Nature article: Why autoimmunity is most common in women (2021)

https://www.nature.com/articles/d41586-021-01836-9

 

I remember Radiolab had a fascinating episode on a possible evolutionary race among the placenta and women's immune system, and how this may have led to increased autoimmunity as women have fewer (or non) pregnancies.

https://www.youtube.com/watch?v=C3CfkTzHcJY

From the description of the episode:

Three scientists dive down into our genome to tackle one question: why do women tend to get autoimmune disease more than men? Multiple sclerosis, lupus, rheumatoid arthritis, even psoriasis — these are diseases in which the body begins to attack itself, and they all have one thing in common: they affect women more than men. Most autoimmune disorders do. And not just by a little bit, often by a lot; in some cases, as much as sixteen times more. But why? On today’s episode, we talk to scientists trying to answer that question. We go back 100 million years, to when our placenta first evolved and consider how it might have shaped our immune system. We dive deep into the genome, to stare at one of the most famous chromosomes: the X. And we also try to unravel a mystery — why is it that for some females, autoimmune disorders seemingly disappear during pregnancy?

ETA: Oh, it's the same person Rhonda Voskul in both the podcast and the Nature article

 

From the second quote in post #1 :

Surely this can't be a new theory? It's one of the most obvious differences between men and women!

 

It depends how the uterus is conceived - as a self contained bucket in which an embryo develops, a female body is just a male body + expandable baby bucket. It's taken a lot of time to actually get away from that perspective, especially toward the understanding that the primate uterus is at the forefront of an evolutionary wave: The immune-modulating pregnancy-specific glycoproteins evolve rapidly and their presence correlates with hemochorial placentation in primates:


"Abstract
Background

Pregnancy-specific glycoprotein (PSG) genes belong to the carcinoembryonic antigen (CEA) gene family, within the immunoglobulin gene superfamily. In humans, 10 PSG genes encode closely related secreted glycoproteins. They are exclusively expressed in fetal syncytiotrophoblast cells and represent the most abundant fetal proteins in the maternal blood. In recent years, a role in modulation of the maternal immune system possibly to avoid rejection of the semiallogeneic fetus and to facilitate access of trophoblast cells to maternal resources via the blood system has been suggested. Alternatively, they could serve as soluble pathogen decoy receptors like other members of the CEA family. Despite their clearly different domain organization, similar functional properties have also been observed for murine and bat PSG. As these species share a hemochorial type of placentation and a seemingly convergent formation of PSG genes during evolution, we hypothesized that hemochorial placentae support the evolution of PSG gene families.

Results
To strengthen this hypothesis, we have analyzed PSG genes in 57 primate species which exhibit hemochorial or epitheliochorial placentation. In nearly all analyzed apes some 10 PSG genes each could be retrieved from genomic databases, while 6 to 24 PSG genes were found in Old World monkey genomes. Surprisingly, only 1 to 7 PSG genes could be identified in New World monkeys. Interestingly, no PSG genes were found in more distantly related primates with epitheliochorial placentae like lemurs and lorises. The exons encoding the putative receptor-binding domains exhibit strong selection for diversification in most primate PSG as revealed by rapid loss of orthologous relationship during evolution and high ratios of nonsynonymous and synonymous mutations.

Conclusion
The distribution of trophoblast-specific PSGs in primates and their pattern of selection supports the hypothesis that PSG are still evolving to optimize fetal-maternal or putative pathogen interactions in mammals with intimate contact of fetal cells with the immune system of the mother like in hemochorial placentation."

 

This is a really interesting article. There is one item I'm not so sure about:

"However, the female system evolved over hundreds of millennia during which people were pregnant for much of their reproductive years. Because people have fewer children now than their ancestors typically did, their bodies do not interact with a placenta as frequently."

The second sentence may be true but the idea that female humans and female human ancestor species were close to being perpetually pregnant seems improbable. Lactation, body weight, nutrition, climate, social stress and numerous other factors can inhibit pregnancy, I don't think we have any sound evidence of female fertility rates in pre history or in earlier editions of Homo, at the very least we must expect substantial variation in fertility between populations measured over millennia. That must include large expanses of time with low pregnancy frequency in some populations.

 

I raised an eyebrow at this part. Homo sapiens as we know them seem to have appeared about 100,000 years ago and only became sedentary in approx. 12,000 BC. Until then, pregnancy would have significantly hindered mobility and thus survival so I have a hard time conceiving how women could have had more children then. However, I accept that 14 millennia (from the beginning of sedentism to nowadays) is a long enough timeframe for genetic changes to have happened.

 

Me being a pedant - Oldest Known Homo Sapiens Fossils Found

I think it's not so much pregnancy itself being the inhibition, rather the care of multiple infants, or otherwise the cost of multiple failed pregnancies. There really isn't the data to deal with this with any certainty but life expectancy in pre history is generally put at an average of 40 years for those attaining adulthood which leaves no space for a grandmother class of post menopausal carers, plus ovulation is heavily diet dependent. Pre historic diets probably only saw a woman becoming child bearing in her late teens, and the idea that a hunter gatherer woman would have 10+ pregnancies over the course of 20 years seems unlikely.

The implications would be that there were many failed pregnancies, very high infant mortality and or very high maternal mortality. All of which in themselves have significant implications for human evolution - and don't seem to mirror our nearest living Great Ape relatives. Analysis of modern hunter gatherer peoples may be unhelpful because they have often been driven to the most challenging environments by agriculturalists or industrialisation - however: Hunter-gatherer energetics and fertility: A reassessment of the !Kung San

"This paper examines the question of why the total fertility rate of the !Kung San hunter-gatherers of the Northern Kalahari desert is as low as 4.69 births. When the intermediate variables involved are examined through the employment of a reproductive equation, it becomes clear that low fecundity is a major issue. Arguments offered previously to explain the low fertility of !Kung women, depending on factors such as nutrition, health status, and lactational practices are insufficient. Drawing upon recent data from sports medicine and endocrinology, I suggest that the pattern of San female energetics in their gathering and subsistence routine has a direct effect upon their fecundity. Such a correlation between activity patterns, endocrine function, and reproductive capacity may also be important for understanding the fertility of other mobile hunter-gatherer groups."

Doesn't of course speak to lifetime number of pregnancies. But still it doesn't suggest that tough lives leave room for extended states of being pregnant.

 

This looks like a mash-mash of very old ideas by young immunologists who have not read much of the literature or thought very hard.

Two things have always seemed relevant to me.

Firstly all the stuff about the foetus being a graft or parasite that might get rejected is garbage because of the two separate circulations. If you consider the immune response as belonging to a specific recirculation domain, which it does, the whole rejection business becomes nonsense. There are some fancy mechanisms to keep the circulations separate in the placenta but that is really all you need as far as I can see. There is a problem with antibody to things like blood group antigens but that is not normally a significant part of a rejection process.

Secondly, antibodies are mostly useful for blocking a second clinical illness from the same microbe. They come too late to save us from the first illness. For men they may not really be necessary - just useful in terms of less time off work hunting and gathering. But newborn babies are uniquely vulnerable because their immune system has not set up a 'quick and dirty' repertoire for starting off specific responses - no T cell selection and little or no microbe-favouring IgM or wide specificity IgG. So for women the crucial role of antibody is as a protection given in the womb and in the milk to the baby. So it makes sense for women to have a souped up antibody system.

The odd thing is that apart from autoantibodies I don't think anyone has been able to show what this souping-up would look like. The total antibody levels are much the same in men and women. Maybe women make high affinity antibody more effectively? Or maybe they do the opposite and focus on broad spectrum antibody specificity? Studying mice seems pointless since people could perfectly well study people, but very little seems to have come out of it in the last sixty years.

 

I wonder why research doesn't focus more on why and how autoimmune diseases that affect both sexes develop in XY men. You'd remove a whole plethora of assumptions and potentially some confounding factors at a stroke.

It wouldn't necessarily make it any easier to get answers, but you'd think it'd be a less cluttered space to work in and would help keep people honest. Specially as we don't seem to have got far by focusing so many questions on why they're more common in XX women.