A variety of proteins are induced in all cells on stress, such as heat shock. This chapter will focus on one particular family of structurally conserved heat shock proteins, which includes GroEL in Escherichia coli, hsp 65 in Mycobacterium bovis, and hsp 60 in mammals. These proteins are termed chaperonins, corresponding to their role in assisting the folding and assembly of other proteins. Members of this family of proteins are highly homologous. For example, there is about 50% identity in amino acid sequence between human hsp 60 and M. bovis hsp 65 (1).
According to the molecular mimicry theory (2,3), it is possible that antibodies and/or T lymphocytes made in the immune response to a specific epitope found on a component of the pathogen may also bind to a structurally similar host determinant, thereby leading to autoimmune damage. This mimicry may be due to the presence of a common epitope in homologous proteins, such as the host and pathogen members of the hsp 60 family. Because of their high degree of homology across the broad stretches of phylogeny and their ubiquity, heat shock proteins represent a logical protein family in which to seek examples of molecular mimicry. As previously reviewed (4), it has been hypothesized that molecular mimicry between host and pathogen heat shock proteins constitutes an important physiological element of the protective immune response, which for unknown reasons may go awry in some individuals and may contribute to autoimmune disease. Another possibility, presented below, is molecular mimicry between unrelated pathogen and host proteins, one of which may be a heat shock protein. Sequence similarities between suspected autoimmune antigens and human hsp 60 have been demonstrated in several autoimmune diseases using an interesting but speculative database search paradigm (5).
One example of heat shock protein autoimmunity is the adjuvant arthritis induced by inoculation of mycobacteria into certain susceptible animals (7). A series of reports focused on a particular T-cell clone derived from rats with adjuvant arthritis that recognized a protein moiety of cartilage proteoglycan; the disease could be adoptively transferred by these cells (7-9). Furthermore, the cells strongly and specifically responded to a peptide comprising residues 180-188 of mycobacterial hsp 65 (10). Interestingly, administration of hsp 65 in oil did not induce arthritis but could provide resistance (11). Corresponding purported examples of arthritis in humans based on cross reactivity with bacterial heat shock proteins have been reviewed (12).
Another prominent example of heat shock protein involvement in autoimmunity is insulin-dependent diabetes mellitus (IDDM) of the nonobese diabetic (NOD) mouse strain. IDDM of NOD mice develops spontaneously (13). Appearance of both T-cell reactivity and antibodies that react with mycobacterial hsp 65 precedes overt diabetic symptoms (14,15). Infusion of T cells obtained from mice with advanced insulitis into prediabetic recipient mice produces severe disease within 1 week (14,15). Furthermore, injection of hsp 65 into 1-month-old mice induced diabetic changes well before disease would have manifested spontaneously (15). In contrast, when 1-month-old NOD mice were vaccinated with attenuated reactive T cells, resistance to both spontaneous and induced diabetes occurred (16). Finally, vaccination with a peptide comprising residues 437-460 of human hsp 60 could prevent diabetes in this model (17).
The earliest reactivity with an autoantigen in NOD mice (18,19) as well as in human IDDM (20,21) is believed to be with the 64-kDa protein, glutamatic acid decarboxylase (GAD). Autoimmunity to GAD may be a necessary but not a sufficient condition for disease progression. The initial insult of GAD immu-noreactivity may lead to a cascade of reactivity with autoantigens, including hsp 60, thereby producing the diabetic state (19). Interestingly, sequence homology between GAD and hsp 60 has been noted (5), and the initial T-cell determinant of GAD (18) is within this region of homology.
In both IDDM of NOD mice and adjuvant arthritis, a correlation between autoimmune disease and members of the hsp 60 family has been suggested. We have hypothesized that an autoimmune response to hsp 60 underlies the neurological symptoms of chronic Lyme disease (22). Rather than molecular mimicry between pathogen and host heat shock proteins or between pathogen heat shock protein and an otherwise unrelated host protein, we have found that antibodies to the 41-kDa flagellin of Borrelia burgdorferi cross react with human hsp 60. Our studies supporting this hypothesis are now presented.
Was this article helpful?