Thursday, January 13, 2011

Just another pretty face

Today's paper is really neat. The group made a mouse mutant that had a craniofacial defect (read: messed-up looking face) and traced it to a molecular/genetic pathway.  And all in 4 jam-packed figures and tons of experiments.  I've got a soft spot in my heart for the craniofacial researchers, so this paper jumped out at me.

The title is a mouthful: The E3 ubiquitin ligase Wwp2 regulates craniofacial development through mono-ubiquitylation of Goosecoid. The 30 second summary: mice lacking the gene Wwp2 have odd faces and jawsf.  Wwp2 gene makes a protein that modifies, and activates another protein, Gsc (Goosecoid) which binds to and activates yet another gene, Sox6, which is a key regulator of normal head and face development.


The Data:
Using a gene trap method of deleting genes in mice, the Wwp2 gene was deleted.  The knock-out mice were born, but had defects in their jaws and face: they had stubby noses and domed heads.  The way they made the mice they could trace where the Wwp2 gene should be expressed by using a blue dye. They traced where the normal expression of Wwp2 was using the dye. They used a gene already known to be involved in craniofacial development, Sox9 as a control, and showed Wwp2 was necessary for normal development.  They looked further into Sox9 and found that when it is overexpressed, Wwp2 expression increases too.  Could they be linked?  They did direct DNA binding assays to show Sox9 does, indeed, bind to and controls the Wwp2 gene.

Great.  Moving on, the Wwp2 gene makes a protein called a ubiquitin ligase.   Wwp2 protein then acts on other proteins to decorate them with these teeny itty bitty modifications called ubiquitin.  Often, this targets the protein for destruction in the proteosome, but not always.  If a long chain of ubiquitin is added, the protein is destroyed; if only one little ubiquitin is added onto the target protein it can change the protein's behavior: increase or decrease it's activity, change where it likes to hang out in the cell, etc.  The first question is what protein does Wwp2 modify.  Second question is what does that modification mean for a cell, particularly one that is trying to become a jaw or nose.

They had a few hints as to the first question.  Based on the protein sequence recognized by Wwp2, they came up with four candidates and quickly showed by direct protein-protein binding assays and ubiquitination detecting assays that Wwp2, indeed, added ubiquitin to the Goosecoid protein (Gsc).  The second question, what does Gsc getting a ubiquitin from Wwp2 do to a cell, took a bit more work.  They looked at if Gsc got sent to the trash can after Wwp2 ubiquitinated it. Nope, Gsc levels stayed the same. It so happens that Gsc is a transcription factor: a protein that acts at the DNA level to change expression of other genes.  They looked at how many ubiquitin molecules Wwp2 added to Gsc, and only one was added, suggesting the ubiquitination of Gsc by Wwp2 changes Gsc's activity on genes at the DNA level.  So begins what probably was a stab in the dark that got lucky.  They did gene expression analysis to determine if the expression of any genes potentially involved with craniofacial development changed when Gsc was modified by ubiquitination.

And in comes Sox6.  Sound familiar?  Well, sort of.  In the same family as Sox9, Sox6 is yet another transcription factor that is involved in cartilage biology... just the sort of thing that could direct normal growth of bone, and also just what, when disrupted, could lead to abnormalities.

The result:
OK, so I skipped some of the experiments.  But really, they were very thorough.  And this post is already getting long. A cool little model from this paper is that the first gene they looked at, Wwp2, is turned on by Sox9, goes on to influence another protein, Gsc, which act on another gene, Sox6 to guide normal head and face development in mice.  Talk about a cascade effect.

There's always more work to be done.  I'd like to see the phenotype of the craniofacial defects be echoed in a Sox6 mutant. This would shed more light if this is the only body region affected by this particular signal cascade.  I've no doubt there's lots of literature on Sox9, but I'd like to see what the instigating signal to the little cells are that kicks off this whole signal sequence.  I mean, not every cell ends up being a part of your nose or cheek, so what sets these cells off on this fate.  And then, linking to humans, are defects in this pathway seen in humans bearing craniofacial anomalies?  And don't you love saying cranifacial anomaly?  I do.

See you next week!

Friday, January 7, 2011

Be careful where you leave your teeth.

Today we're talking about a tidy little paper published a few weeks ago in PNAS.  Genetic evidence for patrilocal mating behavior among Neandertal groups.   No I didn't know what patrilocal meant, but Wikipedia defines it as when a married couple lives near the husband's family.   The whole paper is open access, so you can read the entire thing if you like.  The authors succinctly describe a group of Neandertal remains found in Spain that consist of closely related males with not so closely related females and a few of their offspring.  The authors conclude lots about Neandertal culture: group size hovers at about 9-12 individuals, brothers stick together and wives come from outside, and spacing of children is about 3 years.  How did they decide all of this?  By sequencing mitochondrial DNA.

The Data
A family of 6 adults, 3 adolescents, 2 juveniles and 1 infant were found at the El Sidron site in Spain.  The authors determined them to be a family group of Neandertals based on the excavation and grouping.  They sequenced, and I'm totally blown away by this, the DNA from the teeth of the skeletons.  These bones are about 49,000 years old, people!  They looked for the Y chromosome to verify and/or determine gender and they also sequenced the mitochondrial DNA (mtDNA).  Not all the specimens had a Y chromosome, either  because they were female or because the DNA amplification didn't work.  But all of them yielded positive results for the mtDNA. They examined a couple different regions of mtDNA, where subtle variations, or polymorphisms, are known to exist (see SNPs).   They found evidence of 3 distinct mtDNA lineages, suggesting the 12 individuals came from 3 separate backgrounds.  The adult males came from the first, as did one adult female.  The other 2 adult females came from separate, distinct lineages.  The adolescents, juveniles and infant were mixed, suggesting they were the offspring of the three females.

What does it mean? 
The group drew several clean conclusions from this small data set.  First, the group size.  12 individuals seemed to be the maximum size for a successful group of this heritage, and this group was reaching the maxima.  There were 3 adolescents in the group, so that might speak to when teenagers were considered grown-up enough to venture out on their own.  The males of the group appeared to be related.  The females were outsiders.  So brothers/male cousins stick with eachother, sisters/female cousins go forth to find their mates.  Finally spacing and family planning.  Because each child bears the mom's mtDNA, it was easily possible to determine which kid came from which mom.  And thus, how often a mom had a baby.  And it appears moms had babies every 3 years or so.  This is totally consistent with other evidence (referenced here) showing the normal pattern of nursing and weaning leads moms to regain fertility about 3 years after each birth.

Kind of a lot to learn from one experiment, no?  I'm blown away by a couple things.  First, there's enough intact DNA in these really, really, reallllllly old bones to get good sequence off of it.  Trust me, I've failed at sequencing gobs of purified fresh DNA, so this little technical feat is not lost on me.  This was actually a huge story last year: the entire genome of Neandertals was sequenced (referenced here - an open access paper so have a gander at it).  Pretty friggin cool.  Second, there exists a mtDB - a database of human mitochondrial DNA sequences.  You can look up any of the 1865 human mtDNA sequences completed and banked there, the polymorphic sites.  It might look like a bunch of gobbledy gook, but there's enough information there for anyone with a buccal swab, a PCR machine, and a couple of primers to determine where they fit in the whole mtDNA spectrum. If anyone wants to pitch in, I'm willing to start my own service.

Tuesday, January 4, 2011

You are your mother's mother's mother.

After a halting start, this blog seems to have fallen by the wayside.  So my resolution for 2011 is to do one paper a week.  We'll see how that goes, but in the meantime let's talk about a little round piece of extranuclear genetic material in all of our cells: mitochondrial DNA, or mtDNA.

The mitochondria are the little energy factories within our cells, taking sugar (well, sugar derivitaves really) mixing in oxygen, and making lots of ATP, carbon dioxide, and water.  It's hypothesized they started as a symbiotic bacteria in an ancestor cell looooong ago which got cozy and took up permanent residence.  The mitochondria precursor had it's own circular DNA which liked its new home so much, it  kicked out lots of the genes it no longer needed.  Nature is lazy, so duplicate genes in the chromosomes were deleted from the mtDNA.  Mitochondrial DNA contains some genes essential for making copies of itself (replication), and for a few things that only happen in the mitochondria, just 37 in total.  It's a lean, mean replicatin' machine.

There's lots about mtDNA we can talk about, but let's focus on inheritance.  You inherited 99.999999999% of  your mtDNA from your mother.  And she inherited hers from her mother.  And so on. So much so, a really neat book on the subject was written that traces all of humanity back to seven women, based on mtDNA studies.  Why do we all carry our mother's mtDNA?  When you were just two halves of a cell, the big pillowy egg cell from your mom carried all the good gooey innards whereas the teeny sperm cell from your dad only carried tightly packaged chromosomal DNA - no room for mitochondria, no room for mtDNA.

A big question is why trace lineages from mtDNA?  It could also be done through X or Y chromosomes.   However, the Y chromosome is not a good candidate because, first, only half the population will have a Y chromosome.  Second, the Y chromosome is highly palindromic, meaning lots of the sequence mirrors itself.  Thus, it's a technical challenge to be able to accurately sequence most regions. Perhaps this warrants further exploration another day. The X chromosome would be a great candidate, but it's big.  Much, much bigger than most chromosomes.  Containing about 1098 genes and containing 150,396,262 base pairs (give or take), the X-chromosome is quite a beast.  Getting all researchers to agree to track subtle changes in small regions of such a large hunk of genetic material would be fraught with lots of ego battles, for sure. Not that there's any ego in any of this, right?

So, why mtDNA to trace lineages?  The answer is that humans, like nature, are pretty lazy too.  MtDNA is pretty small - only 16,500 base paris.  And there are just a few regions that can be mutated with no consequence to the genes.  Since these areas mutate at a predictable frequency, about 1 in every 33 generations, (much higher than in chromosomal DNA) these regions can serve as pretty neat molecular clocks.  The mutations give clues to how long it's been since these mutations arose, and when one person's mtDNA is compared another's, or to ancestors, researchers can make guesses as to who is related to who.

Neat?  You bet.  Practical?  Of course.  Stay tuned!