LACK OF EVIDENCE FOR PROTEIN AA REACTIVITY IN AMYLOID DEPOSITS OF LATTICE CORNEAL DYSTROPHY AND AMYLOID

CORNEAL DEGENERATION PETER D. GOREVIC, M.D.

Stony Brook, New York

M E R L Y N M. R O D R I G U E S , M.D. Bethesda, Maryland

JAY H. KRACHMER, M.D. Iowa City, Iowa

C L A I R E G R E E N , M.S. Stony Brook, New York

SHIGEYOSHI FUJIHARA, M.D. Yamaguchi, Japan

AND

GEORGE G. GLENNER, M.D. San Diego, California

Amyloid fibrils occurring in primary and myeloma-associated (AL), secondary (AA), and certain neuropathic hereditary forms of systemic amyloidosis can be distinguished biochemically or immunohistologically as being composed of immunoglobulin light chain, protein AA, or prealbumin respectively. All types of systemic and several localized forms of amyloidosis contain amyloid P component (protein AP). We studied formalin-fixed tissue from eight cases of lattice corneal dystro­phy by the immunoperoxidase method using antisera to proteins AA and AP, to normal serum prealbumin and prealbumin isolated from a case of hereditary amyloidosis, and to light-chain determinants; addi­tional cases were examined by indirect immunofluorescence of fresh-frozen material. We found weak (1:10 dilution) staining with anti-AP, but no reactivity with other antisera. Congo red staining was resistant to pretreatment of sections with potassium permanganate, a character­istic of non-AA amyloid. Two-dimensional gels of solubilized proteins from frozen tissue from two cases of lattice corneal dystrophy resem­bled those obtained from normal human cornea. Western blots of two cases of polymorphous amyloid degeneration and solubilized protein from normal cornea did not react with radioactive iodine-labeled anti-AA or anti-AP with purified protein AP and unfixed protein AA amyloid tissue as controls. We were unable to corroborate the presence of protein AA in the amyloid deposits of lattice corneal dystrophy. Although staining with antiserum to protein AP was demonstrable, the molecular configuration of this protein in stromal deposits remains to be defined.

216 AMERICAN JOURNAL OF OPHTHALMOLOGY 98:216-224, 1984

VOL. 98, NO. 2 LACK OF PROTEIN AA IN AMYLOID DEPOSITS 217

Monospecific antisera to amyloid subu-nit proteins are useful reagents for the immunohistologic definition of various forms of systemic amyloidosis in tissue sections.1,2 Antibodies to amyloid A pro­tein (protein AA) and prealbumin specifi­cally stain deposits in all forms of secon­dary and certain neuropathic hereditary forms of systemic amyloidosis respective­ly, when assessed by immunofluores-cence or immunoperoxidase tech­niques.1 4 In addition to the fibril, all types of systemic, as well as several local­ized, forms of amyloidosis contain amy­loid P component (protein AP), and it has been suggested that the presence of this protein in deposits immunohistologically is as reliable an index of amyloidosis as classic tinctorial properties.5,6

Localized amyloid occurring in the cor­nea includes the disease entities of lattice corneal dystrophy7,8 and polymorphous amyloid degeneration.9 With the excep­tion of the neuropathic syndrome first described by Meretoja, lattice dystrophy is not accompanied by systemic amyloid deposits; autosomal dominant inheritance underscores the importance of genetic factors in the pathogenesis of this disor­der. By contrast, polymorphous amyloid degeneration has no heritability and oc­curs in an older population.9 Congo red binding of both forms of localized corneal

Accepted for publication May 24, 1984. From the Department of Medicine, State Universi­

ty of New York, Stony Brook, New York (Dr. Gorevic and Ms. Green); the Clinical Branch, National Eye Institute, Bethesda, Maryland (Dr. Rodrigues); the Corneal Service, Department of Ophthalmology, University of Iowa, Iowa City, Iowa (Dr. Krachmer); the Department of Medicine, Yamaguchi School of Medicine, Yamaguchi, Japan (Dr. Fujihara); and the Department of Pathology, University of California, San Diego, California (Dr. Glenner). This study was supported by grants GM31866 and AG01973 from the National Institutes of Health and by the HOR Foundation.

Reprint requests to Peter D. Gorevic, M.D., State University of New York, Health Sciences Center, HSC, T-16, 040, Stony Brook, NY 11794.

amyloid has been found to be resistant to pretreatment of tissue sections with di­lute potassium permanganate, a charac­teristic of non-A A amyloidosis.10,11 We tested for protein AA, prealbumin, and protein AP in lattice dystrophy and poly­morphous degeneration immunohistolog­ically and by western blot analysis of solubilized corneal proteins studied by two-dimensional gel electrophoresis using monospecific antisera to these pro­teins.

M A T E R I A L AND M E T H O D S

Rabbit antisera to human protein AA purified from cases of secondary amyloi­dosis were developed independently in two laboratories.1 Antisera to protein AP and prealbumin were obtained by immu­nizing rabbits with purified proteins iso­lated from amyloidotic tissue12 and pooled normal human plasma respectively. Anti-SKO is an antiserum that recognizes a variant prealbumin fragment present in one form of neuropathic hereditary amy­loidosis13 and cross reacts with prealbu­min present in diffuse ventricular senile cardiac amyloidosis (P. D. Gorevic, B. Frangione, and M. Pras, unpublished data). Anti-light chain antisera were ob­tained commercially and anti-Αλ and anti-Ακ to denatured amyloid fibrils were prepared from cases of light chain amyloi­dosis as previously described.1 Goat anti­sera to proteins AA and AP were also purchased.

Sensitivity of Congo red binding to pretreatment of sections with dilute po­tassium permanganate was determined by the method of Van Rijkswick and Van Heusden.10 Transmission electron mi­crography was done as described by Man-nis, Krachmer, and Rodrigues.9

We studied formalin-fixed material from eight cases of lattice dystrophy by the unlabeled antibody1 and indirect im­munoperoxidase methods and fresh-

218 AMERICAN JOURNAL OF OPHTHALMOLOGY AUGUST, 1984

frozen tissue from three cases by indirect immunofluorescence.14 Peroxidase-anti-peroxidase conjugates and horseradish peroxidase-conjugated antisera were pur­chased. Primary antisera were used ini­tially at 1:10 and 1:50 dilutions in 2% bovine serum albumin, and then undilut­ed if no reactivity was found. We also tested commercially available anti-AA and anti-AP antisera on fixed material by indirect immunofluorescence in two cases.15 Controls included normal human cornea, replacement of primary antisera with normal rabbit or normal goat serum or saline, and concomitant testing of pri­mary antisera at dilutions of 1:50 to 1:1,000 on known cases of human secon­dary (protein AA) amyloidosis or senile cardiac amyloidosis.

Frozen samples of cornea obtained at the time of keratoplasty (for lattice corne-al dystrophy) or from eye bank eyes (con­trols and cases of polymorphous amyloid degeneration) were cryostat-sectioned di­rectly or cut from OCT at —20 C. Indi­vidual samples for two-dimensional gel electrophoresis consisted of ten to 20-ΙΟμπι full-thickness sections ( 8 x 1 mm) transferred at —20 C to plastic centrifuge tubes precooled on dry ice. The samples were solubilized in 30 μΐ of a mixture of 9 M urea, 4% nonidet, and 2% mercap-toethanol buffer and frozen and thawed three times. Insoluble material was cen-trifuged out (13,000 rpm for five min­utes). We ran two-dimensional gels, using the system described by Anderson and Anderson16; the isofocusing dimen­sion included ampholines at a pH of 3.5 to 10 and a carbamylation train of creatine phosphokinase as an internal17 standard. The second dimension consisted of a 10% to 20% sodium dodecyl sulfate-polyacrylamide gradient slab gel run with low molecular weight standards. This molecular weight range (5,000 to 25,000 daltons) has proven most useful for the visualization of light-chain (AL),

protein AA, or prealbumin-related amy­loid subunit proteins.18 Individual gels were developed serially with Coomassie and silver stains.

Western blots were run onto amino-benzyloxymethyl paper as described by Renart, Reiser, and Stark.19 Portions of gels for transfer were run at 27 V, 1.2 amp for two hours, and unreacted sites blocked with a mixture of 0.1 M Tris, 10% ethanolamine, and 0.25% gela­tin buffer. Each blot consisted of the 10,000- to 30,000-dalton portions of two-dimensional gels of lattice dystrophy or polymorphous degeneration, normal human cornea, and either purified pro­tein AP12 or single 10-μιη sections of 2-mm blocks of frozen kidney tissue col­lected at autopsy from patients with known human protein AA amyloidosis processed as described. Blots were devel­oped with 1:100 dilutions of antisera fol­lowed by staphylococcal protein A or the IgG fraction of sheep anti-rabbit or anti-goat immunoglobulin labeled with radio­active iodine (125I, 1 or 2 μCi/μg) with Bolton-Hunter reagent. Each blot was incubated with 500,000 cpm of labeled antibody or staphylococcal protein A; autoradiograms were developed after ex­posure times of as much as 30 days.

R E S U L T S

Congo red stains of all cases of lattice dystrophy disclosed stromal and subepi-thelial deposits of amyloid. Transmission electron microscopy showed typical 10-nm fibrils to be the major constituent of the deposits seen by light microscopy (Fig. 1). Congo red binding and birefrin­gence were resistant to pretreatment of sections with dilute potassium permanga­nate (Fig. 1, inset).

Immunoperoxidase staining with anti­sera to protein AA (Fig. 2, inset), preal-bumin, kappa and lambda light chain determinants, prealbumin variant pro­tein SKO, and AL-lambda and AL-kappa

VOL. 98, NO. 2 LACK OF PROTEIN AA IN AMYLOID DEPOSITS 219

Fig. 1 (Gorevic and associates). Lattice corneal dystrophy. Amyloid filaments (asterisk) within the stromal deposit (x 58,500). Inset, Stromal amyloid deposits (asterisk) stained with Congo red after pretreatment with dilute potassium permanganate (hematoxylin and eosin, x 250).

proteins were negative in corneal tissue at all dilutions tested against positive controls in other organs (Fig. 2). Similar­ly, commercially available goat anti-AA antiserum was also negative when tested in two cases by immunoperoxidase stain­ing with a horseradish peroxidase-labeled anti-goat conjugate or by indirect iramu-nofluorescence with fluorescein isothio-

cyanate conjugates of antisera to the pri­mary antiserum on fixed material.15

Indirect immunofluorescence on fresh-frozen tissue from three additional cases of lattice dystrophy was also negative. Anti-AA antisera were tested undiluted and at dilutions of 1:10, 1:50, and 1:500 to exclude the possibility of prozone arti­fact. Controls included positive staining

220 AMERICAN JOURNAL OF OPHTHALMOLOGY AUGUST, 1984

Fig. 2 (Gorevic and associates). Positive control section of fixed autopsy tissue from a case of secondary amyloidosis, reacted with anti-AA antiserum at a 1:200 dilution, demonstrating specific deposition in glomerulus and afferent arteriole (immunoperoxidase reaction with hematoxylin counterstain, x 250). Inset, Corneal section from a case of lattice corneal dystrophy reacted with anti-AA antiserum (1:10 and undiluted) (X250).

of amyloid deposits from cases of known human protein AA amyloid (Fig. 2) and senile cardiac amyloidosis demonstrating specific deposition of stain at antiserum dilutions of 1:200 to 1:500.

Anti-AP gave weak positive staining of lattice dystrophy amyloid at a 1:10 dilu­tion by the peroxidase technique (Fig. 3, inset). This, however, was in marked contrast to specific staining obtained with cases of senile cardiac amyloid tested at dilutions as high as 1:1,000 (Fig. 3). Nor­mal rabbit serum (at the same dilutions) and saline controls were consistently neg­ative.

Two-dimensional gels of solubilized proteins from three normal corneas dis­closed a characteristic map, including

Coomassie and silver-staining spots with the charge and molecular weight coordi­nates of immunoglobulin heavy and light chains, actin, albumin, and trans-ferrin (Fig. 4, top and bottom). Two-dimensional gels of frozen material from two cases of lattice dystrophy did not differ significantly from controls (Fig. 5).

Sufficient frozen material was available from two cases of polymorphous amyloid degeneration for two-dimensional gel analysis and western blotting of replicate gels. Western blots were probed with antisera to protein AA and to protein AP sequentially, and included purified pro­tein P and solubilized proteins from tis­sue sections of known cases of human protein AA amyloidosis as internal con-

VOL. 98, NO. 2 LACK OF PROTEIN AA IN AMYLOID DEPOSITS 221

Fig. 3 (Gorevit and associates). Control section of fixed tissue from a case of diffuse ventricular senile cardiac amyloidosis, reacted with antiserum to protein AP at a 1:500 dilution, demonstrating parenchymal amyloid deposits between muscle fibers (immunoperoxidase reaction with hematoxylin counterstain, x 120). Inset, Cornea! section reacted with antiserum to protein AP at a 1.10 dilution (x 250).

dystrophy also studied by indirect immu-nofluorescence.

Wheeler and Eiferman,11 using the un-labeled antibody technique, found both protein AA (1:75 dilution) and protein AP (1:100 dilution) reactivity in stromal de­posits of three cases of lattice dystrophy studied, with no epithelial staining. This study used commercially available anti-sera and did not include positive controls, but did note blocking of reactivity by absorption of antibodies with purified proteins AA and AP. These results could not be reproduced by Klintworth and associates22 who used antisera obtained from the same commercial source, evalu­ated by indirect immunoperoxidase with a horseradish peroxidase-conjugate to the

trois. In both instances, autoradiograms of both normal cornea and polymorphous degeneration were consistently negative.

DISCUSSION

In 1980, Mondino and associates20 re­ported positive staining with anti-AA (1:2 and 1:10 dilutions) and anti-AP (1:4 and 1:10 dilutions) antisera of a case of lattice corneal dystrophy obtained at keratoplas-ty and studied by indirect immunofluo-rescence. This study did not include ab­sorption controls and disclosed similar positive fluorescence of the epithelial cell layer. A later report21 noted positive staining with anti-AP, but not with anti-AA, of amyloid deposits in the failed corneal graft of a case of familial lattice

222 A M E R I C A N J O U R N A L O F O P H T H A L M O L O G Y AUGUST, 1984

BASIC

30 25

ftCTIN

20

**···**·

ALBUMIN

15

'"-u

10

*'

·*· ·

5

m ~

1

*·»

■ *

*

TRANSFERRIN

I g HEAVY CHAIN

* - * - I g LIGHT CHAIN

Fig. 4 (Gorevic and associates). Two-dimensional gels of solubilized proteins from normal human cor­neas. Top, Positions of albumin, transferrin, and immunoglobulin (Ig) heavy and light chains. Asterisk indicates spots with molecular weight and charge (pi) coordinates of basic corneal protein. Bottom, Rela­tive molecular weight and charge (pi) coordinates. Note positions (1 to 30) of the creatine phosphokinase carbamylation train included as an internal charge standard. Also indicated are the charges of actin and albumin.

primary goat antiserum. Similarly, we were unable to demonstrate protein AA reactivity in stromal amyloid deposits in lattice dystrophy and polymorphous de­generation with monospecific antisera de­veloped independently in two different laboratories by either immunofluores-cence or peroxidase techniques.

Two-dimensional gels of solubilized proteins from two cases of lattice dystro-

Fig. 5 (Gorevic and associates). Two-dimensional gel of one of two cases of lattice dystrophy studied. T K , creatine phosphokinase.

phy and western blot analysis of two-dimensional gels from two cases of poly­morphous amyloid degeneration also failed to show protein spots with the charge and molecular weight coordinates of proteins AA and AP. This method is based on the observation that protein AA, light-chain (AL), and prealbumin-related subunit proteins have molecular weights in the range of 5,000 to 25,000 daltons under dissociating and reducing condi­tions and is applicable to small biopsy specimens, obviating the need for lengthy extraction protocols.18 Its suita­bility for the analysis of corneal proteins, however, is limited by the high fibrous content of this tissue, and we cannot exclude the possibility that amyloid fi­brils in cornea are uniquely resistant to dissociating agents. Further studies will be needed to evaluate for the presence of higher molecular weight forms of protein AP, as has been reported for glomerular basement membrane.23 Our studies, how­ever, make it unlikely that protein AA is a significant constituent of corneal stromal amyloid in a molecular configuration shown for other tissues involved by pro­tein AA amyloidosis.

VOL. 98, NO. 2 LACK OF PROTEIN AA IN AMYLOID DEPOSITS 223

Two-dimensional gels of both normal cornea and lattice dystrophy had spots with molecular weight and charge coordi­nates of serum albumin, IgG heavy and light chains, and transferrin24 confirmed by western blot analysis with monospeci-fic antisera to these proteins (P. D. Gore-vie, unpublished data). An additional grouping was noted with coordinates cor­responding to bovine corneal protein 54, which has been shown to constitute as much as 30% of soluble protein of bovine cornea.25 These findings demonstrate the feasibility of solubilizing corneal proteins directly from frozen tissue and confirm observations by others that several nor­mal serum proteins constitute a signifi­cant percentage of stromal proteins solu­ble under physiologic conditions.26,27

Antiserum to protein AP gave only weak immunoperoxidase staining of stro­mal deposits, and western blots of both lattice dystrophy and normal cornea were negative, with purified protein AP (subu-nit molecular weight, 23,000 daltons) as an internal control. Further studies will be needed to evaluate for additional "hid­den" antigenic determinants, for the full demonstration of protein AP as has been noted after mild trypsinization of tissue sections.28

Immunohistologic studies of cor­nea,21,29 as well as direct extraction of amyloid fibrils from involved kidney,30

have failed to demonstrate protein AA in the familial forms of corneal amyloidosis associated with systemic amyloid depos­its. Since protein A A has so far been demonstrated only in systemic amyloido­sis accompanying chronic inflammatory states and in amyloidosis complicating familial Mediterranean fever, its pres­ence in localized corneal amyloid, but not corneal amyloid with systemic amyloido­sis, requires explanation. Our studies, as well as those of others,22 suggest that direct isolation and biochemical charac­

terization of stromal deposits occurring in the corneal dystrophies will be necessary to establish the relation of localized forms of amyloid to the systemic amyloidoses.

R E F E R E N C E S

1. Fujihara, S., Balow, ). E., Costa, J. C , and Glenner, G. G.: Identification and classification of amyloid in formalin-fixed, paraffin-embedded tissue sections by the unlabeled immunoperoxidase meth­od. Lab. Invest. 43:358, 1980.

2. Shirahama, T., Skinner, M., and Cohen, A. S.: Immunocytochemical identification of amyloid in formalin-fixed sections. Histochemistry 72:161, 1981.

3. Benson, M. D.: Partial aminoacid sequence ho-mology between a heredofamilial amyloid protein and human plasma prealbumin. J. Clin. Invest. 67:1035, 1981.

4. Dalakas, M. C , and Engel, W. K.: Amyloid in hereditary amyloid polyneuropathy is related to pre­albumin. Arch. Neurol. 38:420, 1981.

5. Pepys, M. B., Baltz, M. L., De Beer, F. C , Dyck, R. F . , Holford, S., Breathnach, S. M., Black, M. M., Tribe, C. R., Evans, D. J., and Feinstein, A.: Biology of serum amyloid P component. Ann. N.Y. Acad. Sei. 389:286, 1982.

6. Westermark, P., Shirahama, T., Skinner, M., Noren, P., and Cohen, A. S.: Amyloid P-component (protein AP) in localized amyloidosis as revealed by an immunocytochemical method. Histochemistry 71:171, 1981.

7. Smith, M. E., and Zimmerman, L. E.: Amy­loid in corneal dystrophies. Differentiation of lattice from granular and macular dystrophies. Arch. Oph-thalmol. 79:407, 1968.

8. Klintworth, G. K. : Lattice corneal dystrophy. Am. J. Pathol. 50:371, 1967.

9. Mannis, M. J., Krachmer, J. H., and Rodri-gues, M. M.: Polymorphic amyloid degeneration of the cornea. Arch. Ophthalmol. 99:1217, 1981.

10. Van Rijswick, M. H., and Van Heusden, E. W. G. J.: The potassium permanganate method. Am. J. Pathol. 97:43, 1979.

11. Wheeler, G. E., and Eiferman, R. A.: Immu-nohistochemical identification of die AA protein in lattice dystrophy. Exp. Eye Res. 36:181, 1983.

12. Pepys, M. B., Dash, A. C , Munn, E. A., Feinstein, A., Skinner, M., Cohen, A. S., Gewürz, H., Osmand, A. P., and Panter, R. H.: Isolation of amyloid P component (protein AP) from normal serum as a calcium-dependent binding protein. Lan­cet 1:1029, 1977.

13. Pras, M., Franklin, E. C , Prelli, F. , and Frangione, B.: A variant of prealbumin from amyloid fibrils in familial polyneuropathy of Jewish origin. J. Exp. Med. 154:489, 1981.

14. Johnson, G. D., Holborow, E. J., and Durl-ing, J.: Immunofluorescence and immunoenzyme techniques. In Weir, D. M. (ed.): Immunochemis-

224 AMERICAN JOURNAL OF OPHTHALMOLOGY AUGUST, 1984

try. Oxford, Blackwell Scientific Publications, 1978, vol. 1, pp. 1-30.

15. Dorsett, B. H., and Ioachim, H. L. : A meth­od for the use of immunofluorescence on paraffin-embedded tissues. Am. J. Clin. Pathol. 69:66, 1978.

16. Anderson, N. G., and Anderson, N. L. : Ana­lytical techniques for cell fractions. XXI. Two dimen­sional analysis of serum and tissue proteins. Anal. Biochem. 85:331, 1978.

17. Anderson, N. G., and Hickman, B. L. : Ana­lytical techniques for cell fractions. XXIV. Isoelectric point standards for two dimensional gel electropho-resis. Anal. Biochem. 93:312, 1979.

18. Turner, W. J., Cleveland, A. B., and Gorevic, P. D. : A novel method for the typing of amyloid in small biopsy specimens. Western blot analysis of two dimensional gels. Clin. Res. 31:355A, 1983.

19. Renart, J., Reiser, J., and Stark, G. R.: Trans­fer of proteins from gels to diazobenzyloxymethyl paper and detection with antisera. A method for studying antibody specificity and antigen structure. Proc. Natl. Acad. Sei. U.S.A. 76:3116, 1979.

20. Mondino, B. J., Raj, C. V. S., Skinner, M., Cohen, A. S., and Brown, S. I.: Protein AA and lattice corneal dystrophy. Am. J. Ophthalmol. 89:377, 1980.

21. Mondino, B. J., Rabb, M. F., Sugar, J., Rat, C. V. S., and Brown, S. I.: Primary familial amyloi­dosis of the cornea. Am. J. Ophthalmol. 92:732, 1981.

22. Klintworth, G. K., Ferry, A. P., Sugar, A., and Reed, J.: Recurrence of lattice corneal dystrophy types in the corneal grafts of two siblings. Am. J. Ophthalmol. 94:540, 1982.

23. Dyck, R. F., Lockwood, C. M., Kershaw, M., McHugh, N., Duance, V., Blatz, M. L., and Pepys, M. B.: Amyloid P component is a constituent of normal human glomerular basement membrane. J. Exp. Med. 152:1162, 1980.

24. Anderson, L., and Anderson, N. G.: High resolution two dimensional electrophoresis of human plasma proteins. Proc. Natl. Acad. Sei. U.S.A. 74:5423, 1977.

25. Alexander, R. J., Silverman, B., and Henley, W. L. : Isolation and characterization of BCP 54. The major soluble protein of bovine cornea. Exp. Eye Res. 32:205, 1981.

26. Holt, W. S.,andKinoshita, J. H.: The soluble proteins of the bovine cornea. Invest. Ophthalmol. 12:114, 1973.

27. Allansmith, M. R., Whitney, C , and McClel-lan, B.: Immunoglobulins in the eye. Location, type and amount. Arch. Ophthalmol. 89:36, 1973.

28. Hind, C. R. K., Tennent, G. A., Evans, D. J., and Pepys, M. B.: Demonstration of amyloid A (AA) protein and amyloid P-component (AP) in deposits of systemic amyloidosis associated with renal adenocarcinoma. J. Pathol. 139:159, 1983.

29. Purcell, J. J., Rodrigues, M., Chishti, M. I., Riner, R. N., and Dooley, J. M.: Lattice corneal dystrophy associated with familial systemic amyloido­sis (Meretoja's syndrome). Ophthalmology 90:1512, 1983.

30. Meretoja, J., Hollmen, T., and Penttinen, R.: Partial characterization of amyloid proteins in inherit­ed amyloidosis with lattice corneal dystrophy and secondary amyloidosis. Med. Biol. 56:17, 1978.