Chemokines Explained: The Signals That Direct Immune Traffic

Immunology · Peptide signalling

Chemokines are the roughly four dozen small signalling proteins that tell immune cells where to go. Everything else about the immune system — recognising a pathogen, killing an infected cell, building a lymph node — depends on the right cell arriving in the right tissue at the right moment, and chemokines are the address system that makes that happen.

What are chemokines?

A chemokine is a small secreted protein, generally in the 8–12 kDa range and around 70 amino acids of mature sequence, that directs cell migration. Structurally they are one of the most conserved families in vertebrate biology: nearly all of them fold into the same “IL-8-like” shape — a flexible, unstructured N-terminus, three central beta-strands, and a C-terminal alpha-helix, held together by internal disulfide bonds between conserved cysteines.

That flexible N-terminus is not decoration. It is the part that inserts into the receptor and switches it on, which is why removing or extending just a few N-terminal residues can turn a chemokine into an antagonist. Everything else — the fold, the disulfides — is scaffolding that holds the business end in place.

Chemokines are often lumped in with cytokines, and formally they are a cytokine subfamily. What sets them apart is that they are defined by a shared structural motif rather than a shared function, and that they signal almost exclusively through G protein-coupled receptors. Most other cytokines — interleukins, interferons — use single-pass receptors and JAK/STAT signalling instead.

The four chemokine families, and why the names look like that

Chemokine nomenclature confuses people because there are two overlapping systems. The older names were descriptive and inconsistent: IL-8, MCP-1, RANTES, SDF-1, fractalkine, lymphotactin. In 2000, Zlotnik and Yoshie proposed a systematic scheme built on the one feature every chemokine shares — the arrangement of its first two cysteine residues.

That gives four families. If the first two cysteines sit directly next to each other, it is a CC chemokine. If one other residue separates them, it is CXC. Three residues, CX3C. And if the second and fourth cysteines of the usual four are missing, it is an XC chemokine. Each ligand then gets a number and an L for ligand: CXCL8 is the molecule everyone used to call IL-8, and CCL2 is MCP-1.

The receptors take the same prefix and swap L for R: CC chemokines act on CCRs, CXC chemokines on CXCRs, and so on. A useful rule of thumb is that ligands almost never cross families — a CC chemokine will not activate a CXC receptor — but within a family, promiscuity is the norm. Most receptors bind several ligands and most ligands bind several receptors. Hold onto that fact; it explains a great deal of what follows.

There is also a small set of atypical chemokine receptors (ACKR1–ACKR5, the fifth of which was added to the official nomenclature only in 2025). These bind chemokines but do not trigger conventional G protein signalling. Instead they act as scavengers and shapers, mopping chemokine out of circulation or moving it across cell layers to sculpt gradients.

How a chemokine actually moves a cell

The mental picture most people have — a cell swimming up a soluble concentration gradient — is only half right. Chemokines bind avidly to glycosaminoglycans, the sugar chains decorating endothelial surfaces and extracellular matrix. Immobilised that way, they form a surface-bound gradient rather than a diffusing cloud. This is haptotaxis, and it matters because it keeps the signal spatially precise: a leukocyte in flowing blood does not get prematurely activated by chemokine that has washed downstream.

The signalling itself is standard class A GPCR pharmacology. Chemokine binding couples the receptor to the Gi/Go family, which suppresses cyclic AMP and, more importantly here, launches a phospholipase C arm that raises intracellular calcium. Downstream, CalDAG-GEFI activates the small GTPase Rap1, and Rap1 effectors such as RIAM and RAPL flip integrins from a bent, low-affinity conformation into an extended, high-affinity, clustered one.

That integrin switch is the whole point. Selectins let a leukocyte roll along the vessel wall, but rolling is reversible — the cell drifts on. The chemokine signal is what converts rolling into arrest, after which the cell crawls, finds a junction, and transmigrates into tissue. Delete the chemokine step and the traffic pattern collapses even though every other component still works.

Homeostatic versus inflammatory chemokines

Functionally the family splits in two. Homeostatic chemokines are expressed constitutively and run baseline architecture: CXCL12 retains stem cells in bone marrow via CXCR4; CCL19 and CCL21 acting on CCR7 organise naive T cells and dendritic cells into the correct zones of a lymph node so they can actually meet.

What the human evidence shows

Despite roughly three decades of intensive work, the approved-drug list is short. As of 2026 there are five FDA-approved agents acting on the chemokine system, and every one of them targets a receptor rather than a chemokine.

• Maraviroc (Selzentry, initial US approval 2007) — a CCR5 antagonist, used with other antiretrovirals for CCR5-tropic HIV-1 only.
• Plerixafor (Mozobil, 2008) — a CXCR4 inhibitor used with filgrastim to mobilise haematopoietic stem cells for autologous transplant in non-Hodgkin lymphoma or multiple myeloma.
• Mogamulizumab (Poteligeo, 2018) — a defucosylated anti-CCR4 antibody for relapsed or refractory mycosis fungoides and Sézary syndrome.
• Motixafortide (Aphexda, 2023) — a second CXCR4 inhibitor for stem-cell mobilisation in myeloma.
• Mavorixafor (Xolremdi, 2024) — an oral CXCR4 antagonist for WHIM syndrome, a rare immunodeficiency caused by gain-of-function CXCR4 mutations.

The HIV entry story deserves its own note, because it is the clearest case of a chemokine receptor doing something no one designed it for. HIV-1 uses CD4 plus a chemokine co-receptor to enter cells: CCR5 for R5-tropic strains, CXCR4 for X4-tropic strains, both for dual-tropic virus. People homozygous for a 32-base-pair deletion in CCR5 lack functional surface CCR5 and are strongly resistant to R5 HIV-1 — the observation that made maraviroc possible.

Why blocking chemokines is so hard

Set against those five approvals is a long record of failure, particularly in inflammatory disease. Maraviroc failed in a randomised, placebo-controlled rheumatoid arthritis trial; so did the CCR5 antagonists SCH351125 and AZD5672. In one striking anti-CCR2 study in rheumatoid arthritis, the antibody occupied and reduced free CCR2 on monocytes by 57–94 percent and produced no clinical or synovial improvement whatsoever.

The explanation is the redundancy noted earlier. Work on synovial fluid showed that blocking CCR2 and CCR5 together still failed to stop monocyte chemotaxis, while blocking CCR1 did — because when you close one receptor, the ligands simply signal through another. A network with dozens of ligands, roughly twenty signalling receptors and heavy many-to-many overlap does not have a single load-bearing node to knock out. The successes on the list above are the exceptions that prove it: each targets a setting where one receptor genuinely is non-redundant — a viral co-receptor, a mutated receptor in a rare disease, a marker on a tumour cell, or a single retention signal that only needs to be interrupted for a few days.

Frequently asked questions

Are chemokines peptides or proteins?

The boundary is a matter of convention rather than chemistry. At roughly 70–80 residues, chemokines are larger than what is usually called a peptide and are normally described as small proteins. They are made by ribosomal translation and folding, not by chemical synthesis, and their activity depends on that fold.

What is the difference between a chemokine and a cytokine?

Chemokines are a structurally defined subfamily of cytokines. The practical distinctions are that chemokines share the cysteine motif and the IL-8-like fold, they signal through GPCRs rather than JAK/STAT-coupled receptors, and their defining function is directing cell movement rather than switching on a transcriptional program.

Why do chemokines have two different names?

Historical accident. Ligands were named as they were discovered, often for what they appeared to do (monocyte chemoattractant protein-1) or where they came from. The CCL/CXCL system was introduced in 2000 to replace that with something systematic. Both sets remain in circulation, so CXCL8 and IL-8 are the same molecule.

Can chemokine receptors be targeted with peptide drugs?

Yes, in principle — motixafortide is itself a synthetic peptide antagonist of CXCR4. In practice the same pharmacological questions apply as to any receptor-directed agent: selectivity, whether the ligand is a full or partial agonist or an antagonist, and whether signalling is biased toward G protein or arrestin arms.

Informational only — not medical advice · 21+