Immunoglobulins: Genes

Another post about immunoglobulins! This time it's about gene rearrangements, which I covered briefly for BIOC2001, but now it's time to go into more detail!

Describe multigene organisation of immunoglobulin genes

As you should know by now, antibodies are made up of two identical light chains and two identical heavy chains. Furthermore, these light chains can be either κ or λ. Heavy chains, κ light chains and λ light chains occupy different loci on different chromosomes. Each of these loci contain several gene segments that can combine in different ways to provide a wide variety of antibodies, as mentioned in one of my previous posts. These gene segments are the leader (L), variable (V), joining (J) and constant (C) sequences. Heavy chain genes also have diversity (D) sequences.

Before I get into how the genes are rearranged, I'll quickly run through how they are arranged to begin with.

λ light-chain genes have around 30 leader-variable sequence pairs (each leader sequence has a specific variable sequence that it's always paired with), as well as 4-5 joining-constant sequence pairs (every joining sequence has a constant sequence that it's always paired with).

κ light-chain genes are slightly different in that they also have leader-variable sequence pairs, but not joining-constant sequence pairs. Instead, there are around 5 joining sequences, which all combine with the same constant sequence, as the κ light-chain only has one constant region.

Finally, heavy-chain genes have around 40 leader-variable sequences (you don't need to remember the numbers, by the way), around 23 diversity sequences, around 6 joining sequences and 9 different constant sequences. Each constant sequence corresponds to a different kind of antibody, as mentioned in my previous post about immunoglobulins. Each constant sequence encodes all 3 or 4 domains of the constant heavy chain in question.

Describe variable region gene rearrangements
Describe the mechanism of Ig DNA rearrangements

In most cells of the body, the gene segments encoding antibodies are located far apart from each other. However, in cells that are going to grow up to become B-cells, an irreversible process called somatic recombination moves some of these gene segments closer together.

First, let's look at heavy-chain genes, because I'm pretty sure they get rearranged first. The first step that happens here is that a diversity (D) sequence gets joined to a joining (J) sequence. Just think of it as DJ, as in "It's murder on the dance floor, you'd better not splice this wrong, DJ!" (I'm sorry, I'm terrible at parody lyrics.) The next step is the splicing of a leader-variable pair with the DJ- the VDJ rearrangement. Both of these rearrangements are done courtesy of the RAG-1 and RAG-2 (recombination-activating gene 1 and 2) recombinases. After this, excess mRNA (most notably the bit between the last joining sequence and the required constant sequence) is spliced, a poly-A tail is added and the mRNA leaves the nucleus to be translated. As translation occurs, the L-sequence pulls the growing polypeptide into the lumen of the endoplasmic reticulum before being cleaved off. After all this you finally have a ready-to-go heavy chain!

Light-chain rearrangement is a little bit simpler because they don't have diversity sequences. The only rearrangement here is VJ, which is also catalysed by RAG-1 and RAG-2. After this it goes through all of the processing and translation steps, just like the heavy chain, before it is ready.

When both the light and heavy chains are done, they can then combine. Since there are a lot of different light chains that can be made and a lot of different heavy chains that can be made, there are an enormous number of different antibodies that can be made- and that's not even including some of the other ways in which variability can be introduced!

Another important thing to know about is that there are Recombination Signal Sequences (RSS) flanking the gene segments in both light- and heavy-chain genes. RSS can be one- or two-turns, depending on whether they make up one or two turns of the DNA helix. The rule here is that a one-turn RSS can only join with a two-turn RSS, thereby preventing improper joining of the segments. RSS have both conserved and nonconserved parts: generally a conserved nonamer (9 base pairs) and conserved heptamer (7 base pairs) with around 12-24 nonconserved base pairs in between.

Joining of the segments is not always the same each time. There can be a tiny bit of flexibility in where exactly the RAG-1 and RAG-2 enzymes cleave the DNA, leading to something called "junctional flexibility" that adds to antibody diversity. These rearrangements can be productive or non-productive: non-productive rearrangements have a premature stop codon somewhere in their sequence, while productive rearrangements don't.

One last VERY IMPORTANT note: The DJ, VDJ and VJ rearrangements that I've spoken about above are things that are only done ONCE in the B-cell lifespan. Hence a B-cell will make the same antibody for life. The antibody might change slightly due to mutations and so forth, but not dramatically. Also the antibody can become one of a different class (so IgM can become IgG as I'll discuss later), but it will still recognise the same antigen.

Describe class switching

As I just mentioned, an antibody can switch classes. All B-cells start off producing IgM, so in order to produce the other types of antibodies, they have to do something known as class switching.

Switching between IgM and IgD is relatively straightforward and also reversible as it doesn't actually change anything in the DNA. Essentially the location of the poly-A tail determines whether the antibody generated will be IgM or IgD, and if it will be secreted or membrane-bound. There are four important polyadenylation sites to know about: 1, 2, 3 and 4.

1. Poly-A site 1 is located right after the μ constant regions (IgM). Polyadenylation here causes formation of secreted IgM.
2. After poly-A site 1, there are two "membrane exons" called M1 and M2, required for insertion into the membrane. Poly-A site 2 is located right after these membrane exons, resulting in formation of membrane-bound IgM.
3. Poly-A site 3 is right after δ constant regions and thus results in secreted IgD.
4. Poly-A site 4 follows both the δ constant regions and the membrane exons M1 and M2 (yup, IgD has its own membrane exons).

"New" B-cells favour polyadenylation sites 2 and 4, so naïve mature B-cells will express membrane-bound IgM and IgD. (Immature B-cells express membrane-bound IgM only.) Again, since the generation of membrane IgM and IgD is completely controlled by where RNA is polyadenylated and spliced (a process known as "differential RNA processing"), cells can continue to express both IgM and IgD.

To get IgG, IgE and IgA, further rearrangements of the DNA need to be done. Since this involves splicing out some of the DNA permanently, these rearrangements are irreversible. Basically what happens is that the LVDJ of the heavy-chain gene combines permanently to a constant region other than the ones encoding μ (IgM) and δ (IgD). This occurs when B-cells are stimulated by an antigen, and the process is helped along by an enzyme called activation-induced cytidine deaminase, or AID. (So I guess you can say that the process is AIDed along. Ha. Ha. Ha. I'm sorry.)

The rearrangements of the DNA is facilitated by special regions known as switch sites. Each constant region, aside from IgD, has its own switch site. (IgD doesn't need one because switching from IgM to IgD is completely mediated by differential RNA processing, as described earlier.) When it's time for the DNA to be rearranged, the DNA loops around so that the switch site before the new constant region meets the switch site of the old constant region. The loop is excised, resulting in a shorter piece of DNA that now produces a new class of antibody.

Multiple class switching events can happen in the lifetime of a B-cell, but these events must always proceed in the "forward" direction. To better explain, let me quickly give you the order of constant region genes: μ, δ, γ3, γ1, γ2b, γ2a, ε and α. Any of these genes can be switched for anything to the right of it (well, except for δ, because IgD is a bit different). However, you can't go backwards in the list. Why? Well, when you switch from IgM to IgE, everything in between gets excised out. Hence you can't switch back to IgG, because none of the gamma genes are in the DNA any more.

Describe the generation of antibody diversity

Last point to go over for this topic! I've already touched on a lot of the ways in which antibody diversity is generated, but it's nice to have a summary.

1. There are a lot of genes that encode antibodies, and they can be combined in different ways (i.e. you can have one of 30+ variable sequences, 4-6 joining sequences etc.)
2. Junctional diversity- there's a bit of variation in where exactly RAG-1/2 cleave, resulting in some different sequences.
3. P-nucleotide additions- now here's something that I haven't touched on yet! When RAG enzymes cleave the DNA, this forms a "hairpin" structure in the cleaved ends. When these hairpins are cleaved in order to join up D-J and V-D-J/V-J, complementary nucleotides are added. These latter additions are known as P-nucleotide additions.
4. N-nucleotide additions- another topic I haven't touched on! Aside from P-nucleotide additions, sometimes an enzyme called TdT (terminal deoxyribonucleotidyl transferase- a real mouthful that makes you thankful for the acronym) adds in a few extra random nucleotides to seal up the gap. This adds in a bit more diversity in the binding site.
5. Somatic hypermutations- Just like any other gene, genes encoding antibodies are prone to mutation. However, the mutation rate is 100 000x higher than that in other genes. Mutations are most likely to occur in hypervariable regions, especially in activated B-cells in germinal centres. Somatic hypermutations are responsible for affinity maturation, or the progressive increase in affinity of an antibody for an antigen during the course of an immune response. (And yes, before you ask, some mutations can be detrimental and make an antibody lose its affinity for an antigen. This makes the B-cell less likely to receive survival signals, so "survival of the fittest" means that the better-performing B-cells live on to protect us better.) One last thing you need to know about somatic hypermutations is that they are usually associated with class switching, as both processes require AID enzymes.
6. Finally, as mentioned earlier, there are a wide range of H chains that can pair with a wide range of L chains, giving us a lot of combinations.

Whew! Another post down! There's just one more topic in the upcoming midterm that I haven't blogged about. (The midterm's next Monday- scary to think that it's already halfway through semester!)