B23 Nucleophosmin in Hepatocellular Cancer

Antony Rosen, M.D.

Johns Hopkins

Funded in December, 2002: $300000 for 3 years


back to top

Seeking Means to Block an Autoimmune Response

Autoimmune diseases occur when the immune system mistakes the body’s own cells as “foreign” and attacks them.  Systemic autoimmune diseases, such as systemic lupus erythematosus, systemic sclerosis, and autoimmune myositis, all target a specific “autoantigen” (called B23).  Autoantigens are molecules within the body's own tissues that the immune system misidentifies as foreign and attacks.  There is something about the autoantigen that is intrinsically important in initiating the disease process.  The researchers hypothesize that this intrinsic factor is a unique conformation, a cleavage.  Interestingly, B23 is also present in about one-third of patients with liver cancer, and this provides an opportunity to study the autoantigen in biopsy material from these liver cancer patients.  The investigators suspect the pieces of B23 generated by this cleavage that are identified by helper T cells as “foreign.”  The helper T cells then assist immune B cells in making an antibody to this autoantigen, creating an autoimmune response.

The researchers know how to isolate the B23 autoantigen from liver cancer specimens and can now study immune cell responses to various forms of this autoantigen.  Rosen has already documented several clear differences in the B23 protein as it occurs in liver cancer tissue. The research can now turn to what is likely to be the critical problem in autoantibody formation: how is the helper T cell recruited to help the B cell make the autoantibody?  This research may lead to means to prevent the initial event that lead autoantibodies to attack in systemic sclerosis, dermatomyositis, and myositis, and also to means to enhance the way that treatment is provided to those patients with liver tumor who have this autoantibody.


back to top

B23 Nucleophosmin in Hepatocellular Cancer

The association of a specific autoantibody response with distinct disease phenotypes is observed in both autoimmune diseases and cancer.  Although the mechanisms underlying such an association remain unclear, it is likely that unique properties of disease-specific autoantigens that are expressed in the relevant target cells within the immunizing microenvironment play an important role. 

It has recently been observed that the majority of autoantigens targeted across the spectrum of systemic autoimmune diseases (but not non-autoantigens) are selectively cleaved by the cytotoxic lymphocyte granule protease granzyme B, generating unique fragments not observed during other forms of cell death.  In spite of the strength and specificity of this association, the relevance of susceptibility to cleavage by granzyme B has remained unclear.  Possibilities include a role for cleavage by granzyme B (perhaps altering the subsequent processing of the antigen), or a role for the granzyme B cleavage site itself in influencing antigen loading and presentation. 

Our recent studies using hepatocellular carcinoma (HCC) and the HCC autoantigen, nucleophosmin/B23, as a model system have allowed us to define unique structural features of this molecule expressed exclusively in the relevant disease microenvironment.  Thus, B23 is selectively susceptible to cleavage by granzyme B in extracts of neoplastic, compared to non-neoplastic liver.  Additional studies in vitro and in tumors demonstrated that this increased susceptibility to cleavage by granzyme B arise when B23 translation is initiated at methionine7, generating a unique SDS-stable conformation in which the granzyme B cleavage site loop is exposed. The recent observation that the presence or absence of single protease cleavage sites can completely alter the final processing of that antigen, and determine which peptides are loaded onto MHC class II molecules for presentation to T cells is therefore particularly relevant.

We propose that the unique conformation and protease cleavability of tumor B23 provides the first opportunity to directly address the processing of, and immune responses to, the relevant immunogenic form of an autoantigen, and to define the function of the granzyme B cleavage loop in antigen processing and presentation. The proposed studies will therefore define whether the N-terminally truncated form of B23 (which assumes an SDS-stable oligomeric structure, and exposes an otherwise cryptic protease cleavage site) is processed differently to the wild-type form of the protein (which is expressed under almost all other circumstances).  The different forms of the immunogen will be produced in vitro, and T cell clones recognizing B23 will be generated from patients with HCC and B23 autoantibodies. Further understanding of the mechanism(s) responsible for determining such changes in conformation may have practical implications in both the therapeutic modulation of this response in systemic autoimmune disease and enhancement of this response in tumor immunotherapy. 


back to top
Antony Rosen, M.D.

Antony Rosen, M.D., completed his clinical training at the University of Cape Town in South Africa in 1984, prior to pursuing postdoctoral studies in Immunology at the Rockefeller University in New York. He subsequently did residency in Internal Medicine and fellowship in Rheumatology at Johns Hopkins in Baltimore, where he joined the faculty in 1994. His research focuses on the mechanisms of human autoimmune diseases, with particular emphasis on defining the changes in the structure of autoantigens that occur in the target tissues during cell damage and death.


back to top

Lay Results:
These studies have demonstrated that the form of the B23 molecule found in cancer is strikingly altered in terms of its structure, so that pieces of the molecule previously available are completely hidden, making the molecule resistant to the enzymes that cut it up into the small pieces when it is presented to the immune system. The signature of the cancer form of the molecule therefore appears very different from that of the normal form of the molecule, apparently enabling the immune system to respond. Studies to define the changes that affect this molecule in autoimmune diseases like SLE and scleroderma and to understand how these influence the immune response are presently underway. In the future, preventing these changes may be useful to prevent autoimmunity from beginning in patients at risk.

Scientific Results:
These studies have demonstrated that the tumor form of B23/nucleophosmin, although having a minor deletion of 6 amino acids at the N-terminus, is strikingly different conformationally from the wild-type form of the protein. Deuterium exchange mass-spec (DXMS) studies have demonstrated that the tumor form has several areas of decreased mobility/surface access, particularly around one of the granzyme B cleavage sites. This was precisely the opposite effect of that predicted and will provide critical insights in guiding the next set of experiments.  Numerous autoantigens have been demonstrated to be specific substrates for granzyme B, although the relevance of this property to immunogenicity remains unclear. The direct demonstration that one of the granzyme B cleavage site in B23 becomes obscured in the tumor form of the antigen is therefore of great interest. It suggests that autoantigen conformation around the granzyme cleavage site may play a role in antigen processing and epitope selection in tolerance induction and autoimmunity. Ongoing studies are addressing these effects on antigen processing of various autoantigens which are granzyme B substrates, including several SLE, myositis and RA autoantigens.