Lesson: the bitter paper

Start here: the accident

In 1931 a chemist called Arthur Fox was working with a powdered compound at a DuPont laboratory. Some of it escaped into the air. A colleague across the room complained about the bitter taste. Fox, who was closer to it, tasted nothing at all.

Most people would have shrugged. Fox went and got more of the compound, and started asking everybody he could find to taste it. Some people recoiled. Others could not tell it from a blank strip of paper. It ran in families.

That compound was phenylthiocarbamide — PTC — and the argument that started across that laboratory is still going on, ninety-five years later. Your class can join it in the next forty minutes.

A note on the story: this account is repeated in a great many textbooks and we have not been able to check it against the original record. Treat it as the story the field tells about itself, which is a useful thing for students to notice in its own right.

Do this

1. Ask for a prediction first, before anyone tastes anything. "Put your hand up if you think you will taste something." Write the number on the board. You will come back to it, and the gap between what people expect and what happens is half the lesson.
2. Hand out a control strip. Plain paper, no chemical. Everyone touches it to their tongue. Nobody should taste anything. This matters more than it sounds: without it, some students will report a taste from suggestion alone, and you will have taught them the wrong lesson about evidence.
3. Hand out the test strip. Touch to the tongue, no chewing, no swallowing. Follow the supplier's own instructions — they are on the packet, and they are the ones that count.
4. Count three groups, not two: tasted nothing / tasted something faintly / that was horrible. Write all three on the board.
5. Now ask the question that opens the lesson: why are there three groups when the textbook says there are two?

What the numbers look like

The first thing to notice is the shape. Almost everything you can measure about a class — height, reaction time, how long anyone can hold their breath — makes a single hill: a lot of people in the middle, fewer at each end. This does not. It makes two hills with a valley between them.

That shape is why anyone went looking for a gene. A trait that splits people into two groups is behaving as if something switch-like is behind it, and for seventy years the explanation was the simplest one available: one gene, two alleles, tasting dominant over not tasting.

The part the textbook gets wrong

It is not one gene with two alleles. It is one gene with three variable positions in it, and they are inherited together as a block.

The gene was finally found in 2003. It codes for a bitter taste receptor, and it has three places where the DNA differs between people, each changing one amino acid. Those three positions are close together and travel as a unit — a haplotype — and worldwide there are five of them, of which two are common.

So what a person actually inherits is two haplotypes, one from each parent. Two working copies and the paper is unbearable. Two non-working and it tastes of paper. One of each, and you get the students in the middle — the group the one-gene-two-alleles version has nowhere to put.

The 2003 paper reported that these haplotypes account for between 55% and 85% of the variation in how strongly people taste PTC. That number is worth putting on the board on its own. It is very high for human genetics. It is also not 100%, and the missing part is a real question, not a rounding error.

Compare your class to the world

Your class will not match any of those bars, and the reasons why are the best discussion in this lesson.

The bars are a prediction, calculated from allele frequencies using Hardy–Weinberg — and calculated from only one of the three positions, because that is the one with clean public frequency data. The real trait uses all three.

And a class is a small sample. Thirty people is not a population. If your class comes out at 20% when the prediction says 33%, that is roughly what you should expect from thirty coin flips, and demonstrating that to a class that has just produced the number themselves is worth more than any worksheet on sampling error.

The argument that is still open

Here is the question nobody has settled, and it is a good one to end on because the class can argue about it with the same information the researchers have.

Why do non-tasters still exist? Bitterness is a warning system — a lot of plant poisons are bitter. A broken warning system ought to be selected against. And yet every human population that has ever been tested contains people who cannot taste PTC, at substantial frequencies. Something has kept the non-working version around.

The long-standing answer was balancing selection: the non-working version must be good for something, perhaps letting people eat foods that tasters reject, so both versions survive.

Then in 2016 a study of 5,589 people from 105 populations looked for the signature that natural selection leaves in DNA. It found evidence of something ancient — older than modern humans leaving Africa — but nothing since. Its conclusion was that the modern pattern is explained mostly by where people went and who they had children with, not by selection favouring one version.

That is a genuinely open argument, published in an open-access journal your students can read. Ask them which they find more convincing, and what evidence would settle it. There is no answer at the back of this book.

Questions

To warm up

1. Before anyone tasted anything, how many people predicted they would taste something? How close was the class? Why might people guess wrong about their own tongue?
2. Why did we start with a strip of plain paper? What would have happened without it?
3. Two students disagree about whether they tasted anything. Neither is lying. What does that tell you about measuring a trait?

Core

4. The textbook version says tasting is dominant. Our class had a group in the middle. Does that disprove dominance, or does it mean something more specific? Say exactly what.
5. Three DNA positions sit close together in one gene and are inherited as a block. Why does that make the trait look simpler than it is when you only measure the taste?
6. The haplotypes explain 55–85% of the variation. Give three different things that could account for the rest, and say how you would test one of them.
7. Our class result differs from the predicted bar for our part of the world. Give two reasons that have nothing to do with anybody being unusual.

To stretch

8. Hardy–Weinberg says the share of people with two copies of an allele is its frequency squared. Use the chart to work backwards: what allele frequency gives 33% non-tasters? Now say three assumptions you just made that are not true of any real human population.
9. Design an experiment that would distinguish "natural selection kept the non-working version" from "it is there because of where people migrated". What would you measure? What result would change your mind?
10. This trait was taught as one gene with two alleles for about seventy years, and that was wrong. It was not a careless mistake — it fitted the evidence available. What does that suggest about the things in your textbook now?

Take it further

• One gene, three answers — the same gene turns up in sinus infections and in how well an antibiotic works. Good extension reading.
• rs713598, rs1726866 and rs10246939 — the three positions, one page each, with the sources.
• What is a SNP — if the class needs the vocabulary first.
• Smell and taste — the rest of the sensory genetics on this site.