Figure 1

Figure 1. Protease-resistant prion protein (PrP^res) in the brains of human PrP transgenic mice infected with atypical or zoonotic bovine spongiform encephalopathy (BSE) agents. A) Representative Western blot analysis of PrP^res extracted (for...

The bovine isolates used in this study have been previously described; they all exhibited high infectivity levels in bovine PrP mice (4,7,9). The equivalent of 2 mg of infected bovine brain tissue was injected intracerebrally into tg650 mice. This line of mice overexpresses (≈6-fold) human PrP with methionine at codon 129 (Met¹²⁹) on a Zurich mouse PrP null background and has been shown to be fully susceptible to vCJD agent (13). The resulting transmission data available to date are summarized in the Table. The primary transmission of classical BSE isolates was inefficient as judged by the absence of clear neurologic signs and by Western blot detection of PrP^res in the brain of only 4/25 inoculated mice. The PrP^res banding pattern was essentially similar to that of vCJD (low molecular mass fragments and predominance of diglycoform species; Figure 1).

Secondary passages were performed by using PrP^res-negative or PrP^res-positive individual mouse brains. Every time brain homogenate from an aged mouse (>630 days of age) was inoculated, transmission was observed in >80% of the mice, as determined by clinical signs and PrP^res accumulation. The mean survival time was ≈600 days (Table; additional data not shown). By the third passage, mean survival time approached ≈500 days, as is usually observed with vCJD cases (Table; 13). The vCJD-like PrP^res profile was conserved in all 50 positive brains analyzed (Figure 1). Transmission of L-type isolates to tg650 mice produced markedly different results. First, the 4 L-type isolates induced neurologic disease in almost all mice; survival times averaged 600–700 days (Table). Second, PrP^res accumulated in the brain of all 33 mice analyzed. The molecular profile was distinct from BSE or vCJD; prominent monoglycosylated PrP^res species resembled those found in cattle (Figure 1). Third, we found neither shortening of the survival time on subpassage (1 isolate tested, 2 brains) (Table; data not shown) nor change in PrP^res profile (Figure 1).

In sharp contrast, BSE H-type isolates failed to transmit disease to or even infect tg650 mice. None of the inoculated mice had a detectable level of PrP^res in the brain (22 analyzed). Secondary passages were performed with brains of mice that died at various time points. All inoculated mice survived, and none showed a PrP^res signal in the brain (15 mice analyzed).

Figure 2

Figure 2. Representative histoblots in 4 different anteroposterior sections showing the distribution of disease-specific PrP^res deposits in the brains of tg650 mice infected with bovine spongiform encephalopathy (BSE) or L-type BSE. Panels A–D...

To further compare the behavior of the 3 bovine prions in tg650 mice, we examined the regional distribution and intensity of PrP^res deposition in the brain (14). Histoblot analyses (3 brains per infection) were performed on primary (L-type, H-type) and secondary passages (L-type, H-type, BSE). As shown in Figure 2, L-type and BSE agents showed clear differences according to both the aspect and localization of the PrP deposits. Granular PrP deposits were scattered throughout the brain with BSE, as has been previously observed with vCJD (13). The ventral nuclei of the thalamus, cerebral cortex, oriens layer of the hippocampus, and raphe and tegmentum nuclei of the brain stem were strongly stained. With BSE-L, the staining was finer and essentially confined to the habenular, geniculate, and dorsal nuclei of the thalamus; the lateral hypothalamus; the lacunosom moleculare layer of the hippocampus; the superficial gray layer of the superior colliculus; and the raphe nuclei of the brain stem. Finally, PrP^res could not be detected on brain sections from mice inoculated with H-type isolates (data not shown), thus confirming the Western blot data.

We found that atypical L-type bovine prions can propagate in human PrP transgenic mice with no significant transmission barrier. Lack of a barrier is supported by the 100% attack rate, the absence of reduction of incubation time on secondary passage, and the conservation of PrP^res electrophoretic profile. In comparison, transmission of classical BSE agent to the same mice showed a substantial barrier. Indeed, 3 passages were necessary to reach a degree of virulence comparable to that of vCJD agent in these mice (13), which likely reflects progressive adaptation of the agent to its new host. At variance with the successful transmission of classical BSE and L-type agents, H-type agent failed to infect tg650 mice. These mice overexpress human PrP and were inoculated intracranially with a low dilution inoculum (10% homogenate). Therefore, this result supports the view that the transmission barrier of BSE-H from cattle to humans might be quite robust. It also illustrates the primacy of the strain over PrP sequence matching for cross-species transmission of prions (15).

Extrapolation of our data raises the theoretical possibility that the zoonotic risk associated with BSE-L prions might be higher than that associated with classical BSE, at least for humans carrying the Met¹²⁹ PrP allele. This information underlines the need for more intensive investigations, in particular regarding the tissue tropism of this agent. Its ability to colonize lymphoid tissues is a potential, key factor for a successful transmission by peripheral route. This issue is currently being explored in the tg650 mice. Although recent data in humanized mice suggested that BSE-L agent is likely to be lymphotropic (12), preliminary observations in our model suggested that its ability to colonize such tissues is comparatively much lower than that of classical BSE agent.