Concurrent Neuroborreliosis and Alzheimer’s Disease: Analysis of the Evidence

M. A PAPPOLL4. MD, R. OMAR, MD, PHD, B. SARAN, MD, A ANDORN, MD, M. SUAREZ, MD, C. PAVIA, PHD, A WEINSTEIN, MD, D. SHANK MS, K. DAVIS, MD, AND W. BURGDORFER, PHD

Several recent reports have claimed a possible association be- perimental transmissions of familial AD (Gerstmann- tween Bowelia burgdotferi infection and Alzheimer’s disease Straussler-Skinker syndrome) as a process resembling (AD). Herein, we describe our search for additional evidence of a spongiform and by the occurrence neuroborreliosis in AD. Brain tissue from neuropathologically of neurofibrillary tangles in various neurologic disor- confirmed cases of AD was cultured for B burgdotferi using stan- ders of known infectious etiology.5-g Unfortunately, dard microbiologic methods. Material derived from culture was efforts to transmit sporadic AD or to detect microbial further examined using electron microscopy, direct immunoflu- orescence, and acridine orange fluorescence. Previous studies antigens or nucleic acid sequences in AD brain tissue have shown high titers of antiborrelia antibodies in CSF in all have been disappointing.‘O cases of confirmed neuroborreliosis; therefore, we tested CSF Recently, several reports have claimed a possible from neuropathologically confirmed cases of AD by indirect im- association between burgdoqeri infection of munofluorescence and enzyme-linked immunoassay. In addition, the CNS and AD.“-14 It was proposed that these spi- imprint preparations from AD and control brain tissues were rochetes, as in tertiary neurosyphilis, may invade the studied by direct immunofluorescence using a monoclonal anti- brain, where they remain latent for many years.” borrelia antibody. Finally, a Western blot method was used to The vast significance of this hypothesis prompt- analyze protein extracts from cultures and AD brain tissue for the ed us to search for additional evidence in support of presence of borrelia antigen. Contrary to previous studies, our the association between neuroborreliosis and AD. results do not support an association between infection with B burgdozferi and AD. HUM PATHOL 20:753-757. This is a US gov- ernment work. There are no restrictions on its use. MATERIALS AND METHODS Microbiological Studies Alzheimer’s disease (AD) is a neurologic disorder of unknown etiology characterized by progressive im- Tissues from frontal lobes and hippocampi were se- lected from six autopsy cases of AD, one case of Pick’s dis- pairment of memory and intellectual functions.1.2 ease, one case of Creutzfeldt-Jakob disease, and four neu- Brains from patients with this disease develop intra- rologically normal patients. All cases of AD met the Na- neuron accumulations of abnormal filaments (neuro- tional Institutes of Health consensus criteria for the autopsy fibrillary tangles) and discrete aggregates of degen- diagnosis of AD. I5 All AD patients had a documented clin- erated neuron processes and microglial cells, fre- ical history of chronic ; however, information quently accompanied by deposition of amyloid (senile needed to segregate the AD population into sporadic and plaques).3 AD occurs most often as a sporadic pro- familial subtypes was not available. Tissues were inoculated cess, although a small subgroup is inherited with an into modified Kelly’s mediumI and Spirolate broth (Bec- autosomal dominant mode.4 The infectious hypothe- ton-Dickinson, Cockeysville, MD). The ability of our sys- sis of AD has the support of several research groups. tems to support the growth of borrelia and nonpathogenic treponemes was determined by using control B burgdo$eri This hypothesis is buttressed by some successful ex- isolated from an infected Ixodes tick and a stock strain of denticola. Growth was monitored for up to 6 From the Department of Pathology Montrose, VA Hospital, weeks by using direct acridine orange fluorescence, direct Montrose, NY; the Departments of Pathology, Psychiatry and immunofluorescence, and transmission electron micros- Pharmacology, and the Research Center, New York copy (TEM). Electron microscopy was used because it is of Medical College, Valhalla, NY; the Department of Immunology taxonomic value in spirochetal w0rk.l’ TEM was per- and Rheumatology, Westchester County Medical Center, Valhalla, formed on Formvar-coated grids negatively stained with NY; the Department of Immunology, A&man Hospital and 2% phosphotungstic acid. Some samples were processed by Northeastern Ohio Universities, Rootstown, OH; the Department centrifuging culture vials and embedding the pellets in of Psychiatry, Mount Sinai School of Medicine, New York; and the resin following conventional methods. National Institutes of Health, Rocky Mountain Laboratories, Ham- The primary antibody used in the indirect immunoflu- ilton, MT. Accepted for publication February 6, 1989. Supported in part by a grant from the Veterans Adminis- orescence procedure is a monoclonal antibody directed tration and a grant from the National Institute on Aging against a 30 kd peptide present in B burgdorferi. Character- (#AG05138). ization of this antibody has been published.‘s Address correspondence and reprint requests to M. A. Pap- polla, MD, Chief Laboratory Service, Montrose VA Hospital, Mon- Immunochemistry trose, NY 10548. This is a US government work. There are no restrictions on its Western blot method. The Western blot techniqueI was use. used to detect the possible presence of morphologically or 0046-8 177/89/2008-0006$0.00/O microbiologically undetectable forms of B burgdo@+ in cul- 753 HUMANPATHOLOGY Volume 20, No. 8 [August1989) tures and brain tissue. Samples from culture vials from four available) and from 35 patients with various neurological cases of AD, one case of Pick’s disease, one normal brain, and nonneurological disorders (Table 1). Twenty microli- and control cultures were electrophoresed in gradient poly- ters of appropriate dilutions of spinal fluids were applied to acrylamide gels 2o and electroblotted to nitrocellulose sheets microtiter wells containing B burgdo@ri, incubated for 30 as described.tg An additional control for the method con- minutes at room temperature, washed in PBS, incubated sisted of a preparation of calf brain microtubules known to with fluorescein labeled antihuman y-globulin, washed react with a monoclonal antibody to P-tubulin. Brain pro- again, and mounted. Positive and negative control speci- tein extracts from one case of AD were also studied by this mens were included in each multiwell slide. Samples testing method. positive were titered. All specimens were read by a technol- Indirect immunojluorescence of spinaljluidr. Indirect im- ogist with 15 years of experience with the method. The munofluorescence was performed on spinal fluids from 18 absence or degree of fluorescence was graded 0 (no fluo- neuropathologically confirmed cases of AD (serum was un- rescence), + /- (equivocal), or I+ to 4 + intensity (weak to

TABLE 1. Reacfivities of CSF Samples

FBA-NON Absll:5] FBA-Abs ELISA FTA FTA-Abs

AD 2+ [1:30] +/- (-) AD (-) AD I:; iI; AD 2+ [l:lO] C-J iz; AD (-) C-J C-1 AD (-) C-J C-J AD (-1 C-1 AD 3+ [1:40] E 2+ [1:5] AD 3+ [l:lO] 3-4 + 3-4+ [1:5] AD 4+ [1:20] 3-4 + 3+ [ND] AD AD 11; NY Ll AD ND ND AD ;I; ND ND AD (-) ND ND AD C--J ND ND AD (-1 ND ND AD 11; AD $“-9 r”-9 CLE 2+ [1:5] 1 (+l (-) (-) I:; ;I; Pick’s disease (-) Neuronal dysplasia (-) (-) (-) Chronic renal failure l(+) (+/-l (-) IE +/- l(+) f-1 Lung tumor 3+ [1:40] 2(+) I:; f-1 Transitional cell CA (-1 (-1 Stroke (-1 (-1 Head trauma (-1 I:; Multiple sclerosis C-1 Pneumonia (-1 (-1 Disc disease (-1 Alcoholism I:; CHD I:; I:; ;r; Ovarian carcinoma f-1 Guillian-Barre (-1 R/O meningitis ;I; Bronchopneumonia ;I; ;I; Disc disease ;I; Gastroenteritis 11; I:; Convulsion I:; Vertebral disease ;I; Disc disease (-1 ;I; Disc disease (-1 [I; Bladder carcinoma C-1 Chronic renal failure Neuronal dysplasia

Abbreviations: ND, not done; FBA-Non Abs, fluorescent borrelia antibody test without absorption; FBA-Abs, fluorescent borrelia antibody test absorbed with Treponmu phgedaic as described in Methods section; ELISA, enzyme-linked immunosorbant assay; ETA, fluorescent Treponema antibody test; FTA-Abs, fluorescent Treponema antibody test absorbed with T phagedenic filtrates; CLE, chronic lymphocytic ; IE, infectious endocarditis; R/O, rule out; CHD, coronary heart disease; CA, carcinoma. * Titers are indicated between brackets. NEUROBORRELIOSIS AND ALZHEIMEFS DISEASE (Pappolla et al] strong fluorescence). The samples were also absorbed with filtrates of Treponema pkugedenis (Reiter strain) as standard- ized for serology,21 to remove low titered nonspe- cific group antibodies, which are highly prevalent in the general population. ** Minimally positive control (end-point dilution) serum samples from cases of confirmed Lyme dis- ease remained positive after this step (unpublished obser- vation). Enzyme-linked immunosorbant assay of spinal fluids. The enzyme-linked immunosorbant assay (ELISA) was per- formed by standard methodsz3 using an alkaline phos- phatase p-nitrophenzyl system. Briefly, B burgdorfri were harvested from cultures at logarithmic growth phase. A sol- uble fraction antigen was obtained by sonication and ultra- centrifugation of the organisms. Plates were coated with 0.1 mL of spirochetal antigen at a protein concentration of 5 pg/mL. Results were considered significant when the opti- cal densities were above 3 SD of the mean of 50 normal sera. Direct immunojluorescence. Direct immunofluorescence was performed on imprint preparations from brain tissue (frontal cortex and hippocampi) following conventional methods. We used a fluorescein labeled goat-antimouse IgG (I:50 dilution) and the monoclonal antibody referred to above.

RESULTS Microbiological Studies

Growth of B burgdo$eri or T denticola was consis- tently obtained in all control vials. In the control cul- tures, spirochetes were clearly visualized by all the methods used, and in the case of B burgdorferi; it was 0 positively confirmed by electron microscopy (Fig 1). FIGURE 2. Photograph of nitrocellulose strips after electroblotting No growth of spirochetes was detected in any of the the following preparations: [A) microtubule preparation reacted brains obtained for this study. with a monoclonal antibody directed against P-tubulin (control for technique]; (6) homogenate of lysed 6 burgdorferi organisms re- acted with a monoclonal antibody against the 30 kd peptide as described in text; (C) material from brain cultures of AD brain re- acted with Sclme antibody as B above; (D) lmmunoblotting of brain protein extracts from an AD case reacted against the same mono- clonal antibody as B above. The remaining specimens also failed to show any reacWe bands in a similar manner to C and D. The color reaction was developed using an alkaline phosphatase system, nitrotetrazolium blue was used as the chromogen. Specific bands are detected in A [@-tubulin] and B (30 Kd borrelia peptide] (arrows).

lmmunoblotting

A specific B burgdorferi peptide was identified as a single band in immunoblots from cultures of con- trol spirochetes (Fig 2). Immunoblots from AD and Pick’s disease brain cultures, or AD brain tissue, yielded no reactive bands.

Indirect lmmunofluorescence Assay and ELI.94 9 There were no statistically significant differences between test results on the spinal fluid samples from AD and the control groups tested by both methods (Table 1).

Direct lmmunofluorescence

FIGURE 1. Electron microphotograph of borrelia organisms show- Numerous wavy filamentous structures were ing typical ends and flagellae (arrows). identified in touch imprints from AD brains and in 755 HUMAN PATHOLOGY Volume 20, No. 8 (August 1989) preparations obtained from three normal control of nonspecificity in syphillis serology. Spirochetal brains. These forms resembled spirochetes only group antibodies might gain access to the CSF slightly and were interpreted as artifacts, probably through a disrupted blood brain barrier, known to caused by nonspecific binding of IgG to small myelin- occur in AD.sg In any case, the number of positive ated nerve cell endings. cases in the AD aroun was very, small and not differ- ent from that of”the ‘general population. Although our microbiologic culture system con- DISCUSSION sistently supported growth of B burgdorferi and T den- ticolu. neither trenonemas nor snirochetes could be The cause of AD remains unknown. Yet, sub- detected by direc; immunofluore’scence, electron mi- stantial work needs to be done to completely exclude croscopy, or Western blots of culture material of the possibility of an infectious agent that may be re- brain tissue. All the above methods were successful in sponsible for this disorder. Some experimental and detecting spirochetes or their products in control cul- naturally occurring lines of evidence for an infectious tures. The rationale to include T dmticolu in our stud- etiology are as follows: ies was to exclude the possibility of commensal 1. Cases of familial AD can be transmitted to treponemas misclassified as B burgdoferi in previous chimpanzees as , histologically re- reports of concurrent borreliosis and AD. Our find- sembling Cruetzfeld-Jakob disease (CJ), and the CJ ings do not support those of MacDonald1**13 and virus can be isolated from some cases of familial MacDonald and Miranda,” who reported growth in AD.24 However, the characteristic neuropathologic three of three AD brains. lesions of AD have not been observed in infected an- The role of amvloid denosition in AD has been imals. the focus of intense ‘research, especially since the ge- 2. Neurofibrillary tangles and plaques that netic characterization of the sequences encoding the closely resemble those found in AD occur in several amyloid S-protein.40-44 However, mechanisms lead- neurologic conditions of proven infectious nature.5-g ina’ to amvloid accumulation are still controversial 3. Amyloid plaques similar to those found in AD a& new evidence has shown lack of allelic association can be induced in recipient experimental animals in- for restriction fragment length polymorphisms at cer- oculated with the scrapie agent.25 However, the bio- tain genes and in AD. 44 These data may emphasize chemical composition of scrapie plaque amyloid dif- the importance of environmental (toxic, infectious) fers from the AD amyloid.26 factors. Therefore, the possibility of a different spi- The spirochetal hypothesis of AD was proposed rochete in AD not detectable by the methods used by MacDonald and Miranda,’ l who suggested that B herein should remain open to further work. Because burgdorfti may cause AD. The hypothesis is interest- some forms of amvloid accomnanv chronic infections ing for several reasons. For one, spirochetes can sur- and other reactive process&, additional work is vive latent in host tissue for many years. We know needed to totally exclude the role of an infectious from the late neurosyphilis model that spirochetes organism in this disease. This study was directed to are exceedingly difficult to find in the nervous tissue the detection of the Lyme disease agent, ie, B burg- by histologic methods. This has been claimed as the do+ri, and comprised a relatively small sample of AD reason for not finding this newly discovered spiro- patients. chete in AD.i4 In addition, direct infection of the by B burgdorferi can and does occur. In REFERENCES fact, involvement of the nervous system in Lyme dis- ease is seen in approximately 11% of cases. The neu- 1. Terry RD, Katzman R: Senile dementia of the Alzheimer’s rologic manifestations of Lyme disease may mimic type. Ann Neurol 14:497, 1983 various inflammatory and demyelinating disorders, 2. Cohen CD: Session I: Alzheimer’s disease-The human di- mension, in Katzman R (ed): Banbury Report 15: Biological As- including multiple sclerosis-like syndromes and pects of Alzheimer’s Disease. Cold Spring Harbor, NY, Cold dementia.27-38 Spring Harbor Laboratory, 1983, pp 3-6 Results of our study do not support an associa- 31 Tomlinson BE, Corsellis JAN: Aging and the , in tion between B burgdolferi and AD. Virtually all pre- Adams IH. Corsellis IAN. Duchen LW (eds): Greenfield’s Neuro- viously documented cases of neuroborreliosis have pathology.‘New York, Wiley, 1984, p 951 4. Heston LL, Mastri AR, Anderson VE, et al: Dementia of shown increased immunoglobulin (IgG) titers against the Alzheimer type. Clinical genetics, natural history and associ- B burgdorferi in spinal fluid with titers usually higher ated conditions. Arch Gen Psych 38:1085, 1981 than 1: 160.27928~34~3gIn contrast, most cases of AD 5. Van Bogaert L: Une leuco-encephalite Sclerosante Sub- included in this study were unreactive. Few CSF spec- aigue. J Neurol Neurosurg Psych 8: 101, 1945 6. Corsellis JAN: Sub-acute sclerosing leuco-encephalitis: imens from AD patients were reactive at low titers in Clinical and pathological report of two cases. J Mental Sci 97:570, the nonabsorbed fluorescent borrelia antibody test 1951 (Table 1). However, this reactivity was most likely due 7. Malamud N, Haymake W, Pinkerton H: Inclusion enceph- to the presence of spirochetal group antibodies as in- alitis with cliniconatholotric renort of three cases. Am .,I Path01 dicated by the results obtained following absorption 26:133, 1950 ’ ” ’ 8. Adams RD, Victor M: Viral infections of the nervous sys- with filtrates of T phugedenis (Reiter strain) (Table 1). tem, in Adams RD, Victor M (eds): Principles of Neurology. Mc- Spirochetal group antibodies are highly prevalent in Graw-Hill, New York, 1985, pp 545-568 the population at large** and have also been a source 9. Reyes MG, Poland JJ, Monath TP: St Louis encephalitis. 756 NEUROBORRELIOSISAND AZHEIMER’S DISEASE(Pappolla et al) quantitative histologic and immunofluorescent studies. Arch Neu- Nervous System Disorders, ~013. Geneva, Fondation pour 1’Etude rol38:329, 1981 du Systeme Nerveux (Foundation for the Study of the Nervous IO. Pogo BGT, Casals J, Elizan TS: A study of viral genomes System) 1986, p 91 and antigens in brains of patients with Alzheimer’s disease. Brain 27. Reik L lr, Smith L, Khan A, et al: Demyelating encepha- 110:907, 1987 lopathy in Lym; disease. Neurology 35:267, 1985 - _ 11. MacDonald AB, Miranda JM: Concurrent neocortical bor- 28. Wokke .,IHI. .,, van Giin 1. Elderson A, et al: Chronic forms reliosis and Alzheimer’s disease. Hum Path01 18:759. 1987 of infecto; ‘of the nervous system. Neurology 12. MacDonald AB: Borrelia in the brains of patients dyingI _ 37:1031, 1987 with dementia. JAMA 256:2195, 1986 (letter) - 29. Wokke JHJ, de Koning J, Stanek G: Chronic muscle weak- 13. MacDonald AB: Concurrent neocortical borreliosis and ness caused by Borrelia burgdoferi meningoradiculitis. Ann Neurol Alzheimer’s disease. II. Demonstration of a spirochetal cyst form. 22:389, 1987 Ann NY Acad of Sci (in press) 30. Reik L, Burgdorfer W, Donaldson JA: Neurologic abnor- 14. MacDonald AB: Borreliosis and Alzheimer’s disease. Hum malities in Lyme disease without erythema chronicum nigrans. Am Path01 18:1303, 1987 1 Med 81:73, 1986 15. Khachaturian ZS: Diagnosis of Alzheimer’s disease, con- 31. Reik L, Steere AC, Bartenhagen NH, et al: Neurologic ference report. Arch Neurol42: 1097, 1985 abnormalities of Lvme disease. Medicine (Baltimore) 58:281. 1979 16. Kellv RT: Cultivation and nhvsioloev of relansine fever 32. Pachner AR, Steere AC: The triad of neurdlogic manifes- borreliae, in’Johnson RC (ed): Td BioloG of Para’sitic”Spiro- tations of Lyme disease: Meningitis, cranial neuritis and radiculo- chetes. New York, Academic, 1976, p 87 neuritis. Neurology 35:47, 1985 17. Murray RGE, Hovind-Hougen K, Ritchie AE: Section I: 33. Burgdorfer W, Barbour AC, Hayes SF, et al: Lyme dis- Spirochete morphology, in Johnson RC (ed): Biology of Parasitic ease-a tick borne spirochetosis? Science 2 16: 13 17, 1982 +i;ochetes (chapters 1, 2, and 3). New York, Academic, 1976, pp 34. Waisbren BA, Cashman N, Schell RF: Borrelia 6urgdotfet-i antibodies and amyotrophic lateral sclerosis. Lancet 2:332, 1987 18. Barbour AC, Tessier SL, Todd WJ: Lyme disease spiro- 35. Reik L, Burgdorfer W, Donaldson JA: Neurologic abnor- chetes and ixodes tick spirochetes share a common surface anti- malities in Lyme disease without erythema chronicum migrains. genie determinant defined by a monoclonal antibody. Infect Im- Am J Med 81:73, 1986 mun 41:795, 1983 36. Steere AR: Snirochaetal disease of the CNS. Neurol Clin 1 19. Towbin H, Staeweling T, Gordon J: Electrophoretic trans- 4:207, 1986 fer of proteins from polyacrylamide gels to nitrocellulose sheets 37. Finkel MF: Lyme disease and its neurologic complications. procedure and some applications. Proc Nat1 Acad Sci USA Arch Nemo1 45:99. 1988 76:4350, 1979 38. Kohler J, Kasper J, Kern U, et al: Borrelia encephalomy- 20. Walker JM: Gradiant SDS polyacrylamide gel electropho- elitis. Lancet 2:35, 1986 (letter) resis, in Walker JM (ed): Methods in Molecular Biology. Clifton, 39. Wisniewski HM, Kozlowski PB: Evidence for blood-brain NJ, Humana, 1984, p 57 barrier changes in senile dementia of the Alzheimer’s type (SDAT). 21. Manual of Test for Syphilis. Washington, DC, US Depart- Ann NY Acad Sci 396: 119, 1982 ment of Health, Education and Welfare, 1986 40. Goldgarber D, Lerman MI, McBride OW, et al: Charac- 22. Wilkinson AE, Wiseman CC: The role of sorbent in the terization and-chromosomal localization of a cDNA encoding brain absorbed fluorescent treponemal antibody (FTA-ABS) test. Proc R amvloid of Alzheimer’s disease. Science 235:877. 1987 Sot Med 64:10, 1971 ’ 41. Tanzi RE, Gusella JF, Watkins PC, et al: Amyloid B pro- 23. Voller A, Bidwell DE, Barlett A: Enzyme immunoassays in tein gene: cDNA mRNA distribution and genetic linkage near the diagnostic medicine: Theory and practice. Bull WHO 53:55, 1976 Alzheimer’s locus. Science 235:880, 1987 24. Masters CL, Gajdusek DC, Gibbs CJ Jr: Creutzfeldt-Jakob 42. Robakis NK, Wisniewski HM, Jenkins EC: Chromosome disease virus isolations from the Gerstmann-Straussler syndrome 21q21 sublocalization of gene encoding beta-amyloid peptide in with an analysis of the various forms of amyloid plaque deposition cerebral vessels and neuritic (senile) plaques of people with Alz- in the virus induced spongiform encephalopathies. Brain 104:559, heimer’s disease and Down’s syndrome. Lancet 1:384, 1987 1981 43. Glenner GG, Wong CW: Alzheimer’s disease: Initial re- 25. Bruce ME, Dickinson AC, Fraser H: Cerebral amyloidosis port of the purification and characterization of a normal cerebro- in scrapie in the mouse: Effect of agent shain and mouse genotype. vascular amyloid protein. Biochem Biophys Res Commun 120: Neuropathol Appl Neurobiol 2:471, 1976 885, 1984 26. Merz PA, Kascsak RJ, Robakis NK, et al: Recent studies on 44. St George-Hyslop PH, Tanzi RE, Polinsky RI, et al: Ab- SAF, in Bignami A, Bolis L, Gajdusek DC (eds): Discussions in sence of duplication of chromosome 2 1 genes in famihal and spo- Neurosciences. Molecular Mechanisms of Pathogenesis of Central radic Alzheimer’s disease. Science 238:664, 1987

757