MT1-MMP Expression in First Trimester Placental Tissue Is Up-Regulated in Type 1 Diabetes As a Result of Elevated Insulin and TNF-Α Levels

Diabetes Publish Ahead of Print, published online October 10, 2007

MT1-MMP expression in first trimester placental tissue is up-regulated in type 1 diabetes as a result of elevated insulin and TNF-α levels

Ursula Hiden, Ph.D.1, Elisabeth Glitzner B. Sc.1, Marina Ivanisevic, M.D.2, Josip Djelmis, M.D.2, Christian Wadsack, Ph.D.1, Uwe Lang, M.D.1, Gernot Desoye, Ph.D.1

1Department of Obstetrics and Gynecology, Medical University of Graz, Austria 2Department of Obstetrics and Gynecology, University Hospital, Petrova, Zagreb, Croatia

Running Title: MT1-MMP is elevated in the 1st trimester placenta in T1D

Address for correspondence: Ursula Hiden, M.Sc., Ph.D. Department of Obstetrics and Gynecology, Medical University of Graz Auenbruggerplatz 14, 8036 Graz, Austria E-mail: [email protected]

Received for publication 4 July 2007 and accepted in revised form 3 October 2007.

Abstract

Objective. In pre-gestational diabetes the placenta at term of gestation is characterized by various structural and functional changes. If similar alterations occur in the first trimester has remained elusive. Placental development requires proper trophoblast invasion and tissue remodeling, processes involving matrix- metalloproteinases (MMPs) of which the membrane-anchored members (MT-MMPs) such as MT1-MMPs are key players. Here we hypothesize a dysregulation of placental MT1-MMP in the first trimester of type 1 diabetic pregnancies induced by the diabetic environment.

Research Design and Methods. MT1-MMP protein was measured in first trimester placentas of healthy (n=13) and type 1 diabetic (n=13) women. To identify potential regulators first trimester trophoblasts were cultured under hyperglycaemia and various insulin, IGF1, IGF2 and TNF-α concentrations in presence or absence of signaling pathway inhibitors.

Results. MT1-MMP was strongly expressed in first trimester trophoblasts. In type 1 diabetes placental pro-MT1-MMP was up-regulated, whereas active MT1-MMP expression was only increased in late first trimester. In isolated primary trophoblasts insulin, IGF1, IGF2 and TNF-α up-regulated MT1-MMP expression, whereas glucose had no effect. The insulin effect was dependent on PI3-kinase, the IGF1 effect on MAP-kinase and the IGF2 effect on both.

Conclusions. This is the first study to reporting alterations in the first trimester placenta in type 1 diabetes. The up-regulated MT1-MMP expression in type 1 diabetes may be the result of higher maternal insulin and TNF-α levels. We speculate that the elevated MT1-MMP will affect placental development and may thus contribute to long term structural alterations in the placenta in pre-gestational diabetes.

2 MT1-MMP is elevated in the 1st trimester placenta in T1D

Despite improvement in the expressed in the first trimester placenta quality of metabolic control over the past (11; 12). decades maternal pre-gestational New emerging data show that the diabetes mellitus, in particular type 1 membrane-anchored subfamily of diabetes mellitus (T1D), is still often MMPs, i.e. MT-MMPs, are major associated with a range of maternal and modifiers of the pericellular environment fetal complications (1). In addition to the and key regulators of tumor cell effects on fetal growth the structural and behavior (13). To date, the MT-MMP functional development of the placenta family includes six members i.e., MT1-, is affected. In contrast to the well-known MT2-, MT3-, MT4-, MT5- and MT6-MMP placental changes at term of diabetic (13). MT1-MMP plays an outstanding pregnancies (2-4), the effect of pre- role in tissue remodeling, migration and gestational diabetes on the placenta in invasion (7; 14; 15). Apart from its role the first trimester has remained elusive. in extracellular matrix breakdown and We have recently proposed that a activation of other metalloproteinases diabetic insult early in pregnancy will i.e., MMP2 and MMP13, MT1-MMP is alter long term placental development also able to activate or inactivate and thus result in the observed changes several cytokines and chemokines by at term (5). cleaving their pro-forms e.g., tumor Key processes early in placental necrosis factor-alpha (TNF-α) or active development involve implantation and forms such as of interleukin-8 (IL-8), placentation. These require proliferation, growth-regulated protein alpha (GRO-α) migration and invasion of trophoblast and gamma (GRO-γ) (16; 17). The cells as well as extensive uterine tissue resulting active cytokines may affect remodeling (6; 7). Trophoblast invasion further placental development. is a process that is tightly controlled in MT1-MMP levels can be time and space in a paracrine and regulated at various stages including autocrine manner. A multitude of factors transcription, translation, activity and has been implicated in its control degradation (13; 18). MT1-MMP is including the non-classical MHC class I synthesized as an inactive, 63 kDa molecule HLA-G (8), which is expressed zymogen (pro-MT1-MMP) and, after on a specialized trophoblast transport to the cell membrane, is subpopulation, the extravillous cleaved into the 57 kDa active enzyme trophoblast. at a furin recognition motif. Shedding of Matrix metalloproteinases the membrane-anchored MT1-MMP (MMPs) have been implicated in tissue results in degradation products among remodeling. They form a large family of which the soluble 50 kDa fragment proteolytic enzymes capable of remains active whereas both 44 and 32 degrading extracellular matrix (ECM). kDa products are proteolytically inactive MMPs are members of the metzincin (18). In the human placenta in early group of proteases, which have a gestation, MT1-MMP is highly conserved Met residue and a zinc ion at expressed by various trophoblast their active site (9). Mammalian MMPs subpopulations such as extravillous share a conserved domain structure that invading and non-invading consists of a catalytic domain and an cytotrophoblasts as well as the auto-inhibitory pro-domain. When the proliferating cytotrophoblasts of cell pro-domain is destabilized or removed islands (11). by proteolytic cleavage the active site MMPs in general are becomes available for substrates (10). dysregulated in various diabetes- Various MMP family members are highly associated complications such as

3 MT1-MMP is elevated in the 1st trimester placenta in T1D nephropathy, retinopathy and vascular the last menstrual period and corrected complications (19-22). Notably, MMP2 is after endovaginal ultrasound implicated in diabetes-induced changes examination using published charts (27). (23-25) and is also dysregulated in the The study was approved by the rat placenta in diabetes (26). MT1-MMP institutional review board and ethical is the only MT-MMP member that has committee of the Medical University of been reported to be dysregulated in Graz and informed consent of the diabetes (19; 21; 22). patients. The overall hypothesis tested in this study predicted changes in Isolation of first trimester placental MT-MMP expression trophoblasts associated with pre-gestational Primary trophoblasts were isolated from diabetes. After identification of MT1- and first trimester placentas after pregnancy MT2-MMP as the only membrane-type terminations for psycho-social reasons MMPs present in the placenta we as described previously (28). The cells focused on MT1-MMP because of its thus obtained constitute a mixture of pleiotropic effects. It is predominantly both trophoblast subpopulations i.e., found in the extravillous trophoblast villous and extravillous trophoblasts. All (11). Therefore, we determined cell preparations were subjected to differential expression of MT-MMPs in rigorous immuno-cytochemical the HLA-G expressing invasive versus characterization (28). Trophoblasts were the HLA-G negative, non-invasive tested for viability by measuring human trophoblast subpopulations. We further chorionic gonadotropin (hCG) levels hypothesized that potential changes in secreted into the culture medium (Dade MT1-MMP synthesis are accounted for Behring, Deerfield, IL). Only by the diabetic environment. In order to preparations with a purity ≥ 99% and the identify the diabetes-associated factors characteristic kinetics of hCG secretion causing MT1-MMP dysregulation, in (29) were used. vitro experiments were performed using isolated primary trophoblasts from the Cell Culture first trimester of pregnancy. They were Primary trophoblasts were cultured in treated with insulin, TNF-α and glucose, gelatine-coated plates with DMEM factors with elevated concentrations in (Gibco) supplemented with 2% (v/v) the maternal circulation in diabetes and FCS in a humified atmosphere of 5% their effect on MT1-MMP expression CO2 at 37°C. For growth factor and was determined. Since IGF1 and IGF2 glucose treatment, isolated trophoblasts share some signaling pathways with were seeded in gelatine-coated 24 well insulin they were included in the in vitro plates (50,000 cells per well) and experiments. cultured in DMEM with 2% (v/v) FCS. After 24 h medium was replaced by Research Design and Methods fresh medium supplemented with insulin (0.1, 1 nM; Calbiochem, Merck, First trimester placental samples Darmstadt, Germany), IGF1 (50, 100 After pregnancy termination for psycho- ng/ml; R&D Systems, Minneapolis, MN), social reasons (IR) or missed abortions IGF2 (165, 300 ng/ml; R&D Systems), (MA), tissue samples (Table 1) were TNF-α (10, 25 ng/ml; Sigma) and collected in Medium 199 supplemented glucose (25 mM; Sigma) and the cells with penicillin/streptomycin (Gibco, were cultured for further 48 h. The Invitrogen, Carlsbad, USA), washed experimental levels of insulin as well as immediately in PBS and snap-frozen. IGF1 and IGF2 were chosen as to lie The gestational age was calculated from within the (patho)-physiological range to

4 MT1-MMP is elevated in the 1st trimester placenta in T1D avoid low-affinity binding to others but RNA prepared for hybridization as their specific receptor (30). The described previously (35). For concentrations of insulin (1 nM) and expression analysis cRNA was glucose (25 mM) were chosen as to lie hybridized against Affymetrix (Santa about two to three-fold higher than the Clara, CA, USA) HU133A-chips maternal in vivo concentrations attained according to the manufacturer’s either post-prandially (glucose) or after instructions. Raw data were normalized pharmacological administration (insulin) globally and processed with Microarray (31). The IGF1 level (85 ng/ml) parallels Suite, version 5.0 and Data Mining Tool the maternal IGF1 levels in the first (Affymetrix) software. Annotations were trimester of gestation (32). IGF2 was obtained from NetAffx (Affymetrix) and used in the concentration range that the data screened for membrane showed the highest effect on trophoblast anchored matrix metalloproteinases invasion in vitro (between 125 and 625 (MT-MMPs). ng/ml) (33). The concentration of TNF-α (25 ng/ml) is a standard concentration Isolation of RNA and RT-PCR for used to study the effect of TNF-α (34). various MT-MMPs in first trimester In experiments with inhibition of trophoblasts PI3-kinase or MEK1, cells were pre- Total RNA was isolated from first treated for 5 h with either Wortmannin trimester trophoblasts with Trizole (100 nM; Calbiochem) or U0126 (10 µM; (MRC, Ohio). Primers (Table 2) for Calbiochem) dissolved in DMSO before RPL30 and the MT-MMPs (MT1-MMP, being stimulated with 1 nM insulin for 48 MT2-MMP, MMP16, MMP17, MMP24, h. In un-stimulated control cells and MMP25) were designed using the public insulin-stimulated cells, the same web-page Primer3 volume of DMSO without inhibitors was (http://frodo.wi.mit.edu/cgi- added as vehicle control. bin/primer3/primer3_www.cgi) and purchased from Ingenetix (Vienna, Separation of HLA-G positive and Austria). The primer-pairs included HLA-G negative first trimester splicing sites within the amplicon. trophoblasts Because of its stable expression within Immediately after isolation first trimester the different placental cell types the trophoblasts were separated into mRNA-amount of the ribosomal protein subpopulations expressing or lacking L30 (RPL30) was used as an internal surface HLA-G. Trophoblasts were control (35). Two hundred ng of total incubated with immuno-magnetic beads RNA was used for the one step RT-PCR (Dynabeads M-450) conjugated with kit from Quiagen (Hilden, Germany) anti-HLA-G antibody (MEM-G/9; Abcam, according to the manufacturer’s Cambridge, UK). The unbound, HLA-G instructions. For RPL30 24 cycles and devoid, cells and the cells bound to the for all MT-MMPs 27 cycles were used. beads were separated by applying a For all primer pairs the annealing magnet above the tube. The two cell temperature was 60°C. PCR-products populations thus obtained were were electrophoresed on 3% agarose subsequently cultured in DMEM gels, documented with the Eagle-EyeTM supplemented with 10% (v/v) FCS. system (Stratagene, CA) and quantified with the AlphaDigiDoc 1000 (Alpha Microarray analysis of placental Innotech, CA) software. In order to primary cell RNA validate primer pairs, RT-PCRs were Total RNA from ten first trimester carried out using human RNA from trophoblast preparations isolated from brain, lung and colon as positive different placentas was pooled and 5 µg controls. These gave distinct bands with

5 MT1-MMP is elevated in the 1st trimester placenta in T1D sizes corresponding to the calculated software within the linear range of film size of the amplicon in at least one and camera. organ. In preliminary experiments using first trimester trophoblast RNA the Statistical analysis optimal RT-PCR cycle number for MT1- Statistical analysis used Sigma Stat 3.1 MMP, MT2-MMP and L30 was (Jandel Scientific, San Rafael, CA) determined to lie within the linear range software. After testing for normal of the amplification. distribution (Kolmogorov-Smirnov), the Mann-Whitney U-test for non-parametric Preparation of proteins and western data or a Students t-test was used to blot analysis test for differences in the amounts of Tissue was homogenized and cells were pro-, active and total MT1-MMP protein lysed in buffer containing 0.01 mol/l tris and mRNA of various MT-MMPs. To pH 7.4, 1% SDS, 1 mmol/l Na- test for treatment effects Kruskal-Wallis orthovanadate and Complete protease One Way Analysis of Variance on inhibitor (Roche) mixed with an equal Ranks with Dunn’s Method as post-hoc volume of Laemmli sample buffer test was used. Correlations were (Sigma). Prior to electrophoresis, analyzed by Person Product Moment samples were centrifuged and boiled for Correlation. Significances were 5 min at 99°C. Equal amounts of accepted at a level of P < 0.05. protein, determined according to Lowry were used for SDS-PAGE on a 10 % gel Results (Pierce, Rockford, IL). After electroblotting membranes were blocked Expression of various MT-MMPs in for 1 h with 5 % (w/v) non-fat dry milk first trimester trophoblasts (BioRad, Hercules, CA) and 0.1 % (v/v) Among all six MT-MMPs surveyed by Tween-20 (Sigma) in 0.14 mol/l tris- microarray analysis only MT1-MMP and buffered saline, pH 7.3, at room MT2-MMP were expressed in isolated temperature. This solution was used for primary trophoblasts from the first subsequent washings and as a diluent trimester of gestation (Table 3). This for the antibodies. The membranes were was further confirmed by RT-PCR, incubated overnight at 4°C with which demonstrated a similar amount of antibodies against MT1-MMP MT1-MMP and MT2-MMP expression (Chemicon, Millipore, Bedford, MA; 1 : (Figure 1A, B). Thus, MT3-MMP - MT6- 2000) or β-Actin (Amersham, Little MMP were either absent or present at Chalfont, UK, 1:10,000). After washing, levels below the sensitivity threshold of membranes were incubated with the both methods i.e., microarray and RT- adequate secondary antibody (BioRad; PCR. After separation of the 1:1000) for 1 h at room temperature. trophoblasts in subpopulations Immuno-labeling was visualized using according to their surface HLA-G the SuperSignal CL-HRP Substrate expression, the HLA-G positive System (Pierce). To allow comparison trophoblasts representing the more between gels an internal standard invasive phenotype of trophoblasts had sample was prepared as lysate from a 33 % (p=0.0004) higher expression one first trimester placental tissue to level of MT1-MMP (Figure 1C, D). which all optical densities within each blot were normalized. Membranes were MT1-MMP protein expression in first exposed to Hyperfilm (Amersham) and trimester placental tissue of healthy densitometrically scanned using a digital and T1D women camera and the AlphaDigiDoc 1000 In first trimester placentas complicated by T1D, MT1-MMP expression was

6 MT1-MMP is elevated in the 1st trimester placenta in T1D increased (p<0.0001) by 100% (Figure MMP species (50 and 44 kDa) could be 2). When the two bands of active (57 detected by immunoblotting (not kDa) and pro-MT1-MMP (63 kDa) were shown). The 113% induction (p=0.05) analyzed separately, an up-regulation of caused by insulin was attenuated by both, zymogen (96 %; p=0.0001) and inhibition of the PI3-kinase (PI3K) active proteinase (111 %; p=0.04) was pathway, but not affected by inhibition observed. The ratio between active and with U0126 of MEK1, a central kinase of pro-MT1-MMP did not differ between the ERK1/2 MAP-kinase (MAPK) both groups. No other bands at lower pathway (Figure 5). The IGF1-induced molecular weight i.e., 50, 44 and 32 kDa stimulation was only inhibited by U0126, were detected, which would represent whereas the IGF2-effect was abolished processed MT1-MMP species. by inhibition of both the PI3K and the When the data were stratified ERK1/2 pathway. The effects of insulin, according to the gestational age i.e., IGF1, IGF2 and TNF-α did not differ early (≤ week 8) and late (> week 8) first between isolated primary trophoblasts trimester, active MT1-MMP levels were from early vs. late first trimester (not lower by 60% (p=0.026) in the control shown). samples from late as compared to early first trimester (Figure 3). Because the Discussion T1D samples did not show this change their MT1-MMP levels were higher by This study identified MT1-MMP and 72% (p=0.0018) after week eight in the MT2-MMP as the only members of the first trimester. No change between early membrane-anchored family of MMPs and late first trimester could be found in expressed at high mRNA levels in the total and pro-MT1-MMP expression (not first trimester trophoblast of the human shown). No difference in MT1-MMP was placenta. found between the miscarriages (MA) The prominent expression of and the interruptions (IR) within and MT1-MMP early in gestation suggests a between both study groups (not shown). major role in processes involved in early Neither total nor active or pro- placental development. This notion is MT1-MMP levels correlated with further supported by its predominant maternal HbA1c values. However, the presence in the HLA-G expressing daily insulin dose of the T1D subjects trophoblast subpopulation, which correlated (r=0.626, p=0.04) with the represents the invasive extravillous total MT1-MMP protein levels (not trophoblast. The high MT1-MMP shown). expression in the HLA-G positive trophoblasts is in accordance with Regulation of MT1-MMP expression results of in-situ hybridization in first in first trimester trophoblasts by trimester placental tissue (11). insulin, IGF1, IGF2, glucose and TNF- During the first trimester of α pregnancy the amount of active MT1- When primary first trimester MMP notably decreased in the non- trophoblasts were analyzed by western diabetic control group. The underlying blotting only the 57 kDa band of active mechanism is unclear. The decrease of MT1-MMP could be detected. Insulin, active, but not of total or pro-MT1-MMP IGF2 and TNF-α increased MT1-MMP in the late first trimester suggests a expression in a dose-dependent manner change in pro-MT1-MMP activation. (ANOVA) (Figure 4). IGF1 up-regulation This may result from a reduction of was only significant with Student’s t-test. expression or activity of furin or the pro- Glucose had no effect. In the culture protein convertase PCSK6 that both can supernatant none of the shedded MT1- activate pro-MT1-MMP (36). The

7 MT1-MMP is elevated in the 1st trimester placenta in T1D absence of different MT1-MMP levels or attributed to MT1-MMP it was of special effects in the isolated cells from different interest that placental MT1-MMP gestational weeks indicates stability of expression did not differ between both the intrinsic responsiveness of the cells groups regardless of presence or during the first trimester and suggests absence of maternal diabetes. This changes in placental environment to strongly suggests that MT1-MMP contribute to the decrease in tissue dysfunction is not involved in the MT1-MMP expression. pathogenesis of miscarriages. The period in gestation from Even in the first trimester the week 6 to week 12 is characterized by placenta is a complex tissue comprising an increase in oxygen tension in the several cell types. Here only total tissue intervillous space (37). Analysis of the samples were analyzed and hence the MT1-MMP promoter sequence found no differences between T1D samples and potential binding site of hypoxia controls cannot be directly attributed to inducible factor 1 (HIF-1) that could a specific cell type. However, as in-situ directly up-regulate MT1-MMP hybridization did not detect MT1-MMP expression. However, furin expression is mRNA in the villous stroma in the first up-regulated under hypoxic conditions trimester placenta (11) we conclude by HIF-1 (38). Hence, as a hypothesis, trophoblast cells to account for the lower furin expression under normoxic observed changes of MT1-MMP in conditions in the late first trimester may placental tissue. Therefore, primary result in lower levels of cleaved, active trophoblasts were used for the further in MT1-MMP. This decrease of MT1-MMP vitro experiments. synthesis was absent in the T1D group. In first trimester trophoblasts in To the best of our knowledge this vitro only one MT1-MMP species was is the first study carried out on placental detected which corresponded to active tissue from first trimester diabetic MT1-MMP. This may be the result of pregnancies. Therefore, it is unknown if high trophoblast expression of furin and the oxygen tension in the intervillous PCSK6 found by microarray analysis space is lower in T1D than in non- (Hiden and Desoye, unpublished). In the diabetic pregnancies. Moreover, it is absence of other potential substrates also unknown if utero-placental function such as ECM in cell culture the high associated with MMP activity such as proteolytic activity against pro-MT1- trophoblast migration, invasion or MMP may result in full activation of the uterine tissue remodeling is altered. enzyme. In other tissues MT1-MMP This study did not measure expression is decreased (19; 21) or maternal IGF1, IGF2 and TNF-α levels, increased (22; 39) in diabetes. Hence, because restrictions from the ethical the ultimate effect of the diabetic committee did not permit measurements environment on MT1-MMP expression other than the HbA1c values. Hence, the strongly depends on the specific tissue concentrations for the in vitro as well as on the proportion of factors experiments were chosen from dysregulated in this pathology such as published values for first trimester levels glucose, insulin and TNF-α. of IGF1, IGF2 and TNF-α in maternal Miscarriage is a pregnancy T1D. Under this condition maternal IGF1 problem that may have several and IGF2 serum levels are unchanged underlying causes including inadequate (42). To our knowledge no published trophoblast invasion (40). Pregestational data about serum TNF-α levels of T1D diabetes is associated with an increased women in the first trimester are risk for spontaneous abortions (41). available, but TNF-α is elevated in non- Given the prominent role that is pregnant T1D patients (43).

8 MT1-MMP is elevated in the 1st trimester placenta in T1D

Furthermore, rodents have increased expression in various cells (48; 49). uterine TNF-α expression throughout When PI3-kinase was inhibited by diabetic pregnancy (44). Wortmannin the notably strong effect of The increase of MT1-MMP insulin was reduced. In contrast, the expression in T1D was paralleled by up- IGF1 effect was diminished by inhibition regulation of MT1-MMP in isolated of the ERK1/2 pathway. The IGF2 effect primary trophoblasts after treatment with was absent after inhibition of both insulin and TNF-α, both factors with pathways, which may be accounted for elevated concentrations in the diabetic by the ability of IGF2 to bind to and environment. Insulin shares some activate the IGF1-receptor (IGF1-R) as intracellular signaling pathways with well as the short insulin receptor isoform IGF1 and IGF2. Therefore, both IGFs (50). Thus, the induction of MT1-MMP were included in the study despite their synthesis by IGF2 may be mediated by lack of concentration change in the first the IGF1-R and ERK1/2 as well as by trimester T1D pregnancies. Since the insulin receptor and PI3K. The IGF1 glucose had no effect on MT1-MMP effect in trophoblasts was different from expression, insulin and the pro- tumor cells (49) in which IGF1 inflammatory TNF-α are likely stimulated MT1-MMP expression via the candidates to account for the observed PI3K/Akt pathway, which may reflect changes in placental MT1-MMP changes in tumor cell signaling resulting expression in T1D. The promoter of from malignant transformation. Thus, MT1-MMP includes a binding site for the transcriptional activation of MT1-MMP in nuclear transcription factor SP-1 and trophoblasts can be accomplished by several TIE-like (TGF-β1 inhibitory signaling through different pathways i.e., element like) sequences (45). Both the PI3K-kinase and the ERK1/2 insulin as well as TNF-α can activate pathway. SP-1 (46; 47) and thereby could The consequences of the stimulate MT1-MMP expression. IGF1 diabetes-associated alterations in MT1- and IGF2 also up-regulated MT1-MMP MMP expression and processing are expression in trophoblasts. The unknown. Among all MMPs, MT1-MMP stimulatory IGF1 effect, however, was has a notably broad range of substrates only significant with Student’s t-test. We including ECM components such as assume that ANOVA missed the fibronectin, collagen I-III, laminin 1 and concentration effect as a result from the 5, pro-MMPs such as proMMP2, already maximum stimulation of MT1- proMMP13 as well as cytokines and MMP with the lower IGF1 concentration chemokines including proTNF-α, IL-8 as (50ng/ml). None of the treatments well as GRO-α and -γ (16; 17). resulted in the occurrence of an Therefore, one can picture several additional MT1-MMP species indicating scenarios by which higher levels of that enzyme processing was not MT1-MMP may affect placental affected. The key role of insulin in development: 1) MT1-MMP degrades regulating placental MT1-MMP extracellular matrix; 2) other MMPs are production as found in vitro is activated, which subsequently cleave corroborated further by the correlation of extracellular matrix components; both total MT1-MMP protein with the daily mechanisms may directly affect villous insulin dose of the T1D subjects. differentiation and development; 3) over- Insulin can activate two major activation (proTNF-α) or enhanced signaling pathways i.e., the MAP-kinase degradation and, hence, inactivation (IL- and the PI3-kinase pathway, 8, GRO-α and -γ) of cytokines and respectively. Both pathways can chemokines involved in trophoblast transcriptionally activate MT1-MMP function (16) may further indirectly

9 MT1-MMP is elevated in the 1st trimester placenta in T1D modify cellular processes relevant for dysregulation of placental MT1-MMP placental development. expression is dependent on insulin We propose that some of the levels. well-described structural alterations of the placenta at the end of a T1D Acknowledgements pregnancy may begin already in the first trimester of pregnancy. Obviously, tight The study was supported by grants glycaemic control in the first gestational 10896 (to UH), 10053 and 12601 (to weeks by insulin treatment does not GD) by the Jubilee Fund, Austrian prevent all diabetes-associated changes National Bank, Vienna. in placental development as

10 MT1-MMP is elevated in the 1st trimester placenta in T1D

References 1. Garner P: Type I diabetes mellitus and pregnancy. Lancet 346:157-161, 1995 2. Desoye G, Myatt L: The Placenta. In Diabetes in Women Reece EA, Coustan DR, Gabbe SG, Eds. Philadelphia, Lippincott Williams & Wilkins, 2004 3. Desoye G, Haugel de Mouzon S, Shafrir E: The Placenta in Diabetic Pregnancy. In Textbook of Diabetes and Pregnancy Hod M, Jovanovic L, DiRenzo G-C, deLeiva A, Langer O, Eds. London, Dunitz M, 2003, p. 126-147 4. Desoye G, Kaufmann P: The Human Placenta in Diabetes. In Diabetology of Pregnancy Djelmis J, Desoye G, Ivanisevic M, Eds. Basel, Switzerland, KARGER 2005, p. 94-109 5. Desoye G, Haugel de Mouzon S: The Human Placenta in Gestational Diabetes Mellitus. Diabetes Care 30:S120-S126, 2007 6. Pijnenborg R, Vercruysse L, Hanssens M: The uterine spiral arteries in human pregnancy: facts and controversies. Placenta 27:939-958, 2006 7. Sato Y, Fujiwara H, Zeng BX, Higuchi T, Yoshioka S, Fujii S: Platelet-derived soluble factors induce human extravillous trophoblast migration and differentiation: platelets are a possible regulator of trophoblast infiltration into maternal spiral arteries. Blood 106:428-435, 2005 8. Bainbridge DR: Evolution of mammalian pregnancy in the presence of the maternal immune system. Rev Reprod 5:67-74, 2000 9. Stocker W, Grams F, Baumann U, Reinemer P, Gomis-Ruth FX, McKay DB, Bode W: The metzincins--topological and sequential relations between the astacins, adamalysins, serralysins, and matrixins (collagenases) define a superfamily of zinc- peptidases. Protein Sci 4:823-840, 1995 10. Page-McCaw A, Ewald AJ, Werb Z: Matrix metalloproteinases and the regulation of tissue remodelling. Nat Rev Mol Cell Biol 8:221-233, 2007 11. Bjorn SF, Hastrup N, Lund LR, Dano K, Larsen JF, Pyke C: Co-ordinated expression of MMP-2 and its putative activator, MT1-MMP, in human placentation. Mol Hum Reprod 3:713-723, 1997 12. Isaka K, Usuda S, Ito H, Sagawa Y, Nakamura H, Nishi H, Suzuki Y, Li YF, Takayama M: Expression and activity of matrix metalloproteinase 2 and 9 in human trophoblasts. Placenta 24:53-64, 2003 13. Hernandez-Barrantes S, Bernardo M, Toth M, Fridman R: Regulation of membrane type-matrix metalloproteinases. Semin Cancer Biol 12:131-138, 2002 14. Genis L, Galvez BG, Gonzalo P, Arroyo AG: MT1-MMP: universal or particular player in angiogenesis? Cancer Metastasis Rev 25:77-86, 2006 15. Savinov AY, Strongin AY: Defining the roles of T cell membrane proteinase and CD44 in type 1 diabetes. IUBMB Life 59:6-13, 2007 16. Tam EM, Morrison CJ, Wu YI, Stack MS, Overall CM: Membrane protease proteomics: Isotope-coded affinity tag MS identification of undescribed MT1-matrix metalloproteinase substrates. Proc Natl Acad Sci U S A 101:6917-6922, 2004 17. Sato H, Takino T, Miyamori H: Roles of membrane-type matrix metalloproteinase- 1 in tumor invasion and metastasis. Cancer Sci 96:212-217, 2005 18. Osenkowski P, Toth M, Fridman R: Processing, shedding, and endocytosis of membrane type 1-matrix metalloproteinase (MT1-MMP). J Cell Physiol 200:2-10, 2004 19. Boucher E, Mayer G, Londono I, Bendayan M: Expression and localization of MT1-MMP and furin in the glomerular wall of short- and long-term diabetic rats. Kidney Int 69:1570-1577, 2006

11 MT1-MMP is elevated in the 1st trimester placenta in T1D

20. McLennan SV, Martell SK, Yue DK: Effects of mesangium glycation on matrix metalloproteinase activities: possible role in diabetic nephropathy. Diabetes 51:2612- 2618, 2002 21. Portik-Dobos V, Anstadt MP, Hutchinson J, Bannan M, Ergul A: Evidence for a matrix metalloproteinase induction/activation system in arterial vasculature and decreased synthesis and activity in diabetes. Diabetes 51:3063-3068, 2002 22. Song W, Ergul A: Type-2 diabetes-induced changes in vascular extracellular matrix gene expression: relation to vessel size. Cardiovasc Diabetol 5:3, 2006 23. Death AK, Fisher EJ, McGrath KC, Yue DK: High glucose alters matrix metalloproteinase expression in two key vascular cells: potential impact on atherosclerosis in diabetes. Atherosclerosis 168:263-269, 2003 24. Noda K, Ishida S, Inoue M, Obata K, Oguchi Y, Okada Y, Ikeda E: Production and activation of matrix metalloproteinase-2 in proliferative diabetic retinopathy. Invest Ophthalmol Vis Sci 44:2163-2170, 2003 25. Tsilibary EC: Microvascular basement membranes in diabetes mellitus. J Pathol 200:537-546, 2003 26. Pustovrh MC, Jawerbaum A, Capobianco E, White V, Lopez-Costa JJ, Gonzalez E: Increased matrix metalloproteinases 2 and 9 in placenta of diabetic rats at midgestation. Placenta 26:339-348, 2005 27. Robinson HP: Sonar measurement of fetal crown-rump length as means of assessing maturity in first trimester of pregnancy. Br Med J 4:28-31, 1973 28. Blaschitz A, Weiss U, Dohr G, Desoye G: Antibody reaction patterns in first trimester placenta: implications for trophoblast isolation and purity screening. Placenta 21:733-741, 2000 29. Polliotti BM, Abramowsky C, Schwartz DA, Keesling SS, Lee GR, Caba J, Zhang W, Panigel M, Nahmias AJ: Culture of first-trimester and full-term human chorionic villus explants: role of human chorionic gonadotropin and human placental lactogen as a viability index. Early Pregnancy 1:270-280, 1995 30. Li G, Barrett EJ, Wang H, Chai W, Liu Z: Insulin at physiological concentrations selectively activates insulin but not insulin-like growth factor I (IGF-I) or insulin/IGF-I hybrid receptors in endothelial cells. Endocrinology 146:4690-4696, 2005 31. Lindstrom T, Hedman CA, Arnqvist HJ: Use of a novel double-antibody technique to describe the pharmacokinetics of rapid-acting insulin analogs. Diabetes Care 25:1049-1054, 2002 32. Fuglsang J, Lauszus F, Flyvbjerg A, Ovesen P: Human placental growth hormone, insulin-like growth factor I and -II, and insulin requirements during pregnancy in type 1 diabetes. J Clin Endocrinol Metab 88:4355-4361, 2003 33. Hamilton GS, Lysiak JJ, Han VK, Lala PK: Autocrine-paracrine regulation of human trophoblast invasiveness by insulin-like growth factor (IGF)-II and IGF-binding protein (IGFBP)-1. Exp Cell Res 244:147-156, 1998 34. Li R, Luo X, Archer DF, Chegini N: Doxycycline alters the expression of matrix metalloproteases in the endometrial cells exposed to ovarian steroids and pro- inflammatory cytokine. J Reprod Immunol 73:118-129, 2007 35. Hiden U, Maier A, Bilban M, Ghaffari-Tabrizi N, Wadsack C, Lang I, Dohr G, Desoye G: Insulin control of placental gene expression shifts from mother to foetus over the course of pregnancy. Diabetologia 49:123-131, 2006 36. Yana I, Weiss SJ: Regulation of membrane type-1 matrix metalloproteinase activation by proprotein convertases. Mol Biol Cell 11:2387-2401, 2000 37. Rodesch F, Simon P, Donner C, Jauniaux E: Oxygen measurements in endometrial and trophoblastic tissues during early pregnancy. Obstet Gynecol 80:283-285, 1992

12 MT1-MMP is elevated in the 1st trimester placenta in T1D

38. McMahon S, Grondin F, McDonald PP, Richard DE, Dubois CM: Hypoxia- enhanced expression of the proprotein convertase furin is mediated by hypoxia- inducible factor-1: impact on the bioactivation of proproteins. J Biol Chem 280:6561- 6569, 2005 39. Mondy JS, Anstadt MP, Franga DL, Portik-Dobos V, Hutchinson J, Ergul A: Decreased vascular matrix metalloproteinase abundance in diabetic patients with symptomatic macroangiopathy. Ethn Dis 12:S3-18-22, 2002 40. Myatt L, Cui X: Oxidative stress in the placenta. Histochem Cell Biol 122:369-382, 2004 41. Miodovnik M, Skillman C, Holroyde JC, Butler JB, Wendel JS, Siddiqi TA: Elevated maternal glycohemoglobin in early pregnancy and spontaneous abortion among insulin-dependent diabetic women. Am J Obstet Gynecol 153:439-442, 1985 42. Lauszus FF, Klebe JG, Flyvbjerg A: Macrosomia associated with maternal serum insulin-like growth factor-I and -II in diabetic pregnancy. Obstet Gynecol 97:734-741, 2001 43. Abdel Aziz MT, Fouad HH, Mohsen GA, Mansour M, Abdel Ghaffar S: TNF-alpha and homocysteine levels in type 1 diabetes mellitus. East Mediterr Health J 7:679- 688, 2001 44. Flein A, Kostina E, Savion S, Orenstein H, Shepshelovich J, Ornoy A, Torchinsky A, Toder V: Expression of tumor necrosis factor-alpha in the pregnant uterus of diabetic mice: effect of maternal immunopotentiation. Am J Reprod Immunol 46:161- 168, 2001 45. Lohi J, Lehti K, Valtanen H, Parks WC, Keski-Oja J: Structural analysis and promoter characterization of the human membrane-type matrix metalloproteinase-1 (MT1-MMP) gene. Gene 242:75-86, 2000 46. Horovitz-Fried M, Jacob AI, Cooper DR, Sampson SR: Activation of the nuclear transcription factor SP-1 by insulin rapidly increases the expression of protein kinase C delta in skeletal muscle. Cell Signal 19:556-562, 2007 47. Ryuto M, Ono M, Izumi H, Yoshida S, Weich HA, Kohno K, Kuwano M: Induction of vascular endothelial growth factor by tumor necrosis factor alpha in human glioma cells. Possible roles of SP-1. J Biol Chem 271:28220-28228, 1996 48. Boyd PJ, Doyle J, Gee E, Pallan S, Haas TL: MAPK signaling regulates endothelial cell assembly into networks and expression of MT1-MMP and MMP-2. Am J Physiol Cell Physiol 288:C659-668, 2005 49. Zhang D, Brodt P: Type 1 insulin-like growth factor regulates MT1-MMP synthesis and tumor invasion via PI 3-kinase/Akt signaling. Oncogene 22:974-982, 2003 50. Pandini G, Medico E, Conte E, Sciacca L, Vigneri R, Belfiore A: Differential gene expression induced by insulin and insulin-like growth factor-II through the insulin receptor isoform A. J Biol Chem 278:42178-42189, 2003

13 MT1-MMP is elevated in the 1st trimester placenta in T1D

Tables

Table 1. Characteristics of study subjects

control group T1D group

No of subjects 13 13

Interrupted pregnancies (IR) 9 9

Miscarriages (MA) 4 4 gestational age (wk)

Mean ± SD 8 ± 2 9 ± 2

Range 7 - 12 7 - 12

HbA1c (%)

Mean ± SD n.d. 7.5 ± 2.1

HbA1c values of the control group were not determined (n.d.). The cut off HbA1c value for non-diabetic pregnant women as established by the local clinical laboratory was 6 %.

14 MT1-MMP is elevated in the 1st trimester placenta in T1D

Table 2. Primers used for RT-PCR.

gene forward primer reverse primer

RPL30 CCTAAGGCAGGAAGATGGTG CAGTCTGTTCTGGCATGCTT

MT1-MMP AAGGCACACTTGCTCCTGTT CACTGGTGAGACAGGCTTGA

MT2-MMP GGCCGACATCATGGTACTCT GTCAACGTCCTTCCACTGGT

MT3-MMP GCTGACCCAAGGAAAAATGA CACAAAATTCCCGTCGCTAT

MT4-MMP CTGTACTGGCGCTACGATGA GGCTCTGGTCATGTTGTCCT

MT5-MMP CAAACCCAACATCTGTGACG TAGGTCTTGCCCACAGGTTC

MT6-MMP GATCGATGTGAGGGCAATTT TAGGTCTTCCCGTTCTGTGG

15 MT1-MMP is elevated in the 1st trimester placenta in T1D

Table 3. Microarray analysis of MT-MMP family members expressed in first trimester trophoblasts. Total RNA from ten preparations isolated from ten different placentas was used. Expression of some metalloproteinases was not detectable (ND).

MT-MMP family members Accession Signal number

MT1-MMP (MMP14) NM_004995 2341

MT2-MMP (MMP15) NM_002428 2145

MT3-MMP (MMP16) NM_022564 ND

MT4-MMP (MMP17) NM_016155 ND

MT5-MMP (MMP24) NM_006690 ND

MT6-MMP (MMP25) NM_022468 ND

16 MT1-MMP is elevated in the 1st trimester placenta in T1D

Figure Legends

Figure 1. Expression of MT-MMP mRNAs in human first trimester trophoblasts. RT- PCR revealed that MT1-MMP and MT2-MMP are expressed (A). No bands were detected for MT3-, MT4-, MT5- and MT6-MMP (not shown). Mean mRNA expression levels did not differ between MT1-MMP and MT2-MMP (B). MT1-MMP mRNA is present in HLA-G positive (+) and HLA-G negative (-) trophoblasts (C) with higher expression in the HLA-G positive trophoblast subpopulation (D). The ribosomal protein L30 (RPL30) was used as an internal control; n = 4 trophoblast preparations from different placentas.

Figure 2. MT1-MMP protein expression (mean ± SEM) in first trimester placental tissue from control (n=13) and type 1 diabetic (T1D; n=13) pregnancies. Six representative western blots are shown (A). Prior to data analysis all samples were normalized to the internal control (placenta extract). The graph (B) displays the expression of total MT1-MMP and separate expression of the pro-MT1-MMP zymogen (63 kDa) and the active MT1-MMP (57 kDa).

Figure 3. Box plot depicting expression of active MT1-MMP protein (57 kDa) in early (≤ week 8; n=6) and late (> week 8; n=7) first trimester placental tissue of type 1 diabetic (T1D) and healthy (control) women.

Figure 4. (A) Representative western blot of active MT1-MMP (57 kDa) in first trimester primary trophoblasts (n=5 preparations from different placentas) stimulated for 48 h with insulin (I; 1 nM), TNF-α (25 ng/ml), IGF1 (100 ng/ml), IGF2 (300 ng/ml) and glucose (G; 25 mM) as compared to the untreated control (C). The graphs (B-E) show MT1-MMP protein expression (mean ± SEM) in trophoblasts treated with different concentrations of insulin (0.1; 1 nM), TNF-α (10; 25 ng/ml); IGF1 (50, 100 ng/ml); IGF2 (165; 300 ng/ml). The data are expressed relative to the controls (= 100%). Statistical tests used the raw data. * indicates a significant change (p<0.05) vs. controls using post hoc test, § indicates a significant change using Student’s t- test.

Figure 5. Induction of active (57 kDa) MT1-MMP in primary first trimester trophoblasts after 48 h treatment with insulin (1 nM), IGF1 (100 ng/ml) and IGF2 (300 ng/ml) in the presence or absence of the PI3K inhibitor Wortmannin (100 nM) or the MEK1 inhibitor U0126 (10 µM). The graphs (A-C) display the protein expression (n=4 trophoblast preparations from different placentas; mean ± SEM) compared to the untreated control. P-values refer to differences vs. controls.

17 MT1-MMP is elevated in the 1st trimester placenta in T1D

Figure 1

18 MT1-MMP is elevated in the 1st trimester placenta in T1D

Figure 2

19 MT1-MMP is elevated in the 1st trimester placenta in T1D

Figure 3

20 MT1-MMP is elevated in the 1st trimester placenta in T1D

Figure 4

21 MT1-MMP is elevated in the 1st trimester placenta in T1D

Figure 5