The Molecular Mechanisms of Homeopathy.

With the launch of the 10²³ campaign against homeopathy, it seems appropriate for me to write a bit on that very subject. But where to start – there are so many blogs that talk about the issues that people have with homeopathy.  The ethical reasons for campaigning against it and the slap in the face that homeopathic “medicine” is to scientific thinking have been subject to in-depth coverage. Recently Prof David Colqhoun has finally gained access to teaching materials for a now-defunct BSc in Homeopathy, via a freedom of information request. He is reviewing them systematically on his excellent DC’s improbable science blog.

So with all this covered by other, more eloquent and seasoned bloggers than myself, I have decided to look at homeopathy from my own niche in science – from a structural/molecular biologist’s point of view.

How medicines work.

To understand the issues that science has with homeopathic “remedies”, one needs to know how conventional drugs work. They work by interacting with another molecule in your body. Generally they will either activate or inhibit the activity of protein molecule that is involved in some sort signal transduction pathway, or in the catalysis of a particular chemical reaction. Structural biology, particularly X-ray crystallography is an immensely powerful tool for understanding the nature of these drug: receptor interactions.  For example, crystallography has revealed the nature of interactions between a bacterial ribosome and the antibiotic chlorampenicol and HINI (Swine flu) neuraminidase interaction with anti-influenza drug, Tamiflu. These structures reveal the nature of the interaction at an atomic level. We can see Electrostatic interactions (negatively charged atoms/groups interacting with positively charged atoms/groups) such as salt bridges and hydrogen bonds. We can see aromatic stacking interactions (aromatic groups stacking alongside each other) and other hydrophobic interactions. We can precisely map out the atomic nature of an interaction.

Such information is invaluable in understanding how a drug works. We don’t have this information for every single drug that is used. But in time we will.

A “conventional” drug, eg Tamiflu, consists of a chemically defined active ingredient at a known and measurable concentration. Each adult dose of Tamiflu contains 75 mgs of Tamiflu. Given that Tamiflu has a molecular weight of 312.4 g/mol, each 75mg dose of Tamiflu contains 1.4×10²⁰ (~140,000,000,000,000,000,000) molecules of Tamiflu (0.075/312.4 * Avogadro’s constant).

Your body takes up a known percentage of this drug – this is known as “bioavailability”. In the case of Tamiflu, the bioavailability is ~60% – so of the 1.4×10²⁰ molecules of Tamiflu that you swallow, only 8.6×10¹⁹ get into your blood stream. These then circulate for until broken down/excreted. Tamiflu works by mimicking the natural substrate of influenza virus neuraminidase, sialic acid. By preventing influenza virus neuraminidase binding to endogenous sialic acid presented on the surface of your cells, Tamiflu prevents infection of host cells. Simple.

How Homeopathy works.

Homeopathic remedies are based upon highly dilute samples of a compound that causes a similar effect to the symptoms of a given disease (“like cures like”). For example, if disease X gives me hiccups, and plant extract Y gives me hiccups, I would take a dilute solution of plant extract Y to cure the symptoms of disease X.

We’ll overlook the fact that this is treating a symptom (by causing it?), and not addressing any underlying cause – because homeopathic remedies are “potentised” during their production. The originator of homeopathy was Samuel Hahnemann, who described a process whereby the effect of these cures could be enhanced by serially diluting them until one reached ultra low concentrations. Each dilution step would be accompanied with “succussion” of the solution – ten hard strikes against a soft, often leather-covered object. The wikipedia page on Homeopathy has a good description of this process.

A typical homeopathic remedy has a concentration of 30C, that is to say, 1 part compound in 100³⁰ (so 10⁶⁰). After Hahnemann first devised homeopathy, Avogadro’s constant (number of molecules in a mole of substance) was calculated, and shown to be roughly 6×10²³ (hence 10²³ campaign). So in a 10²³ or 11.5C solution one might reasonably expect to find 1 molecule of compound. The odds of finding a single molecule of active compound of a 30C homeopathic remedy are very roughly 1 in 10,000,000,000,000,000,000,000,000,000,000,000,000 (10 million million million million million million).

When this was pointed out to the homeopaths (that their remedies had zero active ingredients) they happily discovered that the water in which the active ingredient was dissolved in had a memory and not only remembered the active ingredient (doubtless both its molecular structure and pharmokinetic properties), but the water could also pass this memory on to other water molecules. Cool, huh?

Note – the succussion process allows the water to remember the therapeutic ingredient, but not all the other stuff that the water has seen down the years. And given that a molecule or two of the water currently in the coffee I am drinking were likely to have at sometime been intimately associated with an Ebola virus particle or a molecule of the Ricin toxin – I sincerely hope this is the case. The molecular mechanism of succussion, and how it selects for particular solutes, is currently not understood.

But in order for a homeopathic remedy to work, not only does the water need to retain a memory of the active ingredient, it needs to actually mimic it, so it can then interact with the required target molecule to exert an effect. It has to adopt a similar/identical form.

How water interacts with other (bigger) molecules

Crystallographers can show that molecules such as proteins and DNA interact with solvent water atoms in quite specific ways. I’ve even generated a little picture to show precisely that:

The grey surface represents the molecular surface of a particular protein I’m working on. The red spheres represent ordered water molecules that interact with the suface of the protein.

We see these bound water molecules in pretty much every crystal structure we solve, as the vast majority of crystals are formed from aqueous solutions. Indeed, water molecules can become trapped inside proteins and help them fold correctly and bind substrates or metal ions. Water molecules also participate in chemical reactions – often being split to protenate leaving groups.

So the idea that water adopts a structure around another molecule is not that far fetched. So, being inquisitive, scientists tried to determine if this water structure persisted after the molecule was removed. And to their surprise, it did – but for only ~50 femtoseconds (0.00000000000005 seconds – not very long at all – and not much use for a remedy that’s going to sit on the shop shelf for weeks).

And another thing…

For the sake of further enquiry, let’s assume that the peer-reviewed letter to Nature above is wrong (it isn’t), and water can retain a memory of what’s been dissolved in it for much longer than 50 femtoseconds (it can’t).

This isn’t enough – an empty shell of water molecules around a compound that was once there isn’t going to be enough to interact with and engage protein receptors – they actually have to then rearrange and mimic the chemical and physical properties of the molecule that induced the formation of that shell in the first place.

Not only this – this water structure then has to be robust enough to occupy a protein-binding site and fend off other molecules that might normally bind to that site with a high affinity – molecules that will displace water molecules all the time in order to get into that very binding site. The original therapeutic molecule will not only have a particular shape and form – it will have an inherent stiffness and rigidity – water molecules in solution move relatively freely and will shift to surround any molecules dissolved it them – that is why water is such an excellent solvent. But not homeopathic solutions. Oh no.

In order to circumvent this, the energetics of homeopathic remedies must be amazing – water molecules typically interact with each other via low energy, transient “hydrogen bonds” – each hydrogen bond requires 5-30 kJ/mol to break it (depending upon solvent conditions, temperature, etc), but a permanent “covalent” bond that links atoms in a drug molecule typically requires over 300 kJ/mol to break it – i.e. at least over 10 times stronger. And yet the homeopathic shell of water molecules manages to be as strong as this. Utterly fantastic stuff.

The properties of homeopathic solutions are beyond comprehension – and to think that all this was achieved by diluting something to a point where no original molecules remain whilst periodically banging it on something leathery.

Homeopathy is truly unbelievable.

It is the fantastic and unbelievable nature of homeopathic remedies that leads scientists to question their validity as an effective form of therapy. Experiments to investigate this show that homeopathy is no more effective than placebo. When a therapy has no active ingredient, no conceivable mechanism of action, and cannot be demonstrated to be more effective than a placebo, to claim otherwise is both dangerous and immoral. It certainly has no place whatsoever in a public funded health service like the NHS.