Xac isolates and strains collected from different geographical regions (n = 60) were tested to validate all diagnostic assays. Additionally, a collection of type and non-type strains of all pathovars of X. arboricola species, other closely related Xanthomonas species (n = 30), and microorganisms (bacteria and fungi) isolated from symptomatic hazelnut and walnut tissues, i.e. Pseudomonas spp., Pseudomonas avellanae, Sphingomonas spp. and Xanthomonas campestris (n = 46) were included in assays ( Table 1 , Supplementary Table S1 ).

Xanthomonads were grown on yeast extract nutrient agar (YNA) or yeast extract dextrose calcium carbonate (YDC; Schaad et al., 2001) and pseudomonads were cultured on King’s B medium (King et al., 1954) at 28°C for 24 to 48 h. The nine fungal isolates were grown on PDA (potato dextrose agar; Becton Dickinson, Sparks, MD, USA) at 24°C with an 8 h light and 16 h dark photoperiod.

Genomic bacterial DNA was isolated using the Genomic Mini bacterial DNA Purification Kit (A&A Biotechnology, Poland), the DNeasy Mericon Food Kit (Qiagen, Hilden, Germany) or Whole Blood and Tissue kit (Qiagen, Germantown, MD, USA), according to the manufacturer’s instructions. The total fungal DNA was extracted from 100 mg of mycelia scraped from 10-day-old PDA cultures with the GeneMatrix Plant & Fungi DNA Purification Kit (EURx, Gdańsk, Poland) according to the manufacturer’s instructions. The quality and total DNA concentration was estimated with a NanoDrop ND-100 or NanoDrop 2000c (ThermoFisherScientific, Waltham, MA, USA).

DNA sequences from three Xac whole genome shotgun sequencing projects (WGS) (CFBP 1159^PT, CFBP 2565 and NCCB 100457; GenBank WGS prefixes MDEA01, MDSJ01 and APMC02, respectively) was used for comparative genomic analysis. A ‘dual-BLASTn’ comparative genomics pipeline was applied to select 300-bp regions shared among these three target WGS (Schneeberger et al., 2017). After segmentation into 300 bp length fragments, duplicates were removed and Xac unique sequences were selected using BLASTn+ v.2.8.1 (Altschul et al., 1990; Camacho et al., 2009) analysis against the database derived from the three genomes. Regions obtained from this workflow were further checked for Xac specificity using online BLASTn searches against the nr/nt and X. arboricola and Xanthomonas WGS NCBI databases (accessed in July 2019). Finally, Xac-specific DNA markers were also confirmed in three recently released Xac complete genomes (CFBP 1159^PT, CFBP 6600 and Xac 301; GenBank assemblies GCA_905220785.1, GCA_905220805.1, and GCA_905220715.1, respectively; Pothier et al., 2022).

Three Xac-specific regions and their associated primers were given ‘in-house’ names during analyses, the genome context of the regions is illustrated in Figure 1 . These regions were used to design primers for: 1) conventional PCR, 2) qPCR (SYBR^® Green and TaqMan™), and 3) LAMP. The primers for conventional PCR and qPCR were designed using the PrimerSelect program of the LASERGENE package v.9 (DNASTAR, Madison, WI, USA) and Primer3Web v.4.1.0. (Untergasser et al., 2012). LAMP primers were designed using the online platform PrimerExplorer v.5 (Eiken Chemical Co., Ltd, Tokyo, Japan, http://primerexplorer.jp/lampv5e/index.html) also including loop primers (i.e. in total six primers) to speed up the LAMP reaction (Nagamine et al., 2002). Based on the regions selected ( Figure 1 ), ten candidate primer sets were designed for conventional PCR (5, 2 and 3 primer sets based on the “region 2.4”, “PPU54630”, and “target 45” Xac-specific DNA markers, respectively), six for SYBR^® Green I qPCR (3, 1 and 2 primer sets based on the “region 2.4”, “PPU54630”, and “target 45” Xac-specific DNA markers, respectively), two for TaqMan™ qPCR (one primer pair based on the “region 2.4” and one based on the “target 45” Xac-specific DNA markers), and three for LAMP (two primer pairs based on the “region 2.4” and one based on the “PPU54630” Xac-specific DNA markers). The primers for the TaqMan™ qPCR were purchased HPLC purified since this effectively increases the melting temperature (T_m ) for shorter sequences, allowing for an overall shorter amplicon while remaining within temperature requirements. The TaqMan™ probes were designed with a 5′ FAM reporter dye and a 3’ BHQ-1 non-fluorescent quencher. Initially, the specificity of the primers, the TaqMan™ probe, and predicted amplicons to Xac were tested in silico with BLASTn searches against the nr/nt and WGS NCBI databases (accessed in July 2019). All these primer sets were then tested in vitro for specificity, sensitivity, and reproducibility during screening. Depending on the research institutions, primers were synthesized at Genomed S.A. (Warszawa, Poland), Invitrogen (ThermoFisherScientific, Waltham, MA, USA) and MilliporeSigma (Burlington, MA., USA).

The in vitro specificity of all primers was tested with purified genomic DNA of the bacteria and fungi listed in Table 1 (detailed in Supplementary Table S1 ).

To exclude potential non-specific amplification of plant genomic DNA with the primers, total plant DNA was isolated from clean asymptomatic leaves of five hazelnut cultivars (cv. ‘Cosford’, cv. ‘Merveille de Bollwiller’, cv. ‘Garibaldi’, cv. ‘Webb’s Prize Cob’ and cv. ‘Hall’s Giant’) grown in a greenhouse. Total plant DNA was isolated from leaves using the GeneMATRIX Plant & Fungi DNA Purification Kit (EURx, Gdańsk, Poland), as well as the Genomic Mini DNA Extraction Kit (A&A Biotechnology, Gdynia, Poland) to isolate bacterial DNA. Both kits were used according to the manufacturer’s instructions with the following specifications concerning the starting material. To isolate plant DNA: 100 mg from hazelnut leaves were homogenized in liquid nitrogen in a cooled mortar and pestle and transferred to a 2 ml tube before addition of 400 μl lysis buffer L. For bacterial DNA isolation: 100 mg of crushed or cut leaf tissue was placed in 20 ml of PBS buffer, incubated for 1 h at 26°C with shaking (150 rpm), pelleted by centrifugation (5 min at 12,000 × g), and then re-suspended in 100 μl Tris EDTA (TE) buffer.

Three labs participated in the specificity validation of the assays: two assay development laboratories (Poland and Serbia) and one assay testing laboratory (USA).

The reactions were conducted according to the protocols established based on the optimization of all reagents and temperature gradient analysis performed separately for each primer pair. The amplification conditions for all the primers pairs/sets are listed in Table 2 .

Amplification reactions with the four selected primer pairs for conventional PCR were conducted in a Biometra T3000 thermocycler (Biometra, Göttingen, Germany) in Poland, in a Thermo Cycler 2720 (Applied Biosystems, USA) in Serbia, and a Veriti 96-well Thermal Cycler 9902 (Applied Biosystems, USA) in the USA. The total amplification reaction mixtures for primers in 15 μl of volume included: 10 to 15 ng of DNA, 0.4 U of DreamTaq DNA Polymerase (ThermoFisherScientific, Waltham, MA, USA), 1× reaction DreamTaq Green buffer, 0.15 mM each dNTPs and 0.7 mM of each primer. The amplicons obtained in individual reactions for each primer pair were separated in 1.5% agarose gels in 0.5× TBE buffer (0.045 M Tris-boric acid, 0.001 M EDTA, pH 8.0) (Sambrook et al., 1989). To confirm the size of the obtained product O’GeneRuler100-bp DNA Ladder Plus (ThermoFisherScientific, Waltham, MA, USA) was used. Gels were stained in an ethidium bromide solution (0.5 μg ml^-1) and obtained products were visualized under UV irradiation.

SYBR^® Green I qPCRs were conducted in a Bio-Rad CFX96 (Bio-Rad, Hercules, CA, USA) with SsoAdvanced™ Universal SYBR^® Green Supermix (Bio-Rad, Hercules, CA, USA) in Poland or a Mic qPCR Cycler (Bio Molecular Systems, Australia) in Serbia. The reaction mixture in 20 μl of total volume included 1× reaction SYBR^® Green Supermix and 0.5 mM of each primer from the following primer sets: Xac2.4-2RT, Xac45-1RT, Xac45-2RT, Xac2.43RT, and 10 ng of DNA. The PCR programs for all above-listed primers are given in Table 3 . The specificity of amplification products was verified by a melting curve analysis using a progressive denaturation of products at a rising temperature ( Table 3 ). Specific melting temperatures observed are indicated in Table 2 .

The validation of the TaqMan™ qPCR was also done in Poland. The sequence of probes and primers are indicated in Table 2 . Reactions were conducted in a Bio-Rad CFX96 (Bio-Rad, Hercules, CA, USA) using the amplification conditions in Table 3 . The TaqMan™ qPCR assays were carried out in a 10 μl total reaction mixture containing 1 μl of template DNA, 0.25 μl of primers Xac-PPU54630-F and Xac-PPU54630-R (0.25 μM final concentration of each), 0.15 μl of probe Xac-PPU54630-P (0.15 μM final concentration), 1× TaqMan™ Fast Universal PCR Master Mix (Applied Biosystems, USA).

Loop‐mediated isothermal amplifications were performed on a Bio-Rad CFX96 (Bio-Rad, Hercules, CA, USA) in Poland. The reactions mixture carried out in a total volume of 20 μl contained 1× Isothermal Mastermix (OptiGene, Horsham, UK) and primers at the final concentrations as follows: outer primers F3/B3 0.2 μM each, inner primers FIB/BIP 0.8 μM each and loop primers 0.4 μM each. Fluorescence was detected on the FAM channel. The LAMP reaction mixtures were run according to conditions detailed in Table 3 .

The limits of detection (LoD) of all the DNA-template based assays were tested with 10-fold dilutions series prepared in TE buffer using bacterial genomic DNA isolated from pure cultures of CFBP 1159^PT and Xac 301. The dilution series ranged from ~10 ng μl^-1 to 0.1 fg μl^-1 based on the initial concentrations determined with a NanoDrop ND-100 (ThermoFisherScientific, Waltham, MA, USA). Additionally, bacterial genomic DNA was independently extracted from pure bacterial cultures of these two same strains using known bacterial concentrations ranging from ~10⁸ to 10⁰ CFU ml^-1 as described in Kałużna et al. (2016).

For the crude bacterial cell-template based assays, 100 μl of different concentrations of aqueous suspensions of strain Xac 301 were added to 100 mg of crushed/cut fragments of leaves or stems. Then DNA was isolated from these ‘heterogeneous suspensions’ according to the methodology described by Kałużna et al. (2016).

For the PCR-based assays, the efficiency (E) was calculated from the slope (S) of the standard curve generated for each run using the following equation E = 10^{(−1/ S )} with E = 2 corresponding to 100% efficiency (Ramakers et al., 2003).

To test the usefulness of designed primers, positive controls for in planta detection were obtained from artificially inoculated hazelnut cvs. ‘Cosford’ and ‘Merveille de Bollwiller’ (two samples from each cultivar) maintained in a greenhouse, as well as from naturally infected material obtained from orchards (two samples). For artificial inoculation of the hazelnut cultivars, a 48-h culture of Xac 301 grown on YNA medium was suspended in sterile water (10⁸ and 10⁷ CFU ml^-1) and infiltrated into hazelnut leaves with a needleless syringe and/or injected into green shoots using a hypodermic needle (0.7×30mm) attached to a syringe. Four to six weeks post-inoculation, symptomatic plant tissue was harvested. Leaf samples were rubbed for 10 s on both sides with a cotton-swab soaked in 70% ethanol. A sample consisting of three 1-cm² segments including the lesion border was collected, crushed, and suspended in 1 to 2 ml sterile PBS for 15 min. We then tested two DNA extraction methods on the tissue macerate. In the first one, 10 μl of the plant macerate was added to 190 μl of TE buffer, boiled for 10 min at 100°C, and then centrifuged for 5 min at 9,500 × g. In the second approach, 10 μl of the plant macerate was added to 90 μl of TE buffer and total DNA was isolated using the Genomic Mini DNA Extraction Kit (A&A Biotechnology Gdynia, Poland) according to the manufacturer’s instructions. The boiled extract and purified DNA extract were used as templates in molecular assays. To confirm the infection by Xac, especially from naturally infected plant material, bacterial colony isolation was done simultaneously by plating on YNA medium.

The in silico analysis resulted in the detection of a highly conserved, Xac-specific sequence of 300 bp called “target 45” that had no hit with other bacteria in the database. A 2,440 bp genomic region called “region 2.4” encompassing “target 45” was identified after performing the comparative genomic analysis of target 45 in the six Xac whole genomes available ( Figure 1 ). Region 2.4 located on the chromosome corresponds to an insertion in Xac that was not present in other X. arboricola pathovars, such as pv. arracaciae ( Figure 1 ). The annotations for this region varied slightly between the different Xac genomes, but the region contains between two and three singletons that encode hypothetical proteins. A 494 bp singleton located within “region 2.4” in Xac CFBP 1159^PT (locus_tag XaxcyCFBP1159_22010) and annotated as encoding the hypothetical protein PPU54630 was used for the further development of Xac-specific assays.

Out of the candidate primer sets designed for all three detection techniques, a few sets were discarded from further analysis due to the presence of non-specific products that persisted, even after adjusting annealing temperatures. After initial laboratory testing, we focused on validation and testing of six primer pairs for conventional PCR, four primer pairs for qPCR with SYBR^® Green I, two primer pairs for TaqMan™ qPCR, and three primer pairs for LAMP. The sequences of these primer sets are reported in Table 2 . A primer BLASTn analysis of selected primers showed no full similarity to any sequences of bacterial plant pathogens in GenBank in July 2019. This in silico result was also confirmed on 15 May 2023 with a final primer check performed in the course of writing this article.

The genomic DNA of the 60 Xac strains was selectively amplified with all the primers developed for the different assays. No amplification was observed for the bacterial and fungal genomic DNA not belonging to the Xac pathovar ( Table 1 ). Similarly, no amplification was observed with DNA templates obtained from clean, asymptomatic leaves of five hazelnut cultivars using two DNA extraction kits.

The PCR assays using primers designed for conventional PCR gave amplicons ranging from 197 bp to 1,455 bp depending on the primer pairs used ( Table 2 , Supplementary Table S1 ). The six primer sets designed for conventional PCR generated a single amplicon of the size predicted by genome analyses for nearly all the 60 strains of Xac evaluated ( Table 1 ). Although, during validation of conventional PCR reactions on the JL26xx strains of Xac collected in Oregon, amplicons with an unexpected size were observed with Xac strain JL2600 with primers Xac2.4-4 and XacPPU-1. For Xac strain JL2600, the amplicon observed for primer pair for Xac2.4-4 was 1,166 bp instead of 1,455 bp and the amplicon for the primer pair for XacPPU-1 was 1,450 bp instead of 390 bp. The other conventional PCR primer pairs generated the predicted amplicon size for Xac strain JL2600. Conventional PCR reactions for all the other Xac strains in the JL26xx series ( Supplementary Table S1 ) returned the expected amplicon size for each of the primer pairs.

In the SYBR^® Green I qPCR assays, DNA from the Xac strains resulted in a positive reaction. However, non-specific, false-positive results after 28 cycles for a few bacteria not belonging to the Xanthomonas genus were observed when using the primers Xac45-1. Nonetheless, these non-specific amplicons were excluded based on the results of melting curve analysis i.e., having different melting temperature than the target product. The amplicons ranged from 88 bp to 170 bp and melting curve analysis performed on these specific products revealed a single peak characteristic of their already introduced line 492 T_m as reported in Table 2 .

In the TaqMan™ qPCR, two primer sets designed resulted in a positive reaction for the tested DNA from the Xac strains tested ( Table 1 , Supplementary Table S1 ) and no product were observed in case of testing of bacterial and fungal genomic DNA not belonging to the Xac pathovar nor DNA templates obtained from clean, asymptomatic leaves of five hazelnut cultivars.

In the LAMP assays, DNA of the Xac strains gave a positive reaction as expected and no amplification was observed with DNA of other isolates. The T_m of products amplified using the LAMP primers are provided in Table 2 .

Specificity, sensitivity, and efficiency of the Xanthomonas arboricola pv. corylina specific assays based on the organisms evaluated in this study are reported in Table 4 .

The LAMP and both qPCR assays confirmed identity of verified Xac strains. Results for LAMP and qPCR platforms were obtained in less than 1 h.

The sensitivity and detection limit of the Xac target DNA varies not only between the detection systems developed but also depending on the primer sets used. For four primer pairs designed for conventional PCR, 100 fg of genomic DNA generated a visible amplicon with primer pairs Xac2.4-4, and XacPPU-1; ~1 pg genomic DNA for primer pair Xac2.4-1; and 10 pg was detected with primer set Xac45-1. When crude, boiled bacterial cell templates were tested, the LoD was 1.8 × 10¹ CFU per reaction for Xac2.4-1 and XacPPU-1 primer sets, 1.8 × 10⁰ CFU per reaction for Xac2.4-4 primer sets, 1.8 × 10² CFU per reaction for XacPPU54630 primer sets, and 1.8 × 10³ CFU per reaction for Xac45-1 primer sets.

The LoD was lowered by 10¹ when using the primer pairs Xac2.4-1, Xac2.4-4, and XacPPU-1 to detect Xac in plant tissue macerates that contained the pathogen; for the primer pairs XacPPU54630 and Xac45-1 the LoD remained the same with or without plant tissues.

Among qPCR primers designed for SYBR^® Green I, two primer sets (Xac2.4-3RT and Xac45-1RT) detected 1 fg of Xac genomic DNA ( Figure 2 ), however the two other primer sets (Xac2.4-2RT and Xac45-2RT) detected about 10 fg of Xac genomic DNA. When crude boiled bacterial templates were tested, the limit of detection was 1 × 10⁰ CFU per reaction. The same decrease of sensitivity as noticed for conventional PCR (lowered by 10¹) for boiled bacterial preparations and in combinations of plant tissues and bacteria. Parameters of the four qPCR SYBR^® Green I assays are reported in Table 5 .

For TaqMan™ qPCR, the primer sets Xac-reg45 and Xac-PPU54630 detected 80 and 8 pg of Xac genomic DNA, respectively. When crude boiled bacterial templates were tested, the limit of detection was 2 × 10¹ CFU per reaction for primer set Xac-PPU54630 and 2 × 10³ CFU per reaction for primer set Xac-reg45.

When determining the sensitivity of LAMP primers, we detected 1 pg of purified genomic DNA from Xac isolates. When boiled bacterial cell templates were tested, the LoD was 1 × 10⁰ CFU per reaction for the XacPPU-1 primer set and 1 × 10¹ CFU per reaction for the Xac2.4-1 and Xac2.4-2 primer sets. In purified DNA isolated from plant material combined with bacteria, the LoD was 1 × 10³ CFU per reaction for all the primers tested.

The detection of Xac in artificially inoculated plant material was done with four conventional primer sets (Xac2.4-1, Xac2.4-4, XacPPU-1 and Xac45-1), all SYBR^® Green I qPCR (n = 4) and all LAMP primer sets (n = 3). All the primer sets used in the different platforms returned positive results for detection of Xac when the DNA was isolated using the kit procedure. Nonetheless, when the volume of the template of purified genomic DNA (µl or concentration per reaction) significantly increased, detection was decreased. Correspondingly, a one-tenth dilution of the purified genomic DNA template added to plant tissues allowed for consistent detection of Xac.

With all four conventional PCR primer sets, Xac was not detected when the assays were performed on DNA templates obtained via the boiling procedure of plant macerate (plants artificially or naturally infected). Because this was not the case with templates consisting of purified genomic DNA, we suspect that the boiling procedure did not eliminate possible plant inhibitors. The assays also remained negative when a tenfold dilution of the extracts was tested. For the SYBR^® Green I qPCR primer sets, Xac was detected in DNA isolated with both procedures independent of the template DNA concentration. With the LAMP XacPPU-1, Xac2.4-1 and Xac2.4-2 primer sets, Xac was always detected with purified genomic DNA preparations. In case of DNA extracted by boiling, templates with only 0.5 and 1 µl of undiluted extract was detected.