Thursday, May 2, 2013

Accessible Research: Gene survival and death on the human Y chromosome

I opened my mailbox last week, and what should appear before my wondering eyes, but the new issue of Molecular Biology and Evolution. Inside is our recent (and open access!!) paper: Gene survival and evolution on the human Y chromosome. Here's my summary of our work. (Editorial Note: it is so, so much easier to distill down research articles that I haven't spent years of my life on.)

In humans, genetic females have two X chromosomes and genetic males have one X chromosome and one Y chromosome:



You might have noticed from the cartoon above that the human Y is much smaller than the human X. But, it wasn't always this way. Ancestrally, the human X and Y were the same size, and had the same genes. Over time, however, the Y has shrunk, but both the X and Y have also gained some genes. To better understand how the X and Y became so different, and how the evolution of the two sex chromosomes are correlated, we asked three main questions:

1. What has been lost from the Y?
To know which genes were lost, we first had to identify which genes were on the ancestral sex chromosome pair. By comparing the genes on the human X with the genes the X in other species, we identified a set of genes that were likely on the ancestral X chromosome: 600 in total. Then, by searching the Y chromosome for the relics of all of these genes, we identified three classes of sex-linked genes. We should think of each of the 600 ancestral genes as a pair (with one copy on the X, and one on the Y). All of these pairs have a working copy on the human X. Some pairs have a working (functional) copy on the Y, some have a broken copy on the Y (degraded), and some are missing their Y-copy.

Many genes have been lost from the ancestral Y, but a few persist.

So, while some Y-linked genes have survived (I have another paper discussing this), and there have been some unique additions to the Y chromosome, we can see that the Y has lost functional capabilities for 96.83% of the genes that it once shared with the X. Wow!

2. Are there indicators of whether a Y-linked gene will be retained?
We can learn about the evolution of the sex chromosomes by studying differences between classes of sex-linked genes defined above. Specifically we asked, do features of X-linked genes suggest whether their Y-linked partner are retained or lost? In some cases, yes, they do.

First, we found that human X-linked genes with very few changes across mammals were more likely to have a working Y copy. So, if a gene is important enough to survive over long evolutionary time in roughly the same condition across very different species, then it might be very useful to the organism, so it would be important to have that gene in a working form in both males and females in the same species (human).

Second, we looked at expression. Genes can sometimes be "on" (which we would call expressed) or "off" (not expressed), but more often they can fall within a range. It's like a light with a dimmer switch. The light can be turned on very brightly, but can also dimmed to a very low level without being "off".  We found that X-linked genes that were highly expressed (bright) were more likely to have a working Y copy. This might mean that, for these genes, the level of "brightness" or expression is important, so that it is highly beneficial for these genes to be working very hard in both females and in males.

3. Does gene loss on the Y affect the evolution the X?
Okay, so some features of the X-linked partner might predict whether it's Y-linked partner will survive, but is there any feedback from the Y back to the X chromosome? Yes!

Let's think back to that first picture: females have two "big" X chromosomes, while males have one "big" X and one "little" Y. And, I've shown you that the Y chromosome has lost (either because of broken copies, or completely lost) almost 97% of the genes that it once shared with the X. This might lead you to believe that there are more genes expressed in females than in males. But, in many mammals, females silence most of the genes on one of their X chromosomes (X-inactivation), to equalize the dosage of genes expressed between males and females.



Although it has been hypothesized, we showed that the pattern of genes subject to silencing in females among the three classes above is consistent with a process whereby silencing evolves in response to gene loss on the Y chromosome. Moreover, this pattern suggests that some amount of time must pass to allow the signal (that the Y-linked partner is no longer working) to reach the X-chromosome before silencing can occur.

The paper is open access, so if you are curious, you can read it here.

 2013 Apr;30(4):781-7. doi: 10.1093/molbev/mss267. Epub 2012 Dec 4.

Gene survival and death on the human Y chromosome.

Source


16 comments:

Holly Dunsworth said...

This is so cool Melissa. Do you know if the same X (paternal or maternal) is always silenced in a female, in every cell? Or does a particular x-inactivation vary (event to event, cell to cell, tissue to tissue, ...)?

mathbionerd said...

We're still working out the story, Holly. But the short (for a scientist) answer is that:
1. In marsupials, it is always the paternal X that is silenced in every cell.
2. In eutherian mammals the X that is inactivated is randomly the maternal or paternal X (with some funny business during embryogenesis).
3. As far as we can tell, the random X inactivation in eutherian mammals happens in all tissues.

Holly Dunsworth said...

Thanks!

So does this mean that females are more diverse, phenotypically as individuals, than males? (Assuming the X genes are significantly impacting biology over a lifetime of both males and females?)

mathbionerd said...

That is a very cool idea. I don't know about comparing male and female expression, but in addition to the random inactivation of either the maternal or paternal X, there's an extra bit about X-inactivation in human mammals that would lead me to suggest, "yes".

That is, although one of the X chromosomes is silenced, there are some genes that escape inactivation. Moreover, looking at the pattern of X-inactivation across cell lines from different women, there are actually three classes of genes on the X:
1. Those that always escape inactivation
2. Those that are always subject to inactivation
3. Some that escape inactivation in some XX individuals, but are subject to inactivation in others

It is this information that we used to learn about the relationship between Y-gene loss and X-inactivation.

Hm, I should write a few posts about X-inactivation.

Holly Dunsworth said...

And... Does this mean that I'm (as a female) more closely realted to my paternal side than my (nonexistent) brother is since he didn't get an X (and, thus, two "choices" of X genes one from mom AND one from dad to express) from our paternal side?

Holly Dunsworth said...

oops. You're too fast. (No complaints!) My latest comment was before reading the one you just posted.

Holly Dunsworth said...

*related* not *realted*

Holly Dunsworth said...

Okay, now I've gone and geeked out further than logic. He's more closely related to dad because of Y (I know that), so my above question needs to include the following subquestion: Are there MORE genes on the X that he's NOT using from paternal side (cuz they never came) than there are on Y that he is using? (Does this make sense... I'm saying .. are females using more on the X from dad than males are using on the Y from dad?) Sorry if I'm bombarding you... this is so awesomely blowing my mind right now.

mathbionerd said...

Holly, it isn't so clear who is more closely related to your dad, and, like with most things, it comes down to how you define "related".

- If you mean with respect to anatomy, and all pathways are functioning normally, then your brother will have more similar anatomy to your father than you will.

- If you are talking about gene expression profile, both the XY individuals will not be polymorphic for X-linked gene expression, so there will be less variability, potentially, between their expression profiles. But, thinking of expression profiles, there are many autosomal genes that are male-biased or female-biased in their expression, so it depends on whether the androgen-receptor pathway has been initiated to determine which way these genes will sway.
- If you mean actually gene content, then... you probably share more overall gene content with your dad than your brother. Both siblings will have equal autosomal content from each parent, and an X from the mother, but you have your dad's gene rich X, while your brother has your dad's gene poor Y.

Holly Dunsworth said...
This comment has been removed by the author.
Holly Dunsworth said...

Thanks Melissa. I clearly did not take a deep breath and calm down and express my question properly. Maybe I'll pull back and compose my thoughts. Got lots to go with thanks to you... Thanks again for all this!

mathbionerd said...

It's my poor writing skills. In all cases, I meant that you will be more related to your dad than your brother *is related to your dad*. :)

Holly Dunsworth said...

Neato! Thanks again Melissa. Will be bothering you much more in the future. :)

Unknown said...

Cool post! I hope to read the paper soon.
A) Is it possible to date (relatively) the time of a gene's death on the Y? Do certain classes of genes get lost first?
B) are there any cases were a dead gene can be "re-born" as a novel gene? It seems like such fertile genetic material and the dying process is possibly slow enough for a gene co-option event to occur.

Thanks,
Noah

Unknown said...

Cool post! I hope to read the paper soon.

Is it possible to date(relatively) the time of gene death? Do certain classes of genes get lost first?

Are there any examples of a gene "re-birth" on the Y? It seems like there is fertile coding sequence (compared to random sequence) in dead genes that could be co-opted for a new function?

Thanks,
Noah

mathbionerd said...

Noah, I'm actually working on a paper to date the timing of the formation of evolutionary strata. Timing gene death can be approximated, but is pretty difficult with the genes on the Y chromosome because most of them "died" (became pseudogenes) so long ago, that the estimates are very fuzzy. Also, the variation in evolutionary rates between the X and Y add an extra layer of complexity (versus on the autosomes).

I don't yet know of a gene on the Y chromosome that lost functionality, and then regained it, but that certainly could happen. But, just like elsewhere in the genome, it is much easier for something to break than it is to evolve something new.