Follicular dendritic cell dedifferentiation by treatment with an inhibitor of the lymphotoxin pathway dramatically reduces scrapie susceptibility

Transmissible spongiform encephalopathies (TSEs) may be acquired peripherally, in which case infectivity usually accumulates in lymphoid tissues before dissemination to the nervous system. Studies of mouse scrapie models have shown that mature follicular dendritic cells (FDCs), expressing the host prion protein (PrP(c)), are critical for replication of infection in lymphoid tissues and subsequent neuroinvasion. Since FDCs require lymphotoxin signals from B lymphocytes to maintain their differentiated state, blockade of this stimulation with a lymphotoxin beta receptor-immunoglobulin fusion protein (LT beta R-Ig) leads to their temporary dedifferentiation. Here, a single treatment with LT beta R-Ig before intraperitoneal scrapie inoculation blocked the early accumulation of infectivity and disease-specific PrP (PrP(Sc)) within the spleen and substantially reduced disease susceptibility. These effects coincided with an absence of FDCs in the spleen for ca. 28 days after treatment. Although the period of FDC dedifferentiation was extended to at least 49 days by consecutive LT beta R-Ig treatments, this had little added protective benefit after injection with a moderate dose of scrapie. We also demonstrate that mature FDCs are critical for the transmission of scrapie from the gastrointestinal tract. Treatment with LT beta R-Ig before oral scrapie inoculation blocked PrP(Sc) accumulation in Peyer's patches and mesenteric lymph nodes and prevented neuroinvasion. However, treatment 14 days after oral inoculation did not affect survival time or susceptibility, suggesting that infectivity may have already spread to the peripheral nervous system. Although manipulation of FDCs may offer a potential approach for early intervention in peripherally acquired TSEs, these data suggest that the duration of the treatment window may vary widely depending on the route of exposure.     

Sympathetic innervation of lymphoreticular organs is rate limiting for prion neuroinvasion

Transmissible spongiform encephalopathies are commonly propagated by extracerebral inoculation of the infectious agent. Indirect evidence suggests that entry into the central nervous system occurs via the peripheral nervous system. Here we have investigated the role of the sympathetic nervous system in prion neuroinvasion. Following intraperitoneal prion inoculation, chemical or immunological sympathectomy delayed or prevented scrapie. Prion titers in spinal cords were drastically reduced at early time points after inoculation. Instead, keratin 14-NGF transgenic mice, whose lymphoid organs are hyperinnervated by sympathetic nerves, showed reduction in scrapie incubation time and, unexpectedly, much higher titers of prion infectivity in spleens. We conclude that sympathetic innervation of lymphoid organs is rate limiting for prion neuroinvasion and that splenic sympathetic nerves may act as extracerebral prion reservoirs.     

Temporary depletion of complement component C3 or genetic deficiency of C1q significantly delays onset of scrapie

Following peripheral exposure to transmissible spongiform encephalopathies (TSEs), infectivity usually accumulates in lymphoid tissues before neuroinvasion. The host prion protein (PrPc) is critical for TSE agent replication and accumulates as an abnormal, detergent insoluble, relatively proteinase-resistant isoform (PrPSc) in diseased tissues. Early PrPSc accumulation takes place on follicular dendritic cells (FDCs) within germinal centers in lymphoid tissues of patients with variant Creutzfeldt-Jakob disease (vCJD), sheep with natural scrapie or rodents following experimental peripheral infection with scrapie. In mouse scrapie models, the absence of FDCs blocks scrapie replication and PrPSc accumulation in the spleen, and neuroinvasion is significantly impaired. The mechanisms by which the TSE agent initially localizes to lymphoid follicles and interacts with FDCs are unknown. Antigens are trapped and retained on the surface of FDCs through interactions between complement and cellular complement receptors. Here we show that in mice, both temporary depletion of complement component C3 or genetic deficiency of C1q significantly delays the onset of disease following peripheral infection, and reduces the early accumulation of PrPSc in the spleen. Thus, in the early stages of infection, C3 and perhaps C1q contribute to the localization of TSE infectivity in lymphoid tissue and may be therapeutic targets.     

Prions and the lymphoreticular system

Following intracerebral or peripheral inoculation of mice with scrapie prions, infectivity accumulates first in the spleen and only later in the brain. In the spleen of scrapie-infected mice, prions were found in association with T and B lymphocytes and to a somewhat lesser degree with the stroma, which contains the follicular dendritic cells (FDCs) but not with non-B, non-T cells; strikingly, no infectivity was found in lymphocytes from blood of the same mice. Transgenic PrP knockout mice expressing PrP restricted to either B or T lymphocytes show no prion replication in the lymphoreticular system. Therefore, splenic lymphocytes either acquire prions from another source or replicate them in dependency on other PrP-expressing cells. The essential role of FDCs in prion replication in spleen was shown by treating mice with soluble lymphotoxin-beta receptor, which led to disappearance of mature FDCs from the spleen and concomitantly abolished splenic prion accumulation and retarded neuroinvasion following intraperitoneal scrapie inoculation.     

Complement facilitates early prion pathogenesis

New-variant Creutzfeldt-Jakob disease and scrapie are typically initiated by extracerebral exposure to the causative agent, and exhibit early prion replication in lymphoid organs. In mouse scrapie, depletion of B-lymphocytes prevents neuropathogenesis after intraperitoneal inoculation, probably due to impaired lymphotoxin-dependent maturation of follicular dendritic cells (FDCs), which are a major extracerebral prion reservoir. FDCs trap immune complexes with Fc-gamma receptors and C3d/C4b-opsonized antigens with CD21/CD35 complement receptors. We examined whether these mechanisms participate in peripheral prion pathogenesis. Depletion of circulating immunoglobulins or of individual Fc-gamma receptors had no effect on scrapie pathogenesis if B-cell maturation was unaffected. However, mice deficient in C3, C1q, Bf/C2, combinations thereof or complement receptors were partially or fully protected against spongiform encephalopathy upon intraperitoneal exposure to limiting amounts of prions. Splenic accumulation of prion infectivity and PrPSc was delayed, indicating that activation of specific complement components is involved in the initial trapping of prions in lymphoreticular organs early after infection.     

KDEL-tagged anti-prion intrabodies impair PrP lysosomal degradation and inhibit scrapie infectivity

Transmissible spongiform encephalopathy or prion diseases are fatal neurodegenerative disorders characterized by the conversion of the cellular prion protein (PrPC) into the infectious scrapie isoform (PrPSc). We have recently demonstrated that anti-prion intrabodies targeted to the lumen of the endoplasmic reticulum provide a simple and effective means to inhibit the transport of PrPC to the cell surface. Here, we report that they completely block the traffic of mature full-length PrPC molecules, impair prion lysosomal degradation, and interfere with the early phase of scrapie formation. Since anti-prion intrabodies efficiently block PrPSc accumulation in vitro, we investigated whether they could also antagonize scrapie infectivity in vivo. We found that mice intracerebrally injected with KDEL-8H4-NGF-differentiated PC12 cells infected with scrapie neither develop scrapie clinical signs nor brain damage. Furthermore, no protease-resistant PrPSc is detectable in brains of inoculated animals. These results indicate that anti-prion intrabody strategy may be effective against prion infection.     

High prion and PrPSc levels but delayed onset of disease in scrapie-inoculated mice heterozygous for a disrupted PrP gene.

BACKGROUND: It has been proposed that the prion, the infectious agent of transmissible spongiform encephalopathies, is PrPSc, a post-translationally modified form of the normal host protein PrPC. We showed previously that mice devoid of PrPC (Prn-p0/0) are completely resistant to scrapie. We now report on the unexpected response of heterozygous (Prn-p0/+) mice to scrapie infection. MATERIALS AND METHODS: Prn-p0/+, Prn-p0/0 and Prn-p+/+ mice were obtained from crosses of Prn-p0/+ mice. Mice were inoculated intracerebrally with mouse-adapted scrapie agent and the clinical progression of the disease recorded. Mice were sacrificed at intervals, PrPSc was determined as protease-resistant PrP and the prion titer by the incubation time assay. RESULTS: Prn-p0/+ mice, which have about half the normal level of PrPC in their brains, show enhanced resistance to scrapie, as manifested by a significant delay in onset and progression of clinical disease. However, while in wild type animals an increase in prion titer and PrPSc levels is followed within weeks by scrapie symptoms and death, heterozygous Prn-p0/+ mice remain free of symptoms for many months despite similar levels of scrapie infectivity and PrPSc. CONCLUSIONS: Our findings extend previous reports showing an inverse relationship between PrP expression level and incubation time for scrapie. However, contrary to expectation, overall accumulation of PrPSc and prions to a high level do not necessarily lead to clinical disease. These findings raise the question whether high titers of prion infectivity could also persist for long periods under natural circumstances in the absence of clinical symptoms.     

Role of plasminogen in propagation of scrapie

To investigate whether plasminogen may feature in scrapie infection, we inoculated plasminogen-deficient (Plg(-/-)), heterozygous plasminogen-deficient (Plg(+/-)), and wild-type (Plg(+/+)) mice by the intracerebral or intraperitoneal (i.p.) route with the RML scrapie strain and monitored the onset of neurological signs of disease, survival time, brain, and accumulation of scrapie disease-associated forms of the prion protein (PrP(Sc)). Only after i.p. inoculation, a slight, although significant, difference in survival (P < 0.05) between Plg(-/-) and Plg(+/+) mice was observed. Neuropathological examination and Western blot analysis were carried out when the first signs of disease appeared in Plg(+/+) animals (175 days after i.p. inoculation) and when mice reached the terminal stage of illness. At the onset of symptoms, PrP(Sc) accumulation was higher in the brain and spleen of Plg(+/+) and Plg(+/-) mice than in those of Plg(-/-) mice, and these differences were paralleled by differences in the severity of spongiform changes and astrogliosis in the cerebral cortex and subcortical gray structures. Immunohistochemical analysis of the spleens before inoculation did not show any impairment of the immune system affecting follicular dendritic or lymphoid cells in Plg(-/-) mice. Once the disease progressed and mice began to die of infection, differences were no longer apparent in either brains or spleens. In conclusion, our data indicate that plasminogen has no major effect on the survival of scrapie agent-infected mice.     

Copper chelation delays the onset of prion disease

The prion protein (PrP) binds copper and under some conditions copper can facilitate its folding into a more protease resistant form. Hence, copper levels may influence the infectivity of the scrapie form of prion protein (PrPSc). To determine the feasibility of copper-targeted therapy for prion disease, we treated mice with a copper chelator, D-(-)-penicillamine (D-PEN), starting immediately following intraperitoneal scrapie inoculation. D-PEN delayed the onset of prion disease in the mice by about 11 days (p = 0.002), and reduced copper levels in brain by 29% (p < 0.01) and in blood by 22% (p = 0.03) compared with control animals. Levels of other metals were not significantly altered in the blood or brain. Modest correlation was observed between incubation period and levels of copper in brain (p = 0.08) or blood (p = 0.04), indicating that copper levels are only one of many factors that influence the rate of progression of prion disease. In vitro, copper dose-dependently enhanced the proteinase K resistance of the prion protein, and this effect was counteracted in a dose-dependent manner by co-incubation with D-PEN. Overall, these findings indicate that copper levels can influence the conformational state of PrP, thereby enhancing its infectivity, and this effect can be attenuated by chelator-based therapy.     

Tumor necrosis factor alpha-deficient, but not interleukin-6-deficient, mice resist peripheral infection with scrapie

In most peripheral infections of rodents and sheep with scrapie, infectivity is found first in lymphoid tissues and later in the central nervous system (CNS). Cells within the germinal centers (GCs) of the spleen and lymph nodes are important sites of extraneural replication, from which infection is likely to spread to the CNS along peripheral nerves. Here, using immunodeficient mice, we investigate the identity of the cells in the spleen that are important for disease propagation. Despite possessing functional T and B lymphocytes, tumor necrosis factor alpha-deficient (TNF-alpha(-/-)) mice lack GCs and follicular dendritic cell (FDC) networks in lymphoid tissues. In contrast, lymphoid tissues of interleukin-6-deficient (IL-6(-/-)) mice possess FDC networks but have impaired GCs. When the CNSs of TNF-alpha(-/-), IL-6(-/-), and wild-type mice were directly challenged with the ME7 scrapie strain, 100% of the mice were susceptible, developing disease after closely similar incubation periods. However, when challenged peripherally (intraperitoneally), most TNF-alpha(-/-) mice failed to develop scrapie up to 503 days postinjection. All wild-type and IL-6(-/-) mice succumbed to disease approximately 300 days after the peripheral challenge. High levels of scrapie infection and the disease-specific isomer of the prion protein, PrP(Sc), were detectable in spleens from challenged wild-type and IL-6(-/-) mice but not from TNF-alpha(-/-) mice. Histopathological analysis of spleen tissue demonstrated heavy PrP accumulations in direct association with FDCs in challenged wild-type and IL-6(-/-) mice. No PrP(Sc) accumulation was detected in spleens from TNF-alpha(-/-) mice. We conclude that, for the ME7 scrapie strain, mature FDCs are critical for replication in lymphoid tissues and that in their absence, neuroinvasion following peripheral challenge is impaired.     

Follicular dendritic cell of the knock-in mouse provides a new bioassay for human prions

Infectious prion diseases initiate infection within lymphoid organs where prion infectivity accumulates during the early stages of peripheral infection. In a mouse-adapted prion infection, an abnormal isoform (PrP(Sc)) of prion protein (PrP) accumulates in follicular dendritic cells within lymphoid organs. Human prions, however, did not cause an accumulation of PrP(Sc) in the wild type mice. Here, we report that knock-in mouse expressing humanized chimeric PrP demonstrated PrP(Sc) accumulations in follicular dendritic cells following human prion infections, including variant Creutzfeldt-Jakob disease. The accumulated PrP(Sc) consisted of recombinant PrP, but not of the inoculated human PrP. These accumulations were detectable in the spleens of all mice examined 30 days post-inoculation. Infectivity of the spleen was also evident. Conversion of humanized PrP in the spleen provides a rapid and sensitive bioassay method to uncover the infectivity of human prions. This model should facilitate the prevention of infectious prion diseases.     

Oral scrapie infection modifies the homeostasis of Peyer’s patches’ dendritic cells

In transmitted prion diseases the immune system supports the replication and the propagation of the pathogenic agent (PrPSc). DCs, which are mobile cells present in large numbers within lymph organs, are suspected to carry prions through the lymphoid system and to transfer them towards the peripheral nervous system. In this study, C57Bl/6 mice were orally inoculated with PrPSc (scrapie strain 139A) and sacrificed at the preclinical stages of the disease. Immunolabelled cryosections of Peyer’s patches were analysed by confocal microscopy. Membrane prion protein expression was studied by flow cytometry. In Peyer’s patches (PP), dissected at day one and day 105 after oral exposure to scrapie, we observed an increased population of DCs localised in the follicular-associated epithelium. On day 105, PrPSc was found in the follicles inside the PP of prion-infected mice. A subset of Peyer’s patches DCs, which did not express cellular prion protein on their surface in non-infected mice conditions, was prion-positive in scrapie conditions. Within Peyer’s patches oral scrapie exposure thus induced modifications of the homeostasis of DCs at the preclinical stages of the disease. These results give new arguments in favour of the implication of DCs in prion diseases.     

Transgenetic studies implicate interactions between homologous PrP isoforms in scrapie prion replication

Transgenic (Tg) mice expressing both Syrian hamster (Ha) and mouse (Mo) prion protein (PrP) genes were used to probe the mechanism of scrapie prion replication. Four Tg lines expressing HaPrP exhibited distinct incubation times ranging from 48 to 277 days, which correlated inversely with HaPrP mRNA and HaPrPC. Bioassays of Tg brain extracts showed that the prion inoculum dictates which prions are synthesized de novo. Tg mice inoculated with Ha prions had approximately 10(9) ID50 units of Ha prions per gram of brain and less than 10 units of Mo prions. Conversely, Tg mice inoculated with Mo prions synthesized Mo prions but not Ha prions. Similarly, Tg mice inoculated with Ha prions exhibited neuropathologic changes characteristic of hamsters with scrapie, while Mo prions produced changes similar to those in non-Tg mice. Our results argue that species specificity of scrapie prions resides in the PrP sequence and prion synthesis is initiated by a species-specific interaction between PrPSc in the inoculum and homologous PrPC.     

Extraneural prion neuroinvasion without lymphoreticular system infection

While prion infection of the lymphoreticular system (LRS) is necessary for neuroinvasion in many prion diseases, in bovine spongiform encephalopathy and atypical cases of sheep scrapie there is evidence to challenge that LRS infection is required for neuroinvasion. Here we investigated the role of prion infection of LRS tissues in neuroinvasion following extraneural inoculation with the HY and DY strains of the transmissible mink encephalopathy (TME) agent. DY TME agent infectivity was not detected in spleen or lymph nodes following intraperitoneal inoculation and clinical disease was not observed following inoculation into the peritoneum or lymph nodes, or after oral ingestion. In contrast, inoculation of the HY TME agent by each of these peripheral routes resulted in replication in the spleen and lymph nodes and induced clinical disease. To clarify the role of the LRS in neuroinvasion, the HY and DY TME agents were also inoculated into the tongue because it is densely innervated and lesions on the tongue, which are common in ruminants, increase the susceptibility of hamsters to experimental prion disease. Following intratongue inoculation, the DY TME agent caused prion disease and was detected in both the tongue and brainstem nuclei that innervate the tongue, but the prion protein PrP(Sc) was not detected in the spleen or lymph nodes. These findings indicate that the DY TME agent can spread from the tongue to the brain along cranial nerves and neuroinvasion does not require agent replication in the LRS. These studies provide support for prion neuroinvasion from highly innervated peripheral tissues in the absence of LRS infection in natural prion diseases of livestock.     

PrP-dependent association of prions with splenic but not circulating lymphocytes of scrapie-infected mice.

An intact immune system, and particularly the presence of mature B lymphocytes, is crucial for mouse scrapie pathogenesis in the brain after peripheral exposure. Prions are accumulated in the lymphoreticular system (LRS), but the identity of the cells containing infectivity and their role in neuroinvasion have not been determined. We show here that although prion infectivity in the spleen is associated with B and T lymphocytes and to a lesser degree with the stroma, no infectivity could be detected in lymphocytes from blood. In wild-type mice, which had been irradiated and reconstituted with PrP-deficient lymphohaematopoietic stem cells and inoculated with scrapie prions, infectivity in the spleen was present in the stroma but not in lymphocytes. Therefore, splenic B and T lymphocytes can either synthesize prions or acquire them from another source, but only when they express PrP.     

Prolonged follicular helper T cell responses in ME7 scrapie-infected mice

We previously reported that mice intracerebrally inoculated with the mouse-adapted scrapie strain ME7 have markedly diminished T zones in the spleen due to the decreased expression of CCL19 and CCL21. In addition, follicular dendritic cell networks in germinal centers were larger in ME7-infected spleens compared to uninfected spleens. As an extension of that study, we set out to determine how ME7 infection affects spleen structure and follicular helper T (Tfh) cell responses in mice. For this study, mice were intraperitoneally inoculated with brain homogenate of the ME7 inoculum and spleens were analyzed 50, 130, and 200 days after inoculation and compared with those from uninfected mice. The result showed that ME7- infected mice had increased Tfh cell responses which were maintained until end-stage prion disease. Although CD4 T cells decreased in white pulps, they increased in germinal centers, and expressed higher levels of the Tfh-related genes, such as Bcl6, Il21, Cxcr5, Icos, and Pdcd1. In addition, ME7-infected spleens had increased numbers of CD4 memory T cells. These data indicate that although ME7 infection led to impaired splenic white pulp structure, CD4 memory T cells were increased and Tfh cell responses were required and prolonged to provide help for the replication and accumulation of pathogenic prion protein in germinal centers.     

Astrocyte-specific expression of hamster prion protein (PrP) renders PrP knockout mice susceptible to hamster scrapie.

Transmissible spongiform encephalopathies are characterized by spongiosis, astrocytosis and accumulation of PrPSc, an isoform of the normal host protein PrPC. The exact cell types responsible for agent propagation and pathogenesis are still uncertain. To determine the possible role of astrocytes, we generated mice devoid of murine PrP but expressing hamster PrP transgenes driven by the astrocyte-specific GFAP promoter. After inoculation with hamster scrapie, these mice accumulated infectivity and PrPSc to high levels, developed severe disease after 227 +/- 5 days and died 7 +/- 4 days later. Therefore, astrocytes could play an important role in scrapie pathogenesis, possibly by an indirect toxic effect on neurons. Interestingly, mice expressing the same transgenes but also endogenous murine PrP genes propagated infectivity without developing disease.     

Scrapie-infected murine neuroblastoma cells produce protease-resistant prion proteins.

Scrapie and Creutzfeldt-Jakob disease are transmissible, degenerative neurological diseases caused by prions. Considerable evidence argues that prions contain protease-resistant sialoglycoproteins, designated PrPSc, encoded by a cellular gene. The prion protein (PrP) gene also encodes a normal cellular protein designated PrPC. We established clonal cell lines which support the replication of mouse scrapie or Creutzfeldt-Jakob disease prions. Mouse neuroblastoma N2a cells were exposed to mouse scrapie prions and subsequently cloned. After limited proteinase K digestion, three PrP-immunoreactive proteins with apparent molecular masses ranging between 20 and 30 kilodaltons were detected in extracts of scrapie-infected N2a cells by Western (immuno-) blotting. The authenticity of these PrPSc molecules was established by using monospecific antiserum raised against a synthetic peptide corresponding to a portion of the prion protein. Those clones synthesizing PrPSc molecules possessed scrapie prion infectivity as measured by bioassay; clones without PrPSc failed to demonstrate infectivity. Detection of PrPSc molecules in scrapie-infected N2a cells supports the contention that PrPSc is a component of the infectious scrapie particle and opens new approaches to the study of prion diseases.     

Molecular biology and transgenetics of prion diseases.

Considerable progress has been made deciphering the role of an abnormal isoform of the prion protein (PrP) in scrapie of animals and Gerstmann-Str?ussler syndrome (GSS) of humans. Some transgenic (Tg) mouse (Mo) lines that carry and express a Syrian hamster (Ha) PrP gene developed scrapie 75 d after inoculation with Ha prions; non-Tg mice failed to show symptoms after greater than 500 d. Brains of these infected Tg(HaPrP) mice featured protease-resistant HaPrPSc, amyloid plaques characteristic for Ha scrapie, and 10(9) ID50 units of Ha-specific prions upon bioassay. Studies on Syrian, Armenian, and Chinese hamsters suggest that the domain of the PrP molecule between codons 100 and 120 controls both the length of the incubation time and the deposition of PrP in amyloid plaques. Ataxic GSS in families shows genetic linkage to a mutation in the PrP gene, leading to the substitution of Leu for Pro at codon 102. Discovery of a point mutation in the Prp gene from humans with GSS established that GSS is unique among human diseases--it is both genetic and infectious. These results have revised thinking about sporadic Creutzfeldt-Jakob disease, suggesting it may arise from a somatic mutation. These findings combined with those from many other studies assert that PrPSc is a component of the transmissible particle, and the PrP amino acid sequence controls the neuropathology and species specificity of prion infectivity. The precise mechanism of PrPSc formation remains to be established. Attempts to demonstrate a scrapie-specific nucleic acid within highly purified preparations of prions have been unrewarding to date. Whether transmissible prions are composed only of PrPSc molecules or do they also contain a second component such as small polynucleotide remains uncertain.     

High incidence of scrapie induced by repeated injections of subinfectious prion doses

To clarify the mechanisms leading to the development of Creutzfeldt-Jakob disease in some recipients of pituitary-derived human growth hormone (hGH), we investigated the effects of repeated injections of low prion doses in mice. The injections were performed, as in hGH-treated children, by a peripheral route at short intervals and for an extended period. Twelve groups of 24 mice were intraperitoneally inoculated one, two, or five times per week for 200 days with 2 x 10(-5) to 2 x 10(-8) dilutions of brain homogenate containing the mouse-adapted C506M3 scrapie strain. Sixteen control mice were injected once a week for 200 days with a 2 x 10(-4) dilution of normal brain homogenate. Of mice injected in a single challenge with a scrapie inoculum of a 2 x 10(-4), 2 x 10(-5), or 2 x 10(-6) dilution, 2/10, 1/10, and 0/10 animals developed scrapie, respectively. Control mice remained healthy. One hundred thirty-five of 135 mice injected with repeated prion doses of a 2 x 10(-5) or 2 x 10(-6) dilution succumbed to scrapie. Of mice injected with repeated scrapie doses of a 2 x 10(-7) or 2 x 10(-8) dilution, 52/59 and 38/67 animals died of scrapie, respectively. A high incidence of scrapie was observed in mice receiving repeated doses at low infectivity, whereas there was no disease in mice that were injected once with the same doses. Repeated injections of low prion doses thus constitute a risk for development of prion disease even if the same total dose inoculated in a single challenge does not induce the disease.