What the hell is sickle cell? And can CRISPR treat it well? I know a lot of people are heading into finals season – and I was trying to think of the best ways to help – and I’ve always found I learn best when I can connect concepts – to each other – and to the broader world. Like seeing how a random mutation in a gene causing a DNA letter switch can lead to the protein letter switch, which (since different protein letters have different properties) can lead to a malfunctioning protein that interacts with other molecules differently, preventing cells in the body from carrying out their normal functions, and leading to disease symptoms. How this disease is and isn’t inherited, and how scientific findings from multiple fields and diverse topics can coalesce beautifully to provide a radical solution.
When it comes to such stories, there’s perhaps none better than that of sickle cell anemia and other diseases involving the oxygen-carrying protein hemoglobin – broadly categorized into structural hemoglobinopathies (in which the hemoglobin protein itself is messed up) and thalassemias (in which the hemoglobin might be ok, but less of it is made). These diseases cause problems with oxygen transport and in some cases (like sickle cell) lead to clumps of hemoglobin causing abnormally-shaped red blood cells that get stuck in tiny blood vessels, leading to painful sickle cell crises.
Sickle cell anemia is the name given to the disease in which a person has 2 copies of the HbS gene, but other hemoglobin disorders are caused by different mutations, some of which are more problematic than others. But despite having different genetic origins, they may share a treatment opportunity that relies on something they all have, but that they haven’t been using in a while… another form of hemoglobin that gets made in fetuses & infants but then is “switched off.” The cells still have the instructions for making it, and scientists are now testing out using the genetic engineering tool CRISPR/Cas to get the cells to make it in adults.
So this story is particularly relevant (and hopefully helpful & interesting to a wider audience) since the first clinical trials are underway to treat patients with these diseases (1 with sickle cell disease and another with β-thalassemia) with the genetic engineering tool CRISPR (well, actually they treat just their blood-making cells, which they first take out of the body, then they modify them to make this “backup” version of hemoglobin, and stick them back in.) But I’m getting ahead of myself, where to begin?
Hemoglobin & diseases associated with it
Hemoglobin is one of the most important molecules in your body – it’s the protein that carries oxygen throughout your bloodstream, making it essential for life. Lungs are useless if you can’t get the oxygen you breathe in to the places in your body where you need it! Hemoglobin gets its name for the “heme” groups it contains – these groups aren’t made up of protein letters and they’re not in the DNA instructions for the protein – instead these heme groups are “cofactors” – small molecules that bind to the protein and help it carry out its functions. And the heme groups themselves don’t only bind the hemoglobin, they also bind to a metal ion (charged particle) – in this case iron. And that iron binds to oxygen.
The heme is pretty stuck in there. But the oxygen can come and go depending on how much the heme wants it (affinity) and how much oxygen there is. When there’s a lot of oxygen around (like in the blood vessels surrounding the lungs), the heme grabs on. But when that blood reaches areas with lower oxygen concentrations, it starts letting go, so those areas like your fingers and toes get oxygen too. And then when the blood cycles back to the lungs it gets oxygenated again. And it can keep doing this over and over and over, keeping a steady supply of oxygen throughout your body – oxygen that’s needed for things like making ATP (“energy money”) in cellular respiration. As you might imagine, then, problems with hemoglobin can cause system-wide problems with a range of severity depending on the nature of the problem.
There are two main types of diseases involving hemoglobin: structural hemoglobinopathies and thalassemias. The difference? In a structural hemoglobinopathy, the hemoglobin protein is structurally abnormal, whereas in a thalassemia, the hemoglobin itself can be normal but it’s present in decreased levels.
Proteins and their classification
Proteins are made up of chains of building blocks called amino acids that fold into intricate structures. Some proteins are made from a single chain – we call these proteins monomers. Other proteins, oligomers, are made up of multiple chains stuck together. If each chain (subunit) is the same, we call the protein homomeric and if they’re different we call it heteromeric. We further classify oligomers based on how many chains are in the final protein. Hemoglobin (Hb), for instance, is a tetramer (4 chains) made up of 2 dimers (2 chains). Although the dimers are identical, each is made up of 2 different chains, so we classify it as a heteromer. So, if we want to get specific, Hb is a heterotetramer.
The dimers in the hemoglobin adults produce is made up of 1 α subunit and 1 β subunit, making the final product α2β2. When these subunits are all normal, normal adult hemoglobin (HbA) is produced.
You might wonder why I’m speaking in terms of “adult” hemoglobin – this is because fetuses actually produce a different form of hemoglobin, HbF, which has γ subunits instead of β subunits (α2y2). This form of Hb has an increased affinity for oxygen, which allows it to bind oxygen from the mother’s blood even though that blood isn’t highly oxygenated (the fetus doesn’t have direct access to that oxygen-rich blood straight from the mom’s lungs, so it’s “optimized” to latch on to the “leftovers”. After birth, this extra affinity is no longer needed, so the baby gradually (over a few months) switches over to producing HbA. When it comes to hemoglobin disorders, therefore, problems with the α subunits will be present at birth but problems with the β subunit won’t become apparent until these subunits start replacing the γ subunits. We’ll return to HbF later (and hopefully so will patients!)
The instructions for a protein are contained in a gene, written in the DNA language made up of a nucleotide alphabet consisting of 4 bases (A, T, G, & C). Each amino acid within a protein is coded for in that protein’s gene by a 3-nucleotide “word” called a codon. When the DNA of a gene gets altered, it *can* affect the protein output (the levels produced [as in hemoglobinopathies] and/or the structure of the protein itself [as in thalassemias]).
I emphasize *can* because not all genetic mutations even cause protein mutations (and later we’ll see that not all protein mutations cause protein problems). The genetic code is redundant, meaning that multiple 3-base combinations code for the same amino acid. For example, GGG and GGA in a gene both lead to the production of a glutamic acid molecule, so a mutation from GGG to GGA or vice versa won’t affect the protein structure. We call this type of mutation, in which the amino acid is unchanged, “synonymous.”
“Non-synonymous” or “missense” DNA mutations, on the other hand, do change which amino acid gets produced at a certain location on a protein, and this is the type of mutation that causes sickle cell anemia (SCA).
The causal SCA mutation, HbS, is a single point mutation (adenine-to-thymine in the DNA) that causes a change in the amino acid in position 6 in the β-chain. The A->T swap changes the codon for this 6th amino acid from GAG to GTG (GUG in the mRNA copy) – so the protein making machinery (ribosome) sticks in a valine (Val or V) instead of the normal glutamic acid (Glu or E). In biochemistry shorthand, we write this as Glu6Val or E6V.
Another β-hemoglobin mutation is HbE. It involves a single point mutation at amino acid position 26 in the β-chain that changes the amino acid from glutamic acid (Glu or E) to lysine (Lys or K). In biochemistry shorthand, we write this as Glu26Lys or E26K.
Non-synonymous mutations aren’t necessarily harmful. Some mutations can actually be beneficial and can even serve as the basis for evolution of new functions. Other mutations can be disastrous, but many are merely “neutral.” Sometimes, a mutation may be good for some of the protein’s functions but not for others. The effects of the mutation largely depend on where on the protein the change occurs (some regions of the protein are more important than others) and how “different” the swapped amino acid is from the one that was there before (each amino acid has unique biochemical properties, some being quite similar and others being drastically different).
In the case of the HbE mutation, a glutamic acid is replaced by a lysine – even though they can have opposite charges, they’re of similar size & both like hanging out with water (are hydrophilic). So it only causes mild biochemical differences that weaken the interaction between subunits and makes the hemoglobin slightly less stable under certain conditions. Therefore, HbE can be characterized as a hemoglobinopathy.
The HbS mutation, however, causes a more drastic letter-switch. Glutamate is long and charged & hydrophilic, whereas valine is short and stumpy and (most importantly) hydrophobic – it doesn’t want to be near water. But the change is just at that one letter, so the protein (to optimize all the other interactions that are still normal) still folds basically as if the glutamate were there – that position gets situated on the outside of the protein when the protein folds up.
But with the HbS mutation, you now have a hydrophobic letter there (valine), which, finding itself on the outside, facing water, will jump at the chance to hang out with a hydrophobic part of another hemoglobin protein instead of water, and thus they clump together. And, in red blood cells (immature ones are called reticulocytes and mature ones are called erythrocytes), they have a good chance of finding one another because those cells are chock full of hemoglobin.
The tetramers we talked about (2 α and 2 β subunits working together) were normal – that’s how normal hemoglobin hangs out and functions normally – what’s *abnormal* is when these tetramers further group up (polymerize). The polymerization is favored in low the low oxygen so not oxygen-bound state, so as the cells cycle throughout the body (up to 4 times per minute), going from high to low oxygen environments, they keep polymerizing & unpolymerizing over and over, majorly stressing out the cell. 14 of such hemoglobin strands wind together into a helical bundle, and those bundles can join up too. These polymers are awkwardly-shaped, so they make the RBCs awkwardly-shaped too. The RBCs adopt a “sickle” shape that can “clog” blood vessels – especially the really tiny ones – in something called vaso-occlusion. These problems at the small scale can cause large-scale problems – SCDs are characterized by a range of symptoms including acute and severe pain events (“crises”), stroke, and kidney disease.
Mutations can also affect hemoglobin production, which is the “thalassemia” side of things. An example of this is that same HbE mutation we saw before – the structural changes it causes are thought to be less of an issue than the production changes. When a cell wants to make a protein, it first must copy (transcribe) the DNA instructions into messenger RNA (mRNA) that is then translated into protein. The DNA form of the gene has “extra” segments of information (introns) between the parts that contain the amino acid instructions (exons) and these intervening regions are removed in a process called splicing. The splicing machinery knows where on the unedited mRNA to cut because of “splice sites” encoded in the genes.
In addition to changing the 26th amino acid in the chain, the HbE mutation also reveals a hidden (cryptic) splice site that “confuses” the cell when it’s trying to edit the RNA copy of the gene. Less functional mRNA is therefore produced, and less β-globin made, so we classify HbE as a thalassemia in addition to a hemoglobinopathy.
Those are just 2 examples of mutations that can cause hemoglobin diseases. There are many more. Because you get one copy of the hemoglobin genes from each parent, you typically need 2 mutated copies to see an effect, but the mutations don’t have to be the same, so this is a great opportunity to get into genetics…
Different versions of the same gene are called alleles and a person inherits one allele of each gene from each parent – at least for the autosomal (non-sex) chromosomes. If these alleles are the same, the person is said to be homozygous for that gene and if they’re different the person is said to be heterozygous for that gene.
Sometimes, only one faulty copy is enough to cause symptoms (either because there’s not enough of the good stuff made or because the faulty one itself causes problems). We call such diseases “autosomal dominant.” An example is Huntington’s disease. In other cases, one good copy’s enough, the bad copy doesn’t do any harm, and so a person’s fine in the heterozygous state (but not in the homozygous state). We call such diseases “autosomal recessive,” and an example is cystic fibrosis.
Sickle cell disease is what’s referred to as “autosomal co-dominant” which means that both alleles contribute to the ultimate phenotype – so, for example, if you have different mutations in the different copies you get you can have different symptoms. Let’s take a closer look at what this means…
The gene for the β hemoglobin chain is HBB and it’s located on chromosome 11. Scientists give different “nicknames” to the different forms of it.
The allele responsible for HbS (containing the E6V mutation) is referred to as βS. If a person inherits a βS allele from one parent and a normal β allele from the other parent, they are said to carry the sickle cell trait (this heterozygous state is frequently written HbAS). The normal copy of the β chain they have is able to compensate for any reduced production from the abnormal copy and the person is healthy (though they have a 50% chance of passing the abnormal allele onto their children).
If a person is homozygous for βS (they inherit one abnormal copy from each parent, so βS/βS), they are said to have sickle cell disease (SCD). They have a 100% chance of passing on the βS allele (their children will, at a minimum, carry the sickle cell trait).
Similarly, the allele responsible for HbE (containing the E26K mutation) is referred to as βE. If a person inherits a βE allele from one parent and a normal β allele from the other parent, they are said to carry the HbE trait (this heterozygous state is frequently written EA). Like above, the normal copy of the β chain they have is able to compensate for any reduced production from the abnormal copy and the person is healthy (though they have a 50% chance of passing the abnormal allele onto their children).
If a person is homozygous for βE (they inherit one abnormal copy from each parent), they are said to have HbE disease (EE). Although the name has “disease” in it, most people with the condition don’t have any symptoms (though their blood does look slightly different under the microscope) because β globin is still produced, even if at reduced levels, and it functions fairly normally. However, because a person with HbE disease doesn’t have any normal copies, they have a 100% chance of passing on the βE allele (their children will, at a minimum, carry the HbE trait).
More serious problems arise if a person inherits a βE allele from one parent but instead of also inheriting a normal β allele from the other parent, they inherit a β allele with a different, more compromising, mutation. This type of situation is referred to as compound heterozygosity and, in this case, it leads to HbE/β-thalassemia. This second mutation could be one of many mutant β alleles, including ones that prevents any β production from that allele (β0), ones that cause reduced β-globin production (a hypomorphic HBB allele sometimes abbreviated Hbβ+) or one that causes severe structural changes such as βS. People with the HbE trait don’t show symptoms because their normal copy can compensate, but in the case of compound heterozygosity, the second copy can’t compensate and, depending on the severity of this other mutation, the person may suffer from mild to severe thalassemia with symptoms including poor growth, organ damage and anemia (low red blood cell count).
Modifying roles of α-globin
HbE/β-thalassemia is, as the name suggests, a type of β-thalassemia, because the reduced production is of the β-globin chain, but there are also conditions called α-thalassemias, in which insufficient α-globin chain is produced. Unlike β-globin, which is coded for by a single gene (HBB), α-globin is coded for by 2 genes (HBA1 and HBA2), so a person normally inherits 2 α-globin genes from each parent. Sometimes, however, these genes get deleted, so a person can have 1-3 copies (0 copies is fatal before or shortly after birth). If a person only has 1 copy, they have HbH disease (symptoms include anemia, liver and spleen problems, and bone abnormalities) and if they have 2 copies, they have HbH trait (which may cause mild anemia).
Interestingly, decreased α-globin production can actually lessen the severity of symptoms associated with some thalassemias. This is because it is not only the amount of α and β-globin produced, but the ratio between them, that matters. α and β-globin are much more stable in dimers than alone; if the amounts of one are much greater than that of the other (globin-chain imbalance), not all of them will be able to form dimers, so they may aggregate (clump up) and cause problems.
Modifying roles of fetal hemoglobin?
Remember how I said before that fetuses make a “super” version of hemoglobin, HbF, that has a γ subunit instead of a β? Well, it’s super for a fetus dealing with lower “high oxygen” levels but it stops being made after birth because oxygen’s more abundant so it doesn’t need to grab onto it as desperately.
The protein stops being made, but the genetic instructions for making it are still there – written as a gene in chromosome & locked up tight in the nucleus. But it stops being made because another protein, BCL11A, starts being made and it acts as a “brake” to switch off HbF production. If doctors could get patients with hemoglobin problems to make this fetal hemoglobin again, by stopping BCL11A from keeping it “off” this might be able to compensate for the faulty adult hemoglobin. It’s the β form that’s the problem in SCD & β-thalassemias, and it’s the β chain that gets swapped out for a γ in fetal hemoglobin.
Something you might be wondering is – don’t babies stop making it for a reason? Will re-expressing it in adults cause problems? Well, it turns out that some people have mutations in BCL11A that cause them to make HbF even as adults – and they seem perfectly healthy. And people that have SCD but also happen to have a BCL11A mutation has much milder symptoms.
So it seems like disrupting the BCL11A gene “knock it out” could offer a viable strategy. And the editing only needs to be done in the cells that make the blood cells – the hematopoietic stem cells. These stem cells are NOT embryonic stem cells, they’re “adult” stem cells that serve as the source of a continuous supply of mature blood cells. And they are NOT germline cells (those in your eggs or sperm that can get passed down) so any editing done won’t affect any progeny.
So what scientists and doctors are trying out is removing patients’ hematopoietic stem cells, using CRISPR/Cas to edit the BCL11A to tamp down it’s tamping down of HbF and sticking the stem cells back in the patient (after chemotherapy has wiped out their remaining hematopoietic stem cells), where they can repopulate the stem cell stock with cells that make the fetal version that can compensate for the faulty versions. More on CRISPR/Cas here: http://bit.ly/37gbI1w
Scientists and doctors at the NIH Clinical Center in Bethesda, Maryland and at Boston Children’s Hospital have been researching this for several years, and CRISPR Therapeutics and Vertex Pharmaceuticals recently announced data from the 1st company-backed study testing a CRISPR-based therapy in humans. Their treatment is called CTX001 and it seems to be working.
The patient with β thalassemia went from needing an average of 16.5 blood transfusions per year to 0 since receiving it. And the patient with SCA went from an average of 7 vast-occlusive crises to 0. And their adult blood cells, as hoped, were making the fetal hemoglobin! They plan to enroll 45 patients in each study and track them for at least 2 years. I’m excited by the prospects, but also hope that the therapy will become available to all those who need it. 300-400,000 babies are born with it each year, with the highest concentration in sub-Saharan Africa, where resources are often lacking.
If you want to know more about this treatment, here’s a good article http://bit.ly/2Oe8zYu
I hope this post gave you some insight into the wonders and complexities of life on the molecular level that contribute to life on the human level!