Foodborne Transmission of Bovine Spongiform Encephalopathy to Nonhuman Primates

Edgar Holznagel

, Barbara Yutzy, Walter Schulz-Schaeffer, Carina Kruip, Uwe Hahmann, Pär Bierke, Juan-Maria Torres, Yong-Sun Kim, Achim Thomzig, Michael Beekes, Gerhard Hunsmann, and Johannes Loewer

Author affiliations: Paul-Ehrlich-Institut, Langen, Germany (E. Holznagel, B. Yutzy, C. Kruip, J. Loewer); University of Göttingen, Göttingen, Germany (W. Schulz-Schaeffer); German Primate Centre, Göttingen (U. Hahmann, G. Hunsmann); Swedish Institute for Infectious Disease Control, Solna, Sweden (P. Bierke); Centro de Investigación en Sanidad Animal, Madrid, Spain (J.-M. Torres); Hallym University, Anyang, Gyeonggi-Do, South Korea (Y.-S. Kim); Robert-Koch-Institut, Berlin, Germany (A. Thomzig, M. Beekes)

Abstract

Risk for human exposure to bovine spongiform encephalopathy (BSE)–inducing agent was estimated in a nonhuman primate model. To determine attack rates, incubation times, and molecular signatures, we orally exposed 18 macaques to 1 high dose of brain material from cattle with BSE. Several macaques were euthanized at regular intervals starting at 1 year postinoculation, and others were observed until clinical signs developed. Among those who received ≥5 g BSE-inducing agent, attack rates were 100% and prions could be detected in peripheral tissues from 1 year postinoculation onward. The overall median incubation time was 4.6 years (3.7–5.3). However, for 3 macaques orally exposed on multiple occasions, incubation periods were at least 7–10 years. Before clinical signs were noted, we detected a non-type 2B signature, indicating the existence of atypical prion protein during the incubation period. This finding could affect diagnosis of variant Creutzfeldt-Jakob disease in humans and might be relevant for retrospective studies of positive tonsillectomy or appendectomy specimens because time of infection is unknown.

Variant Creutzfeldt-Jakob disease (vCJD) (1) is most likely caused by dietary exposure to bovine spongiform encephalopathy (BSE) prions (2–4). In the United Kingdom, risk for infection with BSE has been considerable, but only 172 cases of vCJD have been documented (5–7). However, the infective dose for oral transmission of the BSE agent to humans is unknown, and incubation times can only be estimated (5,6). In 2001, the European Union funded a risk assessment study in nonhuman primates to estimate the risk for humans exposed to BSE-contaminated food or blood products (8).
A determining factor for susceptibility to BSE prions is a polymorphism for methionine (M) or valine (V) at codon 129 of the human prion protein gene (PRNP). All vCJD cases examined were methionine homozygotes at PRNP codon 129 (129-MM) (9). The overall distribution of PRNP codon 129 genotypes in the general UK population is ≈39% MM, ≈50% MV, and ≈11% VV (7,10). Evidently, persons with a 129-VV genotype can be infected (11), and clinical signs develop after a longer incubation time among those with a 129-MV genotype than among those with a 129-MM genotype (12). However, retrospective analyses of biopsy samples suggest that prevalence of BSE infection is higher among persons who belong to a certain birth cohort and lived in the United Kingdom from 1980 through 1989 (109 cases/million persons [13] to 237 cases/million persons [14]). The reason for the discrepancy between the low number of vCJD cases and higher prevalence of infected persons in the United Kingdom is not known, but the PRNP polymorphism might contribute to this discrepancy as just described. Intriguingly, among hamsters that were orally exposed multiple times to central nervous system (CNS) tissues infected with the scrapie agent, incubation times were significantly prolonged (13). Thus, not only the PRNP polymorphism and the dose but also the mode of transmission might contribute to the development of subclinical cases. However, estimating exposure risks for humans based solely on these results is difficult because of the digestive physiology, life expectancy, and other metabolic parameters of hamsters.

Figure 1

Thumbnail of Schematic diagram of the mature nonglycosylated prion protein and below amino acid sequences of the human and the simian prion polypeptide chain. Homology (198/207 aa) between human and simian mature cellular form of prion protein on the amino acid level is 96%. Large and small arrows indicate major and minor, respectively, proteinase K digestion sites (14). Open arrows indicate digestion sites in type 1 fragments; filled arrows indicate digestion sites in type 2 fragments. The epit

Figure 1. . Schematic diagram of the mature nonglycosylated prion protein and below amino acid sequences of the human and the simian prion polypeptide chain. Homology (198/207 aa) between human and simian mature...

In prion diseases such as CJD, kuru, BSE, scrapie, and chronic wasting disease, the cellular form of prion protein (PrP^C) is thought to be converted into abnormal PrP (PrP^Sc) through a posttranslational event. As a result, PrP^Sc becomes partially resistant to proteases. The misfolded prion protein comprises an N terminal protease-sensitive part followed by a region with variable protease sensitivity and a C-terminal protease-resistant core referred to as PrP^res or PrP27–30 (Figure 1). Limited protease exposure of PrP^Sc in vitro generates nonglycosylated core fragments of 19–21 kDa (14,15), which are used to distinguish 2 major PrP^Sc types by electrophoresis. Type 1 core protein has an apparent molecular mass of 21 kDa. Its primary cleavage site is at residue 82. Type 2 core protein migrates to the 19-kDa region and has a primary cleavage site at residue 97 (16,17). Both types can coexist in a considerable number of sporadic CJD (18,19) and vCJD cases (14). Subtypes and strains can be further characterized by their so-called glycoform profile because the nonobligatory addition of 1–2 sugar chains results in 3 differently glycosylated isoforms in PrP^C and PrP^Sc (non-, mono-, and diglycosylated molecules, referred to as a PrP^res triplet). In sporadic CJD type 2, the monoglycosylated isotype predominates and is referred to as a type 2A signature; whereas, in vCJD, the diglycosylated isoform predominates (1) and is referred to as a type 2B signature (14,16,19). Additional PrP^res fragments have been described, for example the so-called C-terminal fragments of 12/13 kDa (20) and 17 kD (21).
Results of the European Union–funded nonhuman primate risk assessment study, designed to determine the dose at which 50% of macaques will be infected (8), show that a 5-g dose given on 1 occasion infected all macaques. Moreover, multiple exposures to high doses might prolong incubation time. Intriguingly, a non–type 2B PrP^res pattern in CNS tissues of macaques during the preclinical phase indicated the existence of an intermediate prion isoform. This finding might be relevant for retrospective studies of tonsillectomy or appendectomy specimens, because the time point of infection in humans with PrP^res-positive biopsy specimens is not known. As part of the European Union–funded study, we aimed to determine attack rates and incubation times after oral exposure to 5 g or 16 g of BSE-infected brain material in adult cynomolgus monkeys (Macaca fascicularis).