Ultra-efficient PrP(Sc) amplification highlights potentialities and pitfalls of PMCA technology

Abstract

In order to investigate the potential of voles to reproduce in vitro the efficiency of prion replication previously observed in vivo, we seeded protein misfolding cyclic amplification (PMCA) reactions with either rodent-adapted Transmissible Spongiform Encephalopathy (TSE) strains or natural TSE isolates. Vole brain homogenates were shown to be a powerful substrate for both homologous or heterologous PMCA, sustaining the efficient amplification of prions from all the prion sources tested. However, after a few serial automated PMCA (saPMCA) rounds, we also observed the appearance of PK-resistant PrP(Sc) in samples containing exclusively unseeded substrate (negative controls), suggesting the possible spontaneous generation of infectious prions during PMCA reactions. As we could not definitively rule out cross-contamination through a posteriori biochemical and biological analyses of de novo generated prions, we decided to replicate the experiments in a different laboratory. Under rigorous prion-free conditions, we did not observe de novo appearance of PrP(Sc) in unseeded samples of M109M and I109I vole substrates, even after many consecutive rounds of saPMCA and working in different PMCA settings. Furthermore, when positive and negative samples were processed together, the appearance of spurious PrP(Sc) in unseeded negative controls suggested that the most likely explanation for the appearance of de novo PrP(Sc) was the occurrence of cross-contamination during saPMCA. Careful analysis of the PMCA process allowed us to identify critical points which are potentially responsible for contamination events. Appropriate technical improvements made it possible to overcome PMCA pitfalls, allowing PrP(Sc) to be reliably amplified up to extremely low dilutions of infected brain homogenate without any false positive results even after many consecutive rounds. Our findings underline the potential drawback of ultrasensitive in vitro prion replication and warn on cautious interpretation when assessing the spontaneous appearance of prions in vitro.

Figure 1. Vole PMCA: early results.

A) The efficiency of vole substrate to support the amplification of PrP^Sc from different prion sources: 10% infected brain homogenate from hamster, mouse and bank vole with experimental scrapie, deer with natural CWD, cattle with natural BSE, sheep with scrapie and humans with type 1 sCJD were diluted 1∶200 in substrate from voles carrying M109M PrP genotype and amplified for 80 consecutive sonication/amplification cycles. After amplification PMCA material was PK-digested and separated by SDS-PAGE; Western blot was probed with anti-PrP 6D11 primary antibody. B) The putative de novo generation of PrP^Sc from healthy brain homogenates: samples of vole substrate carrying the M109M PrP genotype, (tubes 1 to 4) without any positive inoculum were amplified for 5 consecutive rounds of saPMCA (24 sonication/incubation cycles per round). After the third round the presence of PrP^Sc was observed in 3 out of 4 samples. After the 5 consecutive rounds all samples were positive. After amplification, PMCA material was PK digested and separated by SDS-PAGE; Western blot was probed with anti-PrP D18 primary antibody.

Figure 2. Sensitivity of detection of saPMCA using vole substrate and v586 inoculum.

For each round and seed dilution is reported the number of positive samples over the number of replicates tested. Green boxes indicate that all replicates were positive; yellow boxes indicate that less than 100% of replicates were positive, while boxes are white when no positive replicates occurred. A) Logarithmic dilutions of the seed were prepared in duplicate from 10⁻³ to 10⁻¹⁴ using vole M109M substrate. Eight negative controls were included. Samples were amplified for 7 consecutive rounds of 24 hours. All the samples were analysed by Western blot. Results showed that after the first round, amplifications were observed in dilutions 10⁻³, 10⁻⁴ and in one of two duplicates of 10⁻⁵. After the second round, positive PrP^Sc signals were detected also in 10⁻⁵, 10⁻⁶, and 10⁻⁷ dilutions and in one of the 10⁻⁸ duplicates. After three consecutive rounds, 10⁻⁹ reached positive amplification. Lower dilutions were amplified between the 5^th and 7^th round but, at the same time, also negative controls became positive. B) logarithmic dilutions of v586 in duplicate from 10⁻³ to 10⁻¹⁴ were amplified for 7 consecutive 24-hour rounds using vole M109M substrate. Two groups of 8 unseeded negative controls were included.(groups A and B). All the seeded and unseeded samples were amplified together, but passages of control group A were carried out in a prion-free environment (new laminar flow hood and pipette), while samples from control group B were processed together with the seeded tubes. Results showed similar sensitivity compared to previous experiments, with amplification of 10⁻⁵ dilution after the first round and up to 10⁻⁹ after the second round. After the third round the 10⁻¹⁰, 10⁻¹¹ and 10⁻¹⁴ dilutions along with six out of eight negative samples from control group B become positive. One round later, three out of eight control group A samples also turned positive.

Figure 3. Comparison between PrP^Sc from seeded and unseeded PMCA reactions.

Characterization of PrP^Sc from positive samples of the experiment described in Figure 2A by Western blot, stained with SAF84 primary antibody. The figure shows that samples A (PrP^Sc from 10% brain homogenate used as positive inoculum), B (10⁻⁵ dilution of the inoculum after a single 24-hour round of PMCA) and C (PrP^Sc from an unseeded negative control which was positive after 6 consecutive rounds of saPMCA) share the same apparent molecular weight before or after deglycosylation with PNGase.

Figure 4. Kinetics of PrP^Sc amplification during a single round of PMCA.

Logarithmic dilutions of v586 seed from 10⁻³ to 10⁻¹⁰, were prepared in quadruplicate using M109M vole substrate. Samples were amplified for a single round of PMCA together with unseeded negative controls. The level of amplification was evaluated by western blot in seeded and unseeded samples after 12 hours (panels A and E), 24 hours (panels B and F), 48 hours (panels C and G) and 72 hours (panels D and H). Western blot was probed with SAF84 primary antibody.

Figure 5. Limit of detection of v586 using saPMCA.

Logarithmic dilutions of v586 seed from 10⁻³ to 10⁻¹⁸ were prepared using M109M vole substrate. and were amplified for 7 consecutive 48-hour rounds of saPMCA ( = 96 cycles of incubation/sonication), together with 8 unseeded negative controls (A to H). Positive amplification reached 10⁻⁹ dilution after the 1^st round. In the subsequent rounds all dilutions up to 10⁻⁹ continued to be positive while negative controls and dilutions from 10⁻¹⁰ to 10⁻¹⁸ remained negative. Western blot was stained with SAF84 primary antibody.