How protein-modifying enzymes may connect B cells, T cells and autoantibodies in two very different autoimmune diseases.

Two Very Different Diseases — and a Surprising Similarity

Coeliac disease primarily damages the lining of the small intestine. Rheumatoid arthritis primarily affects the joints, although it can also have wider effects throughout the body. At first, the two diseases may appear to have little in common. Yet both are autoimmune diseases involving an interaction between genetic susceptibility, specialised immune cells, protein-modifying enzymes and antibodies directed against the body’s own molecules.

In a newly published research paper, Professor Ludvig Sollid asks whether the two diseases might share a deeper underlying mechanism: could an enzyme become physically associated with the protein it is modifying, allowing the adaptive immune system to respond to the enzyme and its substrate as one interconnected package?

The idea is well developed in coeliac disease. Sollid explores whether a comparable process could also occur in rheumatoid arthritis.

First, What Are B Cells and T Cells?

B cells and T cells are specialised white blood cells belonging to the adaptive immune response—the part of the immune system that learns to recognise particular targets and can retain a memory of them.

B cells recognise specific substances through receptors on their surface. They can capture these substances, take them inside and display small fragments for other immune cells to inspect. When appropriately activated, some B cells develop into plasma cells that produce antibodies.

Helper T cells examine the fragments displayed by B cells and other antigen-presenting cells. If a helper T cell recognises one of these fragments, it can provide signals that help the B cell survive, multiply and produce antibodies.

This carefully controlled collaboration normally helps the body mount a precise response to infection. In autoimmune disease, however, the conversation can become misdirected, allowing immune cells that recognise the body’s own proteins to participate in a sustained inflammatory response.

Why Do Protein-Modifying Enzymes Matter?

Proteins are assembled from chains of amino acids, but their structure is not necessarily fixed when that assembly is complete. Cells can subsequently make small chemical changes to them. These are called post-translational modifications.

Such modifications are normal and essential. They help regulate how proteins behave, where they are located and how they interact with other molecules. However, a modification can also alter the way a protein fragment is recognised by the immune system.

In coeliac disease and rheumatoid arthritis, two different families of enzymes perform two different modifications:

• In coeliac disease, transglutaminase 2, or TG2, can deamidate gluten peptides.
• In rheumatoid arthritis, peptidylarginine deiminase enzymes, particularly PAD2 and PAD4, can citrullinate proteins.

Neither enzyme is inherently harmful. The question is what happens when the modified protein, the modifying enzyme and susceptible immune cells meet within the wrong immunological context.

The Better-Understood Model: Coeliac Disease

Gluten contains many glutamine-rich protein sequences. Some gluten fragments resist complete digestion and reach the lining of the small intestine. TG2 can modify selected glutamine residues within these fragments through a process called deamidation. This converts glutamine into glutamate and gives the peptide a more negative electrical charge.

In people carrying susceptible HLA molecules—particularly HLA-DQ2 or HLA-DQ8—the deamidated gluten peptides fit especially well into the HLA binding groove. The HLA molecule then displays the gluten fragment to a gluten-reactive CD4 helper T cell. The sequence can be simplified as:

Gluten peptide → modification by TG2 → stronger binding to susceptible HLA → presentation to a helper T cell → adaptive immune activation

Possessing HLA-DQ2 or HLA-DQ8 does not mean that a person will inevitably develop coeliac disease. These molecules create susceptibility, but additional biological and environmental conditions are required.

How Does the Immune Response Turn Against TG2?

A central puzzle in coeliac disease is why the immune system produces antibodies against TG2, one of the body’s own enzymes, when the initiating antigen is dietary gluten. Sollid’s model provides a possible answer.

While TG2 is acting upon gluten, the enzyme and gluten peptide can form an enzyme–substrate complex. This means that the enzyme and the material it is modifying are temporarily—or sometimes more stably—physically connected. A B cell whose surface receptor recognises TG2 can bind and capture the entire TG2–gluten complex. It then takes the complex inside, processes it and displays a gluten-derived fragment on an HLA-DQ molecule.

A gluten-reactive helper T cell recognises the displayed gluten fragment and supplies activating signals to the B cell. The B cell can then multiply and develop into antibody-producing plasma cells. Crucially, the two immune cells do not need to recognise the same component:

• The B cell recognises TG2.
• The T cell recognises gluten.
• The TG2–gluten complex connects them.

This creates a route through which a T-cell response against a foreign dietary protein can provide help to B cells that recognise one of the body’s own enzymes. Those B cells may subsequently produce the anti-TG2 antibodies detected in coeliac blood tests.

Do Anti-TG2 Antibodies Cause the Intestinal Damage?

Anti-TG2 antibodies bind predominantly to extracellular and cell-surface TG2. They are produced in abundance within the coeliac intestinal mucosa and can become deposited around basement membranes, connective tissue and small blood vessels.

They are extremely useful markers of active coeliac disease. However, it would be an oversimplification to say that these antibodies independently attack and destroy the intestinal villi.

Much of the intestinal injury is thought to arise from the wider cellular immune response, including gluten-reactive helper T cells, inflammatory signalling and intraepithelial lymphocytes that damage stressed intestinal epithelial cells.

Anti-TG2 antibodies may still influence the disease by interfering with TG2-related functions such as tissue organisation, epithelial adhesion, repair and blood-vessel formation. Their exact contribution to tissue injury in humans remains less firmly established than their diagnostic importance.

The Rheumatoid Arthritis Side of the Story

Rheumatoid arthritis is a chronic autoimmune disease in which the immune system contributes to persistent inflammation within the joints. The synovium—the tissue lining the joint—becomes inflamed and thickened, and the continuing immune response can eventually damage cartilage and bone.

As in coeliac disease, however, the process involves more than antibodies simply selecting a normal tissue and attacking it. It includes genetic susceptibility, protein modification, antigen presentation and communication between B cells and helper T cells.

Citrullination: A Different Type of Protein Modification

The protein modification most closely associated with rheumatoid arthritis is called citrullination.

It is carried out by a family of enzymes known as peptidylarginine deiminases, or PAD enzymes. PAD2 and PAD4 are particularly relevant to rheumatoid arthritis.

These enzymes convert an amino acid called arginine into another amino acid called citrulline. This removes a positive electrical charge and can alter the protein’s shape, interactions and immune visibility.

Citrullination is not inherently abnormal. It occurs during normal physiological processes and may increase during inflammation, cellular stress and cell death. The presence of citrullinated proteins alone is therefore not sufficient to cause rheumatoid arthritis. The problem arises when the adaptive immune system begins to recognise certain citrullinated proteins as targets.

Antibodies Against Citrullinated Proteins

Many people with rheumatoid arthritis produce anti-citrullinated protein antibodies, commonly abbreviated to ACPAs. These antibodies can recognise citrullinated forms of several proteins rather than one single joint-specific molecule. ACPA testing is clinically useful because these antibodies are strongly associated with rheumatoid arthritis, although not every person with rheumatoid arthritis has them.

Some people with rheumatoid arthritis also produce antibodies against the PAD enzymes themselves—particularly PAD4 and, in some cases, PAD2. This creates the same unusual combination seen in coeliac disease:

• antibodies against proteins that have been modified;
• and antibodies against the enzymes responsible for producing the modification.

Sollid asks whether an enzyme–substrate complex could connect these two immune responses.

Could a PAD–Protein Complex Connect B Cells and T Cells?

While a PAD enzyme is citrullinating a protein, the enzyme and its substrate must come into close physical contact. Sollid proposes that this association may create a PAD–substrate complex that can be captured by a B cell.

A B cell whose receptor recognises PAD could bind the whole complex and take it inside. The B cell could then process the captured material and display a citrullinated peptide from the accompanying protein on an HLA molecule.

A helper T cell that recognises that citrullinated peptide could provide activating signals to the PAD-reactive B cell. The B cell could then multiply and develop into plasma cells producing antibodies against PAD.

The proposed sequence is: PAD enzyme binds a protein → PAD citrullinates the protein → a PAD-reactive B cell captures the complex → the B cell displays a citrullinated peptide → a peptide-reactive helper T cell activates the B cell → anti-PAD antibodies are produced

As in coeliac disease, the B cell and T cell would not necessarily recognise the same part of the complex:

• The B cell recognises the PAD enzyme.
• The T cell recognises a peptide from the associated protein.
• The PAD–protein complex brings the two targets together.

This is the central mechanistic parallel proposed in the paper.

The Role of HLA in Rheumatoid Arthritis

HLA molecules act as molecular display platforms. They hold small protein fragments on the surface of antigen-presenting cells so that T cells can examine them.

In rheumatoid arthritis, particular variants of HLA-DRB1 are among the strongest inherited risk factors. Several of these variants share a similar amino-acid sequence and are therefore described as carrying the shared epitope.

Certain RA-associated HLA-DR molecules can present particular citrullinated peptides to helper T cells. Citrullination may change how a peptide fits within the HLA binding groove or how it is recognised by a T-cell receptor.

This provides another parallel:

• In coeliac disease, susceptible HLA-DQ molecules present deamidated gluten peptides.
• In rheumatoid arthritis, susceptible HLA-DR molecules can present selected citrullinated peptides.

However, carrying one of these HLA variants does not mean that disease is inevitable. Genetic susceptibility creates an opportunity for a particular immune response; it does not determine that the response must occur.

What Is Established—and What Remains a Hypothesis?

Several individual components of the rheumatoid-arthritis model are well established:

• PAD2 and PAD4 can citrullinate proteins.
• ACPAs are strongly associated with a major subgroup of rheumatoid arthritis.
• Some people with rheumatoid arthritis develop antibodies against PAD enzymes.
• Particular HLA-DRB1 variants increase susceptibility to rheumatoid arthritis.
• Citrullinated peptides can be presented to and recognised by T cells.
• B cells play important roles in rheumatoid arthritis, extending beyond antibody production to antigen presentation and immune signalling.

What is not yet established is that PAD–substrate complexes connect all these components in people in precisely the way that TG2–gluten complexes appear to do in coeliac disease.

The coeliac mechanism is supported by direct studies of disease-relevant B cells, T cells, TG2 and gluten. The rheumatoid-arthritis model is an informed mechanistic proposal built from several converging observations, but important links in the complete chain still require direct demonstration.

Is There a Foreign Trigger in Rheumatoid Arthritis?

In coeliac disease, the external antigen is known: gluten. Sollid’s comparison raises a provocative question. Could a foreign protein—perhaps originating outside the body—become citrullinated by a PAD enzyme and provide the T-cell stimulus that helps activate PAD-reactive B cells?

At present, no equivalent of gluten has been established for rheumatoid arthritis. Environmental and microbial influences are being investigated, but the paper’s argument should not be interpreted as proof that one particular food, bacterium or infection causes the disease.

The possibility of an unidentified foreign T-cell antigen is therefore a research hypothesis, not a clinical conclusion.

Why Compare These Two Diseases?

Coeliac disease and rheumatoid arthritis were not paired because they affect the same organ or because one necessarily causes the other. They were compared because they share an unusually specific set of immunological features:

1. Strong associations with particular HLA class II variants.
2. Immune recognition of post-translationally modified peptides.
3. Autoantibodies against the enzymes that create those modifications.
4. Important interactions between B cells and helper T cells.
5. The possibility that an enzyme–substrate complex connects otherwise separate immune targets.

Coeliac disease supplies the better-understood model. Rheumatoid arthritis provides a plausible—but not yet proven—parallel.

The significance of the paper lies not in claiming that the diseases are identical, but in asking whether the immune system may be making the same type of organisational error in two different biological settings.

Paper: Sollid LM. One mechanism, two diseases: Involvement of enzyme–substrate complexes in the pathogenesis of celiac disease and rheumatoid arthritis. Proceedings of the National Academy of Sciences. 2026;123(37).