FOXP2 is known as “the language gene.” When it goes wrong, as it famously did for a family in England, it can cause severe deficits in language ability. In the “KE family" in England, for instance, several family members carry a defective mutation of the gene and have trouble with grammar and writing, and with making the right face and mouth movements for normal speech.  There’s a nice write-up on the gene at “Not Exactly Rocket Science” 

Now this is unusual for a gene: the protein it codes for in humans differs from the chimp version at only 2 of its 700+ amino acids, while the human version of FOXP2 is near-universal in our species and in Neanderthals. We all share the exact same FOXP2 gene on both of our chromosomes; no variability, no matter the population tested. Outside of the “KE family,” we all carry two identical copies of that gene and apparently we need both for full function. Our human genes for eye color, for hemoglobin structure, for immune function, all have incredible variability within and among populations. To have such uniformity across all human populations tells us this FOXP2 gene must be doing something really significant.

We differ by two amino acids from ape and monkey FOXP2, and by only 3 amino acids from mice. While people make much of how much of our genome we share with chimps, it’s obvious that the genes that differ do make us very different in some important ways. Imagine if this were THE gene for language, if those two amino acids were what made our lives so very different from theirs.

Here’s why FOXP2 is interesting to a mouse n monkey vet: with modern technology, sooner or later any interesting human gene is going to be inserted into mice to try to decipher what that gene does. Inserted into mice, the human version of FOXP2 has led to various developmental changes in mouse brains, changes in their learning ability, and changes in the calls baby mice make when separated from mom. This work was only published in 2008; expect lots more to follow.

Mouse vets are accustomed to seeing human genes inserted into mice. Transgenic mice date back to the 1970s. We have three main animal welfare concerns with the current technologies. #1 “Making mice” requires surgeries (vasectomies of some male mice; embryo transfer by surgically exposing the uterus of recipient females; by contrast, egg donor females are simply killed for their eggs). #2 It can take hundreds of births to get mice with the right copies of the new genes in the right place, and to get them to breed true generation after generation. Not only does that require killing lots of mice with the wrong genes, it also means getting tissue for genotyping (i.e., looking for the gene or the protein it encodes in each animal). This technology is sure to improve once smaller tissue samples will do the job, but for now, many mouse gene-jockeys  cut off mouse toes, mouse tail tips, or bits of mouse ear tissue to separate mice with the desired genes from those without. Welfare hot spot #3: what are the effects of the gene in the mouse? Do they make him ill? Are they lethal in utero? “Poor doers” who fail to thrive and grow as sucklings? More prone to disease or pain-sensitivity or anxiety or injury?

In the studies-to-date of mice with human FOXP2 genes, all three problems are possible. The authors who first made these mice don’t tell us what anesthetics and painkillers they might have used (or might not have used) --- always a pet peeve of mine. For all that this technology is standardized, there’s still controversy about whether aggressive peri-surgical pain management for embryo-recipient surrogate mothers disrupts gene-jockeying. Our lab has a study underway to investigate this; stay tuned on that. Second: developing these mice certainly requires amputating tissue for DNA analysis and will take generations of test-and-cull to standardize THE FOXP2 mouse. These are my standard concerns about transgenic mouse production.

And the effects of the gene itself? Depends. The authors at the Max Planck Institute first reported on what happens when you put defective human FOXP2 genes – the ones in the “KE family” – into mice, and found various minor health problems. The really interesting mice, the ones I expect to show up soon in labs around the world, are the ones who got the good human FOXP2 genes. (see Enard et al, Cell 2009 137(5): 961-971) So far, so good: some minor changes in structure and behavior. The changed pup vocalizations don’t even seem to disrupt the mouse mother-pup bond.

New studies have reported what happens to human nerve cells in a culture dish if you add the chimp version of FOXP2. This is new info (November 09). FOXP2 turns out to be an important regulator of other genes, telling some to turn on, some to turn off. It affects the activity of over a hundred other genes, and some of these in turn go on to be regulators of genes further downstream.

What’s next for the mice? Since the advent of transgenic technology in mice in the 1970s, scientists the world around have been good at sharing their animals, so expect pairs of FOXP2human mice to show up in labs everywhere. If it continues to look like FOXP2 and a couple of other key regulator genes are integral to language, expect to see quintuple-GM mice, with 5 or more human genes replacing their mouse counterparts. To develop that pedigree – a mouse with that many genes all where you want them – will probably require breeding and culling thousands before we have stock that are ‘humanized’ at all five loci. How those mice will look, act, and squeak remains to be seen.

Now, I’m a monkey vet as well as a mouse vet, so I have to think: wouldn’t it be really interesting to know what would happen if a monkey or a chimp instead of a mouse got the human FOXP2? That hasn’t been reported yet, and I would have no way of knowing if it’s in progress somewhere. Earlier this year, a group in Japan reported on the first truly transgenic primates, some Pygmy Marmosets with a gene that makes Green Fluorescent Protein (and no, no details shared on how surrogate mothers were cared for, or how egg and sperm donors were treated). Derived from jellyfish, this GFP gene is great for labeling the location of any gene it’s linked to. In the Marmosets, it’s there to show which cells got the foreign gene in an active form, and in the reported work, the gene showed up throughout their bodies, including in the eggs and sperm they produced. This all shows that it’s possible, in these fecund short-gestation primates, to stably introduce a foreign gene into their genome.

If it’s possible to make fluorescent green Pygmy Marmosets, it won’t be long (years, not decades) before we can put genes into other primate species --- Rhesus macaques, chimpanzees. And it’s not far off before we can put human FOXP2 “master genes” into these marmosets, macaques or into Great Apes. And maybe whatever other ‘submaster’ genes turn out to be prime suspects for whatever it is – language or more – that separates us from our primate relatives. And then?

For all the human genes that have been incorporated into genetically-modified mice, I’ve never met a mouse who made me stop and say, “Maybe this mouse is getting TOO human.” But those are genes for immune system chemicals, for drug-metabolizing enzymes, and my job as a vet is to assess the animal’s health and welfare, not to focus on the human bits in the animal. Present-day IACUCs are not trained, constituted or practiced at addressing this particular ethical question. Mouse pups with deep squeaks may not raise too much concern. But what do we do if FOXP2 and a couple of associated genes start dramatically increasing mouse intelligence?

If it can be done, sooner or later it will be, for better or worse. I’m fascinated by, and dreading, the studies on language acquisition in ‘humanized’ FOXP2human monkeys and apes. Chimpanzee research, for instance, has been on the decrease for some time, after an initial flurry of chimp studies in the early years of the HIV epidemic, while monkey numbers hold about study over the years. I can see the great temptation to get to chimp studies as quickly as possible.

Folks whose primary business is human ethics will focus on how human a chimp (or monkey, or mouse) has to be to qualify for special protections. It won’t be as simple as counting how many human genes are in the primate, but which genes, and which functions they confer. A human hemoglobin gene in a chimp probably won’t cause alarm. Genes that alter intelligence, language acquisition, even the throat muscles for speech will surely push chimps to that too-human line.

For us animal people, chimps are already too human, or really, too “chimpy” to experiment on without serious qualms. One concern is whether these studies would lead to new increases in primate numbers in labs. It’s hard to maintain primates in good physical and emotional health in labs, though we certainly try. I would not welcome an upsurge in use of laboratory chimps. My concern is not restricted to the FOX2P gene. If anything, I expect FOXP2human chimps would be nurtured to bring out the most in their language potential. Rather, as with transgenic mice, my animal welfare concerns will depend heavily on what disease-causing genes find their way into transgenic apes and monkeys. 

A chimp with the right T-cells to more closely model human HIV infection. No less a sage then Wikipedia writes: “It is hoped that this will aid research into human diseases that cannot be studied in mice, for example Huntington's disease and strokes.”  These are not trivial diseases, in humans or in mice (where, yes, there ARE studies of these diseases), and they will not be trivial in transgenic monkeys and apes.

Whether or not these animals have enhanced language capabilities, primate transgenesis has potential for more powerful animal research AND for plenty of animal health and welfare challenges.