ENDO1 is expressed at low level in Arabidopsis during plant senescence 31. In order to facilitate the expression of the protein, the ENDO1 ORF was inserted between the 35S promoter and the transcriptional terminator of CaMV in the binary vector pBIN61 (Figure 1a) 30. The construct was transformed into Agrobacterium and co-agroinfiltrated into N. benthamiana leaves in the presence of the p19 protein of tomato bushy stunt virus (TBSV), that prevents the onset of post-transcriptional gene silencing (PTGS) in infiltrated tissues and allows a high level of transient expression (Figure 1a) 32. As negative control the plants were also agroinfiltrated with a construct expressing the green fluorescent protein (GFP).

Ten grams of agroinfiltrated leaves were homogenized and cleared by centrifugation in the absence and in the presence of 30% ammonium sulphate. Ammonium sulphate was then added to 80% final concentration in the supernatant and the pelleted proteins were resuspended in buffer containing 50% glycerol, dialysed against the same buffer and stored at -80°C.

To test whether ENDO1 obtained from this simplified purification protocol is active we tested the protein extract on heteroduplex DNA created from two clones that differ by a single base insertion (Figure 1b). In this experiment we predicted that if the ENDO1 preparation contains a mismatch specific endonuclease, the heteroduplex DNA will be cleaved at the mismatch site releasing two bands of 208 bp and 436 bp (Figure 1b). Duplex DNAs were incubated with dilutions of ENDO1 or GFP protein extracts. ENDO1 led to the predicted digestion product and as expected the GFP control showed undetectable mismatch specific cleavage activity (Figure 1b, lower picture). As shown for His-tag purified ENDO1 30 and CEL I 20, at high concentration (50 to 250 fold dilution, Figure 1b), ENDO1 presented double strand DNAse activity and digested the entire fragments. From this biochemical analysis we concluded that ENDO1 simplified purification led to an active enzyme and the test mutation was detected at dilutions ranging from 500 to 10 000 fold (Figure 1b, upper picture). Based on this analysis and duplicates of this experiment (data not shown), we estimate that the amount of enzyme produced from 10 grams of agroinfiltrated leaf material, and used at 10 000 fold dilution, is enough to carry more than a million mismatch detection tests.

We tested ENDO1 prepared as above in mutation diagnostics in humans. As a proof of concept, we investigated the detection of the factor V G1691A mutation associated with venous thrombosis 3334. Genomic DNA was extracted from individuals harboring or not the G1691A mutation and analyzed using ENDO1 system. To reduce the diagnostic cost, we combined the ENDO1 mismatch detection system with the use of labelled universal primers in the amplification of the target sequence (Table 1). In this analysis we expected that if the 611 bp heteroduplex DNA fragment of factor V is cleaved at the mismatch site, two bands of 300 bp and 311 bp would be released (Figure 2a). Results obtained from both 700 and 800 channels of the LI-COR4300 show that ENDO1 recognizes the G1691A mutation (Figure 2b, lane 1 and 2). No cleavage was observed on PCR product obtained from a control person not carrying the G1691A mutation (Figure 2b, lane 3 and 4). The experiment was repeated three times with similar results (data not shown). As expected the presence of G1691A mutation was confirmed by sequencing in the patient heterozygote for the mutation (Figure 2c). Because mutation detection relies on heteroduplex formation between two alleles, detection of the G1691A mutation at the homozygous state will require mixing genomic DNA from a control individual not harbouring the mutation with that of an individual to be tested.

Human immunodeficiency virus type 1 (HIV-1) present in infected individuals has been described as quasispecies of related but distinct viruses 35363738. When the selective pressure of antiretroviral therapy is exerted on such a population, drug-resistant mutants may emerge and consequently lead to treatment failure 3537. The objective of this work was to assess the HIV-1 quasispecies evolution during different treatments. Direct sequencing of PCR amplified HIV-1 DNA fragment from an infected individual will detect only predominant mutations. Detailed analysis of the HIV quasispecies using sequencing would require the cloning of PCR products and systematic sequencing of individual clones. A list of alternative technologies have been also tested to identify minor HIV drug-resistant populations 3940414243.

Previously we demonstrated that ENDO1 can detect rare alleles in pools of DNA 30. Thus, we tested the use of ENDO1 system as a fingerprinting protocol to assess the HIV-1 quasispecies evolution during different treatments. We focused the analysis on the sequence variation in a DNA fragment of 843 bp coding for the reverse transcriptase gene of HIV-1. Genomic DNAs were extracted from peripheral blood mononuclear cells of an HIV-1 infected patient before (Figure 2d, lane 1) and after 48 weeks of antiretroviral therapy (Figure 2d, lane 2). The reverse transcriptase gene was PCR amplified and analysed using ENDO1. Two different patterns of bands, representing fingerprints of HIV-1 quasispecies at each time, were observed (Figure 2d). Comparing the overall patterns and changes in intensity of particular fragments allowed the identification of mutations that either appeared or were lost upon treatment (Figure 2d, empty arrows). Bands of invariant intensity indicate stable mutations (Figure 2d, filled arrows).

TILLING (Targeting Induced Local Lesions IN Genomes) method combines the induction of a high number of random mutations with mutagens like Ethyl Methane Sulfonate (EMS) and mutational screening systems to discover induced mutations in sequence DNA targets 44. This reverse genetic strategy encompasses all types of organisms as plants, animals, bacteria, without being subjected to the genome size 15161718194546.

To test whether ENDO1, purified using the simplified protocol, in combination with universal primers is efficient in TILLING, we screened 4717 M2 families of Pisum sativum-EMS mutagenized population (Dalmais et al., unpublished data) for induced mutations in the Retinoblastoma-Related gene, RBR. The RBR protein is well-known as a regulator of cell division, differentiation and endoreduplication in plants 47. So far, the use of insertion mutants to investigate the roles of RBR in plant development has been restricted, due to the gametophytic lethality 48. Two DNA fragments of 1078 bp and 972 bp of RBR were PCR amplified using nested PCR in combination with universal primers (Table 1; Figure 3a). PCR products were digested with ENDO1 and analysed on LI-COR4300 gels as described above (Figure 3b). From 1078 bp and 972 bp amplicons we obtained 34 and 16 mutants, respectively (Figure 3c). From this, 3 were intronic, 17 were silent and 30 were missense mutations. No nonsense mutants were recovered, likely because of their gametophytic deleterious effects 48.

To evaluate the robustness of ENDO1 purified using the simplified protocol, in combination with universal primers for mutation detection, RBR TILLING was carried out with conventional method based on Ni-Column-purified ENDO1 and gene specific primers. A similar number of mutants were identified (data not shown). Based on the TILLING of RBR and other targets (data not shown) we concluded that the ENDO1 simplified protocol is suitable for high throughput TILLING screen.

EcoTILLING is a means to determine the extent of natural sequence variation across many germplasms, enabling both SNP discovery and haplotyping 22. This technique is now applied to rice, maize, lotus, poplar 49, Brassica, zebrafish, Drosophila, Caenorhabditis and human 50, indicating its broad applicability.

In Arabidopsis thaliana, we analysed natural sequence variations within a DNA fragment of 456 bp (Figure 4a) using the ENDO1 system (Figure 4c) and sequence analysis (Figure 4b), in two Arabidopsis ecotypes, Columbia and Landsberg. We expected that if the 456 bp heteroduplex DNA fragment, created from Columbia and Landsberg pooled genomic DNAs (lane 1), is cleaved at the mismatch sites, four bands of 325, 274, 182 and 131 bp would be released. Results obtained from both channels 700 and 800 of the LI-COR4300 show that ENDO1 recognizes, with a high specificity the three base pair INDEL and the G to A substitution (Figure 4c, lane 1). No cleavage was observed on PCR product obtained from the control Columbia and Landsberg-homoduplex DNAs (Figure 4c, lane 2 and 3, respectively). The obtaining of four bands indicates that ENDO1 cleaves the DNA fragments partially. This characteristic of ENDO1 enhances the potential of the system, because it is able to detect sequence variation in a DNA fragment carrying multiple mutations in a single reaction.

pBIN35S-ENDOI construct was described previously 52. Agrobacterium transient overexpression in Nicotiana benthamiana was performed as described previously 53 except that the cells were co-infiltrated into N. benthamiana leaves in presence of Agrobacterium harbouring pBIN61-p19 construct 32. Agroinfiltrated plants were incubated for at least 60 to 96 hours in the green house before protein extraction.

Ten grams of agroinfiltrated leaves were homogenized in 10 ml buffer containing 0.1 M Tris-HCl pH 8, 200 μM phenylmethylsulphonyl fluoride (PMSF), 0.125 mM β mercaptoethanol and 10% of glycerol and cleared by centrifugation at 3,000 × g for 30 min. Ammonium sulphate was added to the supernatant to the final concentration of 30% and the samples incubated on ice for 1 hour were centrifuged at 30,000 × g for 30 min. Ammonium sulphate was added to the supernatant to 80% final concentration and the proteins were precipitated by centrifugation as above. The pellet was resuspended in 500 μl of dilution buffer (50 mM Tris-HCl pH 8, Glycerol 50%, 100 μM PMSF), dialysed against the same buffer and stored at -80°C.

In the TILLING screen, target loci were PCR-amplified using nested-PCR and universal primers. The first PCR amplification is a standard PCR reaction using target-specific primers 28. One microlitre of the first PCR served as a template for the second nested PCR amplification, using a mix of gene-specific inner primers carrying a universal M13 tail (Table 1), in combination with M13 universal primers, M13F700 (CACGACGTTGTAAAACGAC) and M13R800 (GGATAACAATTTCACACAGG), labelled at the 5'end with infra-red dyes IRD700 and IRD800 (LI-COR^®, Lincoln, Nebraska, USA), respectively. This PCR was carried out with each primers at 0.1 μM, using the following 2 steps cycling program: 94°C for 2 min, 10 cycles at 94°C for 15 sec, primers specific annealing temperature for 30 sec and 72°C for 1 min, followed by 25 cycles at 94°C for 15 sec, 50°C for 30 sec and 72°C for 1 min, then a final extension of 5 min at 72°C. The PCR amplification of factor V, HIV-1 reverse transcriptase and Arabidopsis BAC 15K9 DNA fragment was carried out using only the second PCR conditions described above using 30 ng of genomic DNA as template. Mutation detection, in non-purified PCR products, was carried out as described previously 30 except that the crude extracted enzyme was used at 10 000 fold dilution and 0.6 μl of ENDO1-digestion products were loaded on the gel using disposable paper membrane combs (The Gel Company, San Francisco, USA).