Proc. Natl. Acad. Sci. USA Vol. 95, pp. 13239–13241, October 1998 Neurobiology

Phenotype-genotype studies in kuru: Implications for new variant Creutzfeldt–Jakob disease

LARISA CERVENA´KOVA´*†,LEV G. GOLDFARB‡,RALPH GARRUTO§,HEE-SUK LEE‡,D.CARLETON GAJDUSEK*, AND PAUL BROWN*¶

*Laboratory of Central Nervous System Studies and ‡Medical Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD 20892; †American Red Cross, J. H. Holland Laboratory, Rockville, MD 20855; and §Department of Anthropology, State University of New York, Binghamton, NY 13902

Contributed by D. Carleton Gajdusek, August 25, 1998

ABSTRACT The PRNP polymorphic (methionine/valine) Kuru, the prototype human transmissible spongiform en- codon 129 genotype influences the phenotypic features of cephalopathy (TSE), was spread by exposure to contaminated transmissible spongiform encephalopathy. All tested cases of human tissues through the practice of ritual endocannibalism, new variant Creutzfeldt–Jakob disease (nvCJD) have been and is by virtue of its oral and/or mucocutaneous route of homozygous for methionine, and it is conjectural whether infection the most appropriate comparison for nvCJD. Here, different genotypes, if they appear, might have distinctive we compare codon 129 genotypes to age at onset and duration phenotypes and implications for the future ‘‘epidemic curve’’ of illness for 92 kuru patients dying in the late 1950s, at the of nvCJD. Genotype-phenotype studies of kuru, the only other height of the kuru epidemic, of which nine cases had been the orally transmitted transmissible spongiform encephalopathy, subject of extensive neuropathological examination (4). might be instructive in predicting the answers to these ques- tions. We therefore extracted DNA from blood clots or sera MATERIALS AND METHODS from 92 kuru patients, and analyzed their codon 129 PRNP We retrieved a total of 92 frozen blood clots and sera from our genotypes with respect to the age at onset and duration of Ϫ illness and, in nine cases, to detailed clinical and neuropa- tissue bank archives, stored at 20°C since collection in Papua, thology data. Homozygosity at codon 129 (particularly for New Guinea in 1957–1958. Genomic DNA was extracted from methionine) was associated with an earlier age at onset and a blood clots after treatment with proteinase K using a standard shorter duration of illness than was heterozygosity, but other phenol-chloroform extraction method and from sera using the clinical characteristics were similar for all genotypes. In the Puregene DNA isolation kit (Gentra Systems, Minneapolis, MN), with the following slight modification: 500 ␮l of plasma nine neuropathologically examined cases, the presence of was centrifuged for 30 sec at 13,000 rpm in a microfuge histologically recognizable plaques was limited to cases car- (Beckman, Palo Alto, CA) and pelleted cells were treated with rying at least one methionine allele (three homozygotes and 300 ␮l of cell lysis solution. Further steps were performed one heterozygote). If nvCJD behaves like kuru, future cases according to the manufacturer’s instructions. (with longer incubation periods) may begin to occur in older The 169-bp fragment of the PRNP gene spanning codons 94 individuals with heterozygous codon 129 genotypes and signal through 150 was amplified by PCR using TaKaRa Ex Taq a maturing evolution of the nvCJD ‘‘epidemic.’’ The clinical polymerase (TaKARA Shuzo Co., Shiga, Japan) and oligonu- phenotype of such cases should be similar to that of homozy- cleotide primers (GenSet Corp., La Jolla, CA) 5Ј- cacccacagtc gous cases, but may have less (or at least less readily identi- agtggaa-3Ј and 5Ј-atagtaacggtcctcatagtca-3Ј under the follow- fied) amyloid plaque formation. ing conditions: an initial denaturation (95°C, 3 min) followed by 35 cycles of denaturation (94°C, 30 sec), annealing (60°C, 30 New variant Creutzfeldt–Jakob disease (nvCJD), the probable sec), extension (72°C, 30 sec), and a final elongation (72°C, 5 consequence of eating beef contaminated with the infectious min). The ATG/GTG polymorphism at codon 129 was agent of bovine spongiform encephalopathy, can be distin- screened in the amplified samples using endonuclease MaeII guished from sporadic cases of Creutzfeldt–Jakob disease by (Boehringer Mannheim) or TaiI (MBI Fermentas, Vilnius, an early age at onset (average, 27 yr), a clinical presentation Lithuania) (5). Digested fragments were resolved in a 3% with psychiatric or sensory symptoms, the absence of periodic MetaPhor agarose gel (FMC BioProducts, Rockland, ME) electroencephalogram activity, a comparatively long duration stained with ethidium bromide. Codon 129 genotypes (MM, (average, 14 mo), and a neuropathological picture featuring an VV, or MV) were determined on the basis of the distinct abundance of amyloid plaques surrounded by halos of spon- restriction patterns seen for each genotype. giosis (‘‘daisy plaques’’) (1, 2). All 26 tested cases have shown Comparisons of genotype frequencies between different age a homozygous methionine genotype at polymorphic codon 129 groups were made by using the two-tailed Fisher’s exact test. of the PRNP gene located on chromosome 20 (ref. 3; R. Will, personal communication). RESULTS It is possible that as time goes on, nvCJD cases may begin to occur with codon 129 genotypes other than methionine ho- The distribution of homozygous methionine, homozygous mozygosity, which could have a sufficiently different clinico- valine, and heterozygous genotypes according to age at onset pathological phenotype to present a problem in recognition of kuru symptoms is shown in Fig. 1. It is evident that compared with the series as a whole, homozygous methionine and diagnosis. patients are overrepresented in the younger age groups, and

The publication costs of this article were defrayed in part by page charge Abbreviations: nvCJD, new variant Creutzfeldt–Jakob disease; TSE, payment. This article must therefore be hereby marked ‘‘advertisement’’ in transmissible spongiform encephalopathy. accordance with 18 U.S.C. §1734 solely to indicate this fact. ¶To whom reprint requests should be addressed at: National Institutes © 1998 by The National Academy of Sciences 0027-8424͞98͞9513239-3$2.00͞0 of Health, Building 36, Room 5B20, Bethesda, MD 20892. e-mail: PNAS is available online at www.pnas.org. [email protected].

13239 Downloaded by guest on September 23, 2021 13240 Neurobiology: Cervena´kova´ et al. Proc. Natl. Acad. Sci. USA 95 (1998)

FIG. 1. Distribution of PRNP codon 129 genotypes according to age at onset of illness in 92 kuru patients.

heterozygous patients are overrepresented in the older age In regard to neuropathology, widespread gliosis occurred in groups. When the genotype frequencies of the group of all cases, associated with a variable pattern of neuronal ab- children and young adolescents (Ͻ15 yr of age) were compared normalities and vacuolation that was unrelated to genotype. with adults Ͼ30 yr of age, the difference between homozygotes However, amyloid plaques were noted only in the four patients and heterozygotes was statistically significant (P ϭ 0.0001), with at least one methionine allele, and not in the five patients Table 1. A less impressive but still statistically significant with only valine alleles. Except for the presence of plaques, the correspondence also exists between the codon 129 genotype lone heterozygote was neuropathologically indistinguishable and the duration of illness, shorter illnesses being associated from either the methionine or valine homozygotes. with homozygosity and longer illnesses being associated with heterozygosity (P ϭ 0.006). DISCUSSION Nine specimens in the series were matched with patients whose brains had earlier been the subject of extensive neuro- One of the more interesting features of the chromosome 20 PRNP gene is the phenotypic influence of polymorphic codon pathological examinations (4). An extract of these data is 129. Although not in itself pathogenic, it has been shown to presented in Table 2, along with clinical summaries and codon influence susceptibility to iatrogenic and sporadic forms of 129 genotypes. Irrespective of the age at onset or duration of TSE and to affect age at onset and duration of illness in familial illness, or of the codon 129 genotype, all nine patients had a TSE; in association with a pathogenic mutation in codon 178, clinical picture dominated by progressive locomotor ataxia the entire disease phenotype is altered (5–11). associated with the shivering tremor characteristic of kuru. Because all cases of nvCJD recognized to date have tested The occurrence of extrapyramidal signs, strabismus, dyspha- homozygous for methionine at codon 129, it is not known gia, and mutism was also unrelated to age at onset or duration whether other genotypes (should they occur) will change the of illness, or to the codon 129 genotype. nvCJD phenotype, possibly even to the point of obscuring its distinctive clinical and neuropathological features. Table 1. PRNP codon 129 genotype frequencies in kuru patients One clue that this may not happen comes from experience Genotype frequency (proportion of total) with iatrogenic Creutzfeldt–Jakob disease, resulting from pe- ripheral injection by contaminated human growth hormone, Total no. ͞ ͞ ͞ which shows that, other than a longer period of latency Met Met Met Val Val Val tested between infection and the onset of symptoms, the clinicopath- Age at onset of illness, yr ological picture of Creutzfeldt–Jakob disease in patients with Ͻ15 14 (0.45) 6 (0.19) 11 (0.36) 31 a heterozygous codon 129 genotype does not differ from that 15–30 10 (0.28) 17 (0.47) 9 (0.25) 36 in patients with either methionine or valine homozygosity (ref. Ͼ30 4 (0.16) 18 (0.72) 3 (0.12) 25 12; D. Dormont, personal communication; M. Preece, per- Duration of illness,* mo sonal communication). Ͻ8 6 (0.29) 4 (0.19) 11 (0.52) 21 An even more appropriate comparison comes from pheno- 8–12 11 (0.42) 10 (0.38) 5 (0.19) 26 type-genotype studies in kuru patients, in whom infection Ͼ12 2 (0.13) 10 (0.67) 3 (0.20) 15 occurred, at least in part, by the oral route. Three recent *Information about duration of illness was incomplete for 30 cases, studies comparing the neuropathological features of kuru and which were excluded from analysis. nvCJD include kuru cases in which the genotype of codon 129 Downloaded by guest on September 23, 2021 Neurobiology: Cervena´kova´ et al. Proc. Natl. Acad. Sci. USA 95 (1998) 13241

Table 2. Summary of clinicopathological phenotypes and codon 129 genotypes in nine kuru patients Clinical signs (all cases had progressive ataxia associated Pathology with tremor) (all cases had severe gliosis in all brain regions) Sex͞age Duration of Abnormal Extra- Cerebral Basal Pons͞ Codon 129 Case at onset illness, mo affect pyramidal Other signs cortex ganglia Thalamus medulla Cerebellum genotype 6F͞68ϩ Rigidity Dysphagia N NV NV NV Val͞Val 7F͞65ϪϪDysphagia, mutism N NV N N Val͞Val 16 M͞78 ϪϪStrabismus, mutism NP NNP NV NVP Met͞Met 8M͞78 ϩ ChAth Myoclonus N NV NV N NV Val͞Val 24 F͞11 12 ϩϪStrabismus P NV NVP NV NVP Met͞Met 11 F͞13 5 ϩ ChAth Strabismus N N NV Val͞Val 10 M͞17 10 ϪϪStrabismus, mutism (N) NV N Val͞Val 2F͞45 9 Ϫ ChAth mutism P NV NP Met͞Met 4F͞50 10 ϩϪDysphagia, mutism P NV NP NV NP Met͞Val ChAth, choreo-athetosis; N, neuronal changes; V, vacuolation; P, plaques.

was determined. In one study, the brains from two valine 2. Zeidler, M., Stewart, G. E., Barraclough, C. R., Bateman, D. E., homozygotes showed a tendency for ‘‘proper plaque forma- Bates, D., Burn, D. J., Colchester, A. C., Durward, W., Fletcher, tion’’ in PrP-stained immunohistochemical sections (13). In N. A., Hawkins, S. A., Mackenzie, J. M. & Will, R. G. (1997) another study, the brain from a valine homozygote showed Lancet 350, 903–907. numerous plaques by both histological and immunohistochem- 3. Zeidler, M., Stewart, G., Cousens, S. N., Estibeiro, K. & Will, ical staining (14). In the third study, brains from three valine R. G. (1997) Lancet 350, 668. and two methionine homozygotes had plaques visible by both 4. Klatzo, I., Gajdusek, D. C. & Zigas, V. (1959) Lab. Invest. 8, histological and immunohistochemical staining, and no geno- 799–847. typic correlations were noticed (15). 5. Goldfarb, L. G., Petersen, R. B., Tabaton, M., Brown, P., In contrast to these studies, our series of nine cases showed LeBlanc, A. C., Montagna, P., Cortelli, P., Julien, J., Vital, C., a distinct correlation between the presence of the methionine Perdelbury, W. W., et al. (1992) Science 258, 806–808. allele and the presence of histologically recognizable amyloid 6. Palmer, M. S., Dryden, A. J., Hughes, J. T. & Collinge, J. (1991) Nature (London) 352, 340–342. plaques. It is possible that these differing results are a conse- 7. Dlouhy, S. R., Hsiao, K., Farlow, M. R., Foroud, T., Conneally, quence of chance or of undefined differences in tissue fixation P. M., Johnson, P., Prusiner, S. B., Hodes, M. E. & Ghetti, B. and staining methods. It is also possible that as a group, kuru (1992) Nat. Genet. 1, 64–67. cases with a valine allele at codon 129 have a comparatively 8. Poulter, M., Baker, H. F., Frith, C. D., Leach, M., Lofthouse, R., reduced potential for plaque formation, and, by analogy, Ridley, R. M., Shah, T., Owen, F., Collinge, J., Brown, J., et al. nvCJD cases with a codon 129 valine allele might not always (1992) Brain 115, 675–685. show the prominent daisy plaque formation observed in me- 9. Brown, P., Cervena´kova´, L., Goldfarb, L. G., McCombie, W. R., thionine homozygotes. Rubenstein, R., Will, R. G., Pocchiari, M., Martinez-Lage, J. F., Our finding that homozygotes have an earlier age at onset Scalici, C., Masullo, C., et al. (1994) Neurology 44, 291–293. (and thus probably a shorter incubation period) than heterozy- 10. Deslys, J.-P., Marce´,D. & Dormont D. (1994) J. Gen. Virol. 75, gotes, but show a generally similar clinical phenotype, is in 23–27. keeping with the effect of codon 129 homozygosity in other 11. Gambetti, P., Parchi, P., Petersen, R. B., Chen, S. G. & Lugaresi, forms of TSE, and, if applicable to nvCJD, would predict (i) E. (1995) Brain Pathol. 5, 43–51. that codon 129 nvCJD heterozygotes will not pose a problem 12. Deslys, J. P., Jaegly, A., Huillard d’Algnaux, J., Moulhon, F., in recognition or diagnosis and (ii) that the appearance of an Billette de Villemeur, T. & Dormont, D. (1998) Lancet 351, 1251. increasing proportion of heterozygotes may signal that the 13. Lantos, P. L., Bhatia, K., Doey, L. J., Al-Sarraj, S., Doshi, R., nvCJD epidemic is already evolving beyond its ‘‘leading edge,’’ Beck, J. & Collinge, J. (1997) Lancet 350, 187–188. and thus provide a more solid foundation for predictive 14. Hainfellner, J. A., Liberski, P. P., Guiroy, D. C., Cervena´kova´,L., modeling studies of its overall extent and duration. Brown, P., Gajdusek, D. C. & Budka, H. (1997) Brain Pathol. 7, 547–554. 1. Will, R. G., Ironside, J. W., Zeidler, M., Cousens, S. N., Estibeiro, 15. McLean, C. A., Ironside, J. W., Alpers, M., Cervena´kova´,L., K., Alperovitch, A., Poser, S., Pocchiari, M., Hofman, A. & Anderson, R. M., Brown, P. W. & Masters, C. L. (1998) Brain Smith, P. G (1996) Lancet 347, 921–925. Pathol. 8, 429–437. Downloaded by guest on September 23, 2021