Antibody specificity is determined by six small loops called complementarity-determining regions (CDRs), located in the variable region at the tip of each antibody arm. Three of these loops sit on the heavy chain and three on the light chain. Together, they form the binding surface (called the paratope) that physically contacts a target molecule, and their unique combination of shape and amino acid sequence is what allows each antibody to recognize one specific target while ignoring everything else.
The Variable Region: Where Specificity Lives
Every antibody is built from two types of protein chains: heavy chains and light chains. Most of each chain is “constant,” meaning it’s identical across antibodies of the same class. But at the very tips of the Y-shaped molecule, each chain has a variable domain (VH on the heavy chain, VL on the light chain) where the amino acid sequence differs dramatically from one antibody to the next. These two variable domains pair up to create a single binding pocket, and that pocket is what gives each antibody its identity.
Within each variable domain, most of the sequence actually folds into a stable scaffold called the framework region. The real action happens in short stretches of highly variable amino acids that form loops projecting from that scaffold. These are the CDRs.
The Six CDR Loops
Each variable domain contains three CDR loops, giving every antibody a total of six. On the heavy chain they’re labeled CDR-H1, CDR-H2, and CDR-H3. On the light chain (which comes in kappa or lambda varieties), they’re CDR-L1, CDR-L2, and CDR-L3. When the heavy and light chain variable domains fold together, all six loops cluster at the tip of the antibody to form the complete binding surface.
These loops vary in both their amino acid sequence and their length. CDR-H1 and CDR-H2 on the heavy chain tend to be relatively constrained, with the most common forms being about eight amino acids long. Light chain CDR2 loops are even shorter, typically just three amino acids. CDR-H3, by contrast, ranges anywhere from 4 to 36 amino acids in length, making it by far the most structurally diverse of the six loops.
Why CDR-H3 Matters Most
Among the six loops, CDR-H3 plays the most important role in determining what an antibody binds. It sits at the center of the binding site, makes the highest number of direct contacts with target molecules on average, and its extraordinary range of lengths and shapes gives it the greatest structural diversity of any CDR. The shape and sequence of CDR-H3 can even influence how the other five loops are positioned, effectively reshaping the entire binding pocket.
This is also why CDR-H3 is the hardest loop to predict computationally. Drug developers trying to design antibodies in silico consider accurate modeling of CDR-H3 the single biggest challenge, precisely because it doesn’t fall into predictable structural patterns the way the other loops often do.
That said, specificity is not a one-loop job. The heavy chain CDRs collectively tend to contribute more binding contacts than the light chain CDRs, but the light chain loops still play a meaningful role. The exact contribution of each loop varies from antibody to antibody depending on the shape and chemistry of the target.
How Heavy and Light Chains Work Together
Specificity doesn’t come from the six loops in isolation. It also depends on how the heavy and light chain variable domains pack against each other, because this packing determines the precise angle and spacing of the CDR loops relative to one another. Research has identified at least two distinct packing arrangements, and the difference between them changes the overall geometry of the binding site enough to influence what size and type of target the antibody recognizes.
Key amino acid positions at the interface between the two chains (particularly positions 41 through 44 and position 36 on the light chain) act as molecular hinges that determine which packing mode the antibody adopts. Swapping even a few of these residues can shift the relative positions of the CDR loops, altering the shape of the binding pocket without changing the loops themselves. So while the CDRs provide the direct chemical contacts with a target, the way the two chains associate acts as a second layer of specificity control.
Framework Regions Shape the Binding Site Indirectly
The framework regions that surround the CDRs are often described as passive scaffolding, but they play a more active role than that suggests. Framework residues form stabilizing bonds with nearby CDR loops, holding them in the correct orientation for binding. In one well-studied broadly neutralizing antibody called PGT121, a framework residue near position 68 on the light chain forms a hydrogen bond with a CDR1 residue at position 27, locking the loop into a specific shape.
Mutations in framework regions can also increase the flexibility of the entire binding site, allowing a single antibody to shift between slightly different conformations. This conformational plasticity lets the antibody accommodate a wider range of target variants, a trait that’s particularly important for broadly neutralizing antibodies against rapidly mutating viruses like HIV. In these cases, framework mutations don’t change which target the antibody recognizes so much as they broaden the range of closely related targets it can bind.
How Specificity Sharpens Over Time
When your immune system first encounters a new pathogen, the antibodies produced are a rough fit for their target. Over the following days and weeks, a process called somatic hypermutation introduces random point mutations into the genes encoding the variable regions, concentrated heavily in the CDR loops. B cells whose mutations happen to improve binding are selected and multiply, while those with neutral or harmful mutations die off.
This cycle of mutation and selection, called affinity maturation, repeats over multiple rounds. The mutations cluster disproportionately in CDR-H1 and CDR-H2 at residues that directly contact the target, fine-tuning the fit between antibody and antigen. The result is antibodies that bind their target hundreds or thousands of times more tightly than the originals.
Somatic hypermutation doesn’t only hit CDRs. Framework mutations accumulate too, and as described above, these can adjust the flexibility or orientation of the binding site in subtler ways. But the CDR mutations are what drive the sharpest gains in binding strength and precision.
Standardized Numbering Systems
Because CDR boundaries matter so much for antibody engineering and drug design, researchers use standardized numbering systems to define exactly where each CDR starts and ends. The most widely adopted is the IMGT system (from the international ImMunoGeneTics information system), which assigns fixed position numbers to the variable and constant domains of every antibody. Under IMGT numbering, CDR1 spans positions 27 to 38, and CDR3 spans positions 105 to 117. Older systems like Kabat and Chothia numbering define slightly different boundaries, which can cause confusion when comparing studies, so checking which system a paper uses is important when reading the literature.