By Sergei A. Grando, M.D., Ph.D., D.Sci.
Professor of Dermatology
University of California Davis
NPF Advisory Board Member
The goal of my research is to develop a safer and more rational treatment for pemphigus. I am deeply concerned that we, as physicians caring for patients with pemphigus, have to accept the risk of severe side effects related to the use of long term, high dose corticosteroid therapy.
Despite recent progress in developing nonhormonal therapy for other autoimmune conditions, the treatment of pemphigus remains largely dependent on corticosteroid hormones. The lack of progress in developing new therapies for pemphigus is ironic because we thought we understood the basic mechanisms responsible for the development of this disease. But, perhaps our understanding was wrong and possibly this misunderstanding has hampered advancement in treatment.
Traditionally, dermatologic research workers have held that pemphigus occurs when one's own antibodies attack and disrupt the desmosomal attachment points of epidermal keratinocytes–the cells comprising the superficial layer of the skin wherein the blisters emerge. It is further believed that these autoantibodies specifically bind to, and block the function of, the desmosomal adhesion molecules of the desmoglein family. Then, when these desmosomal attachment points fail, ker-atinocytes separate allowing interstitial fluid to infiltrate into the space around them. Clinically this is reflected by the formation of fragile blisters which quickly break down leaving the red, raw erosions that we recognize as the typical lesions of pemphigus.
Corticosteroids, such as Prednisone, are thought to control pemphigus by inhibiting the ability of lymphocytes to produce the anti-desmoglein autoantibodies. Although this is the currently accepted view of pemphigus and its treatment, my research leads me to conclude that these traditional beliefs are too simplistic and that they may even be erroneous.
Lesson one: Antibodies to desmoglein 3 may not cause pemphigus vulgaris We have primarily studied the variety of pemphigus known as pemphigus vulgaris. In this disease the antibodies specifically and directly responsible for the development of blisters are currently believed to be directed against the desmosomal adhesion molecule, desmoglein 3. However, I now believe that antibodies to desmoglein 3 may be only incidental and that antibodies directed to other antigens may be more directly responsible for causing the disease. Several lines of evidence, both from our laboratory and from other laboratories, support this belief.
First, desmoglein 3 is localized solely to the desmosomal portion of the keratinocyte cell membrane1, whereas in electron microscopic and immunofluorescent studies, the antibodies present in patients with pemphigus vulgaris bind to many other sites on the keratinocyte cell surface.2, 3 Second, pemphigus vulgaris antibodies bind not only to a 130 kD protein (the molecular weight of desmoglein 3) but also to proteins of various other weights.4-12 Third, in the laboratory, antibodies which bind to a non-desmoglein protein of 66 kD, can cause pemphigus-like blisters when they are injected into mice.9
Fourth, antibodies to desmoglein 3 are found in many relatives of patients with pemphigus but the presence of these antibodies is not associated with the development of blisters in these relatives.13 Fifth, neonatal mice injected with antibodies specifically directed to desmoglein 3 do not develop pemphigus-like lesions as one would expect if these antibodies were the actual cause of the disease.14 Sixth, mice which are bred to have defective15, or absent16, 17, desmoglein 3 do not spontaneously develop pemphigus lesions on the skin as would be expected if this adhesion molecule were solely responsible for the cause of pemphigus vulgaris.
In addition, in my laboratory, we have developed collateral evidence which argues against a causative role for antibodies to desmoglein 3 in the development of pemphigus.18 Specifically, we are able to create pemphigus-like blisters in mice lacking desmoglein 3 ("desmoglein 3 knockout mice") when we inject them with serum from patients with pemphigus vulgaris. Since these mice possessed no desmoglein 3, some other antigen(s) must have been the target of the disease-causing pemphigus antibodies. It could be argued that the antibodies from these patients cross reacted with desmoglein 1 (the antigen said to be responsible for pemphigus foliaceus) and that this cross reaction caused the blisters but we tested for the presence of anti-desmoglein 1 activity in the serum used in these experiments and found none.
Taken altogether, these data indicate that antibodies to a protein other than desmoglein 3 are responsible for the development of blisters in patients with pemphigus vulgaris.
Lesson two: identification of disease-causing antibodies to antigens other than desmoglein 1 and 3 in the serum of patients with pemphigus vulgaris. Recently, our laboratory has identified a keratinocyte-derived, non-desmoglein, protein that can absorb the disease-causing antibody from the serum of patients with pemphigus vulgaris.19 The potential to create blisters in mice can be restored by adding this antibody back to the serum. Both normal and desmoglein 3 knockout mice develop blisters. Since desmoglein 3 knockout mice were bred have no desmoglein 3, and since we documented that there was no anti-desmoglein 1 antibody in the serum, we conclude that the blister-causing antibody is not directed toward a desmoglein protein.
However, this non-desmoglein antibody, when administered alone, is not sufficient to cause blisters in mice. Thus we believe that there are "primary" and "secondary" antibodies which must act cooperatively to cause pemphigus. The primary antibody initiates the pathological process leading to blisters. The secondary antibody is produced by the body in order to get rid of the cell debris left after assault by the disease-causing primary antibody.
We then attempted to observe under the microscope how this non-desmoglein, primary antibody initiates a process of keratinocyte separation from one another and rounding up which is termed "acantholysis" and is unique to pemphigus. When we placed this non-desmoglein antibody with cultured keratinocytes we noted that the keratinocytes began to shed desmosomes and associated small portions of plasma membrane from their surface. This suggests to us that binding of this antibody to the cell surface alters the stability of the keratinocyte plasma membrane, which then breaks up releasing desmosomes into the intercellular space. From these observations we have developed a new hypothesis regarding the pathophysiology of blister formation in pemphigus vulgaris.
Hypothesis one: Blister formation in pemphigus occurs in two phases. First, disease causing, primary, antibodies bind to non-desmoglein antigens. This antibody binding causes injury to the cell membrane and subsequent shedding of desmosomes into the intercellular space. Second, during the next phase, these detached desmosomes expose desmoglein antigens. These antigens then lead to the formation of secondary, anti-desmoglein antibodies. Thus while desmoglein antibodies are not primarily responsible for the blister formation, they may play a role in the scavenging of the defective, shed desmosomes.
Based on this hypothesis, the treatment of pemphigus will be efficient if it can: 1) stop production of primary antibodies; or 2) protect keratinocytes from the deleterious effects of primary antibody binding. The first goal can be achieved by developing "anti-pemphigus vaccine" using as immunogen the keratinocyte protein targeted by primary antibody. To achieve the second goal, one need first to elucidate the mechanism of acantholysis.
Lesson three: Corticosteroid therapy improves pemphigus by a novel pathway. As indicated above, conventional wisdom suggests that corticosteroids such as Prednisone improve pemphigus by suppressing lymphocyte production of disease-causing antibodies. However, in 1983, Swanson and Dahl first demonstrated that methylprednisolone (the activated form of Prednisone) added to skin cultures prevented acantholysis that would otherwise occur when pemphigus antibodies were added to the culture.20 In 1984, the Dr. Ahmed research group reported that similar results can be obtained using the corticosteroid hormone called hydrocortisone.21 Since in either experiment, lymphocytes were not present in the skin cultures treated with corticosteroids, it was clear that the Prednisone was not working by way of acting upon lymphocytes and was instead directly affecting the keratinocytes themselves. This explains why antilymphocyte immunosuppressive drugs such as Imuran (azathioprine), Cytoxan (cyclophosphamide) or Sandimmune (cyclosporin) cannot control pemphigus on their own.
In our laboratory, we now find that corticosteroids stimulate keratinocytes to manufacture the increased amounts of the same protein that is destroyed upon binding by our novel primary pemphigus antibody.22 Normally, this novel keratinocyte protein acts as a receptor binding to a signaling molecule, acetylcholine, which plays essential role in stimulating keratinocytes to maintain their shape and connection to each other. Similar proteins have been found in other tissues and shown to mediate therapeutic effects of corticosteroid by inhibiting the pathways of local inflammation that destabilize the plasma membranes. When a primary antibody parks at the receptor in acetylcholine's place, it initially activates the receptor, but then causes it to stop working due to degradation of the improperly functioning receptor through a process termed receptor "desensitization." What then occurs, the signal to stretch is broken, the desmosomes are shed, and the keratinocyte balls up (acantholysis). These observations led us to formulate a new hypothesis regarding the treatment of pemphigus.
Hypothesis two: Corticosteroids are effective in the treatment of pemphigus, at least in part, because they stimulate replacement of the acetylcholine receptor destroyed by the primary antibody, thus counterbalancing the effect of disease-causing antibody on keratinocytes. This stabilizes the keratinocyte plasma membrane, and, consequently, leads to regrowth of desmosomes. Since this response occurs directly at the keratinocyte level (that is, it is not mediated by way of lymphocytes), we believe that it may be possible to achieve that same effect from the use of non-corticosteroid drugs that can prevent primary antibody binding to keratinocytes.
Testing of this hypothesis is currently underway. We have already identified several nonhormonal drugs that can abolish pemphigus antibody-induced acantholysis in keratinocyte cultures.23 One of them is an acetylcholine-like molecule called pilocarpine. Pilocarpine can mimic both acetylcholine and primary antibody in a sense that it can attach the same protein on the keratinocyte plasma membrane. In contrast to the primary antibody, however, pilocarpine binding should not lead to receptor destruction. The Institutional Review Board (IRB) of the University of California Davis Medical School has approved our request to evaluate the clinical effectiveness of a topically applied gel containing pilocarpine. This medication, PILOPINE HSÒ GEL, is already approved by the Food and Drug Administration for placement in the eyes as part of the treatment of glaucoma.
We are now enrolling in this study patients with pemphigus. To be eligible, patients must have two or more active lesions. Patients who are already receiving other therapy will be allowed to continue with their current treatment. Patients who have not yet started treatment must have disease mild enough so that there is no problem delaying the initiation of other treatment for about two weeks.
Eligible patients participating in the study will need to attend the Dermatology Clinic at UC Davis Medical Center on two occasions. Lesions of pemphigus will be photographed and samples of blood and tissue (skin biopsy) will be taken. Patients will be given two un-labelled, identical appearing gels. One gel will be PILOPINE HSÒ GEL and the other will be a placebo (inactive) gel. Each gel will be applied to a separate lesion on a once daily basis for two weeks. During the two week application period, no changes will be allowed in other medications which are being taken for the treatment of pemphigus. To learn more about this study, you may contact me by letter, by telephone (916-734-6057), or by email (firstname.lastname@example.org), or write:
Sergei A. Grando, M.D., Ph.D., D.Sci.
Professor of Dermatology University of California Davis
Department of Dermatology
Ambulatory Care Center
4860 Y Street, Suite #3400
Sacramento, CA 95817
1. Karpati, S., et al. J Cell Biol 122, 409-15 (1993).
2. Wolff, K. & Schreiner, E. Nature 229, 59-61 (1971).
3. Bedane, C., et al. Arch Dermatol Res 288, 343-52 (1996).
4. Ablin, R.J., et al. J Hyg Epidemiol Microbiol Immunol 13, 321-9 (1969).
5. Shu, S.Y. & Beutner, E.H. J Invest Dermatol 61, 270-6 (1973).
6. Miyagawa, S., et al. Acta Derm. Venereol. 57, 7-13 (1977).
7. Diaz, L.A., et al. J Immunol 124, 760-5 (1980).
8. Murahata, R.I. & Ahmed, A.R. Arch Derm Res 275, 118-23 (1983).
9. Peterson, L.L. & Wuepper, K.D. J Clin Invest 73, 1113-20 (1984).
10. Acosta, E. & Ivanyi, L. Br J Dermatol 112, 157-64 (1985).
11. Lyubimov, H., et al. Israel J Med Sci 31, 42-8 (1995). 12. Joly, P., et al. J Invest Dermatol 108, 469-75 (1997).
13. Mohimen, A., et al. Arch Derm Res 285, 176-7 (1993). 14. Amagai, M., et al. J Clin Invest 90, 919-26 (1992).
15. Allen, E., et al. J Cell Biol 133, 1367-82 (1996).
16. Koch, P.J., et al. J Cell Biol 137, 1091-102 (1997).
17. Montagutelli, X., et al. J Invest Dermatol 109, 324-8 (1997).
18. Nguyen, V.T., et al. Arch Dermatol 134, 971-80 (1998). 19. Nguyen, V.T., et al. J Invest Dermatol 110, 486 (1998). 20. Swanson, D.L. & Dahl, M.V. J Invest Dermatol 81, 258-60 (1983).
21. Jeffes, E.W.d., et al. J Clin Lab Immunol 4, 359-63 (1984).
22. Nguyen, V.T., et al. J Invest Dermatol (submitted).
23. Grando, S.A. & Dahl, M.V. J Eur Acad Dermatol Venereol 2, 72-86 (1993).