Stretching of fetal membranes increases interleukin-8 and collagenase activity

the concentration of

Emad E1 Maradny, PhD, Naohiro Kanayama, PhD, Abdul Halim, MB, BS, Kayoko Maehara, MD, and Toshihiko Terao, PhD

Hamamatsu, Japan

OBJECTIVE" The aim of this study was to determine whether stretcNng of fetal membranes can increase interleukin-8 concentration and collagenase activ]ty. STLIDY DESIöN: Strips of whole fetal membranes, amnion, or muscles of the Iower uterine segment were stretched for 2 or 4 hours. Interleukin-8 and collagenase activity were measured in homogenized control and stretched samples. Immunohistochemical staining for interleukin-8 was carried out. RESULTS: The interleukin-8 concentration increased significantly after the whole fetal membranes were stretched for 2 and 4 hours (p < 0.0007 and 0.001, respectively). Also, stretcNng of amnion and muscles of the Iower uterine segment for 2 and 4 hours increased the concentrat]on of interleukin-8 significantly (p < 0.0007 after 2 and 4 hours, respectively). Collagenase activity was significantly increased after stretching of amnion, amniochorion, and muscles of the Iower uterine segment for 4 hours (p < 0.0007, 0.006, and 0.0007, respectively). After stretching, samples were darkly stained for interleukin-8 compared with control nonstretched samples. CONCLLI81ON: Stretching of amnion, amniochorion, and muscles of the Iower uterine segment increased interleukin-8 production and collagenase activity. (Am J Obstet Gynecol 1996;174:843-9.)

Key words: Amnion, amniochorion, interleukin-8, collagenase activity, stretching, muscles of lower uterine segment

During pregnancy both uterus and fetal membranes

grow and stretch to accommodate the rapid growth of the

fetus. Fetal membranes grow slower than the myome- trium does? After the second trimester the growth offetal membranes almost ceases. 2 Therefore tt?.ey have to

stretch markedly to cope with the rapid growth rate of the uterus. 3

Fetal membranes may play a eentral role in the initia-

tion of parturition. 40verstretching of the uterus and fetal membranes, as in cases of multiple pregnancies and

hydramnios, iS usually accompanied by premature cer~~i- cal ripening and delivery. ~' 6 Mechanieal stretching of

cuhured amnion cells and isolated myometrinm was

found to increase synthesis and release of prostaglandin

E2,: which has a great effect on cervical maturation and uterine contraction, s

The ability of the fetal membranes and the uterine cervix to stretch and resist the intraamniotic pressure

during pregnancy depends on their collagen-rich con-

From the Department of Obstetrics and Gynecology, Hamamatsu Univet~ sity School of Medicine. Received for publication April 18, 1995; revised August 1, 1995; ac- cepted August 11, 1995. Reprint requests: Emad El Maradny, Department of Obstetrics and Gynecology, Hamamatsu Univ«si~y School of Medicine, 3600 Handa- cho, 431-31 Hamamatsu, Japan. Copyright © 1996 by Moslo~-Year Book, Inc. 0002-9378/96 $5.00+ 0 6/1/68535

nective fissue. 1 A defect in the collagen content of fetal

membranes may predispose to premature rupture of

membranes. ~ A marked decrease in the collagen content of the cervix is known to occur during ripening. 1o Also, it

was observed that the collagen content of the fetal mem-

branes decreases significantly during the last 8 weeks of

normal pregnancy/~ The mechanisms underlying col-

lagen changes in fetal membranes are not clear, but they

could be similar to the process that eauses cervical matu- ration. The decrease in the collagen content in the cervix

was attributed to the effect of proteases enzymes, espe- cially collagenolytic enzymes. '~ These enzymes were

found to be prodnced by cervical fibroblasts and neutro- phils, which invade the cervix during ripening. ~3

Interleukin-8 (IL-8) is a newly discovered chemotactic and acUvating factor for neutrophils. 14 It may play an

essential role in cervical ripening and initiation of labor. ~5

IL-8 concentration in amniotic fluid was found to be

increased gradnally in the third trimester of pregnancy? 6

Cultured amnion, chorion, and decidual cells produced IL-8 eonstimtively and in response to other cytokines.17' J~ Human uterine cervix was also found to be able to pro- duce large amounts of IL-8 near term. ~9 IL-8 conld ac-

count for the neutrophil accumulation seen during cer- vital dilatation. Exogenous application of IL-8 was found

to induce cervical ripening through increasing col-

lagenolytic enzymes released by cervical fibroblasts and aceumulated neutrophils? ~

843

844 El Maradny et al. March 1996 AmJ Obstet Gynecol

Fig. 1. Machine used to stretch amnion, amniochorion, and uterine muscles. It has one fixed jaw, whereas other jaw is mobile and connected to spring balance.

70

65

60

55

50

45

4O

35

3O

100

Amnion

Whole fetal membranes I

Myometrium I

I I I I 'I

120 140 160 180 200

% increase in length

Fig. 2. Percentage of increase in length of amnion, arnniochorion, and uterine muscles in relation to traction in grams.

It is unknown which factor(s) potent ia tes the produc-

tion and release o f IL-8. We hypothesized that s t retching

may be one of the factors that control IL-8 and collage-

nase activity. Thus the aim of this research was to study the

s tretching effect o f fetal m e m b r a n e s and muscles o f the

lower u te r ine segment on IL-8 concen t ra t ion and colla-

genase activity.

Material and methods

This research was approved by the research commi t t ee

of Hamamatsu University. Writ ten consent was obta ined

f rom all patients involved in this study.

A muscle biopsy spec imen was col lected f rom the lower

u te r ine segment dur ing elective cesarean sections and

before any clinical signs of labor (n = 15). Cesarean sec-

tions were d o n e because o f previous cesarean sections

(n = 10) or pr imigravid w o m e n with b reech presentat ions

(n = 5). All cases were be tween 38 and 40 weeks' o f gesta-

tion. Multiple specimens f rom the fetal membranes of the

same patients were also col lected immedia te ly after deliv-

ery of the placenta.

Samples were washed thoroughly in warm saline solu-

tion to remove b lood clots and were t ransferred to the

laboratory. Membranes were divided into equal 3 x 1 cm

strips, whereas muscle biopsy specimens were divided

into 2 x I × 0.5 cm strips. In m e m b r a n e biopsy specimens

we used e i ther whole fetal m e m b r a n e s ( amniochor ion

and at tached r e m n a n t materna l decidua) or amn ion

alone after careful separat ion f rom chor ion. Muscle, am-

nion, or whole fetal m e m b r a n e strips were c l amped be-

tween two jaws of a stretching machine (Fig. 1). O n e jaw is

fixed, whereas the o the r jaw is mobile . The mobi le jaw is

a t tached to a spring balance to measure the s tretching

power in grams. Samples were immersed in phosphate-

buf fe red saline solution at 37 ° C. Gradual stretching of

the strips was carr ied out unti l maximal stretching was

Volume 174, Number 3 El Maradny et al. 845 AmJ Obstet Gynecol

1 . 8

1.6

1.4

~ o 1.2

o

~ .4

.2

p = O . O O l i i

* Am. : Amnion * Am-Oh : Amniochorion * myomet. : myometTium

p=O.O007

all p=O.r)o07

p=O,O007

Fig. 3. Concentration of IL-8 in amnion, anmiochorion, and muscles of lower uterine segment mea- sured after 2 and 4 hours of stretching.

q 3 { p=G0007 Cont. : Control | [ ~ 1 Stret. : Stretched

.~ 2.5

O

• N ~ c 2 .~ .~. ~, f p~o.oo6

.5

(cont.) (Stret. 4h) (cont.) (Stret. 4h) (cont.) (Stret. 4h) t j t I u j

Ammon Amniochorion Muscle of lower uterine segment

Fig, 4. Collagenase activity measured in amnion, amniochorion, and uterine muscles before and after stretching for 4 hours,

achieved. The increases in length and stretching power in

grams were calculated for each sample. We stretched the

strips continuously for 2 or 4 hours. T h e n 100 mg (wet

weight) o f tissue was col lected and homogen ized with

1 ml o f phospha te -buf fe red saline solution. After cen-

trifugation of the h o m o g e n i z e d strips, the supernatants

were col lected and kept at - 8 0 ° C until used. T h e IL-8

concent ra t ion and collagenase activity were measured in

nons t re tched control samples and in s tretched samples.

IL-8 was measured with the commercia l ly available kit

Biotra IL-8 enzyme-l inked immunoso rben t assay System

(Amersham, Internat ional p i g Little Chat fon t , Buck-

inghamshire, Uni ted Kingdom). Steps o f measu remen t

were done according to the manufac turer ' s instructions.

Coltagenase activity was measured with highly specific kits

(coltagenase type 1 [matrix meta l loprote inase - t] activity

measurement , Cosmo Bio, Yagai, Tokyo).

Control and su 'etched specimens were fixed in parafor-

maldehyde and e m b e d d e d in paraffin wax. Paraffin sec-

tions were cut at a thickness of 3 to 4 gin. After deparaf-

finization, rehydrated tissue sections were stained by he-

matoxylin and eosin to study the his topathologic changes

after stretching. Also, immunohis tochemica l staining for

IL-8 was carr ied out for control and stretched sections by

846 El Maradny et al. March 1996 AmJ Obstet Gynecol

Fig. 5. A, Normal amnion stained few IL-8. h appears as single layer of smooth cuhoidal ceils. (Original magnification x40.) B, After stretching for 4 hours, amnion cells appeared as long columnar cells with surface projections. Also, the staining of the compact layer below the epithelium was increased. (Original magnification ×40.)

use of a streptavidin-biotin complex-peroxidase kit

(Dako, Calif.). First antibody (antihuman IL-8 antibody)

was purchased from Oncogen Science, New York. Con-

trol sections were subjected to the same methods, except

that primary antibodies were replaced with Tris-buffered saline solution.

Data are shown as mean +__ SD, and the Wilcoxon

signed-ranks test was used for statistical analysis.

Results

Amnion, whole fetal membranes, and muscle strips from the lower uterine segment increased in length in response to stretching for 4 hours. Fig. 2 shows the per- cent increase in length of stretched biopsy specimens in response to traction in grams. Amnion has a remarkable

ability for stretching, compared with whole fetal mem-

branes. High concentrations of IL-8 were found in control

whole fetal membranes, anmion, and muscle biopsy Spec-

imens (0.299 + 0.136, 0,205 _+ 0.097, and 0.111 + 0.044 ng per 100 mg wet.homogenized tissue, respectively). The

concentration of IL-8 in control whole fetal membranes was higher but not significantly different from that in

amnion biopsy specimens. Stretching of amnion cells led to a marked t ime-dependent increase in the IL-8 concen-

tration. After 2 and 4 hours of stretching the homog- enized amnion cells showed a significant increase in 11~8 concentration (p = 0.0007). Also, stretching of the whole fetal membranes induced a significant increase in the IL-8 concentration after 2 and 4 hours (p= 0.0007 and

Volume i74, Nmnber 3 El Maradny et ak 847 AmJ Obstet (;ynecol

Fig. 6. A, Normal amniochorion staiJ~cd for IL-8. (Original magnificado~~ ×20.) B, 32"ter stretchi~lg ~br 4 hours, staining of reticular layer of choHot~ and decidual cells was markedly increased. (Original mag~lification x20.)

0.001, respectively) (Fig. 3). The increased level of 1I.-8

concentration in whole fetal membranes was significa/ltly

higher than in amnion cells after 2 and 4 hours of stretch-

ing (p = 0.0001 and 0.01, respecdvely).

The II.-8 concentration in stretched muscle strips from

the lower uterine segment was increased time depen-

dently. A significant increase in the 1L-8 concentration was found after 2 and 4 hours of stretching (p = 0.0007) (Fig. 3).

Collagenase activity was significantly increased after stretching ofamnion (p = 0.0007). Also, stretching of the

whole fetaI membranes induced a significant increase of collagenase activity compared with the control group (p = 0.006) (Fig. 4). No significant change in collagenase

activity of amnion or whole fetal membranes was found

after 4 hours compared with 2 hours of suetching.

Stretching of muscle strips from the lower uterhm seg-

ment led to a significant increase in co/lagenase activity

compared with nonstretched muscles (p= 0.0007) (Fig.

4). Also, we could E~ot observe significant cha~~ges in

collagenase activity between 2 and 4 hours of stretching,

In histologic sections stained by hematoxyiin and

eosin, normal contro] arrmion cells appeared as a smooth

layer of adherem cuboidal or short cohlmnar cel!s. After

stretching, most of the cells appeared as long columnar cel]s with surface projections. No other obvious changes could be observed after stretching of whole fetal mem-

brane or muscle biopsy specimens, hnmunohistochemi- cal stainhlg for II:8 in amnion cells and whole fetal mem-

branes showed a great difference before and after stretch-

ing (Table I). After stretching, the immtmostaining for

IL-8 was increased in the cytoplasm of amnion cells, indi-

848 El Maradny et al. March 1996 AmJ Obstet Gynecol

Table I. IL-8 immunostaining of normal and

stretched amnion, chorion, and myometrium

Staining Amnion I ChorionIMyometrium

Control _+ + + Stretched ++ +++ ++

caring increased IL-8 production in the cells (Fig. 5).

Also, the staining of the compact layer below amniotic

epithelium was obviously darker than the control

samples. The reticular layer of chorion and decidual cells

showed dark staining for IL-8 after stretching (Fig. 6).

Comment

Fetal membranes do not line the uterine cavity pas-

sively, but they are in a state of continuous stretching and

tension during pregnancy. The maximum stretching and

tension are reached at term, and more stretching will

occur with the start of uterine contractions? It was found that amni0n can be stretched more than the whole fetal

membranes. This observation is similar to normal cervi-

cal dilatation where the chorion, usually torn earlier, and the amnion will form the bag of forewater.

Mechanical stretching induced some histologic

changes in the amnion cells, such as surface projections

and elongation Of the cells. The exact cause of these changes is not clear. Also, mechanical stretching led to an

increase of the IL-8 concentration in amnion, whole fetal

membranes, and the muscles of the lower uterine seg-

ment, This increased concentration of IL-8 was higher in whole fetal membranes than in amnion. This may be due

to increased production of IL-8 by chorion and remnants of maternal deeidua. Increased IL-8 concentration after

stretching of fetal membranes and muscles of the lower

uterine segment may play a role in initiation of cervical

ripening and labor. The concentration of IL-8 in normal amniotic fluid is

very low in the second trimester, 2° whereas in the third

trimester it was markedly increased. A more significant increase in the concentration of IL-8 in amniotic fluid

occurs with labor. ~6 The physiologic role of amniotic fluid IL-8 in parturition has not been determined. Normal noninfected amniotic fluid contains white blood cells, especially during labor, when neutrophils are the domi-

nant cell type. Neutrophil numbers were increased mark-

edly with infectionY Because the fetal membranes are

avascular tissues, neutrophil infiltration should be trans- ferred from the maternal decidua. Butterworth et al. 22 identified the accumulation of neutrophils in placenta

and decidua in normal pregnancy. Because cervical ripening and parturition has been

linked to an inflammatory process, 23 IL-8 may be a signal for neutrophil recruitment and activation in reproduc-

rive dssues during labor. Accumulation of neutrophils in the human cervix is a known phenomenon that occurs during cervical ripening and dilatation] 2 Neutrophils

can produce several kinds of proteases, such as collage-

nase and elastase, that are involved in the degradation of collagen fibers. 24 It was found that the collagen content of

the cervix is markedly decreased during dilatation] ° Col-

lagenase and elastase, enzymes of neutrophil origin, were suspected of playing the essential role in cervical matura-

tion. '3' s~ The collagen content of fetal membranes was

also found to be decreased near term of normal preg-

nancy. 1' A more significant decrease occurred in patients with premature rupture of the membranes. 9 Our results

suggest that the mechanism underlying the decrease in the collagen content of the fetal membranes and cervix

may be similar. This mechanism may be mediated

through increased IL-8 production, neutrophil accumu-

lation, and collagenase activity.

Fibroblasts of fetal membranes are capable of produc-

ing collagenase and other proteases. 2a Moreover, we found that mechanical stretching of fetal membranes

increased their collagenase activity. Previously, we also reported that stretching of amnion leads to increased

production and release of prostaglandin Ez.7

Production of IL-8, collagenase, and prostaglandin E 2 is independent of the viability of the fetus, but only on

stretching of the fetal membranes and lower uterine seg-

ment. Accordingly, successful stretching-induced delivery can be carried out for a dead fetus.

Fetal membranes and muscles of the lower uterine

segment can accommodate stretching up to certain level,

most probably in the second trimester of pregnancy. Af- ter this, more stretching of membranes will lead to a

gradual increase in IL-8 concentration and collagenase

activity. These two factors have a marked effect on cervi-

cal ripening. An inhibitory system for IL-8 and collagenase activity

may also exist during pregnancy. Progesterone was found

to inhibit production and release of IL-8 in cervical tis- sueY Also, we have reported that urinary trypsin inhibi-

tor, which is excreted by the fetal urine in the amniotic fluid, inhibits cervical ripening induced by IL-8. 28 Uri-

nary trypsin inhibitor also decreased production of pros-

taglandin E 2 and prostaglandin F2~ , by uterine muscles (unpublished data). Thus a balance between the effect of stretching and the inhibiting system may be found until

term. Clearly more investigations are needed to verify the

exact role of stretching membranes in initiation of cervi-

cal ripening and labor.

REFERENCES

1. Artal R, Sokol R, Newman M, Burstein A, Stojkov J. The mechanical properties of prematurely and nonprematurely ruptured membranes. ANJ OBSTET Gyu~coc 1976;125:665-71,

Volume 174, Number 3 ~1 Maradny et al. 849 Am .l Obstet Gynecol

2. Alger LS, Pupkin MJ. Etiology of preterm premature rup- ture of the membranes. Clin Obstet Gyneco11986;29:758-70.

3. Parry-Jones E, Priya S. A study of the elasticity, and tension of fetal membranes and the relation of the area of the gesta- tional sac to the area of uterine cavity. BrJ Obstet Gynaecol 1976;83:205-12.

4. Okazaki T, Casey ML, Okita JR, MacDonald PC, Johnston JM. Initiation of human parturition, XII: biosynthesis and metabolism of prostaglandins in human fetal membranes and uterine decidua. AMJ OBSTET G'ZNECOL 1981;139:373-81.

5. Newman RB, Godsey RK, EllingsJM, Campbell BA, Eller DP, Miller MC III. Quantification of cervical changes: relation- ship to preterm delivery in multifetal gestation. AuJ OBSTET G'eNECOL 1991;165:264-71.

6. BuckinghamJC, McElin TW, Bowers VIVI, McVayJ. A clinical study of hydrmnnios. Obstet Gyneeol 1960;15:652-5.

7. Kanayama N, Fukamizu H. Mechanical stretching increases prostaglandin E 2 in cultured human ame.ion cells. Obstet Gynecol Invest 1989;28:123-6.

8. Kloeck FI~Jung H. In vitro release of prostaglandin from the human myometrium under the influence of stretching. A~J OBSTEr G~'ECOL 1973;115:1066-9.

9. Kanayama N, Terao T, Kawashima K, Fujimoto D. Collagen types in normal and prematurely ruptured amniotic mem- branes, ek,~J OBSTEr G~ECOL 1985;153:899-903.

10. Uldbjerg N, Ekman G, Malmstrom A, Olsson K, Uhnsten U. Ripening of the human uterine cervix related to changes in collagen, glycosaminoglycans, and collagenolytic activity. ~x{ J OBSrET G~ECOL 1983;147:662-6.

11. Skinner SJM, Campos GA, Liggins GC. Collagen content of human amniotic membranes: effect of gestation length and premature rupture. Obstet Gynecol 1981;57:487-9.

12. Junqueira LCU, Zugaib M, Montes GS, Toledo OMS, Krisz- tan RM, Shigihara KM. Morphological and histochemical evidence for the occurrence of collagenolysis and for the role of neutrophilic polyn~orphonuclear leukocyte during cervical dilatation. A~J OBSTEr GYNECOL 1980;138:273-81.

13. Osmers R, Rath W, Grill A. Origin of cervical collagenase during parturition. Ax~J OBSTET G'a,'ECOL 1992;166:1455-60.

14. Baggiolini M, Walz A, Kuukel SL. Neutrophil activating peptide-1/interleukin-8 a novel cytokine that activates neu- trophils.J Clin Invest I989;84:1045-9.

15. E1 Maradny E, Kanayama N, Halim A, Maehara K, Suminmto K, Terao T. Interleukin-8 induces cervical ripening in rab- bits. AM J OBSTET G~ECOL 1994; 171:77-82.

16. Laham N, Rice GE, Bishop GJ, Ransome C, Brennecke SR Interleukin 8 concentration in amniotic fluid and periph- eral venous plasma during human pregnancy and parturi- tion. Acta Endoerinol 1993;129:220-4.

17. Traunnan MS, Dudley DJ, Edwin SS, Collmer D, Mitchell MD. Amnion cell biosynthesis ofinterteukin-8: regulatioi~ by inflammatory cytokines. J Cell Physiol 1992; 153:38-43.

18. Kelly RW, Leask R, Calder AA. Choriodecidual production ofinterleukin-8 and mechanism of parturition. Lancet 1992; 339:776-7.

19. Barclay CG, Brennand JE, Kelly RW, Calder AA. lnterleu- kin-8 production by human cerx4x. AM J OBSTET G'ZNECOL 1993;169:625-32.

20. Romero R, Ceska M, Avita C, Mazor M, Behnke E, Lindely I. Neutrophii attractant/activating peptide-1/interleukin-8 in term and preterm parturition. A~I J O~STET G'a,'ECOL 1991; 165:813-20.

21. Romero R, Quintero R, NoresJ, et al. Amniotic fluid white blood cell count: a rapid and simple test to diagnose micro- bial invasion of the amniofic cavity and predict premrm delivery. [0007] 1991;165:821-30.

22. Butterworth BH, Geer IA, Liston WA, Haddad NG.Johnston TA. Immunocytochemical localization ofneutrophil etastase in term placenta decidua and myomerriurn in pregnancy- induced hypertension. BrJ Obstet Gynaeco11991:98:929-33

23. Liggins GC. Cervical ripening as an inflammatory process. In: Ellwood DA, Anderson ABM, eds. The cervix in preg- nancy and labor: clinical and biochemical investigations Edinburgh: Churchill-Lix~ngstone 1981:1-12.

24. Murphy G, ReynoldsJJ, Bertz U, Baggiolini M. Partial puri- fication of collagenase and gelatinase from human polymor- phonnclear teukocytes. BiochemJ 1982:203:209-21.

25. Kanayama N, Terao T. The relationship between gram~Iocyte elastase-Iike activity of cervical mucus and cervical matura- tion. Acta Obstet Gynecol Scand 1991:70:29-34.

26. Soeda Y. A study on tissue inhibitor of metalloproteinases (TIMP) during pregnancy and at delivery. Nippon Sanka Fujinka Gakkai Zasshi 1993;45:340-6.

27. Ito A, Imada K, Sato T, Kubo T, Matsushima K. Mori Y. Suppression of interleukin 8 production by progesterone in rabbit uterine c&rvlx. BiochemJ 1994:301:183-6.

28. Kanayama N, E1 Maradny E, Halim A, Maehara K. Kajiwara Y. Terao T. Urinary trypsin inhibitor suppresses premature cervical ripening. EurJ Obstet Gynecol Reprod Biol 1995: 60:181-6.

Availability of Joum,~AL back issues As a service to o u r subscribers , copies of back issues of the A~XaER[CAN JOURNAL OF

OBSTETRICS AND GYNECOLOGY for the p r e c e d i n g 5 years are m a i n t a i n e d a n d are available for pu rchase f rom the publ isher , Mosby-Year Book, Inc., a t a cost of $14.00 p e r issue. T h e follo~dng quan t i ty d iscounts are available: o n e f o u r t h off o n quant i t ies of 12 to 23 a n d o n e th i rd off o n quant i t i es of 24 or more . Please write to Mosby-Year Book, Inc., Subscr ip t ion Services, 11830 West l ine Indas t r i a l Drive, St. Louis, MO 63146-3318 or call (800)453-4351 or (314)453-4351 for i n f o r m a t i o n o n availability of par t i cu la r issues. If back issues are uavai lable f rom the publ i she l ; p h o t o c o p i e s of comple t e issues are available f rom UMI, 300 N. Zeeb Road, A n n Arbor, MI 48106-1346. Tel.: (313)761-4'700.