How different can ‘identical’ twins be?

I have seen a few news articles recently exclaiming cases of identical (monozygotic) twins with very different appearances. Although this initially seems odd and inexplicable (and apparently news-worthy), when we take a look at the underlying genetics there are actually some relatively simple explanations that can account for these differences and it is not really that rare. Experts call it ‘discordance’ and it probably affects the majority of ‘identical twins to some extent. In this post we are going to take a quick look at how monozygotic twins that, by definition, begin development with identical DNA can end up so different.


What are ‘identical’ twins?

 It is important to differentiate the type of twinning we are discussing here as there are a few different kinds. Each twin type has different developmental origins, which is genetically important. The most common types of twins are ‘fraternal twins’ and ‘monozygotic twins’ but there are other types too that arise in different ways.


  • Fraternal Twins – These twins are commonly known as ‘non-identical’ twins and in terms of their genetic make-up are essentially just siblings. Like siblings, fraternal twins share approximately 50% of their DNA. Fraternal twins are no more likely to share traits or look alike than any other siblings. For fraternal twins to occur the mother must release 2 eggs in the same ovulation cycle. Two individual sperm must then fertilize these separate eggs resulting in the formation of 2 zygotes. For this reason fraternal twins are often referred to as ‘dizygotic twins’.


  • Monozygotic Twins – Commonly known as ‘identical’ twins this is the type of twinning we are mainly interested in for the purposes of this article. Monozygotic twins arise when 1 sperm fertilizes 1 egg creating a single zygote. This single zygote then divides and develops a bit to become an embryo, and then splits in two (for some unknown reason). The end result of this is 2 genetically identical embryos sharing a womb! The length of time between the formation of the single zygote and the splitting into 2 separate embryos will determine whether they share things like the placenta, chorionic and amniotic sac – the later the split the more they share. In fact if the split happens later than about 12 days after fertilisation then conjoined twins will typically result.


  • Other types of twins – less commonly, twins may be neither truly fraternal nor monozygotic. It is thought that so called ‘half identical’ twins (that share 75% of their DNA) can occur when the egg divides (forming what’s called a polar body) just before fertilization, and then a separate sperm fertilises each half of this divided egg. Here there is a pretty unique genetic situation where the resulting fetus will essentially have identical genetic information from the mother, but different genetic information from the father (hence sharing 75% of their DNA). These kids are in between being siblings and identical twins. Another rare form of twining can happen when different eggs get fertilised by sperm from 2 different men leading to dizygotic twins with different fathers. Twins arising in this way would basically be half-siblings and would share approximately 25% of their DNA.


So are monozygotic twins actually genetically identical???

So knowing the developmental origins of monozygotic twins they must all be genetically identical – yes? Well the flawless logic that brings us to arrive at that conclusion is ruined by the (occasionally) flawed biological mechanism by which cells divide and DNA replicates. Although they do indeed start out genetically identical, from the point that the zygote or embryo divides to create two potential humans you can no longer be sure that the DNA contained within the two populations of cells / potential humans is going to remain the same. Although it is not particularly likely that huge changes in the DNA will occur, when we’re talking about the genetic code it’s not always size that matters! In the words of Kirstie Allsop and Phil Spencer it’s ‘location, location, location’ that is really important. The change of a single base (letter in the DNA code) at a vital location can sometimes have pretty significant consequences.

For instance, with the right spontaneous changes in the DNA sequence in specific locations it is entirely possible (although incredibly unlikely) that one twin may have Albinism, whilst the other does not – or any other genetic condition or trait! In fact whilst we’re on unlikely hypotheticals it is probably theoretically possible that with a huge, verging on impossible, number of spontaneous alterations to your genetic code you could essentially be born a Chimpanzee! Someone smarter than me could probably calculate the likelihood of that happening based on mutation frequencies etc. Anyway, I digress.


How does the DNA get altered?

There are a number of alterations that DNA can undergo that can essentially change what the DNA codes for. Most of these changes are largely random events. Although some changes are a lot more rare than others, and a lot of changes will not have much of an effect, it is probably a good bet that in the course development some of these changes will occur (double that chance because we are talking about twins). Changes that alter what the DNA codes for can loosely be grouped into two categories:

  • Genetic Changes – these are alterations that affect the genetic code (the ‘letters’ of DNA) itself. These can be anything from changes in the bases (letters) to massive deletions or additions of sections of the DNA code.
  • Epigenetic Changes – these are changes where the actual genetic code remains the same, but the systems that control whether a gene is ‘on’ or ‘off’ (i.e. whether the gene is expressed or not) are altered. For example, in an individual that originally didn’t produce protein X, an epigenetic change could switch production of protein X on).


What changes to the DNA sequence can occur?

 There are a few different types of changes that actually change the overall genetic sequence the cells have to work with. The largest of these changes in terms of sheer volume of genetic code altered (gained or lost) are changes in chromosome number (known as aneuploidies).


Chromosome Problems:

When the cells divide they are supposed to replicate the chromosome and then split them equally into the two new cells. If this process messes up (via something called ‘non-disjunction’) then some of the cells will have the wrong number of chromosomes. The individual that is eventually born will be what is termed ‘mosaic’ for that mutation (a term used when not all the cells in the body are affected by a genetic variation). If this happens early after the zygote splits then a lot of cells in the individual will have a lot of cells with the mutation. If the error in cell division happens later on in development then it will effect fewer cells and (usually) not be as much as a problem.

Let’s consider an extreme example of this kind of error that could theoretically produce a set of monozygotic twins of different sexes. For this example we will start of with an abnormal zygote that has (due to errors in the development of the sex cells) ended up with two X chromosomes, and in addition has one Y chromosome. This cell has one more Y chromosome than a normal (XX) female, and one more X chromosome than a normal (XY) male. Under normal circumstances (assuming this XXY zygote didn’t get spontaneously aborted by the body as a surprisingly large number of embryos do) this zygote would develop into a male with Klinefelter’s Syndrome (a relatively mild disorder of sexual differentiation). Now let’s consider what might happen if this XXY zygote was to, for whatever unknown reason, split into two as might happen in the case of twins. It would split into two XXY cells right? Well under normal circumstances yes, but what if the sex chromosome numbers get messed up and don’t split equally like we just discussed? If this happens we could then feasibly end up with one (potential) twin that is XX and one that is XY. Through a couple of relatively unlikely errors in divisions we have ended up with a set of monozygotic (aka ‘identical’) twins of different sexes! Pretty good example that demonstrates that not all monozygotic twins will end up genetically the same!


Small, but important changes:

Smaller, but just as important, are the changes that can randomly occur affecting specific bits of the DNA code. These changes have the potential to alter the function of a certain gene. This kind of mutation can occur when the copying process used for DNA gets it wrong. Although the biological processes used to copy the DNA in our cells are mind-bogglingly accurate, when you are copying billions upon billions of bases even an error rate of less than 1 in 100,000 can produce a fair few errors. Most of the time the errors are corrected, but not always. If the error isn’t corrected we can end up with an altered DNA sequence. Depending on what and where the change is the alteration can result in anything from being benign (having no discernable negative effect whatsoever), to causing a serious disease! It may be the case in some developing twins that such an alteration randomly occurs, and this just so happens to change the expression of a gene that has some influence on appearance! Hey presto! The supposedly identical twins are now ‘programmed’ to look different before they have even become a fetus!

There are quite a few different types of mutations that can occur that range from a single base being ‘substituted’ for a different base, to sections of the code being inserted, duplicated, or deleted. Google can explain these if you wish to know more as I am already rambling.


What about that epi-genetics thing you mentioned?

Well remembered! A big part of what makes everyone different from each other is not just contained within the sequence of the genetic codes itself, but also in how that code is modified. Think of epigenetic modifications as the annotations to the text – they basically tell the biological machinery what bits of the code to ‘read’ and use and what to quietly ignore. Well, in actuality they are more like big methyl groups that physically stop the transcriptive machinery doing it’s job on some bit’s of DNA – but I am sticking with my annotation analogy! Anyway, as you might have guessed studies have also show that monozygotic twins can have epi-genetic differences. Like sequence changes these differences can also cause a massive array of potential differences to occur between the twins, including those that affect their appearance. So this is one more mechanism that can make ‘identical’ twins look different!


What about other genetics-y processes? And what about the environment?!?

This is not a comprehensive summary of all the processes that could ever potentially cause differences in the genetics of monozygotic twins from the point the zygote splits in two to the point they are born. Of course, after the twins are born there are a whole shedload of environmental factors to consider too – exposure to the real world really confounds things!


Is that it? Can I stop reading now?

Yes. I am done typing. Go – be free! Hope that was a little bit interesting, or at least distracted you from life for 5 minutes! Please comment, share, send me money etc.




– Non-identical monozygotic twins, intermediate twin types, zygosity testing, and the non-random nature of monozygotic twinning: A review. Geoffrey Machin