Post partum haemorrhage

Maternal mortality rates Wikimedia Commons

In developing countries, loss of blood after childbirth is a major cause of death. The bleeding occurs from blood vessels that rupture when the placenta detaches from the uterine wall. Usually the vessels are closed by a strong contraction of the uterine musculature. Occasionally that mechanism fails.

Recently Elizabeth Abrams and Julienne Rutherford asked what was known about birth in non-human primates. Maternal death from post partum bleeding seemed nowhere near as common as in humans. Why?

Although widening of the uterine arteries does occur in pregnant monkeys, the process extends much deeper in the human uterus. The advantage to humans is that it ensures a greater supply of oxygen to aid development of the fetal brain. Abrams and Rutherford suggested that this advance came with a price tag that included a bigger risk of post partum blood loss.

Guinea pigs mate after giving birth

Pregnant guinea pig Wikimedia Commons

More than a century ago, Hans Strahl understood the importance of studying what happened to the uterus immediately after birth. He noted that guinea pigs and many other rodents mate soon after giving birth. In these species the placenta detaches with minimal damage to the uterus. It can be repaired quickly enough to receive the new embryo as it implants a few days later.

Elephants space their pregnancies

African forest elephant Wikimedia Commons

This is by no means universal. One of the best documented instances is the African elephant. After birth there is extensive bleeding into the uterine cavity. Loosening of the placenta leaves a permanent scar – at autopsy the number of scars in the uterus indicates how often an elephant has been pregnant. However, the interval between pregnancies is 6-8 years so there is more than enough time for the wound to heal. For more on elephant placentation follow this link.

Spoils of the forest

Spolia nemoris is the appropriate title of a paper listing the 1072 uteri collected by Hubrecht during a 19th Century expedition to the Dutch East Indies (Indonesia). They included two hundred specimens of Tarsius.

Phillipine tarsier (Tarsius syrichta) Wikipedia Commons

Ian Redmond recently opined that tarsiers, which can swivel their heads 180^o, look like the creation of a Hollywood special-effects team. They are intermediate between lemurs and lorises on the one hand and monkeys and apes on the other. Tarsiers are easily overlooked in surveys of primate placentation. None of the accounts are recent so they do not readily pop up in literature searches. Hubrecht’s own description can be hard to track down so follow this link and scroll down to Appendix B. A more accessible interpretation of the same material is in J. P. Hill's classic review.

Placenta of the spectral tarsier (Tarsius tarsier) Hubrecht Collection

The exchange area of the tarsier placenta is reminiscent of that in Neotropical primates, with the villi connected by trabecula.

It was on the basis of placentation that Hubrecht suggested tarsiers belonged with monkeys and apes in what are now referred to as haplorrhine primates. This led to a spirited debate in the pages of Science and American Naturalist with Charles Earle, curator of fossils at the American Museum of Natural History. Earle wanted to keep tarsiers with lemurs and lorises (strepsirrhine primates).

Pregnancy-specific glycoproteins

The carcinoembryonic antigen family

Kammerer and Zimmermann BMC Biology 2010, 8:12

Human placenta secretes great quantities of protein including pregnancy-specific beta1-glycoprotein (PSG). The precise biological function of PSG has yet to be resolved, although it is thought to modify maternal immune responses.

PSGs are part of the carcinoembryonic antigen (CEA) family, other members being the CEA cell adhesion molecules (CEACAMs). Primordial CEACAMs occur across the full range of mammals, including marsupials and monotremes (green shading in the figure).

Interestingly, there has been convergent evolution of PSGs in rodents and primates (grey shading). They are not found in dogs or cattle although they may occur in the horse.

These proteins belong to the immunoglobulin gene superfamily. Human PSGs are composed of an immunoglobulin variable-like domain (red circle) and a variable number of constant-like domains (blue circles). As the figure shows, mouse PSGs are built slightly differently. This supports the idea that PSGs evolved separately on the rodent and primate lineages.

Towards the end of human pregnancy, PSG is the most abundant placental protein in maternal blood. It is surprising we know so little about it.

Altricial and precocial neonates

Mouse with newborn litter (P0.5) Courtesy of Peter Bollen

Anyone who has kept mice as pets or in the lab will recognize the tiny pink pups, which are entirely dependent on parental care. They are poorly-developed (altricial) with closed eyes and no hair.

Newborn guinea pigs (one day old) Courtesy of Per Svendsen

Contrast this with the well-developed (precocial) litter of a guinea pig. They need to suckle but can move freely and they have open eyes and a full coat of hair. In a classic survey of 394 mammals, Martin and MacLarnon found that altriciality went with short gestations and large litters whereas precocial neonates came from small litters and long gestations. Some of the species they surveyed fell in between with for example eyes shut at birth and not opening for at least five days.

Where do human beings fit into the spectrum? Babies are clearly more helpless than newborn guinea pigs. Yet singletons are the norm and pregnancy lasts for nine months. As in precocial neonates much organ development occurs before birth in humans. The placenta has to function much longer than in a mouse and as a result acquires additional functions.

Comparative placentation

Dr. Kurt Benirschke

Comparative placentation is an important research area in its own right and as a background to animal models of human placentation. I will not be posting detailed information about the placenta of different mammals. That function is served by a web site maintained by Dr. Kurt Benirschke (above), a retired placental pathologist with strong links to San Diego Zoo. You will find the web site at this link.

The index page currently lists over 150 species and the site is especially strong on primates and artiodactyls (even-toed ungulates). Each entry lists the relevant literature and includes some older papers that may not be found on PubMed or by other search engines.

White-faced saki Wikimedia Commons.

For a few species the entries on this web site constitute the only information available. As an example, one family of Neotropical primates is represented by the white-faced saki (Pithecia pithecia). I know of no other information on placentation in this family (Pitheciidae).

Uterine natural killer cells

Large granular lymphocytes surround a uterine artery in the four-toed

elephant shrew. Courtesy of Dr. Dominic Oduor-Okelo.

Large granular lymphocytes occur in the uterus of many mammals. Often they are associated with uterine blood vessels. In rodents and primates they are known to be natural killer (NK) cells.

Early in human pregnancy, around 70% of the immune cells in the uterus are uterine NK cells. Their phenotype is different from that of circulating NK cells. Their functions include remodelling spiral arteries and regulating trophoblast invasion. Uterine NK cells also occur in the mouse and other muroid rodents, where they are essential for widening the arteries: this fails to occur in transgenic strains that lack NK cells. 

Natural killer cell receptors

NK cells are important in the recognition of self and non-self. Surveillance of the highly polymorphic major histocompatibility complex (MHC) class I antigens (in humans HLA for human leukocyte antigens) involves an equally polymorphic system of killer immunoglobulin-like receptors (KIRs). KIRs seem to be involved in regulating the depth of trophoblast invasion. Peter Parham and Ashley Moffett have just published a review of the topic that can be highly recommended. If you are interested and motivated yet—like me—struggle to understand the interplay of HLAs and KIRs, this is the place to go.

KIR genes are part of the leukocyte receptor complex on chromosome 19. A couple of additional receptors, NKG2D and CD94 (or NKG2A) belong to the killer C-lectin like receptors coded by genes in the natural killer complex on chromosome 12.

Now for what elsewhere has been called “an inconvenient truth.” The mouse does not have an expanded family of KIRs. Indeed, it has only one functional KIR gene and that has translocated to the X chromosome. Muroid rodents instead have chosen to expand a family of lectin-like receptors named Ly49. Apart from being structurally different from KIRs, these receptors bind to a different site on the MHC molecule.

Mismatching of HLA-C variants and KIR haplotypes is associated with increased incidences of recurrent abortion, fetal growth restriction and preeclampsia. This system cannot be explored in the mouse.

The first placental mammal

From O'Leary et al. Science 2013; 339: 662-7

Reprinted with permission from AAAS

When did placental mammals diversify to found the known orders? The old story used to be that an asteroid impact killed all the non-avian dinosaurs, but mammals survived and diversified in the post-impact Palaeogene. More recent molecular studies have challenged this. They place the origin of some orders or most in the Cretaceous. However, a paper in today’s issue of Science asserts the previous version was right all along.

First a bit of terminology: Notwithstanding that every marsupial has a placenta, “placental mammals” refers to the clade comprising all living non-marsupial (and non-monotreme) mammals and their common ancestor, i.e. crown placentals or Placentalia. Some of us use Eutheria to avoid upsetting our Australian colleagues. In the supplementary material from today’s report I am taken to task for such laxity!

What did the revisionists, Maureen O’Leary and colleagues, actually do? Quite a bit. For 86 living and (as far as possible) 40 fossil species they scored 4541 phenomic characters including, importantly, some that relate to reproduction, fetal membranes and placentation. This gave them sufficient information to combine with molecular data without the former being swamped by the latter.

The most important result was that none of the living orders arose before the Cretaceous-Palaeogene boundary – this marks a series of catastrophic events including an asteroid impact. What did survive was a hypothetical placental ancestor. The creature shown above is their best estimate of how that ancestor might have looked.

They also tell us how it reproduced. Its single young was born hairless with eyes closed. The uterus was bicornuate and the chorioallantoic placenta was haemochorial. In the appendices they add that implantation was superficial, amnion formation was by cavitation (a surprise that one), there was a temporary yolk sac placenta and a small allantoic vesicle (written vessel but we get the point).

The new chronology is attractive not least because it is better in agreement with the fossil record. It is no longer necessary to postulate long ghost lineages where an order existed for millions of years without leaving any fossils. However, as the accompanying comment implies, this is unlikely to mark the end of the debate.

Hamilton, Boyd and Mossman

Human implantation site (lacunar stage) Carnegie No. 7700

Little is known about early human development. In vitro fertilization is widespread and with it has come a wealth of information on development to the blastocyst stage. The next step is implantation of the embryo in the uterus and the first stages of placentation. This is not feasible to study for ethical considerations.

  

Therefore our knowledge is derived largely from two data sets collected many years ago. The first was a carefully dated series of embryos obtained at hysterectomy by Arthur T. Hertig and John Rock. Their material was studied at the Department of Embryology of the Carnegie Institute of Washington and now forms part of the Carnegie Collection. A tutorial based on this collection by Allen C. Enders is now available on the web. One of his micrographs appears above.

The second set of data came from from the collections of J. Dixon Boyd at Cambridge and William J. Hamilton of the Charing Cross Hospital. The Boyd Collection is now at the Centre for Trophoblast Research in Cambridge and will shortly be reunited with Hamilton’s Collection.

Hamilton and Boyd wrote an amazing book The Human Placenta (Heffer, Cambridge, 1970). Copies rarely come on the market and command steep prices. Mossman’s book on fetal membranes, mentioned in a previous post, has also gone out of print. So here is a tip. Together, Hamilton, Boyd and Mossman wrote a textbook on Human Embryology that went through four editions. It contains many iconic images of early placentation. At the back is a comprehensive appendix on mammalian fetal membranes by Mossman in full colour. Plenty of second hand copies can be found and most are reasonably priced.

Dixon Boyd and Hamilton each raised four sons. Richard and Robert Boyd recently joined forces to chronicle the contribution of Cambridge academics to fetal and placental research from the 1930’s to the 1960’s. This account makes fascinating reading.

Endogenous retroviruses and the placenta

Turnover of syncytiotrophoblast in a human placental villus

Syncytiotrophoblast lacks inner cell boundaries but contains multiple nuclei. One place where it occurs is the surface of human placental villi. Syncytiotrophoblast has a high turnover rate and is continually replenished from the cytotrophoblast. This requires fusion of cytotrophoblasts with the existing syncytiotrophoblast, a process promoted by syncytins.

Syncytins are genes of retroviral origin that have been incorporated into the genome and are expressed in the placenta where they promote cell fusion to form the multinucleated syncytiotrophoblast. As recently reviewed in Placenta, syncytins have been discovered in primates, muroid rodents (such as the mouse), rabbits and carnivores. The search continues.

A retrovirus is an RNA virus that replicates in a host cell. It uses reverse transcriptase to produce DNA from its own RNA genome. The DNA is then incorporated into the genome of the host cell. This DNA can be transcribed to RNA and translated to viral proteins. Perhaps the best known example is human immunodeficiency virus 1 (HIV-1). Most retroviruses infect somatic cells, but occasionally infection of germ line cells occurs, resulting in creation of an endogenous retrovirus (ERV). The genome of mammals is littered with ERVs. Most have been inactivated in the course of evolution. A few, however, are not only expressed but show evidence of purifying selection.

The envelope (env) genes of retroviruses function to promote fusion of the viral membrane with the plasma membrane of a host cell. Syncytins are derived from env genes and are expressed in the placenta, where they promote fusion of cytotrophoblasts with the syncytiotrophoblast. Thus far six syncytin genes have been discovered including two in the mouse and two in higher primates. These genes are not orthologous so each represents an independent capture from a retrovirus. Yet another example of convergent evolution!

There is more. The envelope protein of retroviruses is immunosuppressive and endogenous env genes may contribute to immune tolerance by the mother of the fetal semi-allograft.

Sweepstakes distribution

The African-Malagasy sweepstakes envisaged by G. G. Simpson

Madagascar has been separated from mainland Africa by 300 miles of ocean since before the Age of Mammals. Any mammal found there today had to cross the ocean and the most likely scenario is that its ancestor got there on a raft of vegetation. The probability of a pair of animals, or perhaps a pregnant female, making that journey and surviving to found a dynasty is extremely low. It is like drawing the winning ticket in a lottery or sweepstakes. But George Gaylord Simpson grasped that over the span of geological time even the improbable becomes possible.

Verreaux’s sifaka (Propithecus verreauxi) Wikimedia Commons

A primate made this journey in the Palaeocene some 57 million years ago (mya). It was the ancestor of the more than 50 species of lemur living today. Tenrecs did not arrive until the Oligocene (29 mya) but they too evolved to fill all sorts of available niches. One would think this rapid radiation would be reflected in differences in placentation. In fact lemurs have a placental type not too different from other strepsirrhine primates – the lorises and bush babies of Asia and Africa. In both groups the placenta is diffuse, with interdigitating villi and an epitheliochorial barrier. Malagasy tenrecs all have the same type of placenta: discoid, labyrinthine and haemochorial. 

The capybara, a caviomorph rodent, Wikimedia Commons

Sweepstakes distribution also occurred from Africa to South America across the South Atlantic. Recently we looked at two groups of mammals that used this route. Caviomorph rodents arrived in the Middle Eocene (41 mya) and the ancestors of New World monkeys by the Late Oligocene (29 mya). Again we found that their subsequent radiation was not accompanied by major changes in placentation. Indeed the placenta of caviomorphs has the same distinctive features as in the related phiomorphs from Africa.

Perhaps the placenta is not quite as plastic as we are inclined to think.

And the hippopotamus in Simpson’s sketch – which predates the one in Dreamworks' Madagascar by 65 years? There are hippopotami in the fossil record of Madagascar, but they left no trace of their placentas.