Field of Science

Elementary; The Science of the Perfect Murder

It was television that inspired me to devise the perfect murder. Not for the usual reasons relating to scientific inaccuracy. This isn’t about watching a CSI unbalance a centrifuge, contaminate a sample or holding a pippette backwards. Small errors like this just don’t bother me anymore. I long ago accepted that most television shows are set in a science fiction fantasy world, where science only works when the gods of plot convenience allow it. It was in this frame of mind of that I watched a recent episode of Elementary (S1e17 “Possibility Two”).

For those of you who aren’t familiar with it, “Elementary” is a procedural cop drama with the twist that it is also a Sherlock Holmes pastiche. It follows the detective (played by Johnny Lee Miller) and Joan Watson (played by Lucy Liu) as they solve crime in modern day New York. Johnny Lee Millers interpretation of Holmes can be described by shouting “QUIRKY” and doing the jazz hands so hard that they violently detach and fly across the room. This is a show that proudly flies the flag and wears the matching underpants of silliness. The very definition of TV to turn your brain off and enjoy. Unfortunately, my brain’s off switch is always slightly faulty  and against all odds, this episode did make me think. I am going to spend the rest of this post exploring these thoughts, so I should warn you that this article contains major spoilers.

This episode kicks off with millionaire Gerald Lydon attempting to get Holmes to reveal that he has hereditary Cerebral Amyloid Angiopathy (CAA). The twist is that he believes that someone gave him this hereditary genetic disease deliberately. To Sherlock, the answer to this particular puzzle is as simple as cluedo. The culprits were Lydon’s parents in the bedroom using the *ahem*. But they have to be ruled out as suspects, because they didn’t have CAA (and it is a dominant genetic trait).

And then Sherlock asks “Have you considered that you may not know your true parentage ?”, and they spend the next forty minutes of the episode tracing Lydon’s real father whilst discussing the degree to which parents are responsible for the genetic inheritance they bestow on their children and the episode finishes with Sherlock explaining the mystery to his dementia ridden client using sock puppets. Just kidding, this wouldn't be a murder mystery procedural drama without an actual murder mystery.

Someone did find a way to give Gerald Lydon CAA, and they did it by using science. In the process of this episode, they frequently invoke fictional geneticists that say “this is totally possible” in a naked attempt to justify the absurdity at the heart of the episode.  Sherlock’s investigations lead him to a slick biotech company where their one jobbing scientist to explain the basis of CAA.
CAA is essentially caused by a an excess build up of beta amyloid protein in capillaries that carry blood within the brain. as these build up, blood is prevented from going where it’s needed, leading to brain cells dying off. As the scientist explains, the hereditary forms can occur through a mutation in the APP gene. This gene encodes the Amyloid precursor protein, which forms the harmful Beta Amyloid that is the cause of CAA.  Later in the episode, that scientist sends the Sherlock a chemical structure, with a note that says something akin to “This is the murder weapon”.

Unfortunately, I’m not an expert on chemistry so I don’t know the identity of this chemical*. In this episode, it acts as a mutagen which specifically interacts with the APP gene to cause the build up of amyloid beta. How it does this is not explained, and I think I am safe in saying that this compound wouldn't actually work in real life**.  But just because this compound is unlikely to give someone CAA, it doesn’t mean that it isn’t possible.
For the past 10 years, people have been developing gene therapies to cure hereditary diseases through manipulating the human genome. But what if we re-purposed the tools used for gene therapy into a murder weapon, in order to cause a disease rather than to cure it ? What if we put ourselves in the shoes of the murderer of this episode. How would we go about giving someone CAA for real ?
Let’s get back to the APP gene, which encodes the Amyloid precursor protein. We want to alter this gene so that it ends up forming as much Beta amyloid protein as possible.   The APP gene, through a clever trick using short RNAs, actually codes for a number of different  APP’s, only a few of which can be broken down to form Beta amyloid. The gene itself is structured into segments (which are called “exons”) which are transcribed depending on the presence or absence of microRNA’s.
It is known that the production of Beta amyloid is related to the presence of one particular section of the APP gene, known as the “Kunitz protein inhibitor” (KPI) is associated with the build up of beta amyloid [1,2].  However, this section of the gene often excluded from APP transcripts by a microRNA known as MiR-124 [3].

When the DNA code is being translated, this MiR-124 binds to the KPI section of the gene, and excludes it from the APP. If we stop MiR-124 from doing this, we increase the amount of APP with the KPI section in it, and in theory this could increase the production of beta amyloid, which would give an individual CAA. How do we do this ?
We can do it by using some nifty chemicals known as Morpholinos. They mimic RNA, except they don’t get broken down as easily, and they can be attached to other chemicals to make sure that they enter cells. We could design a morpholino to mimic the action of MiR-124 if our intention was to prevent someone from getting a build up of Beta amyloid in their brain. But that isn’t what we want to do. We want to stop MiR-124. So we can design a morpholino with a sequence that binds to MiR-124. This would effectively halt its function, and allow for more beta amyloid to build up.
But how do we give this lethal morpholino to a patient without them noticing ?
The murderer from Elementary fortunately solved this one for us. We are told that the murderer targeted three millionaires, and gave the drug intravenously whilst they were staying as patients in hospital. Delivering a drug intravenously allows it to go straight to the blood. Blood goes everywhere in the body, including the brain. While these patients are being infused with the anti-MiR-124 morpholino, Beta amyloid builds up, eventually to the levels needed for them to develop CAA ***.
Now we need to know how much Morpholino to give our murder victims. The main challenge here is the blood brain barrier, which prevents many drugs, including morpholinos, entering the brain. We are going to try to overcome this by simply putting in such a high dose that it overcomes the blood brain barrier.
In studies in rats, researchers have administered a dosage of approximately 300pmol of morpholino per rat brain [1]. Let’s imagine that to get the same effect in humans, all we need to do is increase the dose based on weight****. A rat brain weighs around 2 grams, and a human brain weighs around 1400g.
Anti-MiR-124 dose/Rat Brain= 300pmol/2 g
Human brain= x/1400g
Therefore the dose = 150 x 1400g

210,000pmol of morpholino needs to enter the human brain in order to have the desired effect. I could only find one study that makes quantitative measurements of morpholino concentrations in the brain after an intravenous dose, and it suggests that the dosage of morpholino can be reduced by around a thousand fold [5] *****. So we’re going to have to up that dose to 210,000,000 pmol, per person. 
At this point, one may wonder why anybody would go to this much trouble to kill some rich philanthropists in such a convoluted way ? The answer is as demented as it is ludicrous. The killer is a scientist suffering from CAA. In order to cure himself, he needs more funding. So he decided that the best way to go about this was to deliberately give rich people CAA, which would compel them to donate money to him in order to fight the disease. I will demonstrate the ludicrousness of this plan through the use of simple arithmetic.
10,000,000 pmol of a morpholino costs at least $9,000 dollars (based on numbers from So the total cost of giving one dose of the drug to reach their brain would be around $189,000 dollars. I am going to assume that they were given the drug throughout their stay in hospital. All of these individuals are over the age of fifty, and the average hospital stay for that age group in the United States is 5-6 days. So at most, it would cost $ 1,134,000 to poison one person.
During the course of this episode, not one, but three people have been poisoned successfully in this way, bringing the total budget of this endeavour to $3,402,000. This assumes that it all went according to plan (which science so often doesn’t) and that the murderer only produced enough chemical for three people.  Let’s not forget that the technology for morpholinos already exists . If the murderer wanted to develop some other method to cause this disease, they would have to start from scratch, requiring even more money. This highlights the central contradiction within the plot.
As the scientists often say during the episode, with unlimited money and resources a scientist could develop the technology to give someone CAA.  As I have demonstrated, the killer had plenty of money around to throw at this project. Yet the whole reason behind this convoluted murder scheme is to allow the perpetrator to acquire money.  It’s almost like the murderer has some sort of degenerative brain disease.
In thinking through this murder plot, I am forced to consider the many barriers to actually pursuing it in real life, which I have boiled down to three points
  1. To induce a genetic disease, we need to know a lot about it. I kept coming up against this problem with CAA. It’s not that I couldn’t find a cause, but that there were so many different genes that could play a role that I couldn’t point to any of them being the root cause.
  2. Our technology isn’t there yet. There are a lot of promising developments in gene therapy that are just over the horizon. I’ve spent most of the this post talking about morpholinos, but there are other treatments in the works. Viruses that have been re-purposed to deliver genes are now commonplace in some labs, and advances in nanotechnology may allow for promising developments in the future. Very few of these have ended up in successful human trials, and they are still being tweaked to make them more effective.
  3. Resources. The cost of using this technology immediately puts it out of the hands of a casual amateur. And if you are a professional scientist, you will eventually be called in to explain the half million dollar hole in your lab finances that is designated “convoluted murder plot”.
Our knowledge of genetic diseases is expanding as are the gene therapies that are being developed to combat them. The “Genetic disease” method I have described here may be plausible at some point in the future. But I have to note that the three points I have described are also barriers to curing genetic diseases.
When we do get to the point where our knowledge and technology is advanced enough to manipulate a person’s genome so subtly that we can give them a genetic disease, we will also be at the point where these advancements will also allow for the treatment, prevention and possibly even a cure for it.
It is somewhat amazing that a run-of-the-mill procedural drama catalysed this thought experiment.  But sometimes inspiration can come from the unlikeliest places. Sherlock sums it up the best in the last line of the episode.
  “A good detective knows that every task, every interaction no matter how seemingly banal, has the potential to contain multitudes.”


[1] Matsui T., Ingelsson M., Fukumoto H., Ramasamy K., Kowa H., Frosch M., Irizarry M. &; Hyman B. (2007). Expression of APP pathway mRNAs and proteins in Alzheimer’s disease, Brain Research, 1161 (3) 116-123. DOI:
[2]  Tharp W.G., Lee Y.H., Greene S.M., Vincellete E., Beach T.G. & Pratley R.E. (2012). Measurement of altered AβPP isoform expression in frontal cortex of patients with Alzheimer's disease by absolute quantification real-time PCR., Journal of Alzheimer's disease, 29 (1) 449-457. DOI:
[3] Smith P., Al Hashimi A., Girard J., Delay C. & Hébert S.S. (2011). In vivo regulation of amyloid precursor protein neuronal splicing by microRNAs, Journal of Neurochemistry, 116 (2) 240-247. DOI:
[4]  Fujikawa T., Tamura K., Kawase T., Mori Y., Sakai R., Sakawa K., Yamaguch A., Ogata M., Soya H. & Nakashima K.; (2005). Prolactin Receptor Knockdown in the Rat Paraventricular Nucleus by a Morpholino-Antisense Oligonucleotide Causes Hypocalcemia and Stress Gastric Erosion, Endocrinology, 146 (8) 3471-3480. DOI:
[5] Stein D.A., Huang C.Y.H., Silengo S., Amantana A., Crumley S., Blouch R.E., Iversen P.L. & Kinney R.M. (2008). Treatment of AG129 mice with antisense morpholino oligomers increases survival time following challenge with dengue 2 virus, Journal of Antimicrobial Chemotherapy, 62 (3) 555-565. DOI:


*It looks like a unsaturated aromatic ring linked via three keto groups to branched carbon chains that end in either a napthalene group or an adenine, and these speculations are purely based on the diligent use of google image search.
** If I am wrong about this compound being ineffective, then not only does this cement my inexperience as a chemist, but also worryingly suggests that not only is one of the scriptwriters for the show devising bioweapons, but they are broadcasting them on air for the whole world to see. That’s a scary thought. What will we see next, plans for an Antimatter bomb in the background of Sherlock’s study ?
*** You’ll need to expose the victims to the drug for a long time for it to be effective, as people who are born with the mutation we are attempting to replicate often take an entire lifetime for symptoms to become visible. And this is not to mention that the link between the presence of KPI and high levels of amyloid beta have only really been shown to have a correlational relationship, so this entire method of inducing it could be completely wrong. But let us assume that it works as a method of murder.
**** The actual art of comparing the biological reactions of all the different creatures within the animal kingdom (known as Allometry) is a massive subject which touches on evolution as well as medical science, and boiling it down to a mathematics exercise misses out a lot of its more subtler details.
*****  I experienced severe math fail when reading reference [5]. They administered the drug at 10mg/kg to a mouse. In 200 micrograms of morpholino compound was given to a mouse, equating to approximately 0.26 micromols. This is assuming a mouse weighing 20 g, that a 24 nucleotide morpholino has a similar molecular weight to an RNA oligo of approximately ~ 7500 m.w. They later measured the presence of morpholino in the brain, and found that it hovered at around 0.02 micromols, which would make it look like it goes to the brain very well. However, in this same experiment they detected 9.34 micromols in the liver, which would suggest either that the morpholinos are replicating inside the mouse, or that my math went wrong somewhere. It is almost certainly the second option,  because Figure 6 of that particular paper shows quite clearly that barely any of the morpholino reaches the brain, and my thousand fold guess comes from looking at the graph, and also comparing the amount of morpholino in the brain compared to other organs in table 2 of the paper.

I've written a shorter version of this article on Kinja ( to make it clear, it's that article where I'm self plagiarising, as I wrote it after this one.


  1. Has a serious bug in this episode? Sherlock is trying to set up several possibilities for molecules with plastic models of balls. It does this because it claims that "how not know what elements are in the molecule, it must try all possibilities." But just looking at the image of an assumed molecule anyone with minimal knowledge of chemistry knows that the only elements that exist there are C and H! And the definition of Watson on the molecule is very general: a hydrocarbon poly-cyclic! Of course it is a hydrocarbon and has cyclical shares, only to see the image someone like Sherlock, who even in the same episode deals with acids, and seems to have knowledge in chemistry would know.

    1. As I demonstrated, even someone completely unfamiliar with chemistry can find out the components of this compound using the magic of google image search. The real art of this was to work out where the nitrogen molecules are within these structures.
      If I were to try to hazard a "cannon" explanation, it would be that sherlock as a character can have large and bizarre gaps in his knowledge. In the original books, he is said to have no idea that the earth orbited the sun. I imagine his knowledge of chemistry could manifest a certain level of ignorance as well.
      But realistically, I think Sherlock found this chemistry difficult because the screenwriters found it difficult. And that they wanted him to make a tyrannosaurus out of the carbon molecule playset.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS