Placenta 35 (2014) 229e240
Contents lists available at ScienceDirect
Placenta journal homepage: www.elsevier.com/locate/placenta
Turnover of human villous trophoblast in normal pregnancy: What do we know and what do we need to know? T.M. Mayhew* School of Life Sciences, Queen’s Medical Centre, University of Nottingham, Nottingham NG7 2UH, UK
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 23 January 2014
How the turnover of villous trophoblast is regulated is important for understanding normal and complicated pregnancies. There is considerable accord that syncytiotrophoblast (STB) grows and is refreshed by recruiting post-mitotic cells from the deeper cytotrophoblast (CTB). Nuclei in STB exhibit a spectrum of morphologies and packing densities and, until recently, there seemed to be a consensus that this variation reﬂected a transition from an early undifferentiated CTB-like phenotype to a long preapoptotic and brief apoptotic phase. In these later phases, nuclei are sequestered in clusters (syncytial knots) prior to extrusion as part of normal epithelial turnover. Early in gestation, nuclear clustering and formation of protrusions (syncytial sprouts) also occurs as a preliminary to villous sprouting. Nuclei in these clusters have a CTB-like phenotype and some sprouts may also detach from STB and pass into the uteroplacental circulation. However, this apparent consensus has been challenged and new interpretations of events in the proliferative (CTB), terminal differentiation (STB) and deportation compartments have emerged. Several different types of STB fragment are deported in normal pregnancy: larger multinucleate STB fragments, smaller uninucleate elements with CTB-like morphology, anucleate cytoplasmic fragments, microparticles and nanovesicles. This review identiﬁes points of agreement and disagreement and offers possible avenues of future research. An obvious need is to standardise best practice in several areas including choosing appropriate references for cell cycle phase labelling indices and combining immunolabeling of cell cycle and apoptosis markers (at LM or TEM levels) with designbased stereological estimates of absolute numbers of cells and nuclei in different compartments throughout normal gestation. This would also provide a surer foundation for interpreting results from different research groups and changes in normal and complicated pregnancies. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Villous trophoblast Proliferation Differentiation Turnover Deportation
1. Introduction “The most beautiful thing we can experience is the mysterious. It is the source of all true art and all science. He to whom this emotion is a stranger, who can no longer pause to wonder and stand rapt in awe, is as good as dead: his eyes are closed.” (Albert Einstein, 1879e1955) The conduct of good science and the advancement of its knowledge base depends not only on the application of rigorous study design, sound sampling principles, newer and more accurate technology and the use of precise and unbiased or minimally biased investigative tools. It depends also on the attributes of researchers
* School of Life Sciences, E Floor, Queen’s Medical Centre, University of Nottingham, Nottingham NG7 2UH, UK. E-mail address: [email protected]
0143-4004/$ e see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.placenta.2014.01.011
and their ability and willingness to remain curious, critical, unprejudiced and receptive to different or new ideas. Thomas Huxley is reported to have said that the tragedy of science is the slaying of a beautiful hypothesis by an ugly fact. The measure of a scientiﬁc researcher and the community in which he/she operates is how one deals with ‘ugly facts’. A researcher who is ready to accept the possibility of misinterpretation or error is better than one who always claims exclusivity of the correct interpretation. Sometimes, when extreme views are held, the correct interpretation might lie somewhere in between or require fresh investigations and experiments in order to clarify the situation. It is true to say that some elements of research are inductive, relying on the marshalling of facts to prove a general statement or support an idea. The hallmark of good science is the formulation of a hypothesis and the designing of an experiment that can test it. As the philosopher Karl Popper stated “You cannot prove that all swans are white by counting white swans. But you can prove that not all swans are white by counting one black swan.”
T.M. Mayhew / Placenta 35 (2014) 229e240
Recently, in the biomedical area of villous trophoblast turnover in the human placenta, new interpretations of events in its proliferative and terminal differentiation compartments have emerged and an apparent consensus has been challenged [1e4]. The principal points of contention [1,3] can be summarised as follows: (i) There is little evidence to support the idea that turnover of syncytial nuclei occurs in normal placenta. (ii) There is little evidence that it occurs via an apoptosis-related process. (iii) Epigenetic modiﬁcations, rather than apoptotic events, underlie changes in the heterochromatin content of STB nuclei. (iv) A proportion of STB nuclei remain transcriptionally active throughout gestation. (v) Syncytial knots are more resistant to deportation than syncytial sprouts. (vi) Apoptosis is normally conﬁned to CTB cells and areas of STB associated with ﬁbrin-type ﬁbrinoid. (vii) Apoptotic events seen in CTB may be mistakenly ascribed to STB. This review is an attempt to address the points of agreement and disagreement concerning trophoblast turnover and to offer possible avenues of future research. 2. Growth of placental villi and villous trophoblast through gestation From about mid-gestation, the surface area available for transplacental exchange expands enormously by formation of new terminal villi [5,6]. Early villous sprouting is initiated by cells in regions of CTB and villous stroma with high proliferative activity. Post-mitotic CTB cells are recruited into STB which creates attenuated protrusions called syncytial sprouts and these subsequently acquire a vascularised stroma to become mesenchymal villi. These initial forms develop into a variety of different villous types which include stem, intermediate and terminal villi . Around midgestation, mesenchymal villi transform into mature intermediate villi that generate enormous numbers of well-vascularised and small calibre terminal villi. This process continues to term. These changes are associated with temporal variation in the supply of oxygen and nutrients across gestation, as reviewed recently . Before 10e12 wk, the intervillous oxygen tension is less than 3 kPa and rises to 5e11 kPa before steadying to about 5 kPa at term. Early on, expression of HIF-1a (which regulates responses to oxygen levels) is high in villi and affects cell proliferation and apoptosis. The onset of maternal perfusion of the intervillous space is also accompanied by increases in the activities of antioxidant enzymes . An important component of villi is the trophoblast. Early in gestation, the CTB layer appears continuous but, as gestation progresses, it becomes discontinuous as cells alter shape and disperse more widely [6,9e11]. The apparent sparsity of cells on histological sections is explained by the fact that increases in cell number do not match the expanding trophoblast surface so that CTB cell bodies become more widely separated . Proliferation of CTB cells and incorporation into STB are elements of the growth in villous surface area and both are sensitive to hypoxia which also affects apoptosis in CTB cells . The fact that villous trophoblast has a syncytial compartment is unique for a human epithelium. Whilst it may have evolved to allow invasion of maternal tissues without compromising the integrity of the maternofetal intervascular barrier, a syncytium offers other beneﬁts. For instance, a syncytium allows redistributions of STB cytoplasm to produce locally thick and thin regions. Together with changes in villous surface area and the peripheralisation and dilation of fetoplacental capillaries in the stroma, these improve the efﬁciency of gas and nutrient transfer by passive diffusion by reducing the harmonic mean intervascular distance . STB is also sensitive to oxygen tensions which can induce a form of apoptosis different from that in CTB .
Overall, the increase in surface area outpaces growth in trophoblast volume within which compartments and STB regions respond differently . After 10 wk of gestation, volume growth of CTB lags behind whilst, within STB, growth in volume and surface area of regions of nuclear clustering and in the surface area of sites of de-epithelialisation exceed that of other regions. In addition, deposition of perivillous ﬁbrin-type ﬁbrinoid correlates well with villous surface area. We now examine these events in more detail in the context of trophoblast proliferation, differentiation and loss during normal pregnancy. 3. Villous trophoblast e a continuously renewing epithelium Like cells of the epidermis, intestinal epithelium and bloodforming tissues, villous trophoblast is continuously renewing . CTB cells lie on a basal lamina and exist in one of the phases (G0, G1, S, G2 or M) of the mitotic cell cycle. Overlying CTB is the syncytium. CTB cells divide continuously throughout gestation and some of their progeny are recruited into STB by membrane fusion and cytoplasmic conﬂuence. Estimates of numbers of CTB and STB nuclei from 13 wk to term are consistent with the notion that CTB cells undergo several divisions before fusing into STB . Indeed, fusion into the syncytium makes trophoblastic epithelium uniquely different from epidermis, intestinal and other epithelia where cells retain their individuality as they migrate and differentiate between proliferation and extrusion sites. Once a CTB cell has been incorporated into STB, a tightlycontrolled sequence of differentiation occurs. An early phase of maturation and continuing transcription is followed by one of terminal differentiation in which some nuclei show substantial changes in shape, packing density, chromatin condensation and nuclear envelope integrity and loss of transcriptional activity [16e 18]. 4. CTB e the proliferation and recruitment compartment Although the CTB layer becomes discontinuous during gestation, its cells are not truly independent. Ultrastructural and confocal studies on term placentas  have shown that they possess ﬁne processes which are contiguous with those of surrounding cells. At this stage, CTB cells and their processes cover about 40% of the basal lamina surface  and 10% of trophoblast volume . 4.1. CTB cells and ‘proliferation markers’ Measures of proliferative activity are complicated by the fact that ‘proliferation markers’ are usually applied uncritically. Actually, they mark one or more phases of the cell cycle . Simple counts of mitotic ﬁgures on tissue sections are used to estimate a mitotic index which is taken to represent the fraction of cells in M phase. In fact, this index is not unbiased because it is based on an incorrect implicit assumption, viz. that relative numbers seen on sections equate to relative numbers in 3D. It is also an approximation because it accounts for only some of those phases of mitosis in which mitotic ﬁgures are discernible (e.g. metaphase and anaphase). S phase, which involves synthesis of DNA and the activity of associated proteins, can be labelled also (autoradiographically with tritiated thymidine or immunochemically with bromodeoxyuridine, BrdU) to obtain the fraction of cells or nuclei which are in S phase. Other phase markers include the Ki-67 family (including MIB-1) and proliferation associated nuclear antigen, PCNA. Ki-67 protein is present during all active phases of the cycle (G1, S, G2
T.M. Mayhew / Placenta 35 (2014) 229e240
and M) but not in G0 and this makes it useful as a marker of the growth fraction of a population [19,20]. This is taken to represent the fraction of all cells which are actively cycling. PCNA is less reliable as a ‘proliferation marker’ owing to its involvement in DNA repair and DNA synthesis. Consequently, although it colocalises with Ki-67, PCNA tends to label more cells [21,22]. However, PCNA has a long half-life and, because STB is non-proliferative, these facts have been used imaginatively to provide a measure of the period of incorporation of CTB into STB and as an aid to distinguishing between syncytial sprouts and syncytial knots . Again, these different phase markers produce indices which are approximations because 2D counts on independent sections do not, in general, equate to 3D numbers of cells [23,24]. Comparisons between research groups are made even more difﬁcult when they adopt different references. For instance, a ‘proliferative index’ might refer to the pool of CTB cells or all of the nuclei in villous trophoblast or all of the nuclei in placental villi or in a ﬁxed sectional area [25e29]. Even when estimated unbiasedly in combination with, say, design-based stereological counting tools such as the disector [15,30], ‘proliferation indices’ (morphological or immunochemical) are still only phase-labelling indices. Nevertheless, they are the best estimators we have at present and the real dangers are in wrongly interpreting them . In most labelling studies, the various possible interpretations are not resolved because of reliance on a single phase index. Moreover, if the absolute size of the reference pool (the total number of CTB cells) alters (as it does during gestation ,), a given phase index may not provide a realistic measure of the total number of cells in that phase.
4.2. Quantitative data on CTB cells and the cell cycle Whilst there is some agreement that labelling levels for Ki-67 and PCNA in the whole placenta decline during gestation [25,32,33], ﬁgures for Ki-67 expressed as a proportion of all CTB cells give a different picture. This index varies from about 16% (range 11e22%) in the ﬁrst trimester to 25% (4e38%) at term [25,29,30,34]. In contrast, the S phase index seems to be reasonably constant [22,25] and the mitotic index is low (range 0.5e2.9%), at least in the ﬁrst trimester [22,35]. Attempts have been made to combine alternative phase markers in order to assess the relative lengths of different phases of the cycle. The BrdU:Ki-67 ratio [25,36] represents the fraction of active phases of the cell cycle that are occupied by S phase. This ratio declines during gestation suggesting that S phase is relatively shorter or other phases relatively longer. By combining such estimates with mitotic index, the value of this approach could be extended. However, routine estimates of these ratios will not provide absolute phase durations or measures of absolute compartment size (e.g. total number of CTB cells). What is required is a more systematic and comprehensive analysis of the CTB cycle combining multiple phase-labelling indices with design-based stereological estimates of numbers of cells or nuclei.
Passage through the mitotic cycle is regulated by a set of cyclins, cyclin-dependent kinases (CDKs) and CDK inhibitors whose expression patterns have also been investigated (e.g. Refs. [37,38]). Cyclin E is required for progression from G1 to S phase and is found in about 43% of all CTB nuclei seen on tissue sections. Expression correlates well with that of Ki-67 and the fraction of labelled cells is reasonably constant during gestation . In the ﬁrst trimester , CTB cells express cyclin A (which promotes G1 to S and G2 to M transitions) and, to a lesser extent, cyclin B1 (which promotes progression between S and M phases). They also express the cycle inhibitor p57 which regulates the G1/S transition and S phase completion and might inhibit progression at the G1/S checkpoint in certain CTB cells prior to fusion into STB. All these indices are useful measures of cycle phases and their regulation but, again, they leave us with the task of interpreting relative data. Since cell cycle duration times are not easy to determine, valuable complementary information is offered by estimating cell numbers. 4.3. CTB cells and stereological estimates of their number Disector-based estimates are available for numbers of cells in second and third trimesters [15,30,39]. At 13e15 wk of gestation, there are about 600 million CTB cells and this rises to 5800 million at 37e39 wk (see Table 1 and ). For term placentas, calculations from assumption-based stereological estimates [40,41] suggest a complement of 3100e4900 million cells. A more recent disector estimate, using an antibody against cytokeratin 7, yielded a substantially higher ﬁgure of 86,000 million cells at term . The reason for this substantial disparity is unclear at present. Based on current knowledge from internally-consistent comparative studies, it is clear that cell number increases roughly 10-fold between 13 and 15 wk and term [15,39]. What is the impact of these estimates on cell cycle parameters? Unfortunately, systematic quantitative (rather than semi-quantitative) studies of phase-labelling indices throughout gestation are difﬁcult to ﬁnd. Combining cell numbers  with values of phase indices [22,25,29,30,34,35] suggests that about 84% of CTB cells in the ﬁrst trimester are in Go and 16% are cycling of which 1.7% are in M phase. With a cycle duration of 15 h and S phase of 5.5 h (estimated by autoradiography ), these ﬁgures indicate about 7.9e9.2 h for G1 and G2 and 0.3e1.6 h for mitosis (Table 2). Unfortunately, corresponding estimates for intermediate and term placentas are not available. However, data do indicate that the pool of S phase cells probably expands roughly 9-fold and that of cycling cells 15-fold between ﬁrst and third trimesters. 4.4. CTB cells and their differentiation CTB cells are heterogeneous in phenotype and present various forms identiﬁable on ultrastructural and other criteria . At one extreme are undifferentiated cells and, at the other, cells with nuclei resembling those seen in STB and distinguishable from syncytium only by the presence of apposed plasma membranes.
Table 1 Quantitative data for villous trophoblast compartments in the second and third trimesters (based on ). Baseline values at 13e15 wk of gestation were: Volume of trophoblast, Vtro, 7.3 cm3; number of CTB nuclei, Nctb, 600 million; number of STB nuclei, Nstb, 6200 million. Period 13e15 22e23 29e31 13e15
Nett gain in Vtro to to to to
22e23 29e31 37e39 37e39
wk wk wk wk
10.2 18.2 24.7 53.1
cm3 cm3 cm3 cm3
Gain per wk 1.23 2.43 3.09 2.21
cm3 cm3 cm3 cm3
Nett gain in Nctb
Gain per wk
Nett gain in Nstb
Gain per wk
1400 1400 2400 5200
170 190 300 220
7300 million 18,200 million 26,400 million 51,900 million
880 million 2400 million 3300 million 2200 million
million million million million
million million million million
T.M. Mayhew / Placenta 35 (2014) 229e240
Table 2 Estimates of the sizes of CTB cell pools in ﬁrst and third trimesters. Numbers of cells expressed in millions (% of total). See text for references.
because we have no equivalent data for nuclear loss over the same period.
5. STB e the differentiation compartment
Number of CTB cells
Design-based stereology Ki-67/MIB-1 immunochemistry Ki-67/MIB-1 immunochemistry Counting mitotic ﬁgures in 2D See above Autoradiography in 2D Autoradiography in 2D See above See above
86 (14.3%) 15 h 5.5 h 0.3e1.6 h 7.9e9.2 h
? ? ? ? ?
Once cells have fused into STB, they display a range of nuclear morphologies and packing densities. At term, regions of nuclear clustering account for about 18e30% of trophoblast volume, regions of non-clustering for 50e75% and CTB for the rest [6,14,16]. Prominent in terms of nuclear clustering are regions called syncytial knots and syncytial sprouts. In recent years, terminology regarding these regions has been clariﬁed as have techniques for identifying them [1,2,4,18,55].
Non-cycling cells in G0 Cells in G1, S,G2,M phases Cells in M phase Cells in G1, S, G2 phases Cell cycle duration S phase duration M phase duration G1,G2 duration
Intermediate cells have transitional features [6,9,16]. Undifferentiated cells may constitute a clonogenic or progenitor subpopulation responsible for producing daughter cells some of which differentiate before being recruited into STB [9,15,16]. Undifferentiated cells predominate in early pregnancy and tend to be cuboidal with large euchromatic nuclei (showing a conspicuous nucleolus and nuclear pore complexes) and cytoplasm with few organelles. In contrast, intermediate cells are ﬂatter, the nucleus is more irregular and heterochromatic and the cytoplasm is richer in organelles. In both types of cell, interdigitations of adjacent plasma membranes occur where they lie against the STB and possess desmosomes and other junctional complexes [6,9,16,44]. Differences in cell phenotype raise the possibility that, at some point, cytokinesis is asymmetric and generates different daughter cells . One daughter maintains the proliferative pool and retains contact with the basal lamina whilst the other differentiates into the intermediate form in readiness for recruitment into STB. As mentioned already, a trigger for this might be inhibition of the G1/S transition . However, prior to this point, clonogenic cells probably undergo a few ampliﬁcation divisions  which may be symmetrical and produce identical daughter cells which maintain contact with the basal lamina. In later gestation, this might minimise disruption of the network of CTB cell processes which would otherwise result from cell rounding during mitosis. Presumably, asymmetric division does produce a change in cell conﬁguration but this could facilitate the membrane fusion and cytoplasmic coalescence required for passage of differentiated daughter cells into STB .
4.5. CTB cells and recruitment into the syncytium STB grows and is rejuvenated by recruiting CTB cells. Once intermediate cells have differentiated sufﬁciently, they express connexin-43 in their plasma membranes and display transient gap-junctional communication with each other and with STB. These and other adhesion factors [6,44e46] seem to prepare differentiated cells for fusion into STB. Other factors implicated include ADAM12, phosphatidylserine, syncytins 1 and 2, lipidbased membrane pores and GCM-1 [47e53]. Ultrastructural evidence of fusion includes conﬂuence of cytoplasms and traces, within STB, of plasma membranes with associated desmosomes [9,16,54]. Estimates of numbers of nuclei in CTB and STB indicate that both compartments expand during gestation (Table 1). Recruitment rates into STB increase towards term but may be underestimates
5.1. Syncytial sprouts and knots Syncytial sprouts feature in the formation of new villi during villous sprouting and are particularly common in early pregnancy as outgrowths from the tips and sides of mesenchymal and immature intermediate villi . Some, but not all, later acquire a mesenchymal core and become vascularised. Before vascularisation, they tend to have a long, thin stalk attaching them to the STB. Around mid-gestation, they become more frequent and, towards term, are found mainly in the centres of villous trees and more hypoxic regions where they tend to protrude from the sides of villi. They contain clusters of large, rounded, euchromatic nuclei with little heterochromatin but usually have a prominent nucleolus. The cytoplasm harbours large numbers of free ribosomes and large amounts of rough ER. The attenuated stalk renders sprouts more susceptible to detachment and loss into the maternal intervillous space and, indeed, detached sprouts have been found in term placentas indicating that their deportation probably continues throughout gestation . Cutting independent sections produces microscopical images which make it difﬁcult to distinguish false knots (intervillous bridges and branching points) from true syncytial knots. However, distinction is achievable by combining serial sectioning with immunochemical markers of transcriptional activity and oxidative damage . False knots resemble syncytial sprouts more than they resemble true knots. The latter bulge from the villous surface to varying extents (see below) and increase in frequency during gestation. The signiﬁcance of syncytial knots has been the subject of much speculation  and they been implicated in various processes including senescence, degeneration and adaptation to ischaemic or hypoxic damage [1,4,18,56]. They harbour densely-clustered, irregularly-shaped nuclei with peripheral condensation of chromatin and no obvious nucleoli. By TEM, nuclear proﬁles interlock like jigsaw pieces and may be organised into discrete fascicles each delimited from its neighbours by bundles of intermediate ﬁlaments [4,16]. Nuclei with condensed peripheral chromatin around a central island of euchromatin are also present and tend to be smaller and lack nuclear pores as well as nucleoli. The cytoplasm is relatively small in amount with a paucity of organelles but occasionally demonstrates annulate lamellae [16,49,56]. Though some syncytial knots form gentle bulges at the apical surface of STB, others are more fungiform or, less commonly, have the appearance of a spheroidal bud-like process with a narrow waist at the site of attachment [1,2,4,49,54]. Currently, there is little or no quantitative information available on the relative incidences of these alternative forms. Despite consensus concerning the ultrastructural characteristics of nuclei in syncytial knots, there is ongoing debate over whether or not any of the nuclei are apoptotic and form part of the extrusion component of trophoblast turnover (see below).
T.M. Mayhew / Placenta 35 (2014) 229e240
Table 3 Relative and absolute volumes and surface areas of trophoblast regions between the second and third trimesters (based on ). Volumes are given as cm3 (% of total) and surfaces as cm2 (% of total) at each period of gestation. Variable
Vctb Vstb Vnonc Vskc Vrestc Stot Sden
1.9 9.4 8.7 0.5 0.2 8090 392
2.6 12.4 11.4 0.7 0.3 10,900 325
4.1 31.4 28.1 2.0 1.3 41,700 1630
5.5 41.5 36.0 3.5 2.0 55,900 2740
7.0 59.9 51.0 5.4 3.5 86,800 4390
8.0 73.6 59.3 7.6 6.7 113,000 5880
(17%) (83%) (78%) (4.3%) (1.5%) (100%) (4.8%)
(17%) (83%) (76%) (4.7%) (2.1%) (100%) (3.0%)
(12%) (88%) (79%) (5.6%) (3.7%) (100%) (3.9%)
(12%) (88%) (77%) (7.4%) (4.2%) (100%) (4.9%)
(11%) (89%) (76%) (8.0%) (5.2%) (100%) (5.1%)
(10%) (90%) (73%) (9.3%) (8.2%) (100%) (5.2%)
The early placenta has a relatively greater volume of CTB (Vctb) but relatively small volumes of STB (Vstb), syncytial knots (Vskc) and other regions of nuclear clustering (Vrestc). This state is reversed in term placenta. The fractional volume occupied by regions lacking clustered nuclei (Vnonc) is essentially constant. At 17e20 wk, there is a transient decline in the fraction of total villous surface area (Stot) occupied by denuded regions (Sden)þþ.
5.2. Stereological estimates of regions with and without nuclear clusters Absolute volumes of STB regions containing clustered nuclei, including syncytial knot regions, increase during gestation . At 10e16 wk, trophoblast comprises roughly 2 cm3 of CTB and 9 cm3 of STB (Table 3). There is a relatively greater volume of CTB but smaller relative volumes of syncytial knots and other regions of nuclear clustering (sprouts and false knots). This is consistent with high proliferation and recruitment. At term, the situation is reversed with a relatively smaller volume of CTB and greater relative volumes of syncytial knots and other regions of nuclear clustering. The fractional volume of regions without nuclear clustering is essentially unaltered and the fraction of total villous surface occupied by regions of de-epithelialization drops at 17e20 wk but then increases. These ﬁndings point to changes which create thinner regions of STB, greater volumes of nuclear clustering and some increase in epithelial damage and discontinuity after midgestation. This pattern is consistent with early changes being directed primarily at villous growth and later changes at growth, recruitment, nuclear clustering, loss and damage/repair. 5.3. STB nuclei and stereological estimates of their number Disector estimates of numbers of nuclei (Table 1) vary from about 6200 million at 13e15 wk of gestation to 58,100 million at term . Assumption-based estimates for term placentas are in reasonable agreement (36,000e57,000 million, based on data in Refs. [40,41]). Using the disector to count nuclei classiﬁed on the basis of transcriptional activity, it was found that active nuclei outnumbered inactive nuclei by about 4:1 throughout gestation . This implies that, at the start of the second trimester, about 1240 million are inactive and 4960 million are transcriptionally active. Near term, the corresponding numbers would be 11,620 and 46,480 million nuclei respectively. As with total numbers of CTB nuclei, the ﬁndings indicate an almost 10-fold rise in numbers of nuclei in each category from the end of the ﬁrst trimester to term. They also provide evidence of tight regulation of events within STB. 5.4. STB nuclei and their differentiation The microscopical appearance of nuclei varies not only within STB but across gestation . Early on, nuclei are mostly euchromatic and CTB-like with prominent nucleoli. In the second trimester, there is more heterochromatin and nucleoli are observed less often. At term, some nuclei are heavily heterochromatic and nucleoli are rare. This is consistent with increased relative volumes of syncytial cluster regions during gestation . Recent studies [17,57], have shown that a high proportion of STB nuclei remain transcriptionally active throughout gestation.
Investigations on term placentas [4,9,16,49,56] have distinguished nuclei in clusters from those more dispersed throughout STB. Nuclei in the latter regions vary in ultrastructure from those similar to euchromatic CTB nuclei to others which are irregular, heterochromatic and lack nucleoli. The latter phenotype, together with nuclei which exhibit peripheral condensation of chromatin surrounding a central island of euchromatin but without a nucleolus or nuclear pores, is found also in densely-clustered regions. In clustered and non-clustered regions, nuclei are sometimes associated with annulate lamellae in the cytoplasm. Indeed, the lamellae may represent shed nuclear pores complexes. Such nuclei are extremely unlikely to be functional in transcription or protein synthesis since it is hard to see how they or their RNAs could be synthesised or exported from nucleus to cytoplasm. Studies combining immunochemical labelling for transcriptional activity (RNA polymerase II) with design-based stereology  have conﬁrmed not only that most STB nuclei are transcriptionally active from the second trimester to term but that the fraction is remarkably constant. Furthermore, serial sectioning was used in conjunction with immunochemical visualisation of RNA polymerase II and upstream binding factor (another marker of transcription), PCNA (a marker of recent incorporation into STB) and 8-oxo-deoxyguanosine (a marker of oxidative stress) to distinguish true syncytial knots from false knots and syncytial sprouts . Sprouts and false knots contained transcriptionally active nuclei of recent origin whereas most true knots contained older nuclei (PCNA-negative) which were transcriptionally inactive and positive for oxidative damage. The incidence of true knots was also conﬁrmed to increase across gestation. These ﬁndings support the earlier notion, based on ultrastructural criteria alone [16,54], that nuclei entering STB undergo a sequence of differentiation ending with the phenotype of peripheral chromatin condensation, no nucleolus and no nuclear pores. Because of their resemblance to other nuclei classed as apoptotic on morphological grounds , this later-stage phenotype was described as being ‘apoptotic’ in order to distinguish it from the early-stage euchromatic phenotype and more common intermediate phenotype which was termed ‘pre-apoptotic’ [16,54]. However, though tending to be smaller than the average STB nucleus, the later-stage nuclei did not show ultrastructural evidence of blebbing or fragmentation. Moreover, only a small proportion (less than 1%) of nuclei in clustered and non-clustered regions of STB were found to be positive for terminal dUTP nick-end labelling (TUNEL). 5.5. Apoptosis in STB Death mechanisms are often classiﬁed as being apoptotic or necrotic but forms with resemblances to both (aponecrosis) have been described and some of the underlying molecular players may
T.M. Mayhew / Placenta 35 (2014) 229e240
be shared with autophagy about which little is known in villous trophoblast. Extrinsic and intrinsic pathways are regulated by complexes of molecular interactions that lead to activation of a family of proteolytic caspases [49,59], some of which lead to DNA degradation. Intrinsic apoptosis involves mitochondrial release of cytochrome c whilst extrinsic apoptosis is mediated by membraneassociated receptoreligand interactions. Numerous studies have focussed attention on apoptosis within villous trophoblast (for recent reviews, see Refs. [59,60]). Whilst it is agreed that proliferation is conﬁned to CTB, apoptosis in normal pregnancy has been described in both STB and CTB. It has been observed  that about 0.5% of CTB cells during the ﬁrst trimester show ultrastructural features of apoptosis and, using confocal microscopy, others have noted about 1% of caspase-mediated apoptosis in CTB cells at term . This has led to the suggestion that transit of apoptotic CTB cells or their vesicular fragments across overlying STB might have led to biases in incorrectly ascribing apoptotic labelling of CTB nuclei to STB nuclei. Indeed, apoptosis-like features have been noted in CTB cells which are detached from the basal lamina and seem to be in transit through STB . How big is this misclassiﬁcation bias? For the sake of illustration, let us assume that the frequency of apoptotic nuclei within villi rises from 0.07% in the ﬁrst to 0.14% in the third trimester with about 50% of nuclei residing in the trophoblast . At ﬁrst sight, this difference seems very modest but the total complement of nuclei in villi over this period changes from approximately 17,000e114,000 million . Therefore, the combined data represent an increase in apoptotic nuclei in trophoblast from 6 to 80 million. In the same stages of pregnancy, there are 600 and 5800 million CTB cells respectively of which 0.5e 1% (or 3-58 million) may be apoptotic [3,35]. These ﬁgures suggest that, in the worst case scenario, between half and all of the STB nuclei could be misclassiﬁed. However, these calculations must be treated with caution since different forms of labeling/staining can yield different estimates of apoptotic indices [34,61,62]. Nevertheless, they serve to illustrate the point that misclassiﬁcation will be sensitive to the difference in real labelling indices between CTB and STB. It will be inﬂuenced also by the ability to identify separately the two classes of nuclei. The greater the difference between indices, and the better the ability to distinguish between CTB and STB nuclei, the smaller the bias. As yet, no systematic investigation of actual biases has been undertaken but one result  has indicated that 15% of CTB cells and only 1% of STB nuclei are apoptotic in term placenta. Such differences would imply even greater biases. The misclassiﬁcation bias might be sensitive also to location within STB. It seems reasonable to speculate that biases might be greater when CTB-based apoptotic events lie within STB regions of non-clustered nuclei. In such regions, the lateral resolution afforded by conventional light microscopy, would make it difﬁcult to distinguish a CTB cell in transit from surrounding STB. By contrast, it should be easier to distinguish an apoptotic event in the apical part of a fascicular syncytial knot cluster as belonging to STB. Clearly, there is scope for further investigation using higherresolution immunogold cytochemistry to test these possibilities and to conﬁrm that CTB cells are in transit rather than simply extending their processes into the STB. Another signiﬁcant proposition is that apoptosis in STB occurs only in regions of trophoblast denudation and repair associated with perivillous ﬁbrin-type ﬁbrinoid and is not a normal part of trophoblast turnover [1,3,27]. Applying TUNEL and other markers, apoptotic nuclei have been reported in STB and indices shown to increase between the ﬁrst trimester and term with further increases in post-mature, IUGR and pre-eclamptic organs [61e66]. At least some apoptotic nuclei are seen in CTB cells and syncytial knots in normal and HELLP syndrome pregnancies . Moreover,
there is evidence that levels of Bcl-2 protein (an apoptosis inhibitor) are inversely related to levels of apoptosis and reduced in regions of syncytial knotting [49,62]. A proportion of nuclei within STB display peripheral condensation of heterochromatin enveloping a central euchromatic area. These nuclei lack nucleoli and nuclear pores, are transcriptionally inactive, show evidence of oxidative damage and are particularly conspicuous in syncytial knots [9,16,18]. It has been shown  that nuclear morphology correlates with TUNEL-positivity and other apoptosis markers. In addition, there is evidence of greater TUNEL labelling of nuclei in syncytial knots than syncytial bridges . However, the literature is not consistent in identifying conspicuous TUNEL labelling of the ‘apoptosis-like’ nuclei in syncytial knots. Consequently, it must be concluded that it is unlikely that a high proportion of nuclei in syncytial knots are apoptotic in the sense of being TUNEL-positive. Similar dangers beset studies on apoptosis in STB as they do proliferation in CTB, namely, the unavailability of transit times or estimates of real numbers. These issues are particularly relevant when trying to interpret the impact of changes in labelling indices between normal and complicated pregnancies. For example, apoptosis rates are reported to be greater in IUGR [61,67] but these placentas have smaller-than-normal villous surface areas [68,69] and nuclear complements . Consequently, total numbers of apoptotic nuclei might be smaller despite higher apoptosis indices. 5.6. STB and the extrusion/deportation compartment It has been suggested that sequestration of clustered nuclei not only renders trophoblast less uniform in thickness (making passive diffusion more efﬁcient ,) but is also a prelude to extrusion and deportation of STB fragments [2,16,49,54]. More recently, it has been proposed that the main vehicle of nuclear loss is the syncytial sprout and not apoptotic nuclei in syncytial knots . Indeed, transcriptionally-active multinucleated STB clusters (presumably syncytial sprouts), rich in sFlt protein and mRNA, are found in the maternal circulation and in greater amounts in pre-eclampsia . Notwithstanding this lack of accord on mechanism(s) of origin, syncytial fragments of various sorts are extruded from STB during gestation and then deported from the uteroplacental circulation [72e75]. Deportation involves multinucleated syncytial fragments, microparticles and nanoscale particles. As many as 150,000 nucleated fragments per day may enter the maternal circulation where they vary widely in size (20e200 mm) and nuclear content (1e60) . It is expected that this occurs usually by a budding process which maintains the integrity of trophoblastic epithelium. Occasionally, however, integrity is disrupted by local injury and loss of fragments leads to local denudation of STB. Using anti-cytokeratin antibodies to identify trophoblast, deported fragments in blood from uterine veins were classiﬁed . In normal pregnancy, three different types were found of which the most frequent were large multinucleate STB fragments with densely-packed, heterochromatic nuclei often with a ‘teardrop’ shape. These types were regarded as a mixture of syncytial sprout and syncytial knot derivatives. In pre-eclampsia, syncytial sprouts were predominant and syncytial knots were far less frequent . Less common in normal pregnancy were smaller uninucleate fragments with morphological features of CTB cells and anucleate cytoplasmic fragments. Few large fragments are found normally in peripheral veins because they are cleared in the maternal pulmonary circulation. However, STB microparticles, probably derived from apical microvillous membranes, have been detected in peripheral blood in normal pregnancies together with nanovesicles including exosomes [73,75]. Some of these microparticles are associated with cell-free ‘fetal’ DNA (cffDNA which actually derives from the placenta) and the concentrations of which
T.M. Mayhew / Placenta 35 (2014) 229e240
in maternal blood are inﬂuenced by several factors including STB oxidative stress . Changes in the numbers and sizes of microparticles and nanovesicles may vary in normal and complicated pregnancies and differ in their immunomodulatory effects . Based on the above, estimates have been made of extrusion rates . If each of 150,000 fragments has an average of 30 nuclei, this represents extrusion of about 4.5 million nuclei per day or 32 million per week. However, term placentas contain about 5800 million CTB cells and, to replace the extruded nuclei, this would require only 0.08% of them to divide each day! Again, this estimate is provisional until more detailed quantitative analysis of the structure of deported fragments becomes available. However, they reveal that replacing lost fragments requires a very modest proliferative response which is comfortably within the capacity of CTB cells at term (Tables 1 and 2). Recently , it was proposed that extruded and deported apoptotic syncytial fragments are interiorised by pulmonary vascular endothelial cells without activating them and that this is important in maintaining maternal immune tolerance to fetal tissues. In contrast, clearance of necrotic material in pre-eclampsia stimulates endothelial cell activation and the release of factors that lead to a systemic endothelial dysfunction as part of a wider maternal inﬂammatory response. Though this proposition is attractive, opinion is divided on the nature and origins of the large STB fragments. One view is that some of the nuclei in syncytial knots represent the apoptotic endstage of STB differentiation [14,16,49,78]. A small proportion of knots eventually protrude from the trophoblast surface and are extruded by detachment as part of normal epithelial turnover. Detachment might involve an actin-based excision process [16,54] similar to that which occurs at the isthmus of daughter cells during cytokinesis but conﬁrmation is lacking. Because they contain some apoptotic nuclei, the fragments would not be expected to elicit endothelial and pro-inﬂammatory responses. However, an alternative view [1,3] regards this as unproven arguing that chromatin condensation might be the result of other processes involving acetylation, methylation or phosphorylation. This view also considers it more likely that most nucleus-rich deported fragments are derived from syncytial sprouts as a result of epithelial damage. Consistent with this could be the elevated levels of deported CTB cells in pre-eclampsia since these could signify more severe damage resulting in degeneration and necrosis which trigger vascular endothelial and inﬂammatory responses. It is conceivable that both views are correct. Since deportation involves multiple forms (syncytial knots, syncytial sprouts, false knots, CTB cells, anucleate portions of STB cytoplasm, STB microparticles and nanovesicles), the relative numbers may vary during gestation as they do between normal and complicated pregnancies. Hypoxia and oxidative stress might be common elements in these situations. On the balance of available evidence, it seems likely that the majority of deported multinucleated fragments in preeclampsia derive from syncytial sprouts. Given the infrequent occurrence of apparently budding syncytial knots, the same may be true for normal pregnancy. 5.7. STB damage and repair Other continuously-renewing epithelia are exposed to local forces/factors which cause damage, interfere with normal turnover and have to be repaired. Local injury to trophoblast can occur throughout gestation as a result of mechanical or chemical events. A mechanical factor is the avulsion of large (>10 mm diameter) intervillous bridges [79,80] which correspond to false knots on single sections . SEM studies of denudation sites created by such events show CTB cells scattered on the basal lamina. This is
more likely to occur in an accidental rather than regulated fashion and could be inﬂuenced by the increasing vigour of fetal movements during pregnancy. Similar factors and regional blood ﬂow inequalities may inﬂuence passage of syncytial sprouts into the intervillous space . Apoptosis of STB nuclei may be another way of initiating localised trophoblast denudation but apoptotic and necrotic features have also been seen in cultured STB after hypoxic degeneration followed by regeneration [81e83]. Hypoxia leads to higher levels of apoptosis in CTB cells and failure of fusion into STB, leading the latter to suffer secondary necrosis followed by shedding. Trophoblast denudation sites do show preferential deposition of perivillous ﬁbrin-type ﬁbrinoid , a uteroplacental coagulation product. This indicates that deposition itself is targeted (to limit local damage) and regulated (effected by systemic and trophoblastic haemostatic factors). Fibrin deposits act as a matrix for CT proliferation, syncytialisation and re-epithelialisation . If this occurs where trophoblast on adjacent villi is contiguous, it could initiate formation of intervillous bridges. Therefore, periodic formation and rupture of intervillous bridges may be linked to villous morphogenesis and trophoblast damage and repair throughout gestation. Finally, some sites of STB denudation might arise following extrusion of true syncytial knots [16,49,84]. Syncytial knots vary in morphology from broad but relatively ﬂat clusters bulging slightly towards the intervillous space to more spheroidal bud-like protrusions which look as though they are close to being shed [16,54]. One might expect that the former type would be more resistant, and the latter type more susceptible, to deportation. 6. Trophoblastic epithelium exists in a tightly-regulated steady state As for epidermis and intestinal epithelium, there is substantial evidence that villous trophoblast exists in a steady state which can be described by the relative amounts of CTB and STB or relative numbers of nuclei [12,15]. Nett growth occurs when the steady state favours CTB proliferation and recruitment over extrusion. The steady state may be perturbed not only spatially and temporally (by normal growth, damage and repair) but also as part of adaptive or pathological responses during complicated pregnancies. What are the measures of the trophoblast steady state? Consider the following: (i) Stereological analyses of placentas between the second trimester and term [12,15] show that whilst the number of CTB nuclei rise from 600 to 5800 million, that of STB nuclei rises from 6200 to 58,100 million. Throughout this period, the STB:CTB ratio is relatively constant (about 9:1). More recent analyses based on confocal microscopy have yielded a similar ratio for term placentas . (ii) The volume of trophoblast per CTB nucleus is remarkably constant over this period and amounts to roughly 11,000 mm3 . (iii) Over the same period, the ratio of transcriptionally active:inactive nuclei in STB is about 4:1 . Together, these ﬁndings provide compelling circumstantial evidence that the epithelial steady state is tightly regulated during much of normal gestation. Given this, one might predict that some measure of control would be exercised over STB extrusion as it is over CTB proliferation. Given that the loss of syncytial sprouts and false knots arising from mechanical damage must be essentially stochastic, this seems inconsistent with the notion that these fragments alone account for all trophoblast loss . It also begs at
T.M. Mayhew / Placenta 35 (2014) 229e240
least two questions. First, if true knots contain oxidativelydamaged nuclei which are no longer transcriptionally active, why would they be allowed to accumulate without being extruded and deported? Second, if true knots are marshalled and excised by an actin-based process resembling cytokinesis (see Ref. ), would this not be more suited to a regulated steady state? Quantifying epithelial kinetics requires information about proliferation and extrusion. A reasonable compromise might be to monitor changes in the steady state between proliferative and apoptotic indices, and combine this with a measure of absolute tissue size (e.g. trophoblast volume or complement of nuclei). Some studies have moved in this direction and monitored the relationship between TUNEL-labelled nuclei or M30-labelled regions and Ki67-labelled nuclear pools [27,29,34,49]. Though suggesting changes during gestation and between normal and complicated pregnancies, so far these studies have not included stereological estimates of numbers of nuclei. 7. Discussion and conclusions 7.1. What do we know? On the basis of the current state of knowledge, responses to the original points of contention might be as follows: (i) There is little evidence to support the idea that turnover of syncytial nuclei occurs in normal placenta. This cannot be reconciled with the fact that syncytial sprouts continue to detach and be deported to term. Moreover, deported multinucleate STB fragments sampled from uterine veins contain nuclei whose appearances are reported to resemble those of syncytial knots as well as syncytial sprouts. It seems at least possible that, apart from nuclei in syncytial sprouts, nuclei in aged (syncytial knot), mechanically damaged (false knot) and otherwise damaged (syncytial knot, non-clustered STB denudation sites, CTB) may be lost and deported and, thereby, contribute to epithelial turnover. The proportions of these different deported forms may vary across gestation and in different pregnancy types but this needs to be tested by further quantitative studies. (ii) There is little evidence that it occurs via an apoptosis-related process. There is circumstantial evidence that the ultrastructural features of some nuclei in syncytial knots and elsewhere are apoptosis-like and immunochemical evidence that a small proportion of nuclei in some syncytial knots in normal placentas do label with apoptotic markers. However, there is no clear between-group consistency in ﬁnding conspicuous labelling for TUNEL in nuclei within syncytial knots. Therefore, it seems unlikely that apoptosis within STB plays a major role in determining the nature of the nucleated fragments found in uterine vein samples. However, some apoptotic CTB cells and their vesicular fragments may be lost into the intervillous space after completing their transit across STB. Others might be removed by being interiorised by STB and then degraded within STB by an autophagic pathway [3,35]. Therefore, apoptosis may be implicated in deportation at syncytial, cellular, microparticulate and nanovesicular levels. (iii) Epigenetic modiﬁcations, rather than apoptotic events, underlie changes in the heterochromatin content of STB nuclei. At the moment there is evidence that nuclei in syncytial knots are transcriptionally inactive, have been in STB for a longer duration and show signs (immunopositivity for 8-oxodeoxyguanosine) of oxidative damage to DNA. But we do not have incontrovertible evidence as to whether or not the same
nuclei undergo chromatin modiﬁcations including DNA methylation, histone acetylation, methylation, phosphorylation, sumoylation or ubiquitylation or gene regulation by non-coding RNAs. Nor do we have ﬁrm evidence that, instead, they undergo apoptosis-related processes such as DNA fragmentation (TUNEL) and cleavage of poly(ADPribose) polymerase (PARP) and nuclear membrane proteins such as lamin B [3,49]. (iv) A proportion of STB nuclei remain transcriptionally active throughout gestation. There is compelling evidence not only that this is true but that the underlying process is highly regulated. The ratio of active:inactive nuclei is essentially constant from at least the start of the second trimester. (v) Syncytial knots are more resistant to deportation than syncytial sprouts. The morphological forms of sprouts and knots are consistent with this proposition. Syncytial sprouts and the intervillous bridge form of false knot seem to be lost by a stochastic process involving mechanical forces rather than a controlled process. Syncytial knots with the more spheroidal form seem to occur less frequently than the ﬂatter forms and, consequently, might explain the relative paucity of knot-like derivatives in uterine vein samples. It has been suggested [16,54] that the spheroidal bud-like form of knot might be lost by a more regulated process similar to that mediated by actin ﬁlaments in cytokinesis but more systematic studies are required to test this. However, a recent study failed to ﬁnd actin labelling in association with regions of nuclear clustering . It is, perhaps, worth emphasising that even stochastic processes can produce outcomes similar to those of regulated events! Thus, chromosome partitioning in mitosis is an extremely well-regulated non-stochastic process in which exactly the same numbers are assigned to each daughter cell. However, rough ER is partitioned in HeLa cells by a non-uniform process in which cisternal elements are positioned partly by association with cortical actin and partly with the spindle region . Nevertheless, daughter cells receive roughly equal quantities of rough ER. Similar near-equalities apply to mitotic partitioning of other organelles. (vi) Apoptosis is normally conﬁned to CTB cells and areas of STB associated with ﬁbrin-type ﬁbrinoid. Whilst some studies have shown syncytial knots in normal placentas that are immunopositive for apoptotic markers, others have found no evidence of apoptosis in the absence of ﬁbrin-type ﬁbrinoid. There is evidence that hypoxia increases apoptosis in CTB and, to a lesser extent, STB. Recruitment of CTB into STB is compromised and there is loss of STB fragments and epithelial denudation. On tissue sections, the presence of a region of de-epithelialized STB does not allow us to identify whether the source of the lost region was a syncytial sprout, true or false syncytial knot or attenuated STB with nonclustered nuclei. Indeed, the deposition of perivillous ﬁbrin-type ﬁbrinoid might be a prelude to reepithelialisation and repair whatever the source of the denuded portion of STB. (vii) Apoptotic events seen in CTB may be mistakenly ascribed to STB. It is likely that this occurs and preliminary calculations suggest that the relative bias depends heavily on real rates of apoptosis in the two compartments. The bigger the discrepancy in favour of STB, the smaller the bias. The biases are likely to be large but, at present, we have no hard data which would allow us to estimate relative biases in different regions of STB, across gestation or between normal and complicated pregnancies.
T.M. Mayhew / Placenta 35 (2014) 229e240
7.2. Remaining questions Pertinent to the above are some questions: (i) If loss and deportation involve only syncytial sprouts and these are produced by chance events (mechanical damage), how do we reconcile this with clear evidence of tight regulation as manifested by constancy in the ratios of STB:CTB and transcriptionally active:inactive STB nuclei? (ii) If true knots sequester transcriptionally-inactive, oxidatively-damaged nuclei, what beneﬁt accrues from allowing them to accumulate without being extruded and deported? This raises additional questions about the factors regulating decisions over whether such nuclei are subjected to autophagy instead of deportation. (iii) Why do nuclei with condensed chromatin cluster? If the reason is to improve passive diffusion by creating thick and thin regions of STB, why is it that condensed chromatin predominates? What are the factors inﬂuencing the creation of such nuclear clusters? (iv) STB contains nuclei with ultrastructural features of condensed chromatin, no nucleoli and no nuclear pores. These are found in syncytial knots at high packing densities. If these are not the nuclei being detected by apoptosis markers which are the nuclei? They are not in syncytial sprouts (except perhaps at their tips) and we do not know the extent to which they can be accounted for by misclassiﬁcation of CTB apoptoses. (v) Could the transit across STB of apoptotic CTB cells or membrane-bound CTB cell fragments [3,35] account for some of the cffDNA found in the maternal circulation? Immunochemical analyses have implicated also apoptotic and necrotic regions of STB  but are there other placental sources of cffDNA? (vi) It is reasonable to ask whether apoptosis within STB would be compatible with epithelial viability given that this is a syncytial continuum. Recently , an interesting comparison was drawn with another tissue, skeletal muscle, in which a multinucleated syncytium is associated with uninucleate precursor cells. Apparently, apoptosis rarely, if ever, occurs in the muscle syncytium but is found in the uninucleate cells. If apoptoses do occur (albeit infrequently) in syncytial knots, could the fasciculation provided by cytokeratin intermediate ﬁlaments, coupled with greater Bcl-2 protein levels, help to localise their potentially dangerous effects? (vii) If loss of apoptotic STB fragments maintains maternal tolerance but necrotic fragments trigger endothelial activation, what advantage is derived from not shedding apoptotic STB? Do some STB fragments become apoptotic once they are shed?
Reconstruction of serial physical sections for LM helps to distinguish syncytial sprouts and true knots from false knots . Stereological tools can be used to estimate various structural quantities including numbers of nuclei to understand better CTB proliferation and STB differentiation and transcriptional status [15,17,30]. Although serial physical sectioning for LM has a number of attendant technical problems (section compression, knife score marks, section damage/loss, the need for ﬁducial markers, etc), these can be avoided by optical sectioning using high numerical aperture lenses or confocal microscopy (e.g. see Ref. ). Optical sectioning also has attractions from the stereological perspective. However, even here, there is scope for misinterpretation because of resolution limits . Serial resin sections for LM and TEM offer better resolution than parafﬁn sections for testing whether or not apparently free syncytial knots in the intervillous space are attached to, or detached from, villous trophoblast. It is almost too obvious to state that you cannot count what you cannot identify. This requires not only adequate lateral resolution in the ﬁnal image but the use of techniques to assist identiﬁcation (such as immunochemical labelling) and instruments which improve resolution. These combinations are important for correctly assigning labelling to different classes of nuclei . A powerful and underused technique (at least in placental studies) is the combination of TEM and immunogold cytochemistry for which a coherent set of quantitative approaches is available [89,90]. This combination solves most of the problems associated with visualization versus localisation and has been used to identify TUNELlabelled nuclei in CTB cells . Their use could help to test whether or not the apoptosis-like ultrastructural features seen in certain STB nuclei in syncytial knots correlate with immunopositivity for TUNEL, cleaved PARP or cleaved lamin B [3,49]. There may also be scope for using more recently-developed TEM and SEM imaging modalities such as electron tomography and array tomography [91e93] which offer high-resolution images for serial reconstruction and stereological analysis. (ii) Another advance would be to standardise best practice in other aspects of trophoblast studies. An obvious example is the choice of references for proliferation and other indices. At present, there is no standardisation (e.g. apoptotic indices might refer to the number of STB nuclei, number of trophoblast nuclei, number of all nuclei in villi or a unit area of section). Such discrepancies make it much harder to draw comparisons between research groups. In fact, relating indices solely to the compartment in which they are immediately contained has a practical beneﬁt: it provides estimates of greater precision and this is a desirable property when testing for signiﬁcant differences between experimental groups. Preferably, mitotic phase labelling indices should refer to the total population of CTB cells, apoptotic labelling of CTB cells should refer also to the total population of CTB cells and apoptotic labelling of STB nuclei should refer to the total population of STB nuclei.
7.3. What needs to be done now? (i) A recurring theme in this review has been the importance of undertaking rigorous quantiﬁcation in order to obtain hard (biologically useful) 3D structural data. This implies employing clearly-stated unbiased sampling protocols . Ensuring that quantiﬁcation is 3D is tackled currently in two main ways: (i) by applying reconstruction techniques such as serial physical sectioning or optical sectioning and (ii) by extrapolation using stereological sampling and estimation tools. These approaches are not mutually exclusive but can be, and have been, used together.
Other areas where standardisation might be beneﬁcial include optimising conditions for in vitro cell culture and greater use of villous explants which have certain advantages over primary human CTB cultures in terms of preserving normal architectural features of the tissue [7,43]. (iii) In the case of the CTB compartment, more information is required about the factors controlling transition from a clonogenic cell to an intermediate and then a more differentiated cell and transition from symmetrical to asymmetrical mitoses . Also, we know that expression levels of HIF-1a
T.M. Mayhew / Placenta 35 (2014) 229e240
are high early in gestation but know little about HIF function and HIF levels in CTB at later stages of gestation or the factors which might derive from STB or the villous stroma to inﬂuence CTB. (iv) We need more and better information about the precise nature of STB fragments shed from the placenta and their apoptotic or necrotic nature [74,76] and quantitative data on the different types of fragment (from syncytial to nanovesicular) at different stages of gestation. For example, we might predict that the fraction of syncytial sprouts is relatively greater in early gestation when villous sprouting is more intense. Also, since syncytial knots increase in frequency during gestation, we might predict that their frequency in uterine vein samples also would increase. Furthermore, we have no hard data concerning phases of STB differentiation and loss to match that on phases of the cell cycle. Ultrastructural characteristics of STB nuclei are suggestive of such phases but we have no clear idea of their duration. If a nucleus is neither euchromatic nor displays peripheral condensed chromatin , how long is it in this phase before it displays the end-stage phenotype? Relative incidences suggest that the intermediate phenotype is more common and the phase duration longer. Recent studies  do conﬁrm that nuclei in syncytial knots have been longer in STB than those in syncytial sprouts and this is also consistent with a temporal sequence of differentiation . In the same context, we have little information about the way in which STB is maintained and there has been no systematic analysis of the mechanisms for sequestration and extrusion of nuclei . It has been suggested that STB nuclei are marshalled into syncytial knots from different directions  and, if conﬁrmed, this might explain the fascicular appearance of nuclear groupings within knots as well as the arrangement and phenotypic characteristics of nuclei surrounding knots . (v) We need to ﬁnd and apply markers which might help to distinguish apoptosis in CTB from that in STB. For example, it has been proposed that hypoxia, via p53 protein (which regulates proliferation and maintains genome stability), has different effects on apoptosis and autophagy in CTB and STB . In CTB, there is a relatively greater increase in apoptosis and, in STB, a greater increase in autophagy. (vi) Detailed results obtained from studies on different pregnancy complications were not within the scope of this review. Clearly, the ﬁndings of such studies, particularly those involving pre-eclampsia, are pertinent and shed light on how normal turnover processes are perturbed. These include STB stress and the integrated stress response , the induction of a novel splicing variant of soluble vascular endothelial growth factor receptor 1 in syncytial knot regions  and the pro-inﬂammatory effects of circulating placental microvesicles mediated in part by syncytin 1 [96,97].
Acknowledgements I am grateful for research awards from various funding agencies and the collaborative efforts of, and interactions with, many students and colleagues who have shared an interest in the placenta over the past 30 years. I thank also three anonymous referees who suggested substantial improvements to the earlier drafts of this review. Finally, I dedicate this review to an inspirational morphologist, the late Peter Kaufmann.
References  Burton GJ, Jones CJ. Syncytial knots, sprouts, apoptosis and trophoblast deportation from the human placenta. Taiwan J Obstet Gynecol 2009;48:28e37.  Huppertz B. IFPA award in placentology lecture: biology of the placental syncytiotrophoblast e myths and facts. Placenta 2010;31:575e81.  Longtine MS, Chen B, Odibo AO, Zhong Y, Nelson DM. Caspase-mediated apoptosis of trophoblasts in term human placental villi is restricted to cytotrophoblasts and absent from the multinucleated syncytiotrophoblast. Reproduction 2012;143:107e21.  Coleman SJ, Gerza L, Jones CJP, Sibley CP, Aplin JD, Heazell AEP. Syncytial nuclear aggregates in normal placenta show increased nuclear condensation, but apoptosis and cytoskeletal redistribution are uncommon. Placenta 2013;34:449e55.  Jackson MR, Mayhew TM, Boyd PA. Quantitative description of the elaboration and maturation of villi from 10 weeks of gestation to term. Placenta 1992;13: 357e70.  Benirschke K, Kaufmann P, Baergen R. Pathology of the human placenta. 5th ed. New York: Springer; 2006.  Tuuli MG, Longtine MS, Nelson DM. Review: oxygen and trophoblast biology e a source of controversy. Placenta 2011;32(Suppl. A):S109e18.  Jauniaux E, Hempstock J, Greenwold N, Burton GJ. Trophoblastic oxidative stress in relation to temporal and regional differences in maternal placental blood ﬂow in normal and abnormal early pregnancies. Am J Pathol 2003;162: 115e25.  Jones CJ, Fox H. Ultrastructure of the normal human placenta. Electron Microsc Rev 1991;4:129e78.  Bright NA, Ockleford CD. Cytotrophoblast cells: a barrier to maternofetal transmission of passive immunity? J Histochem Cytochem 1995;9:933e44.  Jones CJP, Harris LK, Whittingham J, Aplin JD, Mayhew TM. A re-appraisal of the morphophenotype and basal lamina coverage of cytotrophoblasts in human term placenta. Placenta 2008;29:215e9.  Mayhew TM, Simpson RA. Quantitative evidence for the spatial dispersal of trophoblast nuclei in human placental villi during gestation. Placenta 1994;15:837e44.  Jackson MR, Joy CF, Mayhew TM, Haas JD. Stereological studies on the true thickness of the villous membrane in human term placentae: a study of placentae from high-altitude pregnancies. Placenta 1985;6:249e58.  Mayhew TM, Barker BL. Villous trophoblast: morphometric perspectives on growth, differentiation, turnover and deposition of ﬁbrin-type ﬁbrinoid during gestation. Placenta 2001;22:628e38.  Simpson RA, Mayhew TM, Barnes PR. From 13 weeks to term, the trophoblast of human placenta grows by the continuous recruitment of new proliferative units: a study of nuclear number using the disector. Placenta 1992;13:501e12.  Mayhew TM, Leach L, McGee R, Wan Ismail W, Myklebust R, Lammiman MJ. Proliferation, differentiation and apoptosis in villous trophoblast at 13-41 weeks of gestation (including observations on annulate lamellae and nuclear pore complexes). Placenta 1999;20:407e22.  Fogarty NME, Mayhew TM, Ferguson-Smith AC, Burton GJ. A quantitative analysis of transcriptionally active syncytiotrophoblast nuclei across human gestation. J Anat 2011;219:601e10.  Fogarty NME, Ferguson-Smith AC, Burton GJ. Syncytial knots (Tenney-Parker changes) in the human placenta. Evidence of loss of transcriptional activity and oxidative damage. Am J Pathol 2013;183:144e52.  Van Dienst PJ, Brugal G, Baak JPA. Proliferation markers in tumours: interpretation and clinical value. J Clin Pathol 1998;51:716e24.  Scholzen T, Gerdes J. The Ki-67 protein: from the known and the unknown. J Cell Physiol 2000;182:311e22.  Blankenship TN, King BF. Developmental expression of Ki-67 antigen and proliferating cell nuclear antigen in Macaque placentas. Dev Dyn 1994;201: 324e33.  Kar M, Ghosh D, Sengupta J. Histochemical and morphological examination of proliferation and apoptosis in human ﬁrst trimester villous trophoblast. Hum Reprod 2007;22:2814e23.  Howard CV, Reed MG. Unbiased stereology: three-dimensional measurement in microscopy. 2nd ed. Abingdon: Garland Science/BIOS Scientiﬁc Publishers; 2005.  Mayhew TM. Stereology and the placenta: where’s the point? e a review. Placenta 2006;27(Suppl. A):S17e25.  Arnholt H, Meisel F, Fandrey K, Löhrs U. Proliferation of villous trophoblast of the human placenta in normal and abnormal pregnancies. Virchows Arch B Cell Pathol 1991;60:365e72.  Smith SC, Price E, Hewitt MJ, Symonds EM, Baker PN. Cellular proliferation in the placenta in normal human pregnancy and pregnancy complicated by intrauterine growth restriction. J Soc Gynecol Invest 1998;5:317e23.  Chan CC, Lao TT, Cheung ANY. Apoptotic and proliferative activities in ﬁrst trimester placentae. Placenta 1999;20:223e7. _  Zadrozna M, Nowak B, Marcinek A, Duc J. Villous trophoblast cell turnover in placentas from preterm pregnancy and pregnancy complicated by intrauterine growth restriction (IUGR). Folia Biol (Kraków) 2010;58:79e83.  Higgins L, Mills TA, Greenwood SL, Cowley EJ, Sibley CP, Jones RL. Maternal obesity and its effect on placental cell turnover. J Matern Fetal Neonatal Med 2013;26:783e8.  Widdows K, Kingdom JCP, Ansari T. Double immune-labelling of proliferating villous cytotrophoblasts in thick parafﬁn sections: integrating immuno-
T.M. Mayhew / Placenta 35 (2014) 229e240
   
  
  
  
  
histochemistry and stereology in the human placenta. Placenta 2009;30:735e 8. Mayhew TM, Huppertz B, Kaufmann P, Kingdom JCP. The ‘reference trap’ revisited: examples of the dangers in using ratios to describe fetoplacental angiogenesis and trophoblast turnover. Placenta 2003;24:1e7. } ﬂer H, Zaccheo D, Castellucci M. Mühlhauser J, Crescimanno C, Kaufmann P, Ho Differentiation and proliferation patterns in human trophoblast revealed by cerbB-2 oncogene product and EGF-R. J Histochem Cytochem 1993;41:165e73. Ishihara N, Matsuo H, Murakoshi H, Laoag-Fernandez JB, Samoto T, Maruo T. Changes in proliferative potential, apoptosis and Bcl-2 protein expression in cytotrophoblasts and syncytiotrophoblast in human placenta over the course of pregnancy. Endocr J 2000;47:517e27. Prusac IK, Tomas SZ, Roje D. Apoptosis, proliferation and Fas ligand expression in placental trophoblast from pregnancies complicated by HELLP syndrome or pre-eclampsia. Acta Obstet Gynecol Scand 2011;90:1157e63. Burton GJ, Skepper JN, Hempstock J, Cindrova T, Jones CJP, Jauniaux E. A reappraisal of the contrasting morphological appearances of villous cytotrophoblast cells during early human pregnancy: evidence for both apoptosis and primary necrosis. Placenta 2003;24:297e305. Babischkin JS, Burleigh DW, Mayhew TM, Pepe GJ, Albrecht ED. Developmental regulation of morphological differentiation of placental villous trophoblast in the baboon. Placenta 2001;22:276e83. Bamberger A-M, Sudah S, Bamberger CM, Schulte HM, Löning T. Expression patterns of the cell-cycle inhibitor p27 and the cell-cycle promoter cyclin E in the human placenta throughout gestation: implications for the control of proliferation. Placenta 1999;20:401e6. Korgun ET, Celik-Ozenci C, Acar N, Cayli S, Desoye G, Demir R. Location of cell cycle regulators cyclin B1, cyclin A, PCNA, Ki67 and cell cycle inhibitors p21, p27 and p57 in human ﬁrst trimester placenta and deciduas. Histochem Cell Biol 2006;125:615e24. Mayhew TM, Wadrop E, Simpson RA. Proliferative versus hypertrophic growth in tissue subcompartments of human placental villi during gestation. J Anat 1994;184:535e43. Teasdale F. Functional signiﬁcance of the zonal morphologic differences in the normal human placenta. A morphometric study. Am J Obstet Gynecol 1978;130:773e81. Teasdale F. Gestational changes in the functional structure of the human placenta in relation to fetal growth. A morphometric study. Am J Obstet Gynecol 1980;137:560e8. Gerbie AB, Hathaway HH, Brewer JI. Autoradiographic analysis of normal trophoblastic proliferation. Am J Obstet Gynecol 1968;100:640e8. Aplin JD. Developmental cell biology of human villous trophoblast: current research problems. Int J Dev Biol 2010;54:323e9. Mayhew TM, Leach L. Aspects of structure and function in human placenta. Curr Obstet Gynaecol 1998;8:2e7. Cronier L, Bastide B, Defamie N, Niger C, Pointis G, Gasc JM, et al. Involvement of gap junctional communication and connexin expression in trophoblast differentiation of the human placenta. Histol Histopathol 2001;16:285e95. Aplin JD, Jones CJ, Harris LK. Adhesion molecules in human trophoblast e a review. I. Villous trophoblast. Placenta 2009;30:293e8. Lyden TW, Ng AK, Rote NS. Modulation of phosphatidylserine epitope expression by BeWo cells during forskolin treatment. Placenta 1993;14:177e86. Adler RR, Ng AK, Rote NS. Monoclonal antiphosphatidylserine antibody inhibits intercellular fusion of the choriocarcinoma line, JAR. Biol Reprod 1995;53:905e10. Huppertz B, Frank H-G, Kingdom JCP, Reister F, Kaufmann P. Villous cytotrophoblast regulation of the syncytial apoptotic cascade in the human placenta. Histochem Cell Biol 1998;110:495e508. Huppertz B, Kadyrov M, Kingdom JC. Apoptosis and its role in the trophoblast. Am J Obstet Gynecol 2006;195:29e39. Lin C, Lin M, Chen H. Biochemical characterization of the human placental transcription factor GCMa/1. Biochem Cell Biol 2005;280:41928e39. Malassine A, Blaise S, Handschuh K, Lalucque H, Dupressoir A, Evain-Brion D, et al. Expression of the fusogenic HERV-FRD Env glycoprotein (syncytin 2) in human placenta is restricted to villous cytotrophoblastic cells. Placenta 2007;28:185e91. Gauster M, Moser G, Orendi K, Huppertz B. Factors involved in regulating trophoblast fusion: potential role in the development of preeclampsia. Placenta 2009;30(Suppl. A):S49e54. Mayhew TM. Villous trophoblast of human placenta: a coherent view of its turnover, repair and contributions to villous development and maturation. Histol Histopathol 2001;16:1213e24. Burton GJ. Deportation of syncytial sprouts from the term human placenta. Placenta 2011;32:96e8. Jones CJ, Fox H. Syncytial knots and intervillous bridges in the human placenta: an ultrastructural study. J Anat 1977;124:275e86. Ellery PM, Cindrova-Davies T, Jauniaux E, Ferguson-Smith AC, Burton GJ. Evidence for transcriptional activity in the syncytiotrophoblast of the human placenta. Placenta 2009;30:329e34. Earnshaw WC. Nuclear changes in apoptosis. Curr Opin Cell Biol 1995;7:337e 43. Heazell AE, Crocker IP. Live and let die e regulation of villous trophoblast apoptosis in normal and abnormal pregnancies. Placenta 2008;29:772e83. Scifres CM, Nelson DM. Intrauterine growth restriction, human placental development and trophoblast cell death. J Physiol 2009;587:3453e8.
 Smith SC, Baker PN, Symonds EM. Placental apoptosis in normal human pregnancy. Am J Obstet Gynecol 1997;177:57e65.  Ishihara N, Matsuo H, Murakoshi H, Laoag-Fernandez JB, Samoto T, Maruo T. Increased apoptosis in the syncytiotrophoblast in human term placentas complicated by either preeclampsia or intrauterine growth retardation. Am J Obstet Gynecol 2002;186:158e66.  Smith SC, Baker PN. Placental apoptosis is increased in post-term pregnancies. Br J Obstet Gynaecol 1999;106:861e2.  Kadyrov M, Kaufmann P, Huppertz B. Expression of a cytokeratin 18 neoepitope is a speciﬁc marker for trophoblast apoptosis in human placenta. Placenta 2001;22:44e8.  Athapathu H, Jayawardana MA, Senanayaka L. A study of the incidence of apoptosis in the human placental cells in the last weeks of pregnancy. J Obstet Gynaecol 2003;23:515e7.  Heazell AE, Sharp AN, Baker PN, Crocker IP. Intra-uterine growth restriction is associated with increased apoptosis and altered expression of proteins in the p53 pathway in villous trophoblast. Apoptosis 2011;16:135e44.  Smith SC, Baker PN, Symonds EM. Increased placental apoptosis in intrauterine growth restriction. Am J Obstet Gynecol 1997;177:1395e401.  Mayhew TM, Ohadike C, Baker PN, Crocker IP, Johnson IR, Mitchell C, et al. Stereological investigation of placental morphology in pregnancies complicated by pre-eclampsia with and without intrauterine growth restriction. Placenta 2003;24:219e26.  Mayhew TM, Wijesekara J, Baker PN, Ong SS. Morphometric evidence that villous development and fetoplacental angiogenesis are compromised by intrauterine growth restriction but not by pre-eclampsia. Placenta 2004;25: 829e33.  Teasdale F. Idiopathic intrauterine growth retardation: histomorphometry of the human placenta. Placenta 1984;5:83e92.  Rajakumar A, Cerdeira AS, Rana S, Zsengeller Z, Edmunds L, Jeyabalan A, et al. Transcriptionally active syncytial aggregates in the maternal circulation may contribute to circulating soluble fms-like tyrosine kinase 1 in preeclampsia. Hypertension 2012;59:256e64.  Johansen M, Redman CWG, Wilkins T, Sargent IL. Trophoblast deportation in human pregnancy e its relevance for pre-eclampsia. Placenta 1999;20:531e9.  Goswami D, Tannetta DS, Magee LA, Fuchisawa A, Redman CWG, Sargent IL, et al. Excess syncytiotrophoblast microparticle shedding is a feature of earlyonset pre-eclampsia, but not normotensive intrauterine growth restriction. Placenta 2006;27:56e61.  Askelund KJ, Chamley LW. Trophoblast deportation part I: review of the evidence demonstrating trophoblast shedding and deportation during human pregnancy. Placenta 2011;32:716e23.  Redman CWG, Tannetta DS, Dragovic RA, Gardiner C, Soothcombe JH, Collett GP, et al. Review: does size matter? Placental debris and the pathophysiology of pre-eclampsia. Placenta 2012;33(Suppl. A):S48e54.  Taglauer ES, Wilkins-Haug L, Bianchi DW. Cell-free fetal DNA in the maternal circulation as an indication of placental health and disease. Placenta 2014;35: S64e8.  Chamley LW, Chen Q, Ding J, Stone PR, Abumaree M. Trophoblast deportation: just a waste disposal system or antigen sharing? J Reprod Immunol 2011;88:99e105.  Huppertz B, Frank H-G, Kaufmann P. The apoptosis cascade e morphological and immunohistochemical methods for its visualization. Anat Embryol 1999;200:1e18.  Burton GJ. Intervillous connections in the mature human placenta: instances of syncytial fusion or section artifacts? J Anat 1986;145:13e23.  Burton GJ. The ﬁne structure of the human placental villus as revealed by scanning electron microscopy. Scanning Microsc 1987;1:1811e28.  Nelson DM. Apoptotic changes occur in syncytiotrophoblast of human placental villi where ﬁbrin type ﬁbrinoid is deposited at discontinuities in the villous trophoblast. Placenta 1996;17:387e91.  Huppertz B, Kingdom J, Caniggia I, Desoye G, Black S, Korr H, et al. Hypoxia favours necrotic versus apoptotic shedding of placental syncytiotrophoblast into the maternal circulation. Placenta 2003;24:181e90.  Heazell AE, Lacey HA, Jones CJP, Huppertz B, Baker PN, Crocker IP. Effects of oxygen on cell turnover and expression of regulators of apoptosis in human placental trophoblast. Placenta 2008;29:175e86.  Mayhew TM, Bowles C, Orme G. A stereological method for testing whether or not there is random deposition of perivillous ﬁbrin-type ﬁbrinoid at the villous surface: description and pilot applications to term placentae. Placenta 2000;21:684e92.  Nelson DM, Crouch EC, Curran EM, Farmer DR. Trophoblast interaction with ﬁbrin matrix. Epithelialization of perivillous ﬁbrin deposits as a mechanism for villous repair in the human placenta. Am J Pathol 1990;136:855e65.  McCullough S, Lucocq J. Endoplasmic reticulum positioning and partitioning in mitotic HeLa cells. J Anat 2005;206:415e25.  Tjoa ML, Cindrova-Davies T, Spasic-Boskovic O, Bianch DW, Burton GJ. Trophoblastic oxidative stress and the release of cell-free feto-placental DNA. Am J Pathol 2006;169:400e4.  Mayhew TM. Taking tissue samples from the placenta: an illustration of principles and strategies. Placenta 2008;29:1e14.  Mayhew TM. Quantifying immunogold localization patterns on electron microscopic thin sections of placenta: recent developments. Placenta 2009;30:565e70.  Mayhew TM. Mapping the distributions and quantifying the labeling intensities of cell compartments by immunoelectron microscopy: progress towards a coherent set of methods. J Anat 2011;219:647e60.
T.M. Mayhew / Placenta 35 (2014) 229e240
 Vanhecke D, Studer D, Ochs M. Stereology meets electron tomography: towards quantitative 3D electron microscopy. J Struct Biol 2007;159:443e 50.  Micheva KD, Busse B, Weller NC, O’Rourke N, Stephen J, Smith SJ. Single synapse analysis of a diverse synapse population: proteomic imaging methods and markers. Neuron 2010;68:639e53.  Kay KR, Smith C, Wright AK, Serrano-Pozo A, Pooler AM, Kofﬁe R, et al. Studying synapses in human brain with array tomography and electron microscopy. Nat Protoc 2013;8:1366e80.  Redman CW, Sargent IL, Staff AC. IFPA Senior Award cLecture: making sense of pre-eclampsia e two placental causes of pre-eclampsia? Placenta 2014;35: S20e5.
 Sela S, Itin A, Natanson-Yaron S, Greenﬁeld C, Goldman-Wohl D, Yagel S, et al. A novel human-speciﬁc soluble vascular endothelial growth factor receptor 1: cell-type-speciﬁc splicing and implications to vascular endothelial growth factor homeostasis and preeclampsia. Circ Res 2008;102:1566e74.  Holder BS, Tower CL, Jones CJ, Aplin JD, Abrahams VM. Heightened proinﬂammatory effect of preeclamptic placental microvesicles on peripheral blood immune cells in humans. Biol Reprod 2012;86:103.  Holder BS, Tower CL, Forbes K, Mulla MJ, Aplin JD, Abrahams VM. Immune cell activation by trophoblast-derived microvesicles is mediated by syncytin 1. Immunology 2012;136:184e91.