Canadian Journal of Physiology and Pharmacology
Cyclo-oxygenase inhibitors for treating preterm labour: what is the molecular evidence?
Journal: Canadian Journal of Physiology and Pharmacology
Manuscript ID cjpp-2018-0380.R1
Manuscript Type: Article
Date Submitted by the 09-Dec-2018 Author:
Complete List of Authors: Urrego, Daniela; University of Calgary Cumming School of Medicine, Physiology and Pharmacology Liwa, Anthony; University of Calgary Cumming School of Medicine, Physiology and Pharmacology; Catholic University of Health And Allied Sciences WeillDraft Bugando School of Medicine, Clinical Pharmacology Cole, William; University of Calgary Cumming School of Medicine, Physiology and Pharmacology Wood, Stephen; University of Calgary Cumming School of Medicine, Obstetrics and Gynecology; University of Calgary Cumming School of Medicine, O'Brien Institute for Public Health Slater, Donna; University of Calgary Cumming School of Medicine, Physiology and Pharmacology; University of Calgary Cumming School of Medicine, Obstetrics and Gynecology
Is the invited manuscript for consideration in a Special Connecting Maternal Fetal Newborn Physiology Issue:
Cyclooxygenase, Preterm Labour, Pregnancy, Myometrium, Fetal Keyword: Membranes
https://mc06.manuscriptcentral.com/cjpp-pubs Page 1 of 42 Canadian Journal of Physiology and Pharmacology
1 Cyclo-oxygenase inhibitors for treating preterm labour: what is the molecular evidence?
2
3 Daniela Urrego1*, Anthony C Liwa1,2, William C Cole1, Stephen Wood3, Donna M Slater1,3*
4
5 1Department of Physiology and Pharmacology, Cumming School of Medicine, University of
6 Calgary, 3330 Hospital Drive NW Calgary, AB, Canada T2N 4N1.
7
8 2Department of Clinical Pharmacology, Weill School of Medicine, Catholic University of Health 9 and Allied Sciences, PO Box 1464, Mwanza,Draft Tanzania. 10
11 3Department of Obstetrics and Gynaecology, Cumming School of Medicine, University of
12 Calgary, 3330 Hospital Drive NW Calgary, AB, Canada T2N 1N4.
13
14 *Corresponding Author:
15 Daniela Urrego
16 Room 280 Heritage Medical Research Building
17 3330 Hospital Drive NW, Calgary, AB, Canada, T2N 4N1
19
20
1 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 2 of 42
21
22
Draft
2 https://mc06.manuscriptcentral.com/cjpp-pubs Page 3 of 42 Canadian Journal of Physiology and Pharmacology
23 Abstract
24 Preterm birth (<37 weeks of gestation) significantly increases the risk of neonatal mortality and
25 morbidity. As many as half of all preterm births occur following spontaneous preterm labour.
26 Since in such cases there are no known reasons for the initiation of labour, treatment of preterm
27 labour (tocolysis) has sought to stop labour contractions and delay delivery. Despite some
28 success, the use of cyclooxygenase (COX) inhibitors is associated with maternal/fetal side effects,
29 and possibly increased risk of preterm birth. Clinical use of these drugs predates the collection
30 of molecular and biochemical evidence in vitro, examining the expression and activity of COX 31 enzymes in pregnant uterine tissues withDraft and without labour. Such evidence is important to the 32 rationale that COX enzymes are, or are not, appropriate targets for the tocolysis. The current
33 study systematically searched existing scientific evidence to address the hypothesis that COX
34 expression/activity is increased with the onset of human labour, in an effort to determine whether
35 there is a rationale for the use of COX inhibitors as tocolytics. Our review identified 40 studies,
36 but determined that there is insufficient evidence to support or refute a role of COX-1/-2 in the
37 onset of preterm labour that supports COX-targetted tocolysis.
38
39 Keywords: Cyclooxygenase, Cyclooxygenase inhibitors, Preterm, Labour, Pregnancy,
40 Prostaglandins, Tocolysis, Myometrium, Decidua, Fetal Membranes
41
42 Introduction
3 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 4 of 42
43 Preterm birth (<37 weeks) is associated with increased risk of neonatal mortality and morbidity,
44 and occurs in 10% of all pregnancies (Purisch and Gyamfi-Bannerman 2017). Of these
45 approximately half are due to the onset of spontaneous labour contractions (Ananth and
46 Vintzileos 2006; Goldenberg et al. 2008). Key treatments for preterm labour have focused on the
47 prevention or inhibition of myometrial contractions (tocolytics), mainly to provide time to
48 administer steroids to aid fetal lung maturation, and transfer to a special neonatal care unit
49 (Vogel et al. 2014; Haram et al. 2015). One tocolytic agent used to treat preterm labour is
50 indomethacin, a non-steroidal anti-inflammatory drug (NSAID) (Zuckerman et al. 1974). 51 Indomethacin decreases production ofDraft the prostaglandins, implicated to be important in the 52 labour process (Wiqvist et al. 1974; Vane and Williams 1973; Olson et al. 1995). The mechanism
53 of action of indomethacin, and other NSAIDs, is the inhibition of cyclooxygenase (COX), a key
54 enzyme in prostaglandin synthesis, thereby reducing prostaglandin production (Figure 1) (Vane
55 1971; Flower and Vane 1974; Vane and Botting 1998). A second COX enzyme, COX-2 (PTGS2,
56 official gene symbol) was subsequently identified and shown to be rapidly inducible by
57 inflammatory mediators, in contrast to COX-1 (PTGS1, official gene symbol) that is generally
58 constitutively expressed (Hla and Neilson 1992; Appleby et al. 1994; Needleman and Isakson
59 1997). The existence of two COX enzymes, each with distinct regulation and function, supported
60 a rationale for selectively inhibiting COX-2 over COX-1, and thus inhibiting inflammatory versus
61 constitutive prostaglandin production (Vane and Botting 1995; Mitchell et al. 1993).
4 https://mc06.manuscriptcentral.com/cjpp-pubs Page 5 of 42 Canadian Journal of Physiology and Pharmacology
62 Indeed, administration of the non-selective COX inhibitor, indomethacin, to women in
63 preterm labour led to a dramatic cessation of uterine contractions (Zuckerman et al. 1974; Gyory
64 et al. 1974). This approach was met with optimism since preterm labour was successfully delayed
65 for up to seven days. However, subsequent studies revealed that indomethacin use is associated
66 with a variety of maternal and fetal side effects, including oligohydramnios, premature closure
67 of the ductus arteriosus, necrotizing enterocolitis and intraventricular hemorrhage (Norton et al.
68 1993; Sood et al. 2011; Van der Veyver et al. 1993). These side effects were thought to be related
69 to COX-1 inhibition, thus the use of COX-2 selective inhibition was explored (Needleman and 70 Isakson 1997; Smith et al. 1996; Vane andDraft Botting 1995). 71 The first clinical report of COX-2 selective inhibition for preterm birth prevention was
72 described by Sawdy et al. (1997). In a pregnant patient with recurrent miscarriage, the COX-2
73 selective inhibitor nimesulide was administered prophylactically from 16 weeks gestation.
74 Nimesulide treatment was stopped at 34 weeks and delivery occurred six days later (Sawdy et al.
75 1997). This single case study report offered new hope, and rationale for the use of COX-2
76 selective inhibitors as tocolytics to delay preterm birth. However, subsequent reports described
77 severe neonatal renal damage after prolonged maternal exposure to nimesulide, suggesting
78 COX-2 is important for normal fetal kidney development and function, and not only involved in
79 pathogenic inflammation (Peruzzi et al. 1999; Balasubramaniam 2000; Magnani et al. 2004). Most
80 concerning, in a clinical trial with women at high risk of preterm delivery, the use of rofecoxib
81 (COX-2 selective inhibitor), not only failed to prevent, but further increased the risk of preterm
5 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 6 of 42
82 birth (Groom et al. 2005). More recent studies have also found that exposure to COX inhibitors,
83 and especially COX-2 selective inhibitors, in late pregnancy increase the risk of premature
84 delivery (Berard et al. 2018). Still, other groups report the possible benefits of low dose aspirin
85 taken prophylactically to reduce the risk of preterm delivery (Andrikopoulou et al. 2018).
86 Current obstetric practice generally suggests limiting indomethacin as a tocolytic, to
87 pregnancies <32 weeks, for temporary (<48 hours) delay of preterm birth (ACOG, 2016). Since
88 COX-inhibition for acute tocolysis is not fully effective, opportunities for further refinement of
89 this approach remain (Reinebrant et al. 2015). There is a paucity of scientific evidence to make 90 solid recommendations on the clinical useDraft of either selective or non-selective COX inhibitors to 91 prevent or treat preterm labour (Khanprakob et al. 2012). To this end, molecular/biochemical in
92 vitro studies sought to determine whether COX enzymes are active and/or upregulated with
93 labour. Studies reporting increased expression and/or activity of either COX-1 and/or COX-2
94 enzyme contrast widely when comparing the gestational tissues, patient groups studied, and
95 methodologies used. That prostaglandins can elicit labour contractions is clear, but whether, or
96 not, inhibition of either COX-1 and/or COX-2, key enzymes in prostaglandin synthesis, provide
97 the best target for tocolysis is uncertain. Therefore, we have systematically evaluated the in vitro
98 scientific literature, to interrogate the evidence that supports or refutes a role of COX-1 and/or
99 COX-2 in the initiation of preterm labour.
100
101 Methods
6 https://mc06.manuscriptcentral.com/cjpp-pubs Page 7 of 42 Canadian Journal of Physiology and Pharmacology
102 Literature Selection. A systematic search was designed to source studies investigating a role of
103 COX-1 and/or -2 in labour, in human pregnant term and preterm tissues. Search terms listed in
104 Table 1 were used in a broad free-text search of original, English language literature to identify
105 articles published prior to January 2018 from MEDLINE and EMBASE databases. This search
106 strategy produced 3958 references (Figure 2). To evaluate the inclusion of articles in the review
107 two investigators (D.U. and A.C.L) independently assessed the articles identified. In cases where
108 a consensus could not be reached, a third evaluator (D.M.S.) was consulted. After the removal of
109 145 duplicate references, the remaining study abstracts were reviewed for inclusion based on 110 whether they studied the involvement ofDraft COX in the onset of labour. Studies that met the initial 111 inclusion criteria were excluded if the study: was not a molecular/biochemical scientific
112 investigation (e.g. review, letter, clinical trial); was conducted on non-human tissues/samples; did
113 not compare between labour and non-labour tissues, or; did not examine human amnion,
114 chorion, decidua, and/or myometrium. In the current review, studies on the placenta were
115 excluded since they focused on COX in the etiology of hypertension in pregnancy rather than
116 spontaneous preterm labour. Eleven studies were identified for inclusion based on these criteria
117 and were used for full-text assessment. Thirty-three additional manuscripts that met inclusion
118 criteria were identified by manually searching reference lists of included articles.
119
120 Data extracted. Relevant data pertaining to COX expression (RNA and protein) and/or enzyme
121 activity was extracted from the included literature. Additional information extracted from each
7 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 8 of 42
122 study, included; methods used, tissues studied, clinical groups from which tissues were obtained
123 and gestational age. Once data extraction was complete, findings from each study were
124 qualitatively represented to summarize trends in COX-1 or COX-2 expression and or activity,
125 comparing: labour and non-labour, preterm and term gestational ages (Tables 2-4;
126 Supplementary Table 1).
127
128 Results
129 Studies on Fetal Membranes. The majority (29/44) of articles identified focused on COX 130 expression or activity in the fetal membranesDraft (amnion/chorion/decidua, examined individually 131 or in combination), obtained and removed from the placenta following vaginal delivery or
132 caesarean section (Figure 3, Supplementary Table 1). Of these studies, total COX enzyme
133 activity was assessed in 15 of the 29 membrane papers (Table 2); none of the enzyme activity
134 papers distinguished between COX-1 and COX-2, rather total COX or prostaglandin production
135 was assessed. Amnion, chorion and decidua were assessed separately or in combination; of the
136 COX enzyme activity papers, nine assessed amnion, two chorion, and three decidua, alone, while
137 one combined amnion-chorion and two chorion-decidua. In amnion, COX enzyme activity was
138 reported higher in; term labour (TL) compared to term non-labour (TNL) in seven studies, and
139 preterm labour (PTL) compared to preterm non-labour (PTNL) in three studies. In the chorion
140 studies, one reported higher activity in TL versus TNL, and one in PTL compared to PTNL. In the
8 https://mc06.manuscriptcentral.com/cjpp-pubs Page 9 of 42 Canadian Journal of Physiology and Pharmacology
141 decidua-only studies no differences were reported. Higher COX activity in TL compared to TNL
142 was reported in the combined amnion-chorion and chorion-decidua (Table 2).
143 Expression of COX-1 and/or COX-2 RNA and/or protein were described in 27 of 29 fetal
144 membrane studies; 18 included RNA expression and 12 protein expression (Figure 3, Table 3).
145 Expression of COX-1 RNA was reported higher in; TL compared to TNL in two amnion studies,
146 and in PTL compared to PTNL in one amnion and one chorion study. All other COX-1 studies
147 reported no labour changes. Expression of COX-2 mRNA was reported higher in: TL compared
148 to TNL in 11 studies (amnion, chorion, decidua alone or chorion-decidua combined), and PTL 149 compared to PTNL in four studies (threeDraft in amnion and one chorion). All other studies reported 150 no change with labour.
151 Changes in COX expression or activity, with gestational age were assessed in nine studies.
152 Five reported an increase in COX expression/activity with increased gestational age and one
153 reported higher COX activity (PGE2 production) in term compared to preterm non-labour. One
154 study by Makino et al. (2007) reported higher preterm versus term enzyme activity, assessed by
155 measuring PGF2 production.
156
157 Studies on Myometrium. Sixteen studies analyzed myometrial COX-1 and/or COX-2 and RNA
158 or protein expression; none reported on COX enzyme activity (Figure 3, Supplementary Table
159 1). Of the seven articles analyzing COX-1 expression, none reported labour or gestational age-
160 associated differences (Table 4). Expression of COX-2 was analyzed in all 15 studies; eight
9 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 10 of 42
161 reported higher COX-2 in TL compared to TNL, and three in PTL compared to PTNL. One study
162 stated higher COX-2 protein in non-labour samples compared to labour, however this study was
163 qualitative not quantitative (Zuo et al. 1994; Table 4). None of the studies on COX-1 reported
164 gestational age-associated changes. Expression of COX-2 was reported higher with gestational
165 age in two studies, and at term compared to preterm in four studies, of which two compared TL
166 to PTL. One study by Mitsuya et al. (2014) reported higher COX-2 expression in PTL compared
167 to TL.
168 169 Discussion Draft 170 Prostaglandins are important mediators in the process of parturition (Challis et al. 2002).
171 Increased prostaglandin levels observed with labour (Durn et al. 2010), and clinical use of
172 prostaglandin E2 and F2 to induce and augment labour (Sanchez-Ramos 2005), provide some of
173 the rationale to lower prostaglandin levels to decrease or slow down the labour process. Thus,
174 the inhibition of prostaglandin synthesis for the treatment of preterm labour progressed. Indeed,
175 the use of indomethacin, a non-selective cyclooxygenase inhibitor, for treating preterm labour
176 was first reported by Zuckerman et al. (1974). Later, following the discovery of the human COX-
177 2 gene (Hla and Neilson, 1992), nimesulide, for the selective inhibition of COX-2 over COX-1, was
178 reported to successfully prevent preterm birth in a single case study (Sawdy et al. 1997).
179 However, inhibition of the COX enzymes for preventing or treating preterm labour is not wholly
180 successful and is not without potential serious side effects. Notwithstanding, prostaglandins
10 https://mc06.manuscriptcentral.com/cjpp-pubs Page 11 of 42 Canadian Journal of Physiology and Pharmacology
181 have a clear role in labour and COX inhibition may still provide a suitable tocolytic target, albeit
182 limited. Questions remain as to which of the COX enzymes are important for prostaglandin
183 production in human pregnancy and labour, and which, if any, of the COX enzymes drive the
184 labour process at term or preterm. It is possible that patients undergoing spontaneous preterm
185 labour are a heterogenous group with different labour etiologies, some of which could therefore
186 benefit from COX-targeted tocolysis over others whose labour is not caused by increased COX-
187 1 or -2. In the current study we have systematically reviewed the literature describing the
188 scientific in vitro evidence for the role of COX-1 and/or COX-2 in human labour to try to address 189 these questions. While animal models suchDraft as mouse and rat have been instrumental in the study 190 of pregnancy and labour, interpreting data pertaining to the role of COX-1/-2 in preterm labour
191 is complicated. Commonly, studies on mouse and rat induce preterm labour by administering
192 lipopolysaccharide endotoxin or interleukin-1, which may directly increase COX-2 expression
193 (Haddad et al. 2006; Nielsen et al. 2016). This complicates the use of animal models as evidence
194 that human spontaneous preterm labour is caused by increased COX-2 expression/activity. By
195 reviewing the literature on human samples we have highlighted some key areas, which we hope
196 will generate discussion as to what further steps can be taken by the scientific community to
197 refine tocolytic strategies, and determine whether there are instances in which inhibiting COX is
198 a justified method of tocolysis.
199 Upon examination of the results generated (Tables 2-4) it is interesting to note the large
200 swathes of grey boxes, indicating that a tissue or clinical subgroup was not investigated. The
11 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 12 of 42
201 earliest paper to meet the inclusion criteria published in 1987 (Bryant-Greenwood et al. 1987),
202 was over a decade after the first use of indomethacin to treat preterm labour (Zuckerman et al.
203 1974). It is also worth noting that less than half of the papers reviewed included preterm labour
204 comparison groups; most papers focused on term labour versus non-labour comparisons from
205 which results were extrapolated to interpret the role of COX in preterm labour.
206 While some reports included in this systematic synthesis demonstrate increased COX-1
207 or COX-2 activity and or expression with the onset of labour, we suggest there remains
208 insufficient evidence to support, or refute, an absolute role of COX-1 or COX-2 in the onset of 209 labour, preterm or term. Of the 44 articlesDraft reviewed (20 of which studied preterm labour), only 210 seven reported significantly higher COX-2 expression with preterm labour in either fetal
211 membranes (four studies) or myometrium (three studies). Higher COX-2 expression was also
212 observed in term labour compared to term non-labour in amnion (nine studies), or myometrium
213 (eight studies), but not significant in all other studies. The variability in reported findings, from
214 study to study, may be accounted for by the: range of scientific techniques employed, types of
215 tissue samples assessed, gestational age range of sample donors, differences in sample retrieval,
216 and clinical phenotype of the preterm births. Future studies on the mechanisms of preterm
217 labour should seek to limit the impact of these factors on the interpretation and reproducibility
218 of results.
219
12 https://mc06.manuscriptcentral.com/cjpp-pubs Page 13 of 42 Canadian Journal of Physiology and Pharmacology
220 Scientific Techniques and Evolution of Measurement Approaches. Techniques for assessing
221 gene expression have continued to evolve, and especially within the time frame of the papers
222 reviewed. As an example, four of the papers described changes in COX-1 expression with labour
223 (Mijovic et al. 1998; Mijovic et al. 1999; Bennett et al. 1992; Bennett et al. 1993). However, the
224 Northern Blot and ribonuclease protection assays used in these studies are less specific, and due
225 to the high sequence homology between the COX-1 and COX-2 genes, the COX-1-targetting
226 cDNA probes used likely detected expression of either or both genes. Comparing these with the
227 highly sensitive and specific technology, such as conventional or real-time, reverse transcription 228 polymerase chain reaction (RT-PCR) utilizedDraft later is difficult. 229 Additionally, over half of the studies identified examined only RNA expression which tells
230 only a small part of the story. To generate active protein, in this case COX enzymes, any RNA
231 transcribed and generated requires subsequent translation. A high level of RNA expression does
232 not absolutely result in increased protein and enzyme activity. Since functional COX enzymes are
233 the targets for tocolysis, studies including protein expression or enzyme activity should be of
234 interest. Adding to the complexity for data interpretation was the description of the newly
235 identified human COX-2 gene (Hla and Neilson, 1992). In early studies the antibodies used for
236 western blot and immunohistochemistry may not have been selective, and it is difficult to discern
237 whether results depict COX-1 or COX-2. The studies identified were generally piecemeal, with
238 none investigating concomitant RNA, protein and enzyme activity in fetal membranes and
239 myometrium, and none selectively distinguishing COX-1 and COX-2 enzyme activity.
13 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 14 of 42
240
241 Types of Tissues Studied. The role of the COX enzymes is likely to be tissue specific. Expression
242 of COX within the myometrium may facilitate increased prostaglandin production and enhanced
243 uterine contractions. In contrast, expression of COX and prostaglandin production in the fetal
244 membranes may facilitate cervical ripening in regions overlaying the cervix and at the site of
245 membrane rupture. Appropriate interpretation of any COX changes with labour should consider
246 the context of tissue location. Over half of the studies reviewed here focused on COX in the
247 amnion which likely plays a different role than the myometrium, chorion or the decidua in labour. 248 The myometrium is the contractile componentDraft of the uterus in labour, and expresses receptors 249 for various prostaglandins that can elicit contractions (Astle et al. 2005) The decidua is in close
250 contact with the myometrium and decidual prostaglandins may act directly on the myometrium.
251 However, it is unclear whether prostaglandins produced by the amnion and chorion act on the
252 myometrium to produce contractions, or whether they act locally to contribute to the weakening
253 of the membranes prior to rupture (McCoshen et al. 1990; Chowdhury et al. 2014).
254
255 Spatio-Temporal Roles of Prostaglandins. Increased COX expression by uterine tissues may
256 not only produce prostaglandins for uterine contraction, but also for the resolution of labour
257 and uterine involution back to a pre-pregnancy state. Emerging evidence suggests that
258 inflammatory resolution involves a prolonged increase in prostaglandin production and COX
259 expression (Rajakariar et al. 2006; Newson et al. 2017). Types of prostaglandins produced, and
14 https://mc06.manuscriptcentral.com/cjpp-pubs Page 15 of 42 Canadian Journal of Physiology and Pharmacology
260 receptors expressed may differ between stages of delivery (i.e. onset of labour, resolution) to
261 modify the effects of prostaglandins (MacDonald and Casey 1993). Such differences may also
262 occur spatially, between the upper and lower segments. Only three papers reviewed here studied
263 paired upper and lower myometrium (Sparey et al. 1999; Sooranna et al. 2006; Tatterstall et al.
264 2008), all of which found higher COX-2 expression in the lower segment.
265 Regionalization of both tissue function and gene expression may also occur in the fetal
266 membranes, as a result of tensile strain distribution through the membrane. Malak and Bell
267 (1994) describe regionalized histological differences at the rupture site compared to the rest of 268 the membrane. Localized differences inDraft membrane strength are accompanied by enhanced 269 collagen remodelling and apoptotic processes that contribute to membrane weakening in
270 preparation for rupture, and precede the onset of labour at term (El Khwad et al. 2005;
271 Chowdhury et al. 2014). Few of the studies reviewed here indicate the location from which fetal
272 membranes were sampled, so there is no ability to discern whether COX-1/-2 expression
273 differences reported truly result from the onset of labour, or from regional variation in sampling.
274 Understanding the influence of both spatial and temporal effects on COX-1/-2 expression in fetal
275 membranes at these different locations, may provide insight into how the prostaglandin pathway
276 plays a role in fetal membrane rupture versus the activation of labour contractions, which should
277 inform decisions on when and whether COX is truly an appropriate target for tocolysis.
278 To accurately characterize the dynamic prostaglandin production at the end of pregnancy
279 it is therefore important to study paired uterine tissues from defined regions and gestational
15 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 16 of 42
280 stages, with and without labour, and to examine not only COX expression, but the expression of
281 other prostaglandin synthetic enzymes, and specific prostaglandin production. This would
282 provide a clearer picture of the spatial and temporal nature of prostaglandin production, that
283 could then be used to clearly identify which, if any, enzyme plays a significant role in the initiation
284 of preterm labour, and which tissues should be targeted for tocolysis.
285
286 Gestational Age Groups. A role for COX in pregnancy and labour is suggested, yet specific roles
287 for COX are unclear. One reason could be the lack of clarity in the gestational ages and 288 classification of labour. In most studies Draftreviewed, term and preterm labour groups were defined 289 by a range of gestational ages, but the composition (mean, SEM) of each group was not always
290 indicated. Term is generally defined as after 37 weeks and below 42 weeks gestation, and
291 preterm as 24 weeks to below 37 weeks gestation. A small number of studies investigated fetal
292 membranes from terminations prior to 24 weeks of gestation (Teixeira et al. 1993; Teixeira et al.
293 1994; Mijovic et al. 1998). The inclusion of such early gestational ages complicates within-study
294 comparisons between preterm labour and non-labour groups, since the ages included in each
295 group may vary, obscuring the effect of labour alone on COX expression. Standardization and
296 clearer delineation of gestational age ranges would be a positive move forward for future studies.
297 For example, defining the subgroups into; very early (24 weeks to <28 weeks), early (28 to <32
298 weeks), moderate (32 to <34 weeks) and late (34 to < 36+6 day) preterm births, will allow
16 https://mc06.manuscriptcentral.com/cjpp-pubs Page 17 of 42 Canadian Journal of Physiology and Pharmacology
299 researchers to account for possible differences that may exist in the physiological processes and
300 etiology of labour at each of these stages (Ananth et al. 2013).
301
302 Clinical Details Provided and Classifications of Labour. A fourth issue identified by the review
303 was the general lack of clinical information. Clinical details recommended to be considered in
304 the study design, and for interpretation of findings include: reasons for caesarean section
305 delivery (i.e. failure to progress, cephalopelvic disproportion, fetal distress, elective, etc.),
306 maternal conditions (i.e. gestational diabetes, hypertension, etc.), and use of drugs prior to 307 delivery (i.e. oxytocin, prostaglandins, corticosteroids,Draft etc.). 308 In many cases, details surrounding the definition and duration of labour were either not
309 described, or defined differently from study to study. Twenty-seven of the 29 reviewed studies
310 on fetal membranes obtained “labour” samples from placentas obtained following vaginal
311 delivery, compared to those obtained from “non-labour” caesarean section. Over half of the
312 studies on vaginally delivered tissues reported positive findings that suggested increased COX
313 expression or activity with the onset of labour. In contrast, only one of the two fetal membrane
314 studies that obtained labour samples by caesarean section found higher COX-2 with preterm
315 labour (Sadovsky et al. 1999; Slater et al. 1999a). Perhaps the higher observed COX expression
316 in labour groups in the studies on vaginally delivered tissue resulted from prolonged exposure
317 of those tissues to the active labouring intrauterine environment. In support of this, Phillips et al.
318 (2014) reported labour ranges from 33 minutes to 17 hours in their study, and a positive
17 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 18 of 42
319 correlation between COX-2 expression and the duration of labour. The increased COX-2 may be
320 a consequence of prolonged exposure to other upstream mediators of labour and labour-
321 associated signaling molecules (e.g. cytokines, chemokines, uterotonins, prostaglandins), which
322 are known to further upregulate COX gene expression. Likewise, other groups have reported
323 differential gene expression in tissues acquired from ruptured versus intact membranes, and with
324 preterm birth phenotypes that may have inflammatory etiologies (Kuc et al. 2012). Without
325 differentiating the samples based on early versus late labour, long versus short labours,
326 spontaneous membrane rupture, and other clinical characteristics that can impact inflammatory 327 gene expression, it is difficult to accuratelyDraft describe the relationship between COX and the onset 328 of labour. Sub-classifying patient groups may provide further insight into the trajectory of COX
329 expression and prostaglandin production leading up to and through labour.
330
331 Study Design and Sample Sizes. Difficulty comparing results from differently defined labour
332 and gestational age groups in each study is further compounded by the relatively small sample
333 sizes in these studies. For example, studies on term tissues included as few as four, and as many
334 as 27 patients per group, while studies on preterm tissues included four to 16 patients per group.
335 These small group sizes and large discrepant group sizes within and between studies complicate
336 data analysis and interpretation, since in most cases it is likely that studies were underpowered
337 to detect meaningful differences between groups. However, it is also possible that the inclusion
338 of certain patient populations in some groups or studies versus others (e.g. earlier gestational
18 https://mc06.manuscriptcentral.com/cjpp-pubs Page 19 of 42 Canadian Journal of Physiology and Pharmacology
339 ages, patients in labour for longer, indications for preterm delivery.) may overestimate the true
340 differences between groups, but were generalized to the broadly defined groups stated. This is
341 especially true of studies in which not all samples were used in every analysis, for undescribed
342 reasons.
343 Additionally, comparisons were made predominantly using t-test and one-way ANOVA
344 (parametric and non-parametric as appropriate), to compare the effects of the presence or
345 absence of labour and term versus preterm. By taking into account more clinical information for
346 each sample, future studies should be poised to conduct statistical analyses that consider more 347 variables that can impact COX expression;Draft for example, clearly defined gestational age groups, 348 rupture of membranes, duration of labour, and cervical dilation. Inclusion of this information in
349 adequately powered studies would allow for more complex analyses of the data using, for
350 example, regression, two-way ANOVA/ANCOVA to inform more meaningful interpretations
351 regarding the trajectory of COX expression through late pregnancy and labour. Design and
352 completion of large, well-powered studies with robustly described comparison groups would
353 provide a clearer view of the effect size of each variable on the expression/activity of COX
354 enzymes, and moving forward potentially allow for meta-analyses to compare the basic science
355 literature on this topic. Such comparisons are required to generate clarity on the circumstances
356 under which COX expression is implicated in preterm labour. The evidence reviewed here is
357 inconclusive, as there is a lack of uniformity in the way patient groups/populations were defined
358 and how data were collected, which precludes the ability to conduct a meta-analysis.
19 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 20 of 42
359
360 The “-Omics” Era. Modern “omic” approaches that simultaneously analyze the expression of
361 thousands of different biochemical products (e.g. RNA, protein) lend themselves well to study
362 designs that can be included in meta-analyses (Sharp et al. 2016). While the current review did
363 not include any meta-analyses, or some “omic” studies not captured with our search criteria,
364 these data approaches should also be considered. Overall, many of the existing gene array, and
365 transcriptome-wide studies inconsistently report significant increases in COX-2 expression with
366 the onset of labour (Supplementary Table 2); only four of the 12 studies identified report 367 labour-associated increases in COX-2 expressionDraft and none report changes in COX-1 expression. 368 Furthermore, many of these studies exhibit similar limitations as the studies we reviewed, such
369 as analysis on small sample sizes, and the use of broadly defined patient groups. Comparison
370 between the studies is also difficult since differences in specificity and precision of assays (e.g.
371 probe-based arrays versus RNAseq) is variable.
372 The first transcriptome-wide study on term labour versus non-labour myometrium was
373 described by Chan et al. (2013), and reported an almost 9-fold increase in COX-2 RNA expression
374 with labour, but did not analyze preterm samples. Though this study incorporated state of the
375 art RNAseq, it along with other “omic” studies are difficult to evaluate in comparison to the rest
376 of the literature, due to the small sample size, and incomplete descriptions of clinical
377 characteristics of the patient groups included. In contrast, a more recent microarray study of
378 labouring and non-labouring myometrium did not find any difference in the expression of COX-1
20 https://mc06.manuscriptcentral.com/cjpp-pubs Page 21 of 42 Canadian Journal of Physiology and Pharmacology
379 or -2 with the onset of labour (Sharp et al. 2016). This study exhibited several strengths that align
380 with the recommendations we have made for research on the etiologies of spontaneous preterm
381 labour moving forward. Notably, Sharp et al. (2016) included larger sample size of well-described
382 matched patient groups to limit the overrepresentation of clinical characteristics in one group
383 versus another. Future research should expand on the work of Sharp et al. (2016), and utilize
384 multiple approaches to detect gene/protein expression/activity in paired tissues from well
385 defined patient populations, in order to appropriately phenotype the cases of preterm labour in
386 which COX enzymes are, and are not causally implicated. 387 Draft 388 Conclusion. In conclusion, the current systematic review did not find sufficient evidence for a
389 role of COX enzymes in the initiation of preterm labour, that would warrant the use of COX-
390 targeted tocolysis. The review emphasizes a lack of detailed information as to the function of
391 COX in the human uterus. Evidence within the publications included herein provide hints that
392 the COX enzymes are suitable targets for tocolysis but certainly not in all cases. Based on recent
393 developments in technology and our understanding that preterm labour is multifactorial,
394 additional research is needed to elucidate (once and for all) whether COX-1 and or COX-2 are
395 upregulated with term or preterm labour. Clear phenotyping and analysis of the different clinical
396 subgroups, and where possible studying all relevant gestational tissues, will provide a clearer
397 picture, and help delineate in what, if any, circumstances COX enzymes are involved in the
398 etiology of spontaneous preterm labour. Though the side effects with COX-targeted tocolytics
21 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 22 of 42
399 will remain, more complete understanding of the onset of preterm labour will help physicians
400 determine whether there are patient populations that stand to benefit from COX inhibition for
401 tocolysis, outweighing the risk of preterm birth. Furthermore, we hope that future research into
402 the mechanisms of spontaneous preterm labour will utilize approaches that simultaneously
403 explore multiple biological targets in different patient subsets, that may then justify different
404 approaches for preterm labour management and tocolysis. In this way, tocolytics can be applied
405 in a more targeted manner, as opposed to expecting that they are a catch-all solution.
Draft
22 https://mc06.manuscriptcentral.com/cjpp-pubs Page 23 of 42 Canadian Journal of Physiology and Pharmacology
407 References
408 References included in the analysis (Tables 2-4 and Supplementary Table 1) denoted by *
409 References included in the (Supplementary Table 2) denoted by **
410
411 **Aguan, K., Carvajal, J.A., Thompson, L.P., Weiner, C. 2000. Application of a functional genomics 412 approach to identify differentially expressed genes in human myometrium during pregnancy and labour. 413 Mol. Hum. Reprod. 6(12): 1141-1145. 414 415 American College of Obstetrics and Gynecologists’ (ACOG) Committee on Practice Bulletins – 416 Obstetrics. 2016. Practice bulletin No. 171: Management of preterm labour. Obstet. Gynecol. 128(4): 417 e155-64. doi: 10.1097/AOG.0000000000001711 418 419 Ananth, C.V., Friedman, A.M., Gyamfi-Bannerman, C. 2013. Epidemiology of moderate preterm, late 420 preterm and early term delivery. Clin. Perinatol. 40(4): 601-10. doi: 10.1016/j.clp.2013.07.001 421 422 Ananth, C.V. and Vintzileos, A.M. 2006. Draft Epidemiology of preterm birth and its clinical subtypes. J. 423 Matern. Fetal. Neonatal Med. 19(12):773-82. doi: 10.1080/14767050600965882 424 425 Andrikopoulou, M., Purisch, S.E., Handal-Orefice, R., and Gyamfi-Bannerman, C. 2018. Low dose 426 aspirin is associated with reduced spontaneous preterm birth in nulliparous women. Am. J. Obstet. 427 Gynecol. Jun 15. pii: S0002-9378(18)30498-8. doi: 10.1016/j.ajog.2018.06.011. [Epub ahead of print] 428 429 Appleby, S.B., Ristimäki, A., Neilson, K., Narko, K., and Hla, T. 1994. Structure of the human cyclo- 430 oxygenase-2 gene. Biochem. J. 302(Pt 3): 723-7. 431 432 *Astle, S., Newton, R., Thornton, S., Vatish, M., and Slater, D.M. 2007. Expression and regulation of 433 prostaglandin E synthase isoforms in human myometrium with labour. Mol. Hum. Reprod. 13(1): 69-75. 434 doi: 10.1093/molehr/gal093 435 436 Astle, A., Thornton, A., Slater, D.M. 2005. Identification and localization of prostaglandin E2 receptors 437 in upper and lower segment human myometrium during pregnancy. Mol. Hum. Reprod. 11(4): 279-87. 438 doi: 10.1093/molehr/gah158 439 440 Balasubramaniam, J. 2000. Nimesulide and neonatal renal failure. Lancet. 355: 575. DOI: 10.1016/S0140- 441 6736(05)73223-2 442 443 *Bennett, P.R., Henderson, D.J., and Moore, G.E. 1992 Changes in expression of the cyclooxygenase 444 gene in human fetal membranes and placenta with labor. Am. J. Obstet. Gynecol. 167(1): 212-6. 445
23 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 24 of 42
446 *Bennett, P.R., Slater, D., Sullivan, M., Elder, M.G., and Moore, G.E. 1993. Changes in amniotic 447 arachidonic acid metabolism associated with increased cyclo-oxygenase gene expression. Brit. J. Obstet. 448 Gynaecol. 100: 1037-1042. 449 450 Berard, A., Sheehy, O., Girad, S., Zhao, J-P., Bernatsky, S. 2018. Risk of preterm birth following late 451 pregnancy exposure to NSIADs or COX-2 inhibitors. Pain. 150:948-955. 452 453 **Bethin, K.E., Nagai, Y., Sladek, R., Asada, M., Sadovsky, Y., Hudson, T.J., et al. 2003. Microarray 454 analysis of uterine gene expression in mouse and human pregnancy. Mol. Endocrinol. 17(8): 1454-1469. 455 456 *Bollopragada, A., Yousseff, R., Jordan, F., Greer, I., Norman, J., Nelson, S. 2009. Term labor is 457 associated with a core inflammatory response in human fetal membranes, myometrium and cervix. Am. 458 J. Obstet. Gynecol. 200(1):104.e1-11. doi: 10.1016/j.ajog.2008.08.032 459 460 *Bryant-Greenwood, G.D., Rees, M.C.P., and Turnbull, A.C. 1987. Immunohistochemical localization of 461 relaxin, prolactin and prostaglandin synthase in human amnion, chorion and decidua. J. Endocr. 114(3): 462 491-496. 463 464 **Bukowski, R., Hankins, G.D., Saade, G.R., Anderson, G.D., Thornton, S. 2006. Labor-associated gene 465 expression in the human uterine fundus, lower segment, and cervix. PLoS Med. 3(6): e169. 466 Draft 467 **Bukowski, R., Sadowsky, Y., Goodarzi, H., Zhang, H., Biggio, J.R., Varner, M., et al. 2017. Onset of 468 human preterm and term birth is related to unique inflammatory transcriptome profiles at the maternal 469 fetal interface. Peer. J. doi:10.7717/peerj.3685 470 471 Challis, J.R.G., Sloboda, D.M., Alfaidy, M., Lye, S.J., Gibb, W., Patel, F.A., et al. 2002. Prostaglandins 472 and mechanisms of preterm birth. Reproduction. 124: 1-17. 473 474 **Chan, Y-W., van den Berg, H.A., Moore, J.D., Quenby, S., Blanks, A.M. 2014. Assessment of 475 myometrial transcriptome changes associated with spontaneous human labour by high-throughput 476 sequencing. Exp. Physiol. 99:510-524. 477 478 **Charpigny, G., Leory, M.J., Breuiller-Fouche, M., Tanfin, Z., Mhaouty-Kodja, S., Robin, P., et al. 2003. 479 A functional genomic study to identify differential gene expression in the preterm and term human 480 myometrium. Biol. Reprod. 68: 2289-2296. doi: 10.1095/biolreprod.102.013763 481 482 *Choi, S.J., Oh, S., Kim, J.H., Roh, C.R. 2007. Changes of nuclear factor kappa B (NF-kB), 483 cyclooxygenase-2 (COX-2) and matrix metalloproteinase-9 (MMP-9) in human myometrium before and 484 during term labor. Eur. J. Obstet. Gynecol. Reprod. Biol. 132(2): 182-188. 485 486 Chowdhury, B., David, A.L., Thrasivoulou, C., Becker, D.L., Bader, D.L., Chowdhury, T.T. 2014. 487 Tensile strain increased COX-2 expression and PGE2 release leading to weakening of the human amniotic 488 membrane. Placenta. 35:1057-1064. 489
24 https://mc06.manuscriptcentral.com/cjpp-pubs Page 25 of 42 Canadian Journal of Physiology and Pharmacology
490 Divers, M.J., Lilford, R.J., Miller, D., Bulmer, J.N. 1995. Cyclo-oxygenase distribution in human placenta 491 and decidua does not change with labour after term or preterm delivery. Gynecol. Obstet. Invest. 39:157- 492 61. doi: 10.1159/000292400 493 494 Durn, J.H., Marshall, K.M., Farrar, D., O’Donovan, P., Scally, A.J., Woodward, D.F., et al. 2010. 495 Lipidomic analysis reveals prostanoid profiles in human term pregnant myometrium. Prost. Leukot. 496 Essent. Fatty Acids. 82: 21-26. doi: 10.1016/j.plefa.2009.11.002 497 498 El Khwad, M., Stetzer, B., Moore, R.M., Kumar, D., Mercer, B., Arikat, S., et al. 2005. Term human fetal 499 membranes have a weak zone overlying the lower uterine pole and cervix before onset of labor. Biol. 500 Reprod. 72(3): 720-726. 501 502 *Erkinheimo, T.L., Saukkonen, K., Narko, K., Jalkanen, J., Ylikorkala, O., Ristimaki, A. 2000. Expression 503 of cyclooxygenase-2 and prostanoid receptors by human myometrium. J. Clin. Endo. Metab. 85(9): 3468- 504 3475. doi: 10.1210/jcem.85.9.6809 505 506 **Esplin, S.M., Manuck, T.A., Varner, M.W., Christensen, B., Biggio, J., Bukowski, R., et al. 2015. 507 Cluster analysis of spontaneous preterm birth phenotypes identifies potential associations among preterm 508 birth mechanisms. Am. J. Obstet. Gynecol. 429 e.1 doi.ord/10.1016/j.ajog.2015.06.011 509 510 Flower, R.J., and Vane, J.R. 1974. InhibitionDraft of prostaglandin synthesis. Biochem. Pharmacol. 23: 1439- 511 1450. 512 513 *Fuentes, A., Spaziani, E.P., O’Brien, W.F. 1996. The expression of cyclooxygenase-2 (COX-2) in 514 amnion and decidua following spontaneous labour. Prostaglandins. 52:261-7. 515 516 *Gaffney, R.C., Rice, G.E., and Brennecke, S.P. 1990. Is human labour triggered by an increase in the 517 rate of synthesis of prostaglandin G/H synthase? Reprod. Fertil. Dev. 2: 603-606. 518 519 *Giannoulias, D., Patel, F.A., Holloway, A.C., Lye, S.J., Tai, H.H., and Challis, J.R. 2002. Differential 520 changes in 15-hydroxyprostaglandin dehydrogenase and prostaglandin H synthase (types I and II) in 521 human pregnant myometrium. J. Clin. Endocrinol. Metab. 87(3): 1345-1352. doi: 10.1210/jcem.87.3.8317 522 523 *Gibb, W. and Sun M. 1996. Localization of prostaglandin H synthase type 2 protein and mRNA in term 524 human fetal membranes and decidua. J. Endocrinol. 150:497-503. 525 526 Goldenberg, R.L., Culhane, J.F., Iams, J.D., and Romero, R. 2008. Epidemiology and causes of preterm 527 birth. Lancet. 371(9606):75-84. doi: 10.1016/S0140-6736(08)60074-4. 528 529 Groom, K.M., Shennan, A.H., Jones, B.J., Seed, P., and Bennett, P.R. 2005. TOCOX- a randomized, 530 double-blind, placebo-controleed trial of refecoxib (a COX-2-specific prostaglandin inhibitor) for the 531 prevention of preterm delivery in women at high risk. Br. J. Obstet. Gynaecol. 112: 725-730. doi: 532 10.1111/j.1471-0528.2005.00539.x 533 534 Gyory, G.Y., Kiss, C.S., Benyo, T., Bagdany, S., Szalay, J., Kurcz, M., et al. 1974. Inhibition of labour 535 by prostaglandin antagonists in impending abortion and preterm and term labour. Lancet. 2(7875): 293.
25 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 26 of 42
536 537 Haddad, R., Gould, B.R., Romero, R., Tromp, G., Farookhi, R., Edwin, S.S., et al. 2006. Uterine 538 transcriptomes of bacteria-induced and ovariectomy-induced preterm labor in mice are characterized by 539 differential expression of arachidonate metabolism genes. Am. J. Obstet. Gynecol. 195: 822-828. doi: 540 10.1016/j.ajog.2006.06.053 541 542 Haram, K., Mortensen, J.H., and Morrison, J.C. 2015. Tocolysis for acute preterm labor: does anything 543 work. J. Matern. Fetal Neonatal Med. 28(4): 371-8. doi: 10.3109/14767058.2014.918095 544 545 **Havelock, J.C., Keller, P., Muleba, N., Mayhew, B.A., Casey, B.M., Rainey, W.E., et al. 2004. Human 546 myometrial gene expression before and during parturition. Biol. Reprod. 72: 707-719. doi: 547 10.1095/biolreprod.104.032979 548 549 *Hirst, J.J., Teixeira, F.J., Zakar, T., and Olson, D.M. 1995a. Prostaglandin H synthase-2 expression 550 increases in human gestational tissues with spontaneous labour onset. Reprod. Fertil. 7: 633-7. 551 552 *Hirst, J.J., Teixeira, F.J., Zakar, T., and Olson, D.M. 1995b. Prostaglandin endoperoxide-H synthase-1 553 and -2 messenger ribonucleic acid levels in human amnion with spontaneous labor onset. J. Clin. 554 Endocrinol. Metab. 80(2): 517-23. doi: 10.1210/jcem.80.2.7852513 555 556 *Hirst, J.J., Mijovic, J.E., Zakar, T., and Olson,Draft D.M. 1998. Prostaglandin endoperoxide H synthase-1 and 557 -2 mRNA levels and enzyme activity in human decidua at term labor. J. Soc. Gynecol. Invest. 5: 13-20. 558 559 Hla, T., and Neilson, K.1992. Human cyclooxygenase-2 cDNA. Proc. Natl. Acad. Sci. U S A. 89(16): 560 7384-8. 561 562 *Johnson, R.F., Mitchell, C.F., Giles, W.B., Walters, W.A., Zakar, T. 2002. The in vivo control of 563 prostaglandin H synthase-2 messenger ribonucleic acid expression in the human amnion at term. J. Clin. 564 Endocrinol. Metab. 87: 2816-23. doi: 10.1210/jcem.87.6.8524 565 566 Khanprakob, T., Laopaiboon, M., Lumbiganon, P., Sangkomkamhang, U.S. 2012. Cyclo-oxygenase 567 (COX) inhibitors for preventing preterm labour. Cochrane Database Syst Rev. 10:CD007748. doi: 568 10.1002/14651858.CD007748.pub2. 569 570 Kuc, P., Laudanski, P., Kowalczuk, O., Chyczewski, L., Laudanski, T. 2012. Expression of selected genes 571 in preterm premature rupture of fetal membranes. Acta Obstet. Gynecol. Scand. 91(8): 936-43. Doi: 572 10.1111/j.1600-0412.2012.01445.x. 573 574 Lee, D.C., Romero, R., Kim, J.S., Yoo, W., Lee, J.H., Mittal, P., et al. 2010. Evidence for a spatial and 575 temporal regulation of prostaglandin-endoperoxide synthase 2 expression in human amnion in term and 576 preterm parturition. J. Clin. Endocrinol. Metab. 95(9): E86-91. doi: 10.1210/jc.2010-0203 577 578 *Leguizamon, G., Smith, J., Younis, H., Nelson, M., Sadovsky, Y. 2001. Enhancement of amniotic 579 cyclooxygenase type 2 activity in women with preterm delivery associated with twins or polyhydramnios. 580 Am. J. Obstet. Gynecol. 87(6): 2816-23. doi: 10.1210/jcem.87.6.8524 581
26 https://mc06.manuscriptcentral.com/cjpp-pubs Page 27 of 42 Canadian Journal of Physiology and Pharmacology
582 MacDonald, P.C., and Casey, M.L.1993.The accumulation of prostaglandins (PG) in amniotic fluid is an 583 aftereffect of labor and not indicative of a role for PGE2 or PGF2 alpha in the initiation of human 584 parturition. J Clin. Endocrinol. Metab. 76(5): 1332-9. doi: 10.1210/jcem.76.5.8496326 585 586 Magnani, C. 2004. Neonatal chronic renal failure with maternal ingestion of Nimesulide as analgesic. Eur. 587 J. Obstet. Gynecol. Reprod. Biol. 116: 244-245. doi: 10.1016/j.ejogrb.2004.02.021 588 589 *Makino, S., Zaragoza, D.B., Mitchell, B.F., Yonemoto, H., and Olson, D.M. 2007. Decidual activation: 590 abundance and localization of prostaglandin F2alpha receptor (FP) mRNA and protein and uterine 591 activation proteins in human decidua at preterm birth and term birth. Placenta. 28: 557-565. doi: 592 10.1016/j.placenta.2006.06.010 593 594 Malak, T.M., Bell, S.C. 1994. Structural characteristics of term human fetal membranes: a novel zone of 595 extreme morphological alteration within the rupture site. Br. J. Obstet. Gynecol. 101:375-386 596 597 *Marques, F.Z., Pringle, K.G. Conquest, A., Hirst, J.J., Markus, M.A., Sarris, M., et al. 2011. Molecular 598 characterization of renin-angiotensin system components in human intrauterine tissues and fetal 599 membranes from vaginal delivery and cesarean section. Placenta. 32:214-221. 600 601 McCoshen, J.A., Hoffman, D.R., Kredenster, J.V., Araneda, C., Johnston, J.M. 1990. The role of fetal 602 membranes in regulating production, transport,Draft and metabolism of prostaglandin E2 during labor. Am. J. 603 Obstet. Gynecol. 163(5 Pt 1): 1632-40. 604 605 *Mijovic, J.E., Zakar, T., Nairn, T.K., and Olson, D.M. 1997. Prostaglandin-endoperoxide H synthase-2 606 expression and activity increases with term labor in human chorion. Am. J. Physiol. 272: E832-840. doi: 607 10.1152/ajpendo.1997.272.5.E832 608 609 *Mijovic, J.E., Zakar, T., Nairn, T.K., and Olson, D.M. 1998. Prostaglandin endoperoxide H synthase 610 (PGHS) activity and PGHS-1 and -2 messenger ribonucleic acid abundance in human chorion throughout 611 gestation and with preterm labor. J. Clin. Endocrinol. Metab. 83(4): 1358-1367. doi: 612 10.1210/jcem.83.4.4692 613 614 *Mijovic, J.E., Zakar, T., Angelova, J., and Olson, D.M. 1999. Prostaglandin endoperoxide H synthase 615 mRNA expression in the human amnion and decidua during pregnancy and in the amnion at preterm 616 labour. Mol. Hum. Reprod. 5(2): 182-187. 617 618 Mitchell, J.A., Akarasereenont, P., Thiemermann, C., Flower, R.J., and Vane, J.R. 1993. Selectivity of 619 nonsteroidal anti-inflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc. 620 Natl. Acad. Sci. U S A. 90(24):11693-7. 621 622 *Mitsuya, K., Singh, N., Sooranna, S.R., Johnson, M.R., Myatt, L. 2014. Epigenetics of human 623 myometrium: DNA methylation of genes encoding contraction-associated proteins in term and preterm 624 labor. Biol. Reprod. 90(5):98, 1-8. doi: 10.1095/biolreprod.113.113209 625
27 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 28 of 42
626 *Mittal, P., Romero, R., Tarca, A.L., Gonzalez, J., Draghici, S., Xu, Y., et al. 2010. Characterization of 627 the myometrial transcriptome and biological pathways of spontaneous human labor at term. J. Perinat. 628 Med. 38: 617-643. doi: 10.1515/JPM.2010.097 629 630 *Moore, S.D., Brodt-Eppley, J., Cornelison, L.M., Burk, S.E., Slater, D.M., and Myatt, L. 1999. 631 Expression of prostaglandin H synthase isoforms in human myometrium at parturition. Am. J. Obstet. 632 Gynecol. 180: 103-109. 633 634 Needleman, P., and Isakson, P.C. 1997. The discovery and function of COX-2. J. Rheumatol. Sup. 49: 635 6-8. 636 637 Newson, J., Motwani, M.P., Kendall., A.C., Nicolaou, A., Muccioli, G.G., Alhouayek, M., et al. 2017. 638 Inflammatory resolution triggers a prolonged phase of immune suppression through COX-1/mPGES-1- 639 derived prostaglandin E2. Cell Rep. 20(13): 3162-75. Doi: doi: 10.1016/j.celrep.2017.08.098 640 641 Nielsen, B.W., Bonney, E.A., Pearce, B.D., Donahue, L.R., Sarkar, I.N. 2016. A cross species analysis of 642 animal models for the investigation of preterm birth mechanisms. Reprod. Sci. 2(4): 482-491 643 644 Norton, M.E., Merill, J., Cooper, B.A., Kuller, J.A. and Clyman, R.I. 1993. Neonatal complications after 645 the administration of indomethacin for preterm labour. N. Eng. J. Med. 329: 1602-1607. doi: 646 10.1056/NEJM199311253292202 Draft 647 648 Olson, D.M., Mijovic, J.E., and Sadowsky, D.W. 1995. Control of human parturition. Sem. Perinatol. 19: 649 52-63. 650 651 Peruzzi, L., Gianoglio, B., Porcellini, M.G., and Coppo, R. 1999. Neonatal end-stage renal failure 652 associated with maternal ingestion of cyclo-oxygenase-type-2 selective inhibitor nimuselide as tocolytic. 653 Lancet. 354: 1615. doi: 10.1016/S0140-6736(99)03105-0 654 655 *Phillips, R.J., Al-Zamil, H., Hunt, L.P., Fortier, M.A. and Lopez Bernal, A. 2011. Genes for 656 prostaglandin synthesis, transport and inactivation are differentially expressed in human uterine tissues, 657 and the prostaglandin F synthase AKR1B1 is induced in myometrial cells by inflammatory cytokines. 658 Mol. Hum. Reprod. 17(1): 1-13. doi: 10.1093/molehr/gaq057 659 660 *Phillips, R.J., Fortier, M.A. and Lopez Bernal, A. 2014. Prostaglandin pathway gene expression in human 661 placenta, amnion and choriodecidua is differentially affected by preterm and term labour and by uterine 662 inflammation. BMC Pregnancy Childbirth 14: 241. doi:10.1186/1471-2393-14-241 663 664 *Price, T.M., Kauma, S.W., Curry Jr, T.E., Clark, M.R. 1989. Immunohistochemical localization of 665 prostaglandin endoperoxide synthase in human fetal membranes and decidua. Biol. Reprod. 41:701-5. 666 667 *Pringle, K.G., Zakar, T., Yates, D., Mitchell, C.M. Hirst, J.J., Lumbers, E.R. 2011. Molecular evidence 668 of a (pro)renin/(pro)renin receptor system in human intrauterine tissue in pregnancy and its association 669 with PGHS-2. J. Renin Angiotensin Aldosterone Syst. 12(3):304-310. 670
28 https://mc06.manuscriptcentral.com/cjpp-pubs Page 29 of 42 Canadian Journal of Physiology and Pharmacology
671 Purisch, S.E., and Gyamfi-Bannerman, C. 2017. Epidemiology of preterm birth. Semin. Perinatol. 41(7): 672 387-391. doi: 10.1053/j.semperi.2017.07.009. doi:10.1053/j.semperi.2017.07.009 673 674 Rajakariar, R., Yaqoob, M.M., Gilroy, D.W. 2006. COX-2 in inflammation and resolution. Mol. Interv. 675 6(4): 199-207. doi: 10.1124/mi.6.4.6 676 677 Reinebrant, H.E., Pileggi-Castro, C., Romero, C.L., Dos Santos, R.A., Kumar, S., Souza, J.P., et al. 2015. 678 Cyclo-oxygenase (COX) inhibitors for treating preterm labour. Cochrane Database Syst Rev. 679 5(6):CD001992. doi:10.1002/14651858.CD001992.pub3 680 681 **Rinaldi, S.F., Makieva, S., Saunders, P.T., Rossi, A.G. Norman, J.E. 2017. Immune cell and 682 transcriptomic analysis of the human decidua in term and preterm parturition. Mol. Hum. Reprod. 683 23(10):708-724. 684 685 *Sadovsky, Y., Nelson, D.M., Muglia, L.J., Gross, G.A., Harris, K.C., Koki, A.K., et al 1999. Effective 686 diminution of amniotic prostaglandin production by selective inhibitors of cyclooxygenase type 2. Am. J. 687 Obstet. Gynecol. 182(2): 370-6. 688 689 Sanchez-Ramos, L. 2005. Induction of labor. Obstet. Gynecol. Clin. North Am. 32: 181-200. doi: 690 10.1016/j.ogc.2004.12.004 691 Draft 692 Sawdy, R., Slater, D., Fisk, N., Edmonds, D.K., and Bennett, P. 1997. Use of a cyclo-oxygenase type-2- 693 selective non-steroidal anti-inflammatory agent to prevent preterm delivery. Lancet. 350: 265-6. doi: 694 10.1016/S0140-6736(05)62229-5 695 696 **Sharp, G.C., Hutchinson, J.L., Hibbert, N., Freeman, T.C., Saunders, P.T.K., Norman., J.E. 2016. 697 Transcription analysis of the myometrium of labouring and non-labouring women. PLoS One. 11(5): 698 e0155413. doi:10.1371/journal.pone.0155413 699 700 *Slater, D., Berger, L., Newton, R., Moore, G., and Bennett, P. 1994. Implication of mRNA binding 701 proteins in the regulation of cyclo-oxygenase in human amnion at term. Biochem. Biophys. Res. Commun. 702 203(1): 67-71. doi: 10.1006/bbrc.1994.2149 703 704 *Slater, D.M., Berger, L.C., Newton, R., Moore, G.E., and Bennett, P.R. 1995. Expression of 705 cyclooxygenase types 1 and 2 in human fetal membranes at term. Am. J. Obstet. Gynecol. 172(1): 77-82. 706 707 *Slater, D., Allport, V., Bennett, P. 1998. Changes in the expression of the type-2 but not the type-1 cyclo- 708 oxygenase enzyme in the chorio-decidua with the onset of labour. Br. J. Obstet. Gynaecol. 105: 745-748. 709 710 *Slater, D., Dennes, W., Sawdy, R., Allport, V., and Bennett, P. 1999a. Expression of cyclo-oxygenase 711 types-1 and -2 in human fetal membranes throughout pregnancy. J. Mol. Endocrinol. 22: 125-130. 712 713 *Slater, D.M., Dennes, W.J., Campa, J.S., Poston, L., and Bennett, P.R. 1999b. Expression of cyclo- 714 oxygenase types-1 and -2 in human myometrium throughout pregnancy. Mol. Hum. Reprod. 5(9): 880- 715 884. 716
29 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 30 of 42
717 *Smieja, Z., Zakar, T., Walton, J, C, and Olson, D, M. 1993. Prostaglandin endoperoxide synthase kinetics 718 in human amnion before and after labor at term and following preterm labor. Placenta. 14: 163-175. 719 720 Smith, W.L., Garavito, R.M., and DeWitt, D.L. 1996. Prostaglandin endoperoxide H synthases 721 (cyclooxygenases)-1 and -2. J. Biol. Chem. 271: 33157-33160. 722 723 Sood, B.G., Lulic-Botica, M., Holzhausen, K.A., Pruder, S., Kellogg, H., Salari, V., et al. 2011. The risk 724 of necrotizing enterocolitis after indomethacin tocolysis. Pediatrics. 128(1):e54-62. doi: 725 10.1542/peds.2011-0265 726 727 *Sooranna, S.R., Grigsby, P.L., Engineer, N., Liang, Z., Sun, K., Myatt, L., et al. 2006. Myometrial 728 prostaglandin E2 synthetic enzyme mRNA expression: spatial and temporal variations with pregnancy 729 and labour. Mol. Hum. Reprod. 12(10): 625-631. doi:10.1093/molehr/gal061 730 731 *Sparey, C., Robson, S.C., Bailey, J., Lyall, F., and Europe-Finner, G.N. 1999. The differential expression 732 of myometrial connexin-43, cyclooxygenase-1 and -2, and Gs alpha proteins in the upper and lower 733 segments of the human uterus during pregnancy and labor. J. Clin. Endocrinol. Metab. 84(5):1705-1710. 734 doi:10.1210/jcem.84.5.5644 735 736 **Stephen, G.L., Lui, S., Hamilton, S.A., Tower, C.L., Harris, L.K., Stevens, A., et al. 2015. 737 Transcriptomic profiling of human choriodeciduaDraft during term labor: inflammation as a key driver of labor. 738 Am. J. Reprod. Immunol. 73:36-55. 739 740 *Tattersall, M., Engineer, N., Khanjani, S., Sooranna, S.R., Roberts, V.H., Grigsby, P.L., et al. 2008. Pro- 741 labour myometrial gene expression: are preterm labour and term labour the same? Reproduction. 135: 742 569-579. doi: 10.1530/REP-07-0461 743 744 *Teixeira, F.J., Zakar, T., Hirst, J., Guo, F., Machin, G., and Olson, D.M. 1993. Prostaglandin 745 endoperoxide H synthase (PGHS) activity increases with gestation and labour in human amnion. J. Lipid. 746 Mediat. 6(1-3): 515-23. 747 748 *Teixeira, F.J., Zakar, T., Hirst, J.J., Guo, F., Sadowsky, D.W., Machin, G., et al. 1994. Prostaglandin 749 endoperoxide-H synthase (PGHS) activity and immunoreactive PGHS-1 and PGHS-2 levels in human 750 amnion throughout gestation, at term, and during labor. J. Clin. Endocrinol. Metab. 78(6):1396-402. 751 doi:10.1210/jcem.78.6.8200943 752 753 *Trautman, M.S., Edwin, S.S., Collmer, D., Dudley, D.J., Simmons, D., Mitchell, M.D. 1996. 754 Prostaglandin H synthase-2 in human gestational tissues: regulation in amnion. Placenta. 17: 239-45. 755 756 Van den Veyver, I.B., and Moise, K.J. 1993. Prostaglandin synthetase inhibitors in pregnancy. Obstet. 757 Gynecol. Survey. 48: 493-502. 758 759 Vane, J.R. 1971. Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. 760 Nat. New Biol. 231(25): 232-235. 761
30 https://mc06.manuscriptcentral.com/cjpp-pubs Page 31 of 42 Canadian Journal of Physiology and Pharmacology
762 Vane, J.R., and Botting, R.M. 1995. A better understanding of anti-inflammatory drugs based on isoforms 763 of cyclooxygenase (COX-1 and COX-2). Adv. Prostaglandin. Thromboxane. Leukot. Res. 23:41-8. 764 765 Vane, J.R., and Botting, R.M. 1998. Mechanism of action of nonsteroidal anti-inflammatory drugs. Am. 766 J. Med. 104(3A):2S-8S. 767 768 Vane, J.R. and Williams, K.I. 1973. The contribution of prostaglandin production to contraction of isolated 769 uterus of the rat. Br. J. Pharmacol. 48(4): 629-639. 770 771 Vogel, J.P., Nardin, J.M., Dowswell, T., West, H.M., and Oladapo, O.T. 2014. Combination of tocolytic 772 agents for inhibiting preterm labour. Cochrane Database Syst. Rev. 11(7):CD006169. doi: 773 10.1002/14651858.CD006169.pub2. 774 775 **Weiner, C.P., Mason, C.W., Dong, Y., Buhimschi, I.A., Swaan, P.W., Buhimschi, C.S. 2010. Human 776 effector/initiator gene sets that regulare myometrial contractility during term and preterm labor. Am. J. 777 Obstet. Gynecol. 202: 474.e1-20 778 779 Wiqvist, N., Bygdeman, M., Gréen, K., and Lundström, V. 1974. Endogenous prostaglandins and the 780 initiation of labor. Acta Obstet. Gynecol. Scand. Suppl. 37:7-16. 781 782 Zuckerman, H., Reiss, U. and Rubinstein, I.Draft 1974. Inhibition of human premature labor by indomethacin. 783 Obstet. Gynecol. 44(6): 787-92. 784 785 *Zuo, J., Lei, Z.M., Rao, CV., Pietrantoni, M., and Cook, V.D. 1994. Differential cyclooxygenase-1 and 786 -2 gene expression in human myometria from preterm and term deliveries. J. Clin. Endocrinol. Metab. 787 79(3): 894-899. doi: 10.1210/jcem.79.3.8077379.
31 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 32 of 42
788 Figure Captions
789
790 Figure 1. Prostaglandin synthesis pathway. Arachidonic acid is cleaved from membrane
791 phospholipids by phospholipase enzymes and converted to a prostaglandin precursor PGH2 by
792 COX-1 and COX-2 enzymes. The PGH2 is converted into bioactive prostaglandins by specific
793 prostaglandin synthases. Prostaglandins such as PGE2 and PGF2 act on their respective receptors
794 in uterine tissues to stimulate cervical ripening and uterine contractions. Clinically relevant
795 inhibitors for the COX enzymes are shown. 796 Draft 797 Figure 2. PRISMA flow diagram of systematic search and article selection for inclusion and
798 exclusion.
799
800 Figure 3. Venn diagram representation of the number of studies on fetal membranes and
801 myometrium that measured and reported RNA, protein and/or total enzyme activity.
32 https://mc06.manuscriptcentral.com/cjpp-pubs Page 33 of 42 Canadian Journal of Physiology and Pharmacology
802
Draft
33
https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 34 of 42
1 Tables
2
3 Table 1. Summary of terms used in the search for basic science literature
Theme Search Term Variants [and / or]
Cyclooxygenase / COX / COX-1 / COX-2 / prostaglandin COX Enzyme endoperoxide synthase / PTGS / PTGS1 / PTGS2 / Prostaglandin H synthase / PGHS / PGHS1 / PGHS2
Gestation Pregnant / pregnancy / premature / labour / labor / term / preterm Amnion / chorion / chorio-deciduaDraft / fetal membranes / decidua / Tissues myometrium / uterus
4 Note: PTGS1 and PTGS2 are the official gene names for the two cyclooxygenase enzymes. Terms within each box were combined in 5 the search by an “or” function; the terms down the column were combined with “and”. Mesh terms in MEDLINE and EMBASE relevant 6 to the terms listed above were included where possible. 7 8 9 10 11
https://mc06.manuscriptcentral.com/cjpp-pubs Page 35 of 42 Canadian Journal of Physiology and Pharmacology
12 Table 2. Total COX enzyme activity in fetal membranes – reported differences with labour. 13
Total COX Activity Labour vs Non-Labour Study Groups Included Amnion Chorion Decidua Gaffney et al. 1990 TNL, TL TL>TNL Bennett et al. 1993 TNL, TL TL > TNL Smieja et al. 1993 PTL, TNL, TL TL > TNL Teixeira et al. 1993 PTNL, PTL, TNL, TL PTL > PTNL, TL > TNL Teixeira et al. 1994 PTNL, PTL, TNL, TL PTL > PTNL, TL > TNL Hirst et al. 1995b TNL, TL nco Mijovic et al. 1997 TNL, TL TL > TNL Mijovic et al. 1998 PTNL, PTL, TNL Draft PTL > PTNL Hirst et al. 1998 TNL, TL nco Mijovic et al. 1999 PTNL, PTL, TNL nco Sadovsky et al. 1999 PTNL, PTL, TNL, TL PTL > PTNL, TL > TNL Slater et al. 1999a PTNL, TNL, TL TL > TNL TL > TNL Leguizamon et al. 2001 PTNL, PTL PTL > PTNL Makino et al. 2007 PTNL, PTL, TNL, TL nco Lee et al. 2010 PTL, TNL, TL TL > TNL nco 14
15 Note: Study groups are defined as preterm non-labour (PTNL), preterm labour (PTL), term non-labour (TNL) and term labour (TL). 16 Results for individual studies are summarized in each row, for the different tissues analyzed in each column. White boxes contain 17 summary of findings for tissue analyzed; grey filled boxes indicate that the tissue was not studied. Merged tissue columns indicate 18 non-homogenous tissue (i.e. amnion-chorion, chorio-decidua). A ‘>’ symbol denotes significant findings of higher COX activity in 19 groups compared. No changes observed denoted by ‘nco’. 20 https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 36 of 42
21
Draft
https://mc06.manuscriptcentral.com/cjpp-pubs Page 37 of 42 Canadian Journal of Physiology and Pharmacology
22 Table 3. Expression of COX-1 and COX-2 in fetal membranes – reported differences with labour.
COX-1 COX-2 Author Groups Included Analysis Amnion Chorion Decidua Amnion Chorion Decidua Bryant-Greenwood et al. 1987 TNL, TL Protein nco nco nco Price et al. 1989 PTNL, PTL, TNL Protein nco nco nco Bennett et al. 1992 TNL, TL RNA TL > TNL nco Bennett et al. 1993 TNL, TL RNA TL > TNL Teixeira et al. 1993 PTNL, PTL, TNL, TL Protein nco Teixeira et al. 1994 PTNL, PTL, TNL, TL Protein nco Slater et al. 1994 TNL, TL RNA nco TL > TNL Divers et al. 1995 PTNL, PTL, TNL, TL Protein nco Slater et al. 1995 TNL, TL RNA nco TL > TNL Hirst et al. 1995a TNL, TL RNA Draftnco nco TL > TNL TL > TNL Hirst et al. 1995b TNL, TL RNA nco TL > TNL Fuentes et al. 1996 TNL, TL Protein TL > TNL nco Gibb and Sun 1996 TNL, TL RNA, Protein nco nco nco Trautman et al. 1996 PTNL, TNL, TL RNA nco nco nco nco nco nco Mijovic et al. 1997 TNL, TL RNA nco TL > TNL Mijovic et al. 1998 PTNL, PTL, TNL RNA PTL > PTNL PTL > PTNL Hirst et al. 1998 TNL, TL RNA nco nco Slater et al. 1998 TNL, TL RNA nco TL > TNL Mijovic et al. 1999 PTNL, PTL, TNL RNA PTL > PTNL PTL > PTNL Sadovsky et al. 1999 PTNL, PTL, TNL, TL Protein nco nco nco PTL > PTNL nco nco Slater et al. 1999a PTNL, TNL, TL RNA, Protein nco nco TL > NL TL > NL Leguizamon et al. 2001 PTNL, PTL Protein nco PTL > PTNL Johnson et al. 2002 TNL, TL RNA TL > TNL Makino et al. 2007 PTNL, PTL, TNL, TL RNA, Protein nco Lee et al. 2010 PTL, TNL, TL Protein TL >TNL nco Pringle et al. 2011 TNL, TL RNA TL > NL nco nco Phillips et al. 2014 PTNL, PTL, TNL, TL RNA nco nco nco nco https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 38 of 42
23
24 Note: Study groups are defined as preterm non-labour (PTNL), preterm labour (PTL), term non-labour (TNL) and term labour (TL). 25 Results for individual studies are summarized in each row, for the different tissues analyzed in each column. The type of measurement 26 made is indicated for each study as either protein, RNA or both. White boxes contain summary of findings for tissue analyzed; grey 27 filled boxes indicate that the tissue was not studied. Merged tissue columns indicate non-homogenous tissue (i.e. amnion-chorion, 28 chorio-decidua). A ‘>’ symbol denotes significant findings of higher COX expression between groups compared. No changes 29 observed denoted by ‘nco’. Draft
https://mc06.manuscriptcentral.com/cjpp-pubs Page 39 of 42 Canadian Journal of Physiology and Pharmacology
30 Table 4. Expression of COX-1 and COX-2 in myometrium– reported differences with labour.
COX-1 COX-2 Author Groups Included Analysis Myometrium Myometrium Zuo et al. 1994 PTNL, PTL, TNL, TL RNA/Protein nco PTNL > PTL Moore et al. 1998 PTNL, PTL, TNL, TL RNA nco nco Sadovsky et al. 1999 PTNL, PTL, TNL, TL Protein nco nco Slater et al. 1999b PTNL, TNL, TL RNA/Protein nco nco Sparey et al. 1999 TNL, TL Protein nco nco Erkinheimo et al. 2000 TNL, TL RNA/Protein TL > TNL Giannoulias et al. 2002 PTNL, PTL, TNL, TL Protein nco nco Choi et al. 2006 TNL, TL Protein TL > TNL Sooranna et al. 2006 PTNL, PTL, TNL, TL RNA PTL > PTL, TL > TNL Astle et al. 2007 PTNL, PTL, TNL, TL DraftRNA TL > TNL Tatterstall et al. 2008 PTNL, PTL, TNL, TL RNA TL > TNL Bollapragada et al. 2009 TNL, TL RNA TL > TNL Marques et al. 2010 TNL, TL RNA TL > TNL Mittal et al. 2010 TNL, TL RNA TL > TNL Phillips et al. 2011 PTNL, PTL, TNL, TL RNA nco PTL > PTL, L > NL Mitsuya et al. 2014 PTNL, PTL, TNL, TL RNA PTL > PTNL 31
32 Note: Study groups are defined as preterm non-labour (PTNL), preterm labour (PTL), term non-labour (TNL) and term labour (TL). 33 Results for individual studies are summarized in each row, for the different tissues analyzed in each column. The type of measurement 34 made is indicated for each study as either protein, RNA or both. White boxes contain summary of findings for tissue analyzed; grey 35 filled boxes indicate gene not studied. A ‘>’ symbol denotes significant findings of higher COX-1 or COX-2 expression between 36 groups compared. No changes observed denoted by ‘nco’.
https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 40 of 42
Figure1
Membrane phospholipids
Phospholipases (PLA) Arachidonic Acid
COX-1 NSAIDs COX-2 (e.g. indomethacin)
Draft Selecitve inhibitors PGH2 (e.g. nimesulide, PG synthases rofecoxib) (mPGES-1, mPGES-2, cPGES)
PGE2 PGF2α
Cervical Ripening, Uterine Uterine contraction contraction
https://mc06.manuscriptcentral.com/cjpp-pubs Page 41 of 42 Canadian Journal of Physiology and Pharmacology Figure2
3958recordsidentifiedthrough databasesearches: 145duplicate • 3791fromMedline recordsremoved 3670didnot • 167fromEMBASE 3813recordsscreened addressthe forinclusion researchquestion 143full-textarticles (IsCOXinvolvedin assessedforeligibility 132studies Draft excluded:theonsetof 11studies •labour?)88wereclinical identifiedfor trials,reviews, inclusion letters,etc. 31additional • 44wereanimal articlesidentified studies through 42studies referencelistsof includedfor includedstudies synthesis
https://mc06.manuscriptcentral.com/cjpp-pubs Canadian Journal of Physiology and Pharmacology Page 42 of 42
Figure3
Fetalmembranes Myometriu m RNA Protein RNA Protein
8 1 4 6 3 3 2 6 3 Draft 2
Enzyme Enzyme Activity Activity
https://mc06.manuscriptcentral.com/cjpp-pubs