Do you remember broccoli-gate? President George H.W. Bush sparked domestic and international outrage with his disdain for the vegetable. Years later, during President Barack Obama’s tenure, broccoli even became embroiled in the health-care debate as a proxy for perceived government overreach.
Why did Bush hate broccoli, while Obama loved it? It may come down to their genetics. To some people, broccoli (and related plants in the genus Brassica like cabbage and mustard, as well as coffee and hoppy beer) taste intensely bitter. For others, it doesn’t.
It turns out there is a stretch of DNA on chromosome 7 that affects how we taste some of the bitter chemicals in broccoli. In biologist lingo, we say that the gene taste 2 receptor member 38 (abbreviated TAS2R38 or T2R38) codes for a protein, and that protein regulates our bitter-taste ability.
While everyone should have a TAS2R38 gene, we don’t all have the exact same version. The different versions are called alleles. Some alleles allow us to taste the bitter compounds; some don’t. Several alleles of this gene are found in Africa (recall that Obama’s father was a Kenyan Luo), but elsewhere there are only two main versions: one for “taster” and one for “nontaster”.
Whether you can taste broccoli bitterness depends on which alleles you have. You inherit half of your autosomal DNA from your mother and half from your father. They each gave you an allele of the TAS2R38 gene. If you got two nontaster alleles, Brassica foods probably don’t taste bitter to you. If you get two taster alleles, broccoli and friends might be less than appealing. And if you get one copy of each allele—one taster and one nontaster—you can taste the bitterness, but only moderately. Scientists call these three states nontasters, supertasters, and tasters, respectively.
I can only guess at the genetic makeups of Presidents Bush and Obama, but I’ll venture that Bush was a supertaster while Obama was not.
You don’t have to guess for yourself, though, because some of the DNA testing companies report your status. And even if they don’t report it, it’s hidden in your raw data file, where you can check it if you know what to look for.
Trait Reports
23andMe includes a “Bitter Taste” report in their Ancestry + Traits test, while AncestryDNA has a $10 add-on feature that includes their “Bitter Sensitivity” report along with several others.
There are three spots in the TAS2R38 gene that differ between the common taster and nontaster alleles. We call those spots SNPs (for single nucleotide polymorphisms), and we give them truly atrocious names like rs10246939, rs1726866, and rs713598. The taster allele has the nucleotides C, G, and G at those respective SNPs, while the nontaster allele has T, A, and C.
You can download your raw data file from your testing company and search it for those SNPs to determine your own status. Depending on the testing company and test version, the file should report two or three of the SNPs. Most people will only need to know their status at one, though, because the SNP triads usually go together.
Not coincidentally, the fact that C, G, and G almost always go together in one allele and T, A, and C almost always go together in the other forms the basis for statistical phasing. Statistical phasing is a computational way of improving DNA matching by inferring which SNPs are together on one chromosomal copy versus the other.
The proportion of the taster allele varies by population. 23andMe has a nifty graph showing how it varies among their customers.
What Does This Have to Do with Covid?
Scientists have known for a while that some taste genes also affect immune response. (Here’s a thought: Are they actually immune genes that affect taste response?) And as a recent scientific study suggests, the ability to taste this particular kind of bitterness may make you more resistant to covid.
The study, published by doctors in Baton Rouge, Louisiana, and Cairo, Egypt, looked at 1,935 healthy people with increased exposure to covid (e.g., health care workers, patients). The researchers gauged the bitter sensitivity of the subjects and tracked them over time. About a quarter of the subjects were nontasters, about half were tasters, and the remaining were supertasters.
After 3 months, 266 (≈14%) of the subjects had tested positive for the virus. Nontasters were about 10 times more likely to test positive and nearly 4 times more likely to be hospitalized. They were also sick for longer (24 days on average) than tasters (14 days) or supertasters (5 days).
Of course, TAS2R38 is not the only factor affecting our susceptibility to covid. Age, underlying health conditions, the degree of exposure, and other genetic factors all play a part. That’s why it’s essential to get vaccinated, regardless of your ability to taste bitterness.
Check Your Status
If you don’t have access to trait reports at 23andMe or AncestryDNA, you can check your TAS2R38 status by looking in your raw data file. Depending on which company you used and when you tested, your file may not include all three SNPs.
Once you download the file (instructions for obtaining it from AncestryDNA are here), open it in a text editor and use the search function to find the SNP names. Here’s what you’re looking for and how to interpret it.
| SNP | nontaster | taster | supertaster |
| rs10246939 | TT | CT | CC |
| rs1726866 | AA | AG | GG |
| rs713598 | CC | CG | GG |
Consider my 23andMe data. The columns are: SNP name, chromosome number (7), position, and my base calls. The files from other companies might look slightly different but will contain the same basic information.
rs10246939 7 141672604 CT rs1726866 7 141672705 AG rs713598 7 141673345 CG
Notice that for each SNP, I have two different base calls, CT for the first SNP, AG for the second, and CG for the third. That’s because I have one taster allele and one nontaster one. I’m a taster, but not a supertaster.
If you have TT, AA, and CC, you are a nontaster, and if you have CC, GG, and GG, you are a supertaster.
If you’ve tested multiple family members, you can sometimes even track the alleles through your tree. My father, for example, is a nontaster, so I know I inherited my nontaster allele from him and my taster allele from my mother. It turns out she’s a supertaster, and her brother is a taster. That tells me that both of their parents (my maternal grandparents) were either tasters or supertasters, and at least one of them was definitely a taster. My husband is a nontaster and so are both of our kids. (One hates broccoli anyway.)
Now, eat your vegetables!
Great article! So much fun. I loved it.
Sincerely,
another geek
So glad you liked it. It was fun to write, too!
What a fun explanation! Of course I had to check. I am a TC; AG; and CC. So a taster for the first two alleles and NonTaster for the 3rd one. Is it a common thing to be different in a category?
How interesting! The two most common versions of TAS2R38 have either TAC or CGG for those three SNPs. You may have a rare version! Yours are probably TAC and CGC. If you’ve tested elsewhere, I’d probably check your raw data file from the other company just to confirm that it’s not a genotyping error. Please let me know!
My brother and I are both the same as @Jamie Longfellow (above): TC, AG, and CC (when tested with ancestry.com; one of us has tested with 23andme, with the same results). And we’re from a parent who, like @Victor (below), also has a C-instead-of-G variation, but theirs is varied from “supertaster”, while ours is varied from “taster”.
Even if you don’t know what to make of this, do you know what search terms I could use to read about this?
I would try “supertaster covid”.
this is a fascinating finding. I have often wondered if there would be a similar finding for sweetness. (Not necessarily covid related). I am one of those people who find things perfectly sweet and others will say “whoa, much too sweet for me”. Any research going on with a sweetness SNP?
Absolutely! AncestryDNA’s trait reports includes “Sweet Sensitivity”, based on a gene called TAS1R3.
Loved this blog. As a child, I was not served broccoli. I never ate it until I married my husband who ate it all the time. I actually love it, but due to health conditions can no longer eat it. I will need to check our DNA raw data.
I’m so glad you enjoyed it!
Awesome topic, DNA relevance, family history and great science writing rolled into one.
Thanks Leah! Sent to my fellow naturopathy students – this must be the really good reason we’re not all eating our broccoli, even though it’s up there on the must have veges list!
So glad you enjoyed it!
Very interesting article! I can’t check our raw data for all three of the SNPs on this laptop. I can’t the USB for my external hard drive plugged into this thing for some reason. But, I was able to check the 23andMe results for David, myself, and our oldest daughter. 23andMe reports on rs713598 for us:
David; CC
Girlie: CG
Me: CG
Since we get one allele from our fathers and one from our mothers, I will assume our daughter got her G from me and her C from her dad, because he couldn’t give a G there if he didn’t have one there himself. Is that a correct assumption?
This means that David is a non-taster and me and Girlie are tasters, right? David and I both love broccoli. Our daughter loved it as a little one. It is one of the few things she would eat after she turned 2 years old. Now, she has an extremely limited diet and it does not involve broccoli at all.
I can’t wait to check the DNA data of our mothers and my father. 🙂
Yes, exactly! David gave her the “C” allele because that’s all he has. You could have given her either “C” or “G”, because you have both, and by chance it ended up being the “G” version. He’s a non-taster and you and your daughter are tasters,
Thank you for confirming that for me. Our daughter found your article quite fascinating, too. 🙂
So glad to hear that!
So I’ve eaten my vegetables and opened my 23andMe file and can’t match what I see with the comment in the article.
I have 1 x taster and 2 x supertaster – and do like broccoli !
100% European ethnicity per 23andMe.
rs10246939 7 141672604 CC
rs1726866 7 141672705 GG
rs713598 7 141673345 CG
You may have a rare allele. How cool! One of your TAS2R38 alleles is the common CGG taster version while the other is apparently CGC. I don’t have any information on how that allele might affect bitter taste. (The other possibility is that you had a genotyping error at rs713598. If you have tested elsewhere, you can check your raw data file from the other company to make sure you got the same results for those SNPs.)
I really like you turning things upside down and suggesting that the gene is actually about immune response and just happens to affect taste.
While my siblings happily eat their broccoli, I am a supertaster. So a reluctant broccoli consumer, while knowing it’s good for me. But this tasting profile helped me when coming up with flavors for therapeutic liquids for children containing naturally bitter antibiotics.
For that resistant child, I recommend steaming until there is still some crunch and then serving with a sauce of butter and honey. Adults are often accepting of charred flavors which themselves are bitter and hide the natural bitterness of broccoli. And sauces containing dried chiles – ancho and such – tend to mask bitterness also, especially the smoked ones.
Now I’m off to look at my DNA and my parent’s (sic) to see what alleles I have and where they came from!
How does this interact with the links on chromosome 4 that that 23 and Me was finding were linked to the loss of taste/smell? It was written about on their blog on June 2nd (not going to paste the link because I don’t want to get flagged by a spam filter!) . 🙂 I find it really interesting that this study used taste test strips to identify supertasters, not genetic testing.
The spam filter is me, so paste away! I can only speculate as to the answer to your question, because even the covid researchers don’t know why patients lose smell/taste. It’s a great question, though! I doubt the loss is related to TAS2R38, because the covid patients who lose taste/smell do so for all tastes and smells, not just bitter. This article from Nature (highly reputable) discusses some reasonable hypotheses: https://www.nature.com/articles/d41586-021-00055-6
Is is possible to have rs10246939 and rs1726866, but not rs713598?
Technically, you have all three of those SNPs. What base calls do you have at each of them?
I’m a bit late to the party but I love this post. I’ve just gone into the raw data files for my partner and myself. And I’ve confirmed why I can order a cocktail containing Campari without there ever being a risk that he will steal it. He’s a super taster and I’m a nontaster who enjoys the slight bitterness that he equates to cough medicine.
I’m tickled that your drinks are safe because of your DNA!