Efficient enrichment cloning of TAL effector genes from Xanthomonas.

Many plant-pathogenic xanthomonads use a type III secretion system to translocate Transcription Activator-Like (TAL) effectors into eukaryotic host cells where they act as transcription factors. Target genes are induced upon binding of a TAL effector to double-stranded DNA in a sequence-specific manner. DNA binding is governed by a highly repetitive protein domain, which consists of an array of nearly identical repeats of ca. 102 base pairs. Many species and pathovars of Xanthomonas, including pathogens of rice, cereals, cassava, citrus and cotton, encode multiple TAL effectors in their genomes. Some of the TAL effectors have been shown to act as key pathogenicity factors, which induce the expression of susceptibility genes to the benefit of the pathogen. However, due to the repetitive character and the presence of multiple gene copies, high-throughput cloning of TAL effector genes remains a challenge. In order to isolate complete TAL effector gene repertoires, we developed an enrichment cloning strategy based on •genome-informed in silico optimization of restriction digestions,•selective restriction digestion of genomic DNA, and•size fractionation of DNA fragments. Our rapid, cheap and powerful method allows efficient cloning of TAL effector genes from xanthomonads, as demonstrated for two rice-pathogenic strains of Xanthomonas oryzae from Africa.

Method overview

In our previous studies, a few tal genes were isolated from African strains of Xanthomonas oryzae. Among them, only tal5 from MAI1 and talC from BAI3 have been characterized as major virulence TAL effectors [1,2]. These studies relied on screening of genomic DNA cosmid libraries to isolate and sequence tal genes, which is a laborious and time-consuming process. Due to the presence of multiple, very similar tal genes in X. oryzae strains and due to their highly repetitive character, PCR amplification is not feasible [3]. Recently, a method based on size fractionation of restriction-digested genomic DNA was developed, which made use of two conserved BamHI restriction sites, one at the ATG start codon and another one approximately 150 bp upstream of the stop codon [4]. Size-fractionated BamHI fragments (1.5–7.5 kb), covering most of the tal gene sequence, were cloned into pUC19, followed by transformation into Escherichia coli. Dot blot hybridization revealed that 115 out of 3000 clones contained a tal gene (i.e. less than 4%), which was further confirmed by Southern blot analyses and Sanger DNA sequencing. Here, we improve this method by making use of the extreme similarity among repeat DNA sequences and the strong conservation of the N- and C-terminal regions to identify frequently cutting restriction enzymes that do not cut within any of the tal genes. Complete combinatorial digestion of genomic DNA with BamHI and two additional frequent cutters for counter-selection, followed by size fractionation of DNA fragments allows rapid, cheap and efficient cloning of tal gene BamHI fragments from xanthomonads, as demonstrated with two African strains of X. oryzae.

In silico combinatorial restriction digestion

Using BioEdit (http://www.mbio.ncsu.edu/bioedit/page2.html), we identified 54 restriction enzymes that would not cleave in any of 71 tal gene BamHI fragments from X. oryzae that were retrieved from GenBank. From this analysis, two restriction enzymes, ApaLI (GTGCAC) and SfoI (GGCGCC), were selected for further analyses. Genome sequences of nine X. oryzae strains were then used for in silico combinatorial restriction digestion using Microsoft Office software (Microsoft Word and Microsoft Excel) (Supplemental Table 1). First, each genome sequence was converted into a consecutive list of BamHI fragments in Word. Upon conversion from text to table format, all virtual DNA fragments were transferred to an Excel spreadsheet. In Excel, fragment sizes and absence/presence of ApalI and SfoI sites were scored using appropriate formulas. DNA fragments were sorted by the absence/presence of the two restriction sites and the size of virtual BamHI fragments. tal gene-related BamHI fragments were identified by TBLASTN searches. Finally, data were re-analyzed assuming isolation of 2– 5 kb DNA fragments prior to cloning.

Assuming that only DNA fragments flanked by two BamHI overhangs would be cloned, but none of the fragments that contain only one BamHI overhang (which could nevertheless occur by head-to-tail ligation of two fragments flanked by one BamHI overhang and one overhang from the other two enzymes), one would expect to obtain between 10% and 20% of tal gene BamHI fragments without size fractionation and between 50% and 90% of tal gene BamHI fragments with size fractionation (2–5 kb DNA fragments). However, our simulation revealed that this procedure would miss to isolate tal-gene related DNA fragments from truncated tal genes (two in PXO86, one in KACC10331, one in BAI11), which nevertheless are of functional relevance [5,6]. Most importantly, this method should allow isolation of BamHI fragments from all full-length tal genes, corresponding to the full functional TALome that induces the expression of resistance or susceptibility genes.

Bacterial strains, plasmids and growth conditions

The bacterial strains used in this study were Escherichia coli DH5a (Stratagene, La Jolla, CA, USA) and X. oryzae strains BAI3 and MAI1 [7]. E. coli bacteria were cultivated at 37 °C in lysogenic broth (LB), X. oryzae strains at 28 °C on PSA medium (10 g peptone, 10 g sucrose, 1 g glutamic acid, 16 g agar, l−1 H2O). Antibiotics were used at the following concentrations: gentamicin, 20 μg/ml.

Plasmids were introduced into E. coli by electroporation and into X. oryzae by biparental conjugation using E. coli strain S17-1. The plasmid used to clone tal gene BamHI fragments, pSKX1, was obtained upon digestion of pSXK1-talC [2] with BamHI and re-ligation, leading to a construct where the translational start codon with an overlapping BamHI site (ATGGATCC, BamHI site underlined) of talC is fused to 151 base pairs corresponding to the 3′ end of the talC open reading frame downstream of the second BamHI site of talC.

DNA extraction

Total genomic DNA of two African X. oryzae strains was prepared using the midi-prep Qiagen® genomic DNA preparation protocol for bacteria using Genomic-tips 100/G (QIAGEN SAS, Courtaboeuf, France). The total genomic DNA was resuspended in 500 μl of double-destilled H2O, aliquoted into several Eppendorf tubes and stored at −30 °C. Plasmid DNA wase extracted using QIAprep Spin Miniprep Kit or QIAGEN Plasmid Kit for midiprep (Qiagen SAS).

Enrichment cloning of tal gene BamHI fragments

20 μg of genomic DNA were digested with BamHI-HF, SfoI and ApaLI in CutSmart® Buffer (New England Biolabs SAS, Enry, France) at 37 °C overnight. The High-Fidelity version of BamHI was used to avoid star activities upon prolonged incubation. Digestion products were purified using QIAquick PCR Purification Kit (Qiagen SAS) for column purification or using the QIAquick Gel Extraction Kit (Qiagen SAS) to isolate 2–5 kb DNA fragments. DNA fragments were cloned into BamHI-digested pSKX1, which was dephosphorylated by TSAP thermosensitive alkaline phosphatase (Promega, Charbonnières-les-Bains, France) and purified using the QIAquick PCR Purification Kit (Qiagen SAS). Upon ligation with T4 DNA ligase (Promega) at 4 °C overnight, DNA was transformed into E. coli. Transformants were plated onto LB-gentamycin plates and bacterial colonies were screened by PCR for the presence of a cloned tal gene BamHI fragment by polymerase chain reaction (PCR).

PCR screening of bacteria for cloned tal gene BamHI fragments

To assess the presence and orientation of cloned tal gene BamHI fragments, two primers pairs were designed manually and used in PCR assays. Firstly, primers PthXo1-nt-Fw1 (5′-GCAGCTTCAGCGATCTGCTC) and PthXo1-nt-Rev2 (5′-TCAGGGGGGCACCCGTCAGT) were used to amplify a DNA fragment of 590–660 bp corresponding to the highly conserved N-terminal region of all X. oryzae TAL effectors (size variation is due to an in-frame deletion in a few tal genes). Amplification was carried out with an initial denaturation step of 5 min at 95 °C, 30 cycles of 30 s at 95 °C, 30 s at 62 °C, and 45 s at 72 °C, and a final elongation step of 10 min at 72 °C. Secondly, to determine the orientation of the cloned tal gene BamHI fragment in pSKX1, primer AvrXa7-Ct-Fw2 (5′-GCGTTGGCCGCGTTGACCAA), which anneals to the 3′ end of the cloned tal gene BamHI fragment, and pSKX1-Rev (5′-GGGCACCAATAACTGCCTTA-3′), which anneals to the plasmid vector pSKX1, were used to amplify a DNA fragment of 904 bp when a tal gene BamHI fragment is inserted in frame with the 3′ portion of the vector-borne talC fragment. Amplification was carried out with an initial denaturation step of 5 min at 95 °C, 30 cycles of 30 s at 95 °C, 30 s at 64 °C, and 1 min at 72 °C, and a final elongation step of 10 min at 72 °C. The empty vector, pSKX1, and a plasmid containing the tal5 gene from strain MAI1, pSKX1-tal5, served as negative and positive control, respectively.

DNA sequencing of cloned tal gene BamHI fragments

Plasmid DNA from positive colonies in the first PCR assay was digested with BamHI to estimate the insert size and thus, the expected number of repeats, via gel electrophoresis in a 1.2% agarose gel. Repeat regions were Sanger sequenced from both sides with two oligonucleotide primers: forward (5′-GCCGGATCAGGGCGAGATAACT) and reverse (5′-CACTGACGGGTGCCCCCCTGAA). Plasmid DNA from in-frame clones, as revealed by the second PCR assay, were Sanger sequenced from both sides of vector using primers pSKX1-For (5′-GGCACGACAGGGTTTTCCCGAC) and pSKX1-Rev (5′-GGGCACCAATAACTGCCTTA).

Restriction fragment length polymorphism analyses

To confirm that the two counter-selection restriction enzymes, ApaLI and SfoI, do not cut within tal gene BamHI fragments, we performed restriction fragment length polymorphism (RFLP) analyses. 3 μg of single (BamHI) or triple (BamHI, ApaLI and SfoI) digested genomic DNA of the African X. oryzae strains BAI3 and MAI1 were electrophoretically separated on 0.8% agarose gels in 0.5 × TBE (Tris-Boris-EDTA) buffer at 25 V for 84 h at 4 °C. Upon DNA fragments were transferred to an Amersham Hybond-N + membrane by Southern blotting (GE Healthcare, Vélizy-Villacoublay, France). A 663-bp fragment of the tal5 gene was PCR-amplified and digoxigenin-labeled with DIG-High Prime (Roche Diagnostics, Meylan, France) using primers pthXo1-nt-Fw1 and PthXo1-nt-rev2, as described above.

Efficiency of tal gene BamHI fragment enrichment cloning

As a proof-of-concept, we applied our strategy to two African strains of X. oryzae, BAI3 and MAI1 [7]. 20 μg of genomic DNA of MAI1 was digested with BamHI only or in combination with ApaLI/SfoI, respectively. The digestion products were aliquoted and subjected to two independent purification treatments, i.e. column (total fragments) and gel purification (2 to 5 kb fragments). Both sets of DNA fragments were ligated into plasmid pSKX1. tal gene-harboring colonies were rarely detected in both single BamHI digestion treatments (yielding 0.6% and 0.8% for column purification and gel purification, respectively) (Supplemental Table 2). Upon combinatorial digestion, column purification resulted in 10 tal gene-positive clones among 324 tested colonies (3.1%). Strikingly, counter-selection digestion combined with size fractionation resulted in 28 out of 89 tal gene-positive colonies (31.5%), as revealed by PCR screening. Cloned DNA fragments of all 28 positive colonies were sequenced from both sides using primers that anneal to the plasmid vector, thus confirming the presence of tal gene BamHI fragments in all tal gene-positive colonies. Similar results were obtained when the enrichment cloning was repeated for strain MAI1 (25.0% tal gene-positive colonies) and applied to strain BAI3 (26.9% tal gene-positive colonies) (Supplemental Table 2), thus demonstrating superiority of the new method over previous cloning approaches.

Additional information

Background information and significance of the method

Strains of Xanthomonas spp. cause important diseases of many economically important crop and ornamental plants. In most cases, pathogenicity depends of a set of type III effector proteins, which are injected into host cells via a molecular syringe, the type III secretion system (T3SS). Among all the type III effectors, one class is of particular interest for Xanthomonas: the Transcription Activator-Like (TAL) effectors. TAL effectors are conserved in many Xanthomonas spp. and some of them have been shown to significantly contribute to pathogenicity [8]. Upon injection via the T3SS, TAL effectors localize into the host nucleus to directly or indirectly activate the expression of specific host genes [9,10]. Among them are so-called susceptibility (S) genes, the induction of which promotes bacterial colonization in the affected plant tissues and/or development of disease symptoms [11].

Plant gene induction by TAL effectors depends on their central repeat region as well as C-terminally located nuclear localization signals and an activation domain [8]. The central repeat region is composed by almost identical tandem repeats (typically ranging from 33 to 35 amino acids) where residues at positions 12 and 13 are hypervariable, also referred to as repeat variable di-residues (RVDs). As demonstrated by the TAL effector-DNA binding code, the string of RVDs of each TAL effector determines the DNA sequence (or effector binding element, EBE) to which it binds [12,13]. Upon refinement of the code, various in silico platforms were developed that allow prediction of TAL effector target genes in complex plant genomes [[14], [15], [16]]. In order to understand the collective function and evolution of TAL effector genes, there is a need to isolate and sequence complete repertoires of TAL effectors (e.g. TALomes) from multiple strains of Xanthomonas. Yet, efficient cloning was hampered by the facts that the genes have a highly repetitive structure and that most strains contain multiple copies of TAL effectors, thus limiting the usefulness of PCR-based approaches [3]. Recently, long-read, single molecule, real-time (SMRT), a.k.a. PacBio sequencing technology emerged as a new strategy for full TALome sequencing [[17], [18], [19]]. Yet, functional studies require the molecular cloning of tal genes. To facilitate such studies, several pipelines were developed that allow the assembly of artificial or designer tal genes [20]. However, these pipelines rely on prior knowledge of the tal gene sequences and they do not generate the authentic tal gene sequences, which do not only vary in the RVD codons. Here, we obtained the full TALome of strain MAI1 in only two weeks without prior knowledge of the tal gene sequences. Remarkably, this protocol can easily be parallelized and the obtained clones in the expression vector pSKX1 can directly be used in pathogenicity assays. Consequently, our new enrichment cloning procedure is expected to spur TALome research by allowing medium-throughput TALome cloning.

Despite the fact that the procedure was developed for strains of X. oryzae, it can easily be adapted to other xanthomonads as long as the sequence for a few tal genes is known, thus allowing to find appropriate restriction enzymes for counter-selection. Before use of new enzyme combinations it is recommended to perform Southern blot analyses comparing single and multiple digested DNA samples in order to avoid enzymes that would cleave within the tal gene BamHI fragments.