Functionally significant SNP MMP8 promoter haplotypes and preterm premature rupture of membranes (PPROM)

Functionally significant SNP MMP8 promoterhaplotypes and preterm premature rupture ofmembranes (PPROM)

Hongyan Wang1, Samuel Parry1, George Macones1, Mary D. Sammel2, Pedro E. Ferrand1,

Helena Kuivaniemi3, Gerard Tromp3, Indrani Halder4, Mark D. Shriver4, Roberto Romero3

and Jerome F. Strauss III1,*

1Center for Research on Reproduction and Women’s Health and 2Center for Clinical Epidemiology and Biostatistics,

University of Pennsylvania, Philadelphia, PA 19104, USA, 3Perinatology Research Branch, NICHD, Hutzel Hospital,

Detroit, MI 48201, USA and 4Department of Anthropology, The Pennsylvania State University, University Park,

PA 16802, USA

Received July 23, 2004; Revised and Accepted September 3, 2004

Matrix metalloproteinase 8 (MMP8), an enzyme that degrades fibrillar collagens imparting strength to the fetalmembranes, is expressed by leukocytes and chorionic cytotrophoblast cells. We identified three singlenucleotide polymorphisms (SNPs) at 2799C/T, 2381A/G and 117C/G from the major transcription startsite in the MMP8 gene, and determined the functional significance of these SNPs by analyzing their impactupon MMP8 promoter activity and their association with preterm premature rupture of membranes(PPROM). The minor alleles 117 (G) and 2381 (G) were in complete linkage disequilibrium. A promoter frag-ment containing the three minor alleles had 3-fold greater activity in chorion-like trophoblast cells (BeWo,JEG-3 and HTR-8/SVneo) compared with the major allele promoter construct. Electrophoretic mobility shiftassays revealed differences in BeWo nuclear protein binding to oligonucleotides representing the 2381and 2799 SNPs, suggesting that the minor alleles have reduced transcription factor binding. A case–controlstudy of African-American neonates using allele-specific primers revealed a statistically significant asso-ciation between the three minor allele haplotype, which displays the highest MMP8 promoter activity introphoblast cells, with PPROM with an odds ratio (OR) of 4.63 (P < 0.0001), whereas the major allele promoterappeared to be protective (OR 5 0.52, P < 0.0002). None of the minor alleles were individually associated withPPROM. These findings demonstrate the functional significance of SNP haplotypes in the MMP8 gene andassociations with obstetrical outcomes.

INTRODUCTION

Extensive extracellular matrix (ECM) remodeling is animportant process in several phases of human parturition,including cervical ripening, fetal membrane rupture and pla-cental detachment from the uterus (1–3). The fetal membranesare a complex multilaminate tissue composed of the amnionand chorion. These two closely adherent tissues consist ofseveral cell types, including epithelial cells, mesenchymalcells and cytotrophoblast cells. The tensile strength of thefetal membranes depends on the integrity of these cells andtheir associated ECM (4–6). Preterm premature rupture of

the membranes (PPROM), defined as spontaneous rupture ofthe membranes before 37 weeks of gestation, is a significantobstetrical complication associated with increased risk forintrauterine infection (7). PPROM is the leading identifiablecause of preterm delivery and its complications (8). Histori-cally, obstetricians have attributed fetal membrane rupture tophysical stress. However, familial clustering and ethnic differ-ences in the incidence of PPROM (9), which cannot beaccounted for by socioeconomic status (10), suggest thatgenetic factors contribute to the risk of PPROM.

The matrix metalloproteinases (MMPs) are a family ofproteolytic enzymes that degrade the main protein components

Human Molecular Genetics, Vol. 13, No. 21 # Oxford University Press 2004; all rights reserved

*To whom correspondence should be addressed at: Center for Research on Reproduction and Women’s Health, University of Pennsylvania, 1354 BRBII/III, 421 Curie Boulevard, Philadelphia, PA 19104, USA. Tel: þ1 2158980147; Fax: þ1 2155735408; Email: [email protected]

Human Molecular Genetics, 2004, Vol. 13, No. 21 2659–2669doi:10.1093/hmg/ddh287Advance Access published on September 14, 2004

by guest on Novem

ber 8, 2015http://hm

g.oxfordjournals.org/D

ownloaded from

http://hmg.oxfordjournals.org/

of the ECM (11). Consequently, MMPs are widely assumed toplay a central role in the remodeling of the cervical and fetalmembrane ECM throughout gestation and preceding parturi-tion. Also, we previously reported that polymorphisms in theMMP1 and MMP9 promoters that increase promoter activityare associated with PPROM (12,13), suggesting that geneticvariation in MMP genes contributes to the risk of adverseobstetrical outcomes.

MMP8 (also known as collagenase-2 or neutrophil colla-genase) is a member of the MMP family. It is a glycoproteinthat is synthesized as a zymogen. Activation of MMP8requires autolytic removal of 80 amino acids from the N-termi-nus (14,15). MMP8 cleaves fibrillar collagens, such as collagentypes I, II, III, V and XI, as well as non-fibrillar collagens,including collagen types IX, XII and XIV (16). Cleavage ofthe triple helical collagen molecule by MMP8 changesthe stability and solubility properties of the collagen, resultingin denaturation in cleavage products that are subsequentlydegraded by other MMPs (so-called gelatinases) (17–21).

MMP8 is released from leukocytes during chemotacticstimulation in vitro and in response to inflammatory condi-tions in vivo (22). Although MMP8 was thought to be expres-sed exclusively by neutrophils, its expression has recentlybeen detected in a number of other cell types, includingchondrocytes, fibroblasts, corneal epithelial cells, endothelialcells, smooth muscle cells and cytotrophoblasts of the chorion(23–29), which raises the possibility of its involvement infetal membrane rupture. Indeed, MMP8 concentrations arehigh in the amniotic fluid of patients with intra-amniotic infec-tion, preterm labor and PPROM (30,31). Elevated amnioticfluid MMP8 levels are also strongly associated with neonataldeath and other adverse neonatal outcomes (32).

We undertook the present study to determine whether thereis genetic polymorphism in the MMP8 promoter, whetherthe polymorphism has functional significance and whether itis associated with PPROM. The association study was con-ducted in an African-American population because of thehigher incidence of PPROM in this ethnic group (8–10). Inthis report, we demonstrate for the first time that there arefunctionally significant polymorphisms in the MMP8 gene;that the impact of these variants on MMP8 promoter functionis influenced by cell context and that there is an associationbetween the minor allele haplotype with increased promoteractivity in chorion-like cytotrophoblast cells and PPROM inan African-American population.

RESULTS

Identification of the MMP8 transcription start site

50 RACE analysis indicated that the major transcription startsite in the MMP8 gene is at 22 relative to the transcriptionstart site designated in the sequence deposited in GenBank(accession no. AF059679). Of seven sequenced PCR products,six ended at nucleotide 22 and one at þ1. Transcription startsites do not share extensive sequence homology, but in �50%of transcription initiation sites, the first base of the mRNA isan A flanked on either side by pyrimidines. The start site forthe MMP8 mRNA identified in our 50 RACE is at an Aflanked 30 by a C. Most promoters have a TATA box, usuallylocated about 25 bp upstream from the transcription startpoint. A TATA-like sequence, TTTAAA, is located 26–31 bpupstream of the putative transcript start site. A CAAT box,another characteristic promoter feature, is located at basepair 249 to 252. The structure of the MMP8 promoter isshown in Figure 1. In the process of performing the 50 RACE,we identified a novel MMP8 mRNA splice variant, which isthe result of the insertion of a 67 bp sequence from a crypticexon (exon 1a) located in the 2 kb intron between exons 1 and2 (see later).

Identification of SNPs in the MMP8 promoter

Analysis of promoter fragments amplified by PCR from32 unrelated African-Americans identified single nucleotidepolymorphisms (SNPs) at positions 2799C/T (minor ‘T’allele detected in 11 subjects), 2381A/G (minor ‘G’ alleledetected in seven subjects) and þ17C/G (minor ‘G’ alleledetected in seven subjects) with an allele frequency .5%.The allele frequencies and carrier rates of SNP haplotypeswere determined from genotypes of 216 African-Americancontrols (Table 1). The 2381 and þ17 alleles appeared tobe in complete linkage disequilibrium. When a subjecthad the minor ‘G’ allele at 2381, the þ17 position was invari-ably the minor ‘G’ allele. Thus, there are four possible SNPhaplotypes including the major allele promoter 2799C/2381A/þ17C, one minor allele promoter 2799T/2381A/þ17C, two minor alleles promoter 2799C/2381G/þ17Gand three minor alleles promoter 2799T/2381G/þ17G.

Figure 1. Genomic structure of the promoter and the adjacent sequences of the human MMP8 gene. The promoter of the gene is depicted as a thick line,the exons as striped boxes and the introns as thin lines. SNPs C/T (2799), A/G (2381) and C/G (þ17), as well as CAAT box (249 to 252), putativeTATA box (TTTAAA) (228 to 233) and the transcription start site are shown. The translation start site is indicated in exon 1. Between exons 1 (ends atþ198) and 2 (begins at þ199), there are two other potential exons of 91 bp (close to exon 2) and 67 bp (next to exon 1) within intron 1.

2660 Human Molecular Genetics, 2004, Vol. 13, No. 21

by guest on Novem



ownloaded from


Functional significance of the MMP8 promoter alleles

We examined the activities of MMP8 promoter fragmentscontaining different experimentally detected SNP haplotypestransfected into different host cells that express the MMP8gene endogenously (Table 2). Each promoter fragmentdemonstrated activity that was substantially greater thanthe pGL3 empty vector. Moreover, when the promoter frag-ments were cloned into the pGL3 vector in the opposite orient-ation, all promoter function was lost and relative luciferasevalues were similar to those found with the empty vector(data not shown). Cell host-dependent differences in MMP8promoter activity were observed. A promoter fragmentcontaining the three minor alleles (2799T/2381G/þ17G)showed 2–3-fold greater activity in the three chorion-like cytotrophoblast cells, BeWo, HTR-8/Svneo and JEG-3cells, compared with the major allele promoter construct(2799C/2381A/þ17C) and the other promoters with one ortwo minor alleles. However, in U937 leukocytes cells, theminor allele promoter had �34% of the activity of themajor allele promoter. These promoters had similar activitiesin THP-1 monocytes/macrophage cells.

We analyzed the binding of nuclear extract (NE) proteinsprepared from BeWo, THP-1 and U937 cells to oligonucleo-tides representing the different 2381 and 2799 alleles byelectrophoretic mobility shift assays (EMSA) (Fig. 2). Specificbinding of proteins was identified through cross-competitionusing unlabeled major allele, minor allele and non-specificoligonucleotide probes. The binding pattern for proteinsfrom the different cell lines to the 2381A/G and 2799C/Toligonucleotides differed. NE from BeWo cells producedone complex (arrow) with the 2381A probe, and showed signi-ficantly less binding to the minor allele 2381G probe. Similarly,decreased binding was observed with NE from U937 cellswhen we compared the binding to the major 2381A alleleprobe (arrow) and the minor 2381G allele probe (arrow).Weak binding to the common 2381A allele probe (arrow)and little binding to the minor 2381G allele probe was foundwith THP-1 cell NE.

NE from BeWo cells produced two complexes on binding tothe common 2799C allele probe and the minor 2799T alleleprobe (Fig. 2). The binding to the lower mobility complex wasequivalent (arrow), but the higher mobility complex showedhigher affinity for the common 2799C allele probe (arrow).Only one complex was formed by extracts from U937 cellswith similar binding (arrow) to the common 2799C and the

minor 2799T allele probes. No specific binding was observedwith NE prepared from THP-1 cells. Collectively, the EMSAresults revealed differences in nuclear protein binding to the2381A/G and 2799C/T oligonucleotides,suggesting that the minor alleles have reduced BeWo cell tran-scription factor binding. The nature of the proteins binding tothe different oligonucleotides remains to be determined.

Identification of a new MMP8 mRNA splice variantand its expression pattern

The nucleotide sequence of the MMP8 cDNA (GenBankaccession no. NM_002424) encodes a protein of 467 aminoacids, with a secretory signal sequence of 20 residues followedby the prodomain of 80 residues. In the course of performing50 RACE to identify the MMP8 transcription start site, weidentified three different cDNAs from THP-1 and U937 cellsby RT–PCR using an antisense primer corresponding to bp2392 to 2413 of the MMP8 cDNA sequence deposited inGenBank together with the universal primer from the ClontechSMART RACE kit as the sense primer. A number of PCRproducts were recovered and sequenced. Besides fragmentsmatching the MMP8 cDNA sequence in GenBank, and apreviously reported splice variant of the MMP8 transcriptwith an additional 91 bp insertion encoded by a cryptic exonin the intron between exons 1 and 2 (33), we found a newsplice variant containing a 67 bp insertion, also resultingfrom a cryptic exon in the intron between exons 1 and 2(Fig. 3). The þ91 and þ67 bp insertions separate bp 173and 174 of the coding sequence and both inserted sequencesin the first intron are flanked by donor and acceptor splicejunctions that obey the GT/AG rule (33,34). The þ91 andþ67 bp splice variants encode a short peptide that startswith the reference sequence initiation codon and containsthe leader sequence of the reference MMP8 protein. A newopen reading frame beginning with Met 85 of the referencesequence would yield a protein lacking the signal sequence,and thus would be expected to be an intracellular enzyme (33).

Quantitative RT–PCR was used together with primer pairsfor amplification of each splice variant to define the MMP8mRNA expression patterns in human chorion, BeWo cellsand the U937 and THP-1 cell lines (Fig. 4). Chorion wasrelatively enriched with all the MMP8 splice variants, butthe 0 bp form predominated. BeWo cells express equivalentlevels of the 0 bp variant and þ67 bp variant without detect-able þ91 bp splice variant. The MMP8 mRNA variantswere expressed in a similar pattern in THP-1 cells, whereasin U937 cells the 0 bp insertion predominated.

To determine whether the þ67 and þ91 bp splice variantsare translated yielding detectable amounts of protein, weperformed western blot analysis on chorion tissue extractsand the extracts from BeWo, U937 and THP-1 cells (Fig. 5).The MMP8 proenzyme has a molecular weight of �85 kDaand the active enzyme is 64 kDa. The protein encoded by theinsertion splice variants is theoretically 9 kDa smaller than theproenzyme. When equivalent amount of total protein fromamnion, chorion, BeWo cells, THP-1 cells and U937 cellswere loaded on gels with amnion extract as a negative control,chorion samples contained abundant MMP8 proenzyme; BeWocells contained active MMP8 with no detectable proenzyme;

Table 1. MMP8 promoter SNP allele frequencies

SNP African-American (%)

Controls (n ¼ 216) PPROM (n ¼ 168)

2799C 69.8 75.9T 30.2 24.1

2381A 92.6 94.2G 7.4 5.8

þ17C 92.6 94.2G 7.4 5.8

Human Molecular Genetics, 2004, Vol. 13, No. 21 2661

by guest on Novem



ownloaded from


and THP-1 and U937 cells had only trace amounts of activeform (Fig. 5). In no case was a 76 kDa immunoreactiveprotein detected. These findings demonstrate that chorion isenriched in MMP8 and that the proteins encoded by þ91and þ67 bp splice variants are either not very abundant orrapidly processed to active enzyme.

The minor MMP8 promoter allele haplotype isassociated with risk of PPROM and the common alleleMMP8 haplotype is protective

To determine if the functional differences in MMP8 promoteractivity for the different SNP haplotypes could contributeto differences in ECM degradation, we performed a case–control study to test the association between the MMP8SNPs and PPROM. The study was conducted in an African-American population because PPROM is 2–4 times moreprevalent in this ethnic group (8–10). The study was alsofocused on the genotype of the offspring based on the hypo-thesis that the genotype of the extraembryonic tissues (fetalmembranes) represents the primary determinant of risk ofpremature rupture of the membranes.

The demographic characteristics of the 216 controls, neo-nates born at term from normal pregnancies, and the 168cases, neonates from pregnancies complicated by PPROM,are shown in Table 3. There were no significant differencesin maternal age, gravidity and parity, but the length of gesta-tion and birth weight were significantly lower in the PPROMgroup, as expected.

Our case–control study of neonatal genotypes and PPROMrevealed no significant association between the individualSNPs and PPROM (Table 1). However, an analysis based oncarriage of the three minor allele haplotype (2799T, 2381G,þ17G) that confers increased MMP8 promoter activity incytotrophoblasts demonstrated that it was significantly morefrequent in PPROM (8.04% of cases compared with 1.85%of controls) with an odds ratio (OR) ¼ 4.63 [P , 0.0001;95% confidence intervals (CI) 2.01, 11.94] (Table 4). Conver-sely, homozygosity for the major allele MMP8 promoter hap-lotype appeared to be protective against PPROM (P , 0.0002;CI: 0.362, 0.751; Table 4).

As the urban African-American population from which oursubjects were drawn is heterogeneous, we performed analyses

to determine if population stratification could have affectedour findings. The impact of population admixture was alsoconsidered for subjects for whom genotypes were availableto evaluate (171 cases of PPROM and 208 controls). Forthis analysis, estimates of ancestry [proportion/probabilityAfrican-American (Nigeria, Sierra Leone and CentralAfrican Republic) versus European (combined British, Irish,German and Spanish populations)] were made for cases andcontrols and for each subject, and the impact of the geneticadmixture was evaluated using multiple logistic regressionanalysis. For the cases and controls, there was no significantdifference in ancestry using a dihybrid model (% Africanancestry: cases, 0.847 + 0.141; controls, 0.827 + 0.149,P ¼ 0.487). There was also no evidence of confounding dueto ancestry in the multiple logistic regression analysis, as theOR estimates decreased by at most 3% for the associationfor the three minor SNP haplotype and PPROM (adjustedOR ¼ 4.48; CI: 1.90, 11.75).

DISCUSSION

The ECM is recognized as a key regulatory component in cell-ular physiology, providing an environment for cell division,differentiation, migration and invasion, and in some cases, cellsurvival and cell death (35). ECM turnover and homeostasisare highly regulated and the catabolism is due, in part, to theaction of a specific class of proteolytic enzymes known asMMPs. Currently, the MMP family encompasses four broadclasses: collagenases, gelatinases, stromelysins and membrane-type enzymes. Collagenases, which degrade the fibrillarcollagens, include MMP1, MMP8 and MMP13. The ECM offetal membranes and uterus (cervix and lower uterinesegment), in particular collagen and elastin, is degraded in prep-aration for delivery. Each of the three collagenases may partici-pate in this process with the enzymes originating from differentcell types. In the amnion, epithelial and mesenchymal cells aresources of MMP1 and MMP13, whereas the cytotrophoblastof the chorion and invading leukocytes are the primary sourceof MMP8. Chorion tissue, as shown here and in our previouswork, is a rich source of MMP8. Because the chorion is adherentto the amnion, chorion-derived MMP8 would presumably haveready access to the fibrillar collagens of the amnion.

Table 2. Genotype-dependent MMP8 promoter activities

Construct and genotype Cell host

Chorion-related cells Macrophage Leukocytes

BeWo HTR-8/SVneo JEG-3 THP-1 U937

pGL3 basic 1.00 + 0.08a 0.30 + 0.02a 0.12 + 0.01a 0.17 + 0.05a 0.11 + 0.02a

2799C/2381A/þ17C 4.86 + 0.31b 9.62 + 0.61b 0.87 + 0.40b 4.02 + 0.63b 1.34 + 0.18b

2799T/2381A/þ17C 5.53 + 0.38b 9.44 + 0.82b 0.79 + 0.03b 5.13 + 0.77b 1.32 + 0.15b

2799C/2381G/þ17G 7.11 + 1.12b 12.35 + 0.89b 1.04 + 0.05b 4.41 + 0.25b 0.85 + 0.14b

2799T/2381G/þ17G 14.66 + 1.12c 21.24 + 1.76c 2.40 + 0.11c 4.61 + 0.52b 0.46 + 0.06a

The four experimentally determined possible haplotypes of the MMP8 promoter were cloned into the pGL3 basic vector. The relative Photinusluciferase activities + SE (n ¼ 4 separate experiments for each cell host), standardized to Renilla luciferase of all constructs including the emptycontrol vector were compared using Tukey–Kramer test. Values with different letters are significantly different (P , 0.05) from each other withineach cell host.


by guest on Novem



ownloaded from


The significance of the different MMP8 mRNA splicevariants with respect to ECM catabolism remains to be clari-fied. The splice variants containing insertions of crypticexons encoded in the intron between exons 1 and 2 appearto be intracellular proteins, on the basis of the absence of asignal sequence, but their role in intracellular proteolysis isa matter of speculation (33). Our findings demonstrate thatin the chorion, the MMP8 transcript encoding the enzymewith an intact leader sequence predominates, which is con-sistent with a role for cytrophoblast-derived MMP8 in fetalmembrane ECM metabolism.

Because of the important role of MMPs in ECM turnover,we have been interested in the question of whether geneticvariation in MMP promoter regions could contribute tofamilial clustering of preterm birth as well as differences inthe incidence of preterm birth among various ethnic groups.Our past investigations have indicated that specific poly-morphisms in the MMP1 and MMP9 promoters are associatedwith risk of PPROM in African-Americans (12,13). Thepresent study suggests that the haplotype with three minoralleles of the MMP8 promoter is also associated withPPROM. Interestingly, this haplotype conferred greater

Figure 2. EMSA of nuclear proteins from different cell types: (A) BeWo; (B) THP-1; (C) U937 to the oligonucleotides containing MMP8 2381A/G alleles (upperpictures) and 2799C/T alleles (lower pictures). Every left seven lanes show NE binding with wild-type2381 (A) or2799 (C) probe; every right seven lanes show NEbinding with mutant 2381 (G) or 2799 (T) probe; NE: nuclear extract; þ: present; 2 : empty; : increased amount of competitor or cross-competitor probes.


by guest on Novem



ownloaded from


promoter activity only when the constructs were introducedinto cells resembling the chorion cytotrophoblasts. In U937and THP-1 cells, this haplotype did not show increasedpromoter activity, and it was in fact reduced in U937 cells.The basis for the cell-specific responses is not entirely clear.However, EMSA with NE derived from BeWo cells didreveal a complex with the 2799C/T oligonucleotide thatwas not detected in EMSAs with NE from U937 and THP-1cells. The minor allele oligonucleotide was bound withlower affinity, suggesting that the protein producing thishigher mobility complex could be a transcription factorwhose binding determines promoter function. Because themutant allele appeared to have lower affinity for this protein,it is possible that the protein is a transcription repressor andthat reduced binding results in increased promoter activity.Notably, the haplotype with minor alleles of the three SNPswas significantly more common in PPROM. This raisesthe possibility of interactions among different cis elements.Unfortunately, there have been no prior functional analysesof the human MMP8 promoter, and the cis elements importantfor MMP8 transcription and the transcription factors that bind

to them have not been identified. This is, to the best of ourknowledge, the first exploration of human MMP8 promoterfunction.

To determine the possible identities of the proteins respon-sible for the high mobility complex seen only with BeWo cellNE, we searched various databases for transcription factorbinding sites. This search indicated that the 2799 oligo-nucleotide with the SNP in the middle contains a potentialbinding site for CBG-02 protein (36), whereas 2381 oligo-nucleotide contains a potential binding site for GATA-1(http://www.genomatix.de and http://www.ifti.org). However,at this juncture it is a matter of speculation as to whetherCBG-02, GATA-1 or other factors play roles in modulatingMMP8 transcription, especially because the two transcriptionfactors previously noted have not yet been reported to beexpressed in chorion trophoblast cells.

There are a number of important caveats that must be keptin mind when interpreting the results of association studies.First, the findings may be influenced by population stratifi-cation. This is a particularly important concern when thereis heterogeneity in the population under investigation, as is

Figure 3. Nucleotide sequence and genomic structure of MMP8 around the 67 bp insertion (exon 1a). (A) Nucleotide sequence of the 50 350 bp of the MMP8cDNA and the sequence of the 67 bp insertion. The nucleotide sequence of the 67 bp insertion is shown in the box. Numbering follows the published sequence.The potential stop codon TGA is indicated. (B) Genomic organization of MMP8 gene encodes three alternatively spliced transcripts. The schematic shows spli-cing patterns that generate MMP8 and MMP8 with 67 bp insertion transcripts. Open boxes represent exons and lines represent introns. Numbers indicatethe positions in the MMP8 cDNA. Consensus splice donor/acceptor sites are underlined.


by guest on Novem



ownloaded from


the case with African-Americans. Recruitment of cases andcontrols from the same communities and recruiting a largenumber of subjects mitigates this concern to some extent.Moreover, an analysis of potential population stratificationusing 29 markers selected for significant differences amongAfrican and European populations revealed no evidence forstratification influencing the major conclusions of this study.Second, association does not mean linkage and it is possiblethat the MMP8 SNP haplotype found to be associated withPPROM is also in linkage disequilibrium with another SNPor mutation in a gene other than MMP8. MMP8 lies on

chromosome 11 in a cluster of other MMP genes includingMMP1, 3, 7, 10, 12, 13, 20 and 26. Thus, it is possible thatour genetic study implicates an MMP other than MMP8. Aswe had previously studied a polymorphism in the MMP1 pro-moter and PPROM, we examined the possibility that the21607 2G polymorphism associated with PPROM was inlinkage disequilibrium with the MMP8 minor allele haplotype.It was not, making it unlikely that the significant associationwe observed between the MMP8 minor allele promoter haplo-type and PPROM is due to a MMP1 promoter variation. Third,our findings in an African-American population may not begeneralizable to other ethnic groups.

In conclusion, we have identified three SNPs in the MMP8gene 50 region and a haplotype that confers increased MMP8promoter activity in chorion-like cytotrophoblast-like celllines. This haplotype is associated with increased risk ofPPROM. Collectively, these findings suggest that genetic vari-ation at the MMP8 locus can contribute to adverse eventslinked to ECM breakdown.

MATERIALS AND METHODS

Subjects

Subjects in this study were African-American women andtheir neonates receiving obstetrical care at the Hospital ofthe University of Pennsylvania, Philadelphia, and HutzelHospital, Detroit. The study was approved by the respectiveInstitutional Review Boards, and written informed consentwas obtained from mothers before collection of the samples.

Figure 5. Western blot analysis of MMP8 in amnion tissue, chorion tissue,BeWo cells, THP-1 cells and U937 cells. (A) Shorter exposure time; (B)longer exposure time; (C) b-actin protein to assess equality of proteinloading. No detectable MMP8 protein in amnion samples (negative control).Chorion samples contain abundant 85 kDa (latent form) and trace amount of64 kDa (active form) MMP8 protein. BeWo cells contain 64 kDa proteinbut no detectable 85 kDa protein. THP-1 and U937 cells have similar traceamounts of 64 kDa protein only. A non-specific band appears at 60 kD inamnion and, to a lesser extent, chorion.

Figure 4. Quantitative RT–PCR amplification of different splice variants of human MMP8 gene with exclusive primer pairs in chorion tissue and following celltypes. (A) THP-1 cells; (B) U937 cells; (C) BeWo cells; (D) human chorion; real-time PCR (red line: þ91 bp; green line: þ67 bp; yellow line: 0 bp insert) areshown above and the gel analysis of PCR products below.


by guest on Novem



ownloaded from


Control samples (n ¼ 216) were obtained from singletonpregnancies delivered at term of mothers with no priorhistory of PPROM or preterm labor. Cases of PPROM(n ¼ 168) were defined as neonates from pregnancies compli-cated by rupture of membranes prior to 37 weeks of gestation.The diagnosis of membrane rupture was based on pooling ofamniotic fluid in the vagina, amniotic fluid ferning patternsand a positive nitrazine test. Patients with multiple gestations,fetal anomalies, trauma, connective tissue diseases andmedical complications of pregnancy requiring induction oflabor were excluded. The chi-square test was used to deter-mine the significance of the association between MMP8promoter alleles and PPROM. The OR and 95% CI werealso determined.

Mapping of the MMP8 transcription start site

Mapping of the transcription start site of the MMP8 genewas a prerequisite for studies of MMP8 promoter function.We employed SMART technology (37) for mapping thetranscription start site, based on direct sequencing of theSMARTTM (switching mechanism at 50 end of RNA template)(38) RACE products. Total RNA or mRNA was preparedfrom cultured THP-1 and U937 cells with TRIzolw or usingthe poly(A) tract. Using 50 RACE CDS primer and Power-ScriptTM reverse transcriptase from the SMARTTM RACEcDNA amplification kit (Promega), cDNA was synthesizedin a 10 ml volume according to the manufacturer’s instruc-tions. The cDNA was then diluted with Tricine–EDTAbuffer into 100 ml, and 2.5 ml of the diluted sample was

used as a template for a 50 ml PCR in the presence of 5 ml0.4 mM universal primer mix (from the kit) and 1 ml of20 mM gene-specific primer (50-AGGTCAAGTTAGTGCGTTCCCAC-30). Advantagew 2 enzyme mixture and reagents(Clontech) were used. Amplification was performed in amodel 9600 thermal cycler (Applied Biosystems) as follows:five cycles of 5 s at 958C and 3 min at 728C, and then fivecycles of 5 s at 948C, 10 s at 708C and 3 min at 728C, followedby 27 cycles of 5 s at 948C 10 s at 688C and 3 min at 728C.PCR products were purified on QIAquickTM spin columns(QIAgen) and subjected to sequencing using the gene-specificprimer as a sequencing primer. The position of the transcrip-tion start site was detected at the point of abrupt loss ofsequence identity between the RACE product and the 50

genomic sequence.

Identification of SNPs and genotyping of theMMP8 promoter

DNA was isolated from umbilical cords, cord blood or neo-natal cheek swabs by digestion with proteinase K and extrac-tion with conventional reagents or by an alternative isolationmethod using the BioRobot 9604 (QIAgen) (39). On thebasis of the human MMP8 promoter sequence deposited inGenBank (accession no. AF059679), a promoter fragmentwas amplified by PCR with a forward primer sequenceof 50-CTGTTGAAGGCCTAGAGCTGCTGCTCC-30 (corres-ponding to bp 2872 to 2846) and a reverse primer50-CATCTTCTCTTCAAACTCTACCC-30 (corresponding tobp þ74 to þ96) yielding a 968 bp product. PCR wasperformed in a 50 ml reaction volume containing 100 nggenomic DNA, 0.5 pmol of each primer, 0.2 mM dNTPs,1�reaction buffer, 2.5 mM MgCl2 and 2.5 units AmpliTaqGoldTM DNA polymerase (Perkin–Elmer) in a 9600 GeneAmp PCR thermal cycler (Perkin–Elmer Life Sciences).After initial denaturation at 948C for 5 min, PCR was per-formed for 35 cycles of denaturation at 948C for 45 s anneal-ing at 568C for 45 s and extension at 728C for 1 min followedby a final 10 min elongation at 728C.

MMP8 promoter fragments were amplified from 32 unre-lated individuals and subjected to DNA sequence analysis toidentify polymorphisms. The PCR products were cloned intothe TOPO TA cloning vector (PCR2.1-TOPO, Invitrogen)and four to five separate clones derived from each subject’sPCR amplification were sequenced. Three SNPs (2799C/T,2381A/G and þ17C/G) were identified in more than twosubjects in this screen. These SNPs were subjected to furtheranalysis.

Genotypes of the MMP8 promoter were determined by twodifferent methods: restriction fragment length polymorphism(RFLP) and GeneScan analysis. For RFLP, the PCR productswere digested with either restriction endonuclease Sfc I (NewEngland Biolabs) for genotyping of the 2799C/T SNP orrestriction endonuclease Dde I (New England Biolabs) forthe þ17C/G SNP. The digested PCR products were separatedin 1% agarose gels. The 2799C allele is digested by Sfc Iyielding fragments 894 and 74 bp; whereas the 2799T alleleis not cleaved by Sfc I. The þ17C allele is digested by Dde Iyielding fragments 799, 90 and 79 bp; whereas the þ17Gallele yields fragments of 878 and 90 bp.

Table 3. Demographic characteristics and clinical outcomes of indexpregnancies

Controls(n ¼ 216)

PPROM(n ¼ 168)

P-value

Maternal age (years) 26 + 4.6 25 + 5.8 0.63Gravidity 3.1 + 1.9 3.4 + 2.1 0.34Parity 1.4 + 1.5 1.5 + 1.6 0.55Gestational age at delivery (weeks) 39.1 + 1.3 31.9 + 2.8 ,0.0001Birth weight (g) 3309 + 483 1937 + 502 ,0.0001

Values presented are the means + SD.

Table 4. Frequencies of the triple minor and triple major SNP MMP8 promoterhaplotypes and PPROM

MMP8 promoter genotype Controls (n ¼ 216) PPROM (n ¼ 168)

No. % No. %

Haplotype2799T/2381G/þ17G

8 1.9 27 8.0

P-value ,0.0001Odds ratio 4.6395% Confidence interval 2.01–11.94

Haplotype2799C/2381A/þ17C

361 83.6 244 72.6

P-value ,0.0002Odds ratio 0.5295% Confidence interval 0.36–0.75


by guest on Novem



ownloaded from


For GeneScan, the purified 968 bp PCR products wereanalyzed with an ABI 3100 genetic analyzer using GeneScansoftware in order to genotype the SNP at 2381A/T and toverify the results of RFLP analysis. We used HPLC-purifiedprimers with a sequence of 50-AGCCAGAGACTCAAGTGGGAGACTACCATGCAGAGCC-30 (36 nt) for genotypingthe 2799C/T SNP; 50-CTCCACATACAATGAGGGAGG-30

(21 nt) for 2381A/G SNP; and 50-GCTGTGAGTGACACATGATGCTGTGAAC-30 (28 nt) for þ17C/G SNP. TheRFLP results for the 2799C/T and þ17C/G SNPs werecompletely concordant with the GeneScan analysis.

We also designed allele-specific primers based on the 2799and þ17 SNPs to directly determine the haplotypes of thestudy population. The two forward primers (A and B) areclose to 2799 SNP and only varied at the last nucleotide witheither the major 2799 ‘C’ or the minor 2799 ‘T’:50-AGTGGGAGACTACCATGCAGAGCC/T-30. The tworeverse primers (C and D) are close to þ17 SNP and onlyvaried at the last nucleotide with either the major þ17 ‘C’or the minor þ17 ‘G’: 50-TTCCCTGGCGAGCACCCTGAC/G-30. We used one set of four pairs of primers (AC,AD, BC and BD) to amplify DNA samples using conditions:5 min denaturation at 948C, followed by 24 cycles of 948C,1 min; 658C, 1 min and 728C, 1 min, and finally 10 minelongation at 728C.

Assessment of population structure

As many African-American populations have substantialadmixture and this admixture is not evenly distributedthroughout the population (40), there is the chance that someof the observed association could be the result of admixturestratification. To control for the possibility that admixture stra-tification may be the source of the association, we have typed29 ancestry informative markers (AIMs), which are particu-larly useful for calculating gene flow between West Africanand West European populations (41). These markers werethen used to calculate the individual biogeographical ancestry(BGA) levels of the persons in the study in the context of thetwo primary parental populations (West African and WestEuropean) using parental allele frequencies (42) and themaximum likelihood as first described in Hanis et al. (43).These BGA estimates were then used as conditioning variablesin the logistic regression analyses to control for any effectsthat admixture stratification could be having on the phenotype.This secondary analysis was conducted using subject leveldata, where the subject was considered to have the MMP8 hap-lotype of interest (exposed) if they were heterozygous orhomozygous for the haplotype. Next, using logistic regres-sion methods, the unadjusted OR for the MMP8, PPROMassociation was estimated. Next, adjusted OR estimates werecomputed by incorporating the admixture estimates intothe model. The resulting OR is an average of the MMP8,PPROM association, over subjects with like genetic profiles.

Construction of promoter-reporter plasmids

To determine whether the 2799C/T, 2381A/T and þ17C/GSNPs influence transcription of the MMP8 gene, we obtained

a 968 bp fragment from 2872 to þ96 bp of the MMP8promoter, amplified using forward and reverse primers withthe indicated sequences (forward primer: 50-CTGTTGAAGGCCTAGAGCTGCTGCTCC-30 reverse primer: 50-CATCTTCTCTTCAAACTCTACCC-30). A mutagenesis kit (Stratagene)was used to create the targeted genotypes with a uniformbackbone sequence. Promoter fragments containing eachof the haplotypes that we identified (major allele, 2799C/2381A/þ17C; one minor allele, 2799T/2381A/þ17C; twominor alleles, 2799C/2381G/þ17G and the three minoralleles: 2799T/2381G/þ17G) were cloned into the pGL3

vector (Promega), which contains the firefly luciferase geneas a reporter. The DNA sequences of the promoter constructswere confirmed prior to use and three different plasmidpreparations for each construct were tested.

Cell culture and transfection

BeWo, HTR-8/SVneo and JEG-3 cells were cultured inDulbecco’s modified Eagle’s medium (DMEM) and THP-1and U937 cells were cultured in RPMI 1640 medium. Themedia were supplemented with 10% fetal bovine serum andantibiotics (100 IU/ml penicillin G, 100 IU/ml streptomycinsulfate, 0.25 mg/ml amphotericin B; Gibco/BRL). All cellswere maintained at 378C in a water-saturated atmosphereunder 5% CO2 in air.

For transfection, 10 � 105 BeWo cells, 50 � 105 HTR-8/SVneo cells, 30 � 105 JEG-3 cells, 80 � 105 THP-1 cellsand 5 � 105 U937 cells were seeded in individual wells of a12-well culture plate. Cells were transfected using FuGENETM

6 transfection reagent (Roche) with 0.5 mg of the pGL3 vectorcontaining a 968 bp MMP8 promoter fragment coupled to thefirefly luciferase reporter gene. In each transfection, 25 ngpRL-TK (Promega), a control plasmid expressing Renillareniformis luciferase, was used to correct for transfectionefficiency. We changed the medium 12 h post-transfectionand continued the culture for an additional 24–36 h beforecollecting cells for the luciferase assays.

Luciferase assay and statistical analysis

After 36–48 h culture, the transfected cells were broken usinglysis buffer and 20 ml aliquots of supernatant were thenassayed for luciferase activity using the Dual-LuciferaseReporter Assay System (Promega) in a luminometer (LumatLB 9507, Berthold). Promoter activities were expressed asthe ratio between Photinus luciferase and Renilla luciferaseactivities. Significant differences in activities among the diff-erent promoter constructs were evaluated using the Tukey–Kramer test with P , 0.05 considered as significant.

Preparation of NE and EMSA

Nuclear proteins were extracted as described previously (44).The following double-stranded oligonucleotide probes (SNPidentified in bold) were constructed: 2381A sense: 50-ACAATGAGGGAGGATAAGTACAGAG-30; 2381A antisense:50-CTCTGTACTTATCCTCCCTCATTGT-30; 2381G sense:50-ACAATGAGGGAGGGTAAGTACAGAG-30; 2381G anti-sense: 50-CTCTGTACTTACCCTCCCTCATTGT-30; 2799C


by guest on Novem



ownloaded from


sense: 50-CCATGCAGAGCCTATAGTAGCTCC-30 2799Cantisense: 50-GGAGCTACTATAGGCTCTGCATGG-30;2799T sense: 50-CCATGCAGAGCTTATAGTAGCTCC-30;2799T antisense: 50-GGAGCTACTATAAGCTCTGCATGG-30;sense unrelated competitor: 50-ATGCTGTGAACCTCAGGGTGCTCG-30; antisense unrelated competitor: 50-CGAGCACCCTGAGGTTCACAGCAT-30. The double-stranded syn-thetic oligonucleotides were labeled with T4 polynucleotidekinase (Invitrogen) and [g232P]-ATP. The EMSA binding reac-tion was mixed in 1�binding buffer (Promega) with 10 mg ofnuclear protein, 1 � 105 c.p.m of 32P-labeled double-strandedoligonucleotide probe (1 ng) with or without unlabeled com-petitor probe in a total volume of 10 ml. Reaction mixtureswere incubated at room temperature for 30 min and thensubjected to 8% PAGE at 250 V for 4.5 h. The dried gels werethen exposed to X-ray film.

Quantitative real-time PCR for MMP8 mRNAsplice variants

Total RNA was extracted from the cultured BeWo cells, THP-1 cells and U937 cells using TRIzoL reagent (Invitrogen).PolyATtractw mRNA Isolation System (Promega) was usedto isolate mRNA from total RNA according to the manufac-turer’s instructions. Two micrograms of total RNA or 150 ngmRNA was reverse transcribed to single-strand cDNA usingoligo (dT) primer, RNasin inhibitor and Moloney murine leu-kemia virus reverse transcriptase (Clontech) as described bythe manufacturer. Quantitative real-time PCR was performedto compare the levels of the MMP8 mRNA splice variantswith þ91, þ67 and 0 bp insertions from cryptic exons inthe intron between exons 1 and 2 using primers designedwith the Primer Express software package that accompaniesthe Applied Biosystems 7700 sequence detector (Perkin–Elmer Life Sciences). The forward primer 50-AAGATCATGTTCTCCCTGAAGACG-30 was shared by the þ91 andþ67 bp variants. The reverse primer for the þ91 bp variantwas selected within the extra þ91 bp region (50-GCATCAGTGCAGTTCCTCTTTTT-30). The reverse primer for theþ67 bp variant (50-GGCTGGGAAGTCCAAGATCAG-30)was selected from sequence in the þ67 bp insertion. Thesize of PCR product for þ91 bp variant was 170 bp, and167 bp for the þ67 bp variant. We designed the forwardprimer (50-GAGGACAGAAAGAAAGCCAGGAG-30) andreverse primer (50-AACTTTTCCAGGTAGTCCTGAA-30) toamplify a 165 bp PCR product for the 0 bp variant. Allreverse primers, of which the one for þ0 bp variant spannedexons 1 and 2, gave exclusive amplification of the þ91,þ67 and þ0 bp variants. The real-time PCR used a 300 nM

concentration of each primer and 12.5 ml of 2� SYBRGreen PCR Master Mix (Applied Biosystems). The PCRproducts were also analyzed on 1% agarose gels to displaythe 170, 167 or 165 bp products.

Western blotting for MMP8 protein

RIPA buffer (45) was used for lysis of freshly frozen humanchorion samples, amnion tissue samples, cultured BeWocells, THP-1 cells and U937 cells. Western blotting was

carried out as previously described (46). We loaded 66 mgprotein from each sample and detected MMP8 with ananti-human MMP8 monoclonal antibody (Research Diagnos-tics, Inc.), which recognizes both latent (85 kDa) and active(64 kDa) forms of the enzyme. The blots were also probedfor b-actin to assess protein loading.

ACKNOWLEDGEMENTS

This work was supported by HD34612 (J.F.S.) TW006197(P.E.F.) and the Bill and Melinda Gates Foundation (J.F.S.).

REFERENCES

1. Rajabi, M.R., Dean, D.D., Beydoun, S.N. and Woessner, J.F., Jr (1988)Elevated tissue levels of collagenase during dilation of the uterine cervixin human parturition. Am. J. Obstet. Gynecol., 159, 971–976.

2. Bryant-Greenwood, G.D. and Yamamoto, S.Y. (1995) Control ofperipartal collagenolysis in the human chorio-decidua. Am. J. Obstet.Gynecol., 172, 63–70.

3. Tsatas, D., Baker, M.S. and Rice, G.E. (1999) Differential expression ofproteases in human gestational tissues before, during and afterspontaneous-onset labor at term. J. Reprod. Fertil., 116, 43–49.

4. Aplin, J.D., Campbell, S. and Allen, T.D. (1985) The extracellular matrixof human amniotic epithelium: ultrastructure composition and deposition.J. Cell. Sci., 79, 119–136.

5. Aplin, J.D. and Campbell, S. (1985) An immunofluorescence study ofextracellular matrix associated with cytotrophoblast of the chorion leave.Placenta, 6, 469–479.

6. Bell, S.C. and Malak, T.M. (1997) Structure and cellular biology of fetalmembranes. In Elder, M.G., Lamont, R.F. and Romero, R. (eds), PretermLabor. Churchill Livingstone, Philadelphia, pp. 401–455.

7. Premature rupture of membranes. ACOG Technical Bulletin. AmericanCollege of Obstetricians and Gynecologists, Washington, DC, 115, April,1988.

8. Parry, S. and Strauss, J.F., III (1998) Premature rupture of the fetalmembranes. New Engl. J. Med., 338, 663–670.

9. Hoffman, J.D. and Ward, K. (1999) Genetics factors in preterm delivery.Obstet. Gynecol. Surv., 54, 203–210.

10. Adams, M.M., Read, J.A., Rawlings, J.S., Harlass, F.B., Sarno, A.P. andRhodes, P.H. (1993) Preterm delivery among black and white enlistedwomen in the United States Army. Obstet. Gynecol., 81, 65–71.

11. Brickerhoff, C.E. and Matrisian, L.M. (2002) Matrix metalloproteinases: atail of frog that became a prince. Nat. Rev. Mol. Cell. Biol., 3, 207–214.

12. Fujimoto, T., Parry, S., Urbanek, M., Sammel, M., Macones, G.,Kuivaniemi, H., Romero, R. and Strauss, J.F., III (2002) A singlenucleotide polymorphism in the matrix metalloproteinase-1 (MMP-1)promoter influences amnion cell MMP-1 expression and risk for pretermpremature rupture of the fetal membranes. J. Biol. Chem., 277,6296–6302.

13. Ferrand, P.E., Parry, S., Sammel, M., Macones, G.A., Kuivaniemi, H.,Romero, R. and Strauss, J.F., III (2002) A polymorphism in the matrixmetalloproteinase-9 promoter is associated with increased risk of pretermpremature rupture of membranes in African Americans. Mol. Hum.Reprod., 8, 494–501.

14. Devarajan, P., Mookhtiar, K., van Wart, H. and Berliner, N. (1991)Structure and expression of the cDNA encoding human neutrophilcollagenase. Blood, 77, 2731–2738.

15. Hasty, K.A., Pourmotabbed, T.F., Goldberg, G.I., Thompson, J.P.,Spinella, D.G., Stevens, R.M. and Mainardi, C.L. (1990) Humanneutrophil collagenase. A distinct gene product with homology to othermatrix metalloproteinases. J. Biol. Chem., 265, 11421–11424.

16. Curry, T.E. and Osteen, K. (2003) The matrix melloproteinase system:changes, regulation, and impact throughout the ovarian and uterinereproductive cycle. Endocrine Rev., 24, 428–465.

17. Woessner, J.F., Jr (1998) Matrix metalloproteinases. In Parks, W.C. andMecham, R.P. (eds), The Matrix Metalloproteinase Family. AcademicPress, San Diego, pp. 1–14.


by guest on Novem



ownloaded from


18. Nagase, H. (1997) Activation mechanisms of matrix metalloproteinases.Biol. Chem., 378, 151–160.

19. Lauer-Fields, J.L., Tuzinski, K.A., Shimokawa, K., Nagase, H. andFields, G.B. (2000) Hydrolysis of triple-helical collagen peptide modelsby matrix metalloproteinases. J. Biol. Chem., 275, 13282–13290.

20. Murphy, G., Stanton, H., Cowell, S., Butler, G., Knauper, V., Atkinson, S.and Gavrilovic, J. (1999) Mechanisms for promatrix metalloproteinaseactivation. Acta Pathol. Microbiol. Immunol. Scand., 107, 38–44.

21. Saarinen, J., Kalkkinen, N., Welgus, H.G. and Kovanen, P.T. (1994)Activation of human interstitial procollagenase through direct cleavageof the Leu83–Thr84 bond by mast cell chymase. J. Biol. Chem., 269,18134–18140.

22. Tschesche, H. (1995) Human neutrophil collagenase. Methods Enzymol.,248, 431–449.

23. Arner, E.C., Decicco, D.P., Cherney, R. and Tortorella, M.D. (1997)Cleavage of native cartilage aggrecan by neutrophil collagenase (MMP-8)is distinct from endogenous cleavage by aggrecanase. J. Biol. Chem., 272,9294–9299.

24. Cole, A.A., Chubinskaya, S., Schumacher, B., Huch, K., Szabo, G.,Yao, J., Mikecz, K., Hasty, K.A. and Kuettner, K.E. (1996)Chondrocyte matrix metalloproteinase-8. Human articularchondrocytes express neutrophil collagenase. J. Biol. Chem., 271,11023–11026.

25. Cole, A.A. and Kuettner, K.E. (1995) MMP-8 (neutrophil collagenase)mRNA and aggrecanase cleavage products are present in normal andosteoarthritic human articular cartilage. Acta Orthop. Scand., 266 (suppl),98–102.

26. Hanemaaijer, R., Sorsa, T., Konttinen, Y.T., Ding, Y.L., Sutinen, M.,Visser, H., van Hinsbergh, V.W., Helaakoski, T., Kainulainen, T.,Ronka, H., Tschesche, H. and Salo, T. (1997) Matrix metalloprotein-8 isexpressed in rheumatoid synovial fibroblasts and endothelial cells.J. Biol. Chem., 272, 31504–31509.

27. Herman, M.P., Sukhova, G.K., Libby, P., Gerdes, N., Tang, N.,Horton, D.B., Kilbride, M., Breitbart, R.E., Chun, M. and Schonbeck, U.(2001) Expression of neutrophil collagenase (matrix metalloproteinase-8)in human atheroma: a novel collagenolytic pathway suggested bytranscriptional profiling. Circulation, 104, 1899–1904.

28. Li, D.Q., Shang, T.Y., Kim, H.-S., Solomon, A., Lokeshwar, B.L. andPflugfelder, S.C. (2003) Regulated expression of collagenases MMP-1, -8and -13 and stromelysins MMP-3, -10 and -11 by human corneal epithelialcells. Investig. Ophthalmol. Vis. Sci., 44, 2928–2936.

29. Arechavaleta-Velasco, F., Marciano, D., Diaz-Cueto, L. and Parry, S.(2004) Matrix metalloproteinase-8 is expressed in human chorionduring labor. Am. J. Obstet. Gynecol., 190, 843–50.

30. Maymon, E., Romero, R., Pacora, P., Gomez, R., Athayde, N., Edwin, S.and Yoon, B.H. (2000) Human neutrophil collagenase (matrixmetalloproteinase 8) in parturition, premature rupture of themembranes, and intrauterine infection. Am. J. Obstet. Gynecol., 183,94–99.

31. Angus, S.R., Segel, S.Y., Hsu, C.D., Locksmith, G.J., Clark, P.,Sammel, M.D., Macones, G.A., Strauss, J.F., III and Parry S. (2001)Amniotic fluid matrix metalloproteinase-8 indicates intra-amnioticinfection. Am. J. Obstet. Gynecol., 185, 1232–1238.

32. Maymon, E., Romero, R., Chaiworapongsa, T., Berman, S.,Conoscenti, G., Gomez, R. and Edwin, S. (2001) Amniotic fluid matrixmetalloproteinase-8 in preterm labor with intact membranes. Am. J.Obstet. Gynecol., 185, 1149–1155.

33. Hu, S.I., Klein, M., Carozza, M., Rdiske, J., Peppard, J. and Qi, J.S. (1999)Identification of a splice variant of neutrophil collagenase (MMP-8).FEBS Lett., 443, 8–10.

34. Mount, S.M. (1982) A catalogue of splice junction sequences. Nucl. AcidsRes., 10, 459–472.

35. Birkedal-Hansen, H., Moore, W.G.I. and Bodden, M.K. (1993) Matrixmealloproteinases: a review. Crit. Rev. Oral Biol. Med., 4, 197–250.

36. Albanese, C., Kay, T.W., Troccoli, N.M. and Jameson, J.L. (1991)Novel cyclic adenosine 30,50-monophosphate response element in thehuman chorionic gonadotropin beta-subunit gene. Mol. Endocrinol.,5, 693–702.

37. Tabansky, I. and Nurminsky, D.I. (2003) Mapping of transcription startsites by direct sequencing of SMARTTM RACE products. BioTechniques,34, 482–486.

38. Zhu, Y.Y., Machleder, E.M., Chenchik, A., Li, R. and Siebert, P.D. (2001)Reverse transcriptase template switching: a SMART approach forfull-length cDNA library construction. BioTechniques, 30, 892–897.

39. Strauss, W.M. (1994) Current protocols in molecular biology. In Ausubel,J.F., Brent, R., Kingston, R.E., Moore, D.D., Smith, J.A., Seidman, J.G.and Struhl, K. (eds), Preparation and Analysis of DNA. WileyInterscience, New York, pp. 2.2.1–2.2.3.

40. Pfaff, C.L., Parra, E.J., Bonilla, C., Hiester, K., McKeigue, P.M.,Kamboh, M.I., Hutchinson, R.G., Ferrell, R.E., Boerwinkle, E. andShriver, M.D. (2001) Population structure in admixed populations:effects of admixture dynamics on the pattern of linkage disequilibrium.Am. J. Hum. Genet., 68, 198–207.

41. Shriver, M.D., Parra, E.J., Dios, S., Bonilla, C., Norton, H., Jovel, C.,Pfaff, C., Jones, C., Massac, A., Cameron, N. et al. (2003) Skinpigmentation, biogeographical ancestry and admixture mapping.Hum. Genet., 112, 387–99.

42. Parra, E.J., Kittles, R.A., Argyropoulos, G., Pfaff, C.L., Hiester, K.,Bonilla, C., Sylvester, N., Parrish-Gause, D., Garvey, W.T., Jin, L. et al.(2001) Ancestral proportions and admixture dynamics in geographicallydefined African-Americans living in South Carolina. Am. J. Phys.Anthropol., 114, 18–29.

43. Hanis, C.L., Chakraborty, R., Ferrell, R.E. and Schull, W.J. (1986)Individual admixture estimates: disease associations and individual risk ofdiabetes and gallbladder disease among Mexican-Americans in StarrCounty, Texas. Am. J. Phys. Anthropol., 70, 433–441.

44. Schreiber, E., Matthias, P., Muller, M.M. and Schaffner, W. (1989)Rapid detection of octamer binding proteins with ‘mini-extracts’, preparedfrom a small number of cells. Nucl. Acids Res., 17, 6419.

45. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning,2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

46. Watari, M., Watari, H., DiSanto, M.E., Chacko, S., Shi, G.P. andStrauss, J.F., III (1999) Pro-inflammatory cytokines induce expressionof matrix-metabolizing enzymes in human cervical smooth muscle cells.Am. J. Pathol., 154, 1755–1762.


by guest on Novem



ownloaded from


Functionally significant SNP MMP8 promoter haplotypes and preterm premature rupture of membranes (PPROM)

Documents