasion

1986;

Plasminogen

and

for

lagen degradation metalloproteinases and

7. Dano

SH.

49:636-649.

LA. Tumor Res

needs

is specific

selective proteolysis of basement [54]. Thus, it is likely that type

in vivo,

the from

of epithehial

parently involves brane components

mation

de-

material is esrat testis, the

completely encloses As the cords grow

continuity

broken.

is the

[53]. Basement membrane formation. In the developing

1983;

6. Lions

allowing germ cells and move toward

However,

of type N collagenase highest at early times

that levels nases are

length,

Invest

tumor cells degrade basement membrane type N V collagen [5, 52]. It is possible that in the semitubule, type N collagenases may be involved in

cells and and type niferous

LA. Rao CN, Barsky

5. Lions

cer

963

1978;

75:4417-4421.

production cell

in adult

products.

rabbit

Proc

NatI

1980;

MA, Byers SW, Suarez-Quian

metalloproteinases CA, Kleinman

IlK,

in testicular Dym

cell cul-

M. Extracellular

regulates

development

Sertoli

cell

in vitro.J

differentiation,

testicular

cord

formation,

and germ

ma-

cell Cell Biol 1985; 101:1511-1522. 32. Skinner MK, Fritz lB. Identification of a non-mitogenic paracrine factor involved in mesenchymal.epithelial cell interactions between testicular peritubular cells and Sertoli cells. Mol Cell Endocrinol 1986; 44:85-97. 33. Heussen C, Dowdle EB. Electrophoretic analysis of plasminogen activators in trix

964

SANG

polyacrylamide

gels

strates. Anal

Biochem

containing 1980;

sodium

sulfate

dodecyl

and copolymerized

sub-

M, Turpeeniemi-Hu;anen

K, Liotta

tein

with

FEBS

UI(

40. Towbin Natl

H, Staehelin

41.

Acad

Green

gels

1986;

and

collagenase

a pro-

stromelysin.

proteins

during

J. Electrophoretic sheets:

of the head

the assembly transfer

procedures

of proteins

and some

from

of

poly-

applications.

of macrophages

CB. Aggregation antibody

Proc

to the

hyaluronate

and

receptor.

tibroblasts Exp Cell

Res

activities

capillary

J,

EJ, Gavnilovic endothelial

is regulated

Z. Secretion

Werb cells

II. Expression

by endogenous

inhibitors.

and

remodeling

behavior

phase.

Anat

B, Bauer EA. Stimulation of collagenase cytokine-evidence for pretranslational E, Tong

JA, Balekjian

crete

an inhibitor

of Sertoli

cells in sparse

Rec 1988;

220:11-21.

synthesis

by a 20,000-Dalregulation. J Biol them

PSFIB.

Rat testicular

of plasminogen

activator

peritubular

activity.

Biol

cells Reprod

in culture 1988;

se

38:359-

371.

G, Cawston TE, Reynolds B. An inhibitor of collagenase from human amniotic fluid-purification, characterization and action on metalloproteinases. BiochemJ 1981; 195:167-170. 48. Murphy G, Reynolds .5, Werb 1 Biosynthesis of tissue inhibitor of metalloproteinases by human fibroblasts in culture. J Biol Chem 1985; 260:3079-3083. 49. Cawston TE, Galloway WA, Mercer E, Murphy G, Reynolds B. Purification of rabbit bone inhibitor of collagenase. Biochem J 1981; 195:159-165. 50. Overall CM, Wrana JL, Sodek J. Independent regulation of collagenase, 72-kDa progelatinase, and metalloendoproteinase inhibitor expression in human fibro47. Murphy

51.

blasts

by transforming

Curry

TE Jr,

cation

and

growth

Sanders

SL, Pedigo

54. Pelliniemi

gonad.

S, Glaser

Extracellular

Biol

1987; B, Lions

Chem

1989;

RS, Wilson

264:1860-1869.

EA, Vernon

inhibitor

activity

1981;

J, Grund in testicular

MW.

Identifi.

in human

ovar-

123:1611-1618.

LA. Basement

collagen:

membrane

cells. Science (Wash DC) 1983; L. Development of sexual dimorphophism

Genet

U, Paranko matrix

Estes

endothelial

U, Lauteala Hum

J

NG,

Endocrinology

by migrating

dation

factor-s.

of metalloproteinase

characterization

ian follicular fluid. 52. Kalebic T, Garbisa

of collagenase

261 :2810-2813.

munofluoresence

Hettle

onic

Chem

the transitional

1985; 260:2080-2085. 46.

of metalloproJ Biol

Fritz lB. Topography

AHC,

during

53. Pelliniemi MJ, Clark

by stimulated

stromelysin

PS, Choi

45. Johnson-Wint ton epithelial

227:680-685.

76:4350-4354.

G, Underhill

Banda

Plau JL, Michael 842.

1979;

by amonoclonal GS,

teinases

43.

to interstitial

H, Tryggva-

recognize

IV collagenase

178:224-232.

Flerron and

to nitrocellulose

SJ, Tarone

1988;

1970;

T, Gordon

Sci USA

is inhibited 42.

of structural

T4. Nature

acrylamide

homology

Krutzsch

WG,

to type

233:109-113.

Cleavage

bacteriophage

antibodies

sequence

1988:

Lett

39. Laemmli

T, Stetler-Stevenson

IA. Monocional

limited

44. Tung culture

1356. son

AL.

102:196-202.

34. Hettle JA, Wailer EK, Fritz IS. Hormonal stimulation alters the type of plasminogen activator produced by Sertoli cells. Biol Reprod 1986; 34:895-904. 35. Stetler-Stevenson WG, Krutzsch HC, Liotta LA. Tissue inhibitor of metalloproteinase (TIMP-2)-a new member of the metalloproteinase inhibitor family. J Biol Cheni 1989; 264:17374-17378. 36. Chen J-M, Chen W-T. Fibronectin-degrading proteases from the membranes of transformed cells. Cell 1987; 48:193-203. 37. Stetler-Stevenson WG, Krutzsch HC, Wacher MP, Margulies IMK, Lions LA. The activation of human type IV collagenase proenzvme. J Biol Chem 1989; 264:135338. Hohvta

F

degra-

221:281-283.

in the embry-

58:64-67.

SK, Frojdman differentiation.

K, Foidart Ann

J-M, Lakkala-Paranko

NY Acad

Sci 1984;

T.

438:405-

416.

AF. Retardation

of fading

byp-phenylenediamine.

and

J

enhancement

Histochem

of intensity

Cytochem

1983;

of im31:840-

55. Tung PS, Fritz lB. Extracellular influence the rate and pattern 113:119-134.

matrix

components

of Sertoli

and

cell migration

testicular peritubular

in visro. Dev

cells

Biol 1986;