Novel PCR-Based Detection Methods for the Lettuce Bacterial Leaf Spot Pathogen, Xanthomonas hortorum pv. vitians Morinière et al., 2020

Abstract

Bacterial leaf spot in lettuce is a sporadic but devastating disease that threatens lettuce production worldwide. Severe outbreaks have resulted in up to 100% crop loss, and even smaller outbreaks can cause a significant yield loss, as the affected tissue must be removed from lettuce heads prior to their sale. The pathogen, Xanthomonas hortorum pv. vitians (Xhv), has at least three races, with each defined by the disease or resistance phenotype it elicits in lettuce cultivars and accessions. Specific molecular detection of Xhv would facilitate the work of clinicians, growers, seed companies, and researchers in the lettuce industry. We present an Xhv-specific touchdown PCR method and progress toward race-specific methods. We used an alignment of 18 Xhv strains and 9 closely related, non-target strains to identify pathovar- and race-specific gene clusters as targets for PCR primers. We evaluated the specificity first using in silico methods and then empirically using a collection of Xanthomonas strains. Our protocol demonstrated Xhv-specific detection from two sample types, including genomic DNA extracts and bacterial suspensions. Additional research is required to refine the race-specific protocols.

1. Introduction

Bacterial leaf spot in lettuce poses a significant threat to the 4.1-billion-dollar lettuce industry in the United States [1]. Most of its large-scale production occurs in California, Arizona, and Florida, and this disease has been detected in these areas and worldwide where lettuce is grown [2,3]. Nearly all market lettuce is cultivated for its leaves, which are damaged in a bacterial leaf spot infection due to the disease’s symptoms: water-soaking, chlorosis, and small necrotic spots that later join to form large lesions [4]. Following an outbreak, crop loss and the removal of infected leaves from the surviving lettuce heads result in a sizable reduction in both the sellable yield and product value. Cool, wet weather favors disease development, and although sporadic, outbreaks have been shown to affect up to 100% of field-grown lettuce [5].

The successful management of any plant disease depends on the timely and accurate identification and detection of the pathogen. Bacterial leaf spot in lettuce is caused by a bacterium, Xanthomonas hortorum pv. vitians (Xhv, formerly X. campestris pv. vitians or Xcv) [6]. Previously, a conventional PCR was developed by Barak et al. [7]. This method using B162 primers resulted in target amplification from Xcv strains but not from the other X. campestris strains tested, suggesting that the protocol was Xcv-specific. However, Xcv and other members of X. campestris have since been transferred to X. hortorum as Xhv, X. hortorum pv. carotae, X. hortorum pv. cynarae, X. hortorum pv. gardneri, X. hortorum pv. hederae, X. hortorum pv. pelargonii, and X. hortorum pv. taraxaci [6,8]. Of the closely related pathovars, Barak et al. [7] only included X. hortorum pv. carotae and X. hortorum pv. pelargonii as outgroup strains during the development of their method. In this manuscript, we demonstrate that in addition to Xhv, the B162-based conventional PCR amplifies several other X. hortorum pathovars and closely related strains, though not X. hortorum pv. carotae and X. hortorum pv. pelargoni. This lack of specificity could lead to false positive results when seeking to identify or detect Xhv. We sought to develop a conventional PCR that would replace this method for the identification and detection of Xhv.

Similarly, Xhv can be detected using the loop-mediated isothermal amplification (LAMP) assay developed by Dia et al. [9], but its detection is not limited to Xhv strains. Instead, this method targets the clade containing Xhv, X. hortorum pv. gardneri, and X. hortorum pv. cynarae [9]. The continued lack of an Xhv-specific method among the previously published detection methods prompted our research to fill this technical gap.

The phenotypic variation among Xhv strains is an important consideration in the development of detection methods. Xhv consists of three distinct groups of strains capable of triggering a host hypersensitive response (HR), each in a different lettuce cultivar or accession [10,11]. The HR is a plant defense strategy; after the recognition of a pathogen, programmed cell death of the affected tissue is initiated to limit the spread of infection. It has been hypothesized that this variation in plant reactions to Xhv strains, used to define the three groups as “races”, is caused by genetic variation. This claim has been supported by both a multilocus sequence analysis [10] and a genome-wide SNP variant analysis [12]; in both cases, the Xhv strains were clustered by race. A set of race-specific PCR-based detection methods would need to target sequences common to strains within a single known race and absent from non-target, related strains.

Following this strategy, we developed touchdown PCR protocols based on the whole genome alignments of 18 Xhv strains and 9 closely related strains. The primers were designed to target gene clusters present in all of the Xhv strains but absent from all of the others. We originally designed a conventional PCR, but we found in testing subsets of strains that the touchdown method improved the specificity of the reactions. We then tested our protocol using a collection of 96 Xhv isolates from around the world representing the three known races, the 9 other closely related strains within X. hortorum and X. hydrangeae, and 25 plant pathogenic type strains within Xanthomonas. Once we observed the desired Xhv specificity with this collection, we evaluated our method using two Xhv isolates from a 2018 bacterial leaf spot outbreak in lettuce in Pennsylvania, which were correctly detected using our touchdown method. The same process was used to develop methods for race-specific detection; we observed mixed results due to the absence of target amplification, or the presence of non-target amplification, but here, we share our progress and insights for future development.

2. Results

2.1. The Specificity of the Previously Published Detection Methods

When we tested the B162 primer set developed by Barak et al. for Xhv-specific detection [7], we found that while it resulted in the amplification of the targeted 700-base-pair (bp) fragment from the Xhv strains (BS0347, race 1; ICMP 1408, race 2; BP5181, race 3), as expected, it also amplified the same fragment from four non-target yet closely related strains: X. hortorum pv. garderi (CFBP 8163^PT), X. hortorum pv. cynarae (CFBP 4188^PT), X. hortorum pv. taraxaci (CFBP 410^PT), and an X. hortorum strain isolated from radicchio (BP5178) (Figure 1). The target fragment was not amplified from the other closely related strains, X. hortorum pv. hederae (CFBP 4925^T), X. hortorum pv. pelargonii (CFBP 2533^PT), and X. hortorum pv. carotae (CFBP 7900).

These results were corroborated by our in silico data. The B162 amplicon encoded a glycosyl hydrolase of family 3 that may play a role in starch hydrolysis, toxin degradation, and virulence promotion [13], and when we compared it to the NCBI whole genome database, it shared high sequence homology with a region in the whole genomes of the X. hortorum pvs. vitians, gardneri, cynarae, and taraxaci strains, as well as with a region in another close relative, X. campestris pv. nigromaculans, a pathogen of the Eurasian root vegetable and otherwise invasive weed known as greater burdock [14] (Supplemental Figure S1). Such high sequence homology could explain the ability of the B162 primers to bind to the off-target X. hortorum strains’ genomic DNA; further, these results suggested that X. campestris pv. nigromaculans could also be detected using this method. There were other X. campestris, X. arboricola, and Xyllela fastidiosa strains among the matches with the 700 bp B162 amplicon sequence, but they had significant variation that would likely prevent primer binding and subsequent target amplification from these strains.

Our laboratory and in silico experiments demonstrated that the conventional PCR procedure using the B162 primers does not have sufficient specificity to distinguish Xhv from several other agriculturally significant Xanthomonas plant pathogens, including many closely related X. hortorum pathovars. These results revealed the need for an improved, pathovar-specific protocol for Xhv detection that would remove the risk of false positive results in agricultural settings where multiple plant pathogens may be present.

2.2. A Pangenome Analysis Revealed Xhv Pathovar- and Race-Specific Gene Clusters

Pangenome alignment revealed gene clusters that were unique to the Xhv strains and not present in any of the genomes from the other Xanthomonas strains included in the alignment. Several other gene clusters appeared to be race-specific (encoded by all strains of a single Xhv race but not by the Xhv strains of the other races). A full list of the identified gene clusters is given in Supplemental Table S1, but our in silico evaluations of specificity using NCBI’s BLASTs led us to select gene clusters 3906, 4021, 4381, and 4980 to evaluate their specificity for Xhv, Xhv race 1, Xhv race 2, and Xhv race 3, respectively.

We chose gene cluster 3906 (GC3906) for Xhv specificity. For this gene cluster, we found no matches with NCBI’s database of conserved protein domains collected from most known forms of life. Comparing the gene cluster sequence to the NCBI BLAST database revealed that it had a 100% identity match with a hypothetical protein in Xhv but only a 90–98% identity match with the same protein in X. arboricola, X. nasturtii, X. campestris, X. phaseoli, and X. translucens (Supplemental Figure S2). Due to the observed sequence variation in all of the non-Xhv matches, it appeared to be a good candidate for the specific detection of Xhv strains. Comparisons of this cluster with the NCBI whole genome database revealed hits for the same Xhv strains that we included in our pangenome alignment, as expected, but also for several other X. hortorum strains that lack pathovar designations. X. hortorum pv. gardneri Xc69, which was re-classified from X. arboricola based on the average nucleotide identity (ANI) and genome-based DNA-DNA (gDDH) hybridization data [15], also appeared as a match. However, the ANI and gDDH are typically used for species determination and are not precise enough to distinguish between pathovars within the same species. While this strain may be a member of X. hortorum, its pathovar designation remains unclear. Two Xhv strains of unknown races and all of the known Xhv race 2 strains for which we had previously sequenced whole genomes [12] included the same SNP in this gene cluster, while all of the other matches described above were identical. All additional matches were not from Xhv strains and had significant sequence variation, including strains of X. nasturtii, X. campestris, and X. hortorum pv. cynarae. These results indicated that an Xhv-specific PCR-based detection method could be designed using the GC3906 gene cluster if the primers were designed to bind to segments that were conserved in all Xhv strains but showed variation in non-target strains.

Gene clusters that were unique among the Xhv strains to one of the three races were not necessarily absent from the related X. hortorum pathotype strains or other reference strains. Any race-specific detection method would need to be used in tandem with a pathovar-specific protocol for Xhv detection. For race 1, we investigated gene cluster 4021 (GC4021). We found a 100% identity match with an Xhv helix–turn–helix transcriptional regulator domain which is known to be activated in bacteria as a response to chemical stressors [16] (Supplemental Figure S3). This cluster matched with all Xhv race 1 strains for which we had previously obtained whole genome sequences [12] and no other Xhv strains. Many X. hortorum pv. pelargonii strains and one X. gardneri strain also matched but with sequence variations and truncations that could be exploited during the primer design to develop a Xhv race 1-specific reaction.

Neither GC4381 nor GC4980 had any matches with conserved domains, and each had a 100% identity match with different hypothetical proteins from Xhv (Supplemental Figures S4 and S5). GC4381 had identical sequence matches to all four Xhv race 2 strains that we had sequenced previously [12] and two Xhv strains of unknown races that we predict to also be Xhv race 2 strains. GC4980 had identical sequence matches to the two X. hortorum pv. vitians race 3 strains that were sequenced previously [12]. We therefore utilized GC4381 and GC4980 in an attempt to develop race 2- and race 3-specific primers, respectively.

2.3. The Development of an Xhv-Specific Detection Method

Many primer sets were designed to target the Xhv-specific gene clusters identified through the pangenome alignment, and while all of the others were eliminated during empirical testing due to the detection of off-targets, GC3906-152 (so called after the targeted gene cluster and amplicon size) proved to be an effective target for Xhv-specific detection (

Table 1. Primer sets developed for the specific detection of Xhv and its three known races.

Primer Set	Forward Sequence	Reverse Sequence	Amplicon Size (bp)	Target Strains
GC3906-152	GTTCGGTCGCCATTTCGATG	AGATAACCTCCCAGACCGCT	152	Xhv
GC4021-112	GGTGGCCTACTTTCATGCGA	GAGCAAGCCCTTCACAAGGT	112	Xhv race 1
GC4381-178	TATGATGCGGCACACAACCT	CGTATTGCGGTGCGAACTTT	178	Xhv race 2
GC4980-138	TCACTCAAAAGCCCACCCTC	ACATTCCTCGGCTATCCCCT	138	Xhv race 3

). The target fragment was amplified from 99% (95 out of 96) of the Xhv strains tested, including the pathotype strain CFBP 8686^PT (

Table 2. A list of the bacterial strains that were included in this study.

Organism	Strain *	Other Strain IDs	Host of Isolation	Known or Hypothesized Race	GC3906-152 Detection	GC4021-112 Detection	GC4381-178 Detection	GC4980-138 Detection	Origin	Collector or Citation
X. hortorum pv. vitians	BP5172 ‡	Xav 98-37 2/01	Lactuca sativa	Xhv race 1	Yes	No	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0339 ‡	Salinas 2/01	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0340 ‡	Xav 98-23 2/01	Lactuca sativa	Xhv race 1	Yes	No	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0347 ‡	Xcv 5/01	Lactuca sativa	Xhv race 1	Yes	Faint	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BP5176 ‡	Xcv 5/01	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BP5177 ‡	“Edge A”	Lactuca sativa	Xhv race 1	Yes	No	No	No	CO, USA	S. Koike
X. hortorum pv. vitians	BP5179 ‡	“Daniel Rom”	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BP5182 ‡	“Moreno Let”	Lactuca sativa	Xhv race 1	Yes	Yes	Yes	No	Santa Maria, CA, USA	J. Barak
X. hortorum pv. vitians	NCPPB 4058 ‡	N/A	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	UK	H. Stanford
X. hortorum pv. vitians	CFBP 8686PT ‡	LMG 938PT, NCPPB 2248PT, MR20213PT	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	Zimbabwe	[6,8]
X. hortorum pv. vitians	BP5191 ‡	VT111	Lactuca sativa	Xhv race 1	Yes	No	No	No	Canada	V. Toussaint
X. hortorum pv. vitians	BP5192 ‡	Xcv-2	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	CA, USA	C. T. Bull
X. hortorum pv. vitians	ICMP 1408 ‡	PDDCC 1408, Burkholder XL5	Lactuca sativa	Xhv race 2	Yes	No	Yes	No	Ithaca, NY, USA	W. H. Burkholder
X. hortorum pv. vitians	ICMP 4165 ‡	LMG 7508, PDDCC 4165	Lactuca sativa	Xhv race 2	Yes	No	Yes	No	New Zealand	H. J. Boesewinkel
X. hortorum pv. vitians	BS3127 ‡	VT106	Lactuca sativa	Xhv race 2	Yes	No	Yes	No	Canada	V. Toussaint
X. hortorum pv. vitians	BP5194 ‡	917	Lactuca sativa	Xhv race 2	Yes	No	Yes	No	OH, USA	[18]
X. hortorum pv. vitians	BS2861 ‡	“Christy BuLet 2”	Lactuca sativa	Xhv race 3	Yes	No	No	Yes	King City, CA, USA	S. Koike and Rianda
X. hortorum pv. vitians	BP5181 ‡	“Christy BuLet 3”	Lactuca sativa	Xhv race 3	Yes	No	No	Yes	King City, CA, USA	S. Koike and Rianda
X. hortorum pv. vitians	BS0313	A674-2B (e1)	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0341	C 5/20/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0342	Xcv 5/20/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0343	Xcv 5/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0345	Xcv 5/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0346	Xcv 5/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS2849	“Mike Lombard Reeves”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2850	“Mike Lombard Harden”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2852	“Daniel Rom”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2855	“Frank Let”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2857	“Frank Let”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2858	John DeCarli Rom1-1	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2859	John DeCarli Rom1-2	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2870	“Moreno Let”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Santa maria, CA, USA	S. Koike
X. hortorum pv. vitians	BS2871	“Keller Rom”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2872	“Keller Rom”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2873	“Greg greenleaf”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2874	“Greg greenleaf”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2875	“Greg redleaf”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2876	“Greg let”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2908	“Matt romaine”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Watsonville, CA, USA	S. Koike
X. hortorum pv. vitians	BS2994	ICMP 6656, DAR 30547	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	New South Wales, Australia	R. Fitzell
X. hortorum pv. vitians	BS2996	ICMP 6735, Watson B2578	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	Yes	No	Palmerston North, WI, New Zealand	D. R. W. Watson
X. hortorum pv. vitians	BS2997	ICMP 7423, Watson J2928.	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Patumahoe, AK, New Zealand	D. R. W. Watson
X. hortorum pv. vitians	BS3050	LMG 8688; ICMP 6461; Watson D2538	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	New Zealand	D. R. W. Watson
X. hortorum pv. vitians	BS3051	LMG 8690; Hill AS3997; ICMP 6857	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	New Zealand	D. R. W. Watson
X. hortorum pv. vitians	BS3054	CFBP 3980, Audusseau 11.72	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Vaucluse, France	C. Audusseau
X. hortorum pv. vitians	BS3056	CFBP 3996 Audusseau 17.09	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Isère, France	C. Audusseau
X. hortorum pv. vitians	BS3128	105 VT107	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Canada	V. Toussaint
X. hortorum pv. vitians	BS3131	119 VT41	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Canada	V. Toussaint
X. hortorum pv. vitians	BS3132‡	B07-007, ID200707A	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	Canada	V. Toussaint
X. hortorum pv. vitians	BS3272	L11	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	Yes	FL, USA	[2]
X. hortorum pv. vitians	BS3300	Xcv-4	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	CA, USA	C. T. Bull
X. hortorum pv. vitians	BS3301	Xcv-5	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	CA, USA	C. T. Bull
X. hortorum pv. vitians	BS3303	Xcv-7	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	CA, USA	C. T. Bull
X. hortorum pv. vitians	BS3304	Xcv-8	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	CA, USA	C. T. Bull
X. hortorum pv. vitians	BS3306	Xcv-10	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	CA, USA	C. T. Bull
X. hortorum pv. vitians	BS0301	10S7-2 (a2)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0302	10TB7 (a4)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0304	10TB9 (a6)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0305	10TB12-1 (a7)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0309	S4-1(K) (c5)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0310	S5-2 (c6)	Lactuca sativa	Hypothesized Xhv race 1	Yes	Faint	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0311	S5-2(K) (c7)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0314	A674-4B (e2)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0316	10TB9 (e4)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0318	QR71B (e8)	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	HI, USA	A. Alvarez
X. hortorum pv. vitians	BS0335	Xav 98-05 2/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0337	Xav 98-67 2/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0338	Xav 98-76 2/01	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0542	10S7-2 (a2)	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS0543	“Spot A”	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS2946	Julio Rodrigues Neto IBSBF 1553 Embrapa K 532	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Brazil	I. M. G. Almeida
X. hortorum pv. vitians	BS2998	ICMP 7465, IBSBF 325, C.F. Robbs: ENA2008	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Brazil	C. F. Robbs
X. hortorum pv. vitians	BS3035	NCPPB 970, ICPB XL 102, Thornberry 1-49	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	USA	H. H. Thornberry
X. hortorum pv. vitians	BS3036	NCPPB 1839, ICPB XV169, Robbs ENA-250	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Brazil	A. P. Viegas
X. hortorum pv. vitians	BS3041	NCPPB 3663, Neto ISBF 473	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No		J. R. Neto
X. hortorum pv. vitians	BS3042	NCPPB3931, Sellwood/Wilson A6520/1	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	The UK	J.E. Sellwood and J.K. Wilson
X. hortorum pv. vitians	BS3047	LMG 7453; ATCC 11525; Burkholder XL3; CNBP 500; Dye YA3; ICMP 337; ICPB XL3; NCPPB 992; PDDCC 337	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	USA	H. Anderson
X. hortorum pv. vitians	BS3049	LMG 7510; Fahy DAR30526; ICMP 6655; PDDCC 6655	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Australia	P. Fahy
X. hortorum pv. vitians	BS3053	CFBP 3973, Audusseau 11.08	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No		C. Audusseau
X. hortorum pv. vitians	BS3055	CFBP 3983, Audusseau 14.28	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	Isère, France	C. Audusseau
X. hortorum pv. vitians	BS3130	118 VT25	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No		V. Toussaint
X. hortorum pv. vitians	BS3271	L43	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	FL, USA	C. T. Bull
X. hortorum pv. vitians	BS3526	B55	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	CA, USA	[18]
X. hortorum pv. vitians	BS3527	B57	Lactuca sativa	Hypothesized Xhv race 1	Yes	No	No	No	CA, USA	[18]
X. hortorum pv. vitians	BS0344 ‡	Xcv 5/01	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	J. Barak
X. hortorum pv. vitians	BS2851 ‡	“Mike Lombard Harden”	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	Salinas, CA, USA	S. Koike
X. hortorum pv. vitians	BS2909	“Matt romaine”	Lactuca sativa	Hypothesized Xhv race 1	Yes	Yes	No	No	Watsonville, CA, USA	S. Koike
X. hortorum pv. vitians	BS3528 ‡	B59	Lactuca sativa	Xhv race 1	Yes	Yes	No	No	CA, USA	[18]
X. hortorum pv. vitians	BS3034 ‡	NCPPB 2969, ICPB XL6	Lactuca sativa	Xhv race 2	Yes	No	Yes	No	USA	W. H. Burkholder
X. hortorum pv. vitians	BS3043 ‡	NCPPB 4033, Sahin/Miller 700a	Lactuca sativa	Xhv race 2	Yes	No	Yes	No	USA	F. Sahin
X. hortorum pv. vitians	BS3126 ‡	99 VT101	Lactuca sativa	Xhv race 2	No	No	Yes	No	Isère, France	C. Audusseau
X. hortorum pv. vitians	BS3529 ‡	906	Lactuca sativa	Xhv race 2	Yes	No	Yes	No	OH, USA	S. Miller
X. hortorum pv. vitians	BS3531 ‡	923	Lactuca sativa	Xhv race 2	Yes	No	No	No	OH, USA	[18]
X. hortorum pv. vitians	BS3532 ‡	924	Lactuca sativa	Xhv race 2	Yes	No	No	No	OH, USA	[18]
X. hortorum pv. vitians	BS2860	Christy BuLet 1	Lactuca sativa	Hypothesized Xhv race 3	Yes	No	No	Yes	King City, CA, USA	S. Koike
X. hortorum pv. vitians	BS2863	Christy BuLet 4	Lactuca sativa	Hypothesized Xhv race 3	Yes	No	No	Yes	King City, CA, USA	S. Koike
X. hortorum pv. vitians	BP4476	N/A	Lactuca sativa	Hypothesized Xhv race 1	Yes	Not tested	Not tested	Not tested	PA, USA	This study
X. hortorum pv. vitians	BP4477	N/A	Lactuca sativa	Hypothesized Xhv race 1	Yes	Not tested	Not tested	Not tested	PA, USA	This study
Pseudomonas viridiflava	BP4478	N/A	Lactuca sativa	N/A	No	Not tested	Not tested	Not tested	PA, USA	This study
Pseudomonas allivorans	BP4479	N/A	Lactuca sativa	N/A	No	Not tested	Not tested	Not tested	PA, USA	This study
X. hortorum	BP5178	N/A	Cichorium intybus (radicchio)	N/A	No	No	No	No	Salinas, CA, USA	[17]
X. hortorum pv. hederae	CFBP 4925T	ICMP 453T, NCPPB 939T, LMG 733T	Hedera helix (English ivy)	N/A	No	No	No	No	USA	[8,19,20]
X. hortorum pv. taraxaci	CFBP 410PT	ATCC 19318PT, NCPPB 940PT, LMG 870PT	Taraxacum kok-sahgyz (Russian dandelion)	N/A	No	No	No	No	Ithaca, NY, USA	[8,19,20,21]
X. hortorum pv. pelargonii	CFBP 2533PT	ICMP 4321PT, LMG 7314PT, NCPPB 2985PT	Pelargonium peltatum (pelargonium)	N/A	No	No	No	No	Auckland, New Zealand	[20,22]
X. hortorum pv. gardneri	CFBP 8163PT	LMG 962PT, ATCC19865PT, NCPPB 881PT, PDCC 1620PT	Solanum lycopersicum (tomato)	N/A	No	No	No	No	Yugoslavia	[23,24]
X. hortorum pv. cynarae	CFBP 4188PT	ICMP 16775PT	Cynara scolymus (artichoke)	N/A	No	No	No	No	France	[25]
X. hortorum pv. carotae	CFBP 7900	M081	Daucus carota (carrot)	N/A	No	No	No	Faint	Hungary	[8,20,26]
X. campestris pv. coriandri	CFBP 8452PT	LMG 687PT, ATCC 17996PT, ICMP 5725PT, NCPPB 1758PT, PDDCC 5725PT	Coriandrum sativum (coriander)	N/A	No	No	No	No	India	[8,20,27]
X. hydrangeae	LMG 31884T	Van Vaerenbergh gbbc963T; GBBC 2123T, CCOS 1956T	Hydrangea arborescens (smooth hydrangea)	N/A	No	Not tested	Not tested	Not tested	Flanders, Belgium	[28]
X. axonopodis	CFBP 4924T	ATCC 19312T, CFBP 2156T, ICMP 50T, ICPB Xa103T, LMG 538T, LMG 982T, NCPPB 457T	Axonopus scoparius (carpet grass)	N/A	No	Not tested	Not tested	Not tested	Colombia	[29]
X. campestris	CFBP 5251T	ATCC 33913T, CFBP 2350T, CFBP 5241T, DSM 3586T, ICMP 13T, LMG 568T, Labo 11405T, NCPPB 528T	Brassica oleracea var. gemmifera (Brussels sprout)	N/A	No	Not tested	Not tested	Not tested	The UK	[30,31]
X. translucens	CFBP 2054T	ATCC 19319T, ICMP 5752T, ICPB XT2T, LMG 876T, NCPPB 973T	Hordeum vulgare (barley)	N/A	No	Not tested	Not tested	Not tested	USA	[8,32]
X. theicola	CFBP 4691T	ICMP 6774T, LMG 8684T	Camellia sinensis (tea tree)	N/A	No	Not tested	Not tested	Not tested	Japan	[8]
X. arboricola	CFBP 2528T	ATCC 49083T, ICMP 35T, LMG 747T, NCPPB 411T	Juglans regia (walnut)	N/A	No	Not tested	Not tested	Not tested	New Zealand	[8,33]
X. cucurbitae	CFBP 2542T	ICMP 2299T, LMG 690T, NCPPB 2597T	Cucurbita maxima (squash)	N/A	No	Not tested	Not tested	Not tested	New Zealand	[8,34]
X. alfalfae	CFBP 7686T	ATCC 11765T, ICPB 10701T, ICPB XA 121T, LMG 495T	Medicago sativa (alfalfa)	N/A	No	Not tested	Not tested	Not tested	India	[35,36]
X. sacchari	CFBP 4641T	LMG 471T	Saccharum officinarum (sugarcane)	N/A	No	Not tested	Not tested	Not tested	Guadeloupe, France	[8]
X. melonis	CFBP 4644T	IBSBF 68T, ICMP 8682T, LMG 8670T, NCPPB 3434T	Cucumis melo (melon)	N/A	No	Not tested	Not tested	Not tested	Brazil	[8]
X. hyacinthi	CFBP 1156T	ATCC 19314T, ICMP 189T, LMG 739T, NCPPB 599T	Hyacinthus orientalis (hyacinth)	N/A	No	Not tested	Not tested	Not tested	Netherlands	[8,37]
X. phaseoli	CFBP 8462T	ATCC 49119T, LMG 29033T	Phaseolus vulgaris (common bean)	N/A	No	Not tested	Not tested	Not tested	NE, USA	[38,39]
X. bromi	CFBP 1976T	ICMP 12545T, LMG 947T	Bromus carinatus (bromegrass)	N/A	No	Not tested	Not tested	Not tested	France	[8]
X. euvesicatoria	CFBP 6864T	ATCC 11633T, DSM 19128T, ICMP 109T, ICMP 98T, NCPPB 2968T	Capsicum frutescens (wild chili pepper)	N/A	No	Not tested	Not tested	Not tested	USA	[23]
X. vasicola	CFBP 2543T	ICMP 3103T, LMG 736T, NCPPB 2417T	Sorghum vulgare (sorghum)	N/A	No	Not tested	Not tested	Not tested	New Zealand	[8,40]
X. cassavae	CFBP 4642T	ICMP 204T, LMG 673T, NCPPB 101T	Manihot esculenta (cassava)	N/A	No	Not tested	Not tested	Not tested	Malawi	[8,41]
X. populi	CFBP 1817T	ATCC 51165T, ICMP 5816T, ICPB XP 240T, LMG 5743T	Populus x canadensis cv. Regenerata (Canadian poplar)	N/A	No	Not tested	Not tested	Not tested	Noyon, Oise, France	[42,43]
X. vesicatoria	CFBP 2537T	ATCC 35937T, CFBP 4645T, ICMP 63T, LMG 911T, NCPPB 422T	Lycopersicon esculentum (tomato)	N/A	No	Not tested	Not tested	Not tested	New Zealand	[8,44]
X. pisi	CFBP 4643T	ATCC 35936T, ICMP 570T, ICMP 570T, LMG 847T, Labo 13356T, NCPPB 762T	Pisum sativum (pea)	N/A	No	Not tested	Not tested	Not tested	Japan	[8,45]
X. dyei	CFBP 7245T	DSM 110537T, ICMP 12167T, NCPPB 4446T	Metrosideros excelsa (New Zealand Christmas tree)	N/A	No	Not tested	Not tested	Not tested	Omahanui, Bay of Plenty, New Zealand	[46]
X. citri	CFBP 3369T	ATCC 49118T, LMG 9322T	Citrus aurantifolia (key lime)	N/A	No	Not tested	Not tested	Not tested	FL, USA	[38,47,48]
X. albilineans	CFBP 2523T	ATCC 33915T, ICMP 196T, LMG 494T, NCPPB 2969T	Saccharum officinarum (sugarcane)	N/A	No	Not tested	Not tested	Not tested	Fiji	[49,50,51]
X. perforans	CFBP 7293T	DSM 18975T, NCPPB 4321T	Solanum lycopersicum L. (tomato)	N/A	No	Not tested	Not tested	Not tested	FL, USA	[23]
X. codiae	CFBP 4690T	ICMP 9513T, LMG 8678T	Codiacum variegatum (croton)	N/A	No	Not tested	Not tested	Not tested	FL, USA	[8]

* Strains in bold were used in the whole genome sequence alignments to identify Xhv-specific and race-specific gene cluster targets for PCR-based detection. Xhv strains marked with ‡ have had their race confirmed via HR screening [10,11], while those lacking the ‡ symbol have been hypothesized to belong to their race based on a multilocus sequence analysis (MLSA).

; Supplemental Figure S6). However, the target fragment was not amplified from strain BS3126, a race 2 strain from France. The target fragment was not detected from any of the other X. hortorum pathotype strains, X. hydrangeae (LMG 31884^T), X. campestris pv. coriandri (CFBP 8452^PT), or X. hortorum from radicchio (BP5178) [17], or any of the 25 Xanthomonas species type strains included in this study, resulting in 100% exclusivity (Figure 2; Table 2). No amplification was observed for any of the sterile water controls.

2.4. Evaluating the Xhv-Specific Detection Method Using Environmental Samples

To evaluate whether the Xhv-specific PCR protocol designed could be used to identify Xhv from novel outbreaks, four isolates collected from a 2018 outbreak of BLS in lettuce in Pennsylvania were investigated. The Xhv-specific touchdown protocol using the GC3906-152 primer set amplified fragments from only two of the four strains. The expected 152 bp band was amplified for BP4476 and BP4477 but not for BP4478 and BP4479 (Figure 3). As expected, the positive control strain, Xhv race 1 BP5172, produced the expected 152 bp band, and the sterile water control did not. Strains BP4476 and BP4477 were identified as Xhv using the MLSA, which placed them within the clade containing the Xhv race 1 strains (Figure 4). Analysis of the 16S rRNA subunit sequence revealed that the two other isolates, BP4478 and BP4479, were members of Pseudomonas viridiflava (99.9% identity and 100% query cover) and Pseudomonas allivorans (100% identity and 100% query cover), respectively.

2.5. Progress Toward Xhv Race-Specific Detection

Many gene clusters were also evaluated as candidate targets for Xhv race-specific detection using our touchdown PCR method. The best-performing primer sets, those with the greatest number of target strains detected and the least number of non-targets, were selected and are shown in Table 1. These included GC4021-112, GC4381-138, and GC4980-178 for Xhv race 1-, 2-, and 3-specific detection, respectively. When used with our touchdown method, GC4021-112 resulted in strong amplification of a 112 bp fragment from 48 strains (59%) and its faint amplification from 2 strains (2%) out of the 82 strains that either demonstrated HR elicitation in the L. sativa cultivar Little Gem, leading to an Xhv race 1 designation, or clustered with the known Xhv race 1 strains in the MLSA, leading to the prediction that they belonged to Xhv race 1 (Table 2; Supplemental Figure S7). Among the thirty-four strains that were not detected using this method, two strains were known members of race 1 and were included in the pangenome alignments used to identify the race-specific gene clusters. This indicates that the whole genome sequence assemblies for these strains, BP5172 and BP5177, may not be representative of the true gene cluster sequence, possibly due to the limited ability of short reads to resolve certain types of genetic variations, such as long repetitive sequences. Such a consequence of the short-read assemblies may also have been true for the other twenty-eight known or predicted Xhv race 1 strains that did not show amplification of the 112 bp fragment. Long-read sequencing will be needed to evaluate these hypotheses and improve the positioning of the primers to improve the specificity or to identify stronger targets for race-specific detection. However, the protocol did successfully exclude all Xhv race 2 and 3 strains tested; none of these strains produced the 112 bp fragment, and this was expected, as the gene cluster GC4021 was absent from these strains’ alignments.

The same touchdown PCR protocol was used to evaluate the GC4381-178 primer set for Xhv race 2-specific detection. Among the race 2 strains tested, 8 out of 10 (80%) produced the expected amplicon, excluding BS3531 and BS3532 (Table 2; Supplemental Figure S8). The two strains that did not produce the amplicon were not included in the pangenome alignment used to identify the race-specific gene clusters because they lacked available whole genome sequence data. It is possible that these two strains, known to belong to race 2 due to their elicitation of an HR in L. serriola PI491114, do not encode GC4381 or that they encode a sufficient variation in this cluster to prevent primer binding. Other unexpected results included the detection of two known Xhv race 1 strains, one that was part of the pangenome alignment that showed GC4381 to be absent (BP5182) and one predicted to belong to Xhv race 1 based on the MLSA data (BS2996). It is possible that these strains retain GC4381 but that it is somehow inactive and unable to elicit an HR in L. serriola PI491114. These hypotheses will be investigated as we continue to improve the protocol; long-read sequencing will be needed to help resolve the conflicting results between the in silico sequence analyses and the empirical HR testing.

The touchdown PCR protocol used with the GC4980-138 primer set for Xhv race 3 detection showed amplification of the 138 bp target fragment in one known Xhv race 3 strain and three Xhv race 3 strains whose identities were predicted based on the MLSA data (Table 2; Supplemental Figure S9) [10]; these are all of the Xhv race 3 strains currently in curation. Non-target amplification of the 138 bp fragment occurred for X. hortorum pv. carotae CFBP 7900, though it was faint, which was part of our pangenome alignment, and for an Xhv strain predicted to belong to race 1 due to the MLSA results (BS3272). In combination with this protocol, the use of the described pathovar-specific protocol could eliminate any concern about the Xhv non-target amplification. HR screening using the BS3272 strain will be necessary to confirm its placement in Xhv race 1. However, no other Xhv race 1 strains, nor any Xhv race 2 strains, were amplified using this primer set. Again, long-read sequencing is expected to provide the resolution needed to confirm the appropriate targets, but these results represent major steps toward Xhv race-specific detection.

3. Discussion

In previous studies, amplification using the B162 primer pair was used as the initial screening for Xhv isolates [52]. The touchdown PCR protocol using the GC3906-152 primer set can now be used for Xhv pathovar-specific detection from unknown samples. The specificity of our protocol represents an improvement over previous methods using the B162 primer set [7] designed to detect Xhv and LAMP assays [9] designed to detect all members of X. hortorum species. In addition to Xhv strains, the B162 primer set also amplifies the target 700 bp amplicon in X. hortorum pv. garderi, X. hortorum pv. cynarae, X. hortorum pv. taraxaci, and an X. hortorum strain isolated from radicchio. The two pathovars used as the outgroup strains to develop this protocol, X. hortorum pv. carotae and X. hortorum pv. carotae, were not amplified in our assays. The LAMP assay as designed by Dia et al. [28] detects an entire clade of X. hortorum species, including Xhv, X. hortorum pv. gardneri, and X. hortorum pv. cynarae.

The Xhv-specific primer set described here and the touchdown PCR protocol detected 99%, or 95 out of 96, of the known Xhv strains tested (including the pathotype strain CFBP 8686^PT) and did not detect any of the 33 outgroup strains (demonstrating 100% exclusivity), including 9 closely related X. hortorum and X. hydrangeae strains and 23 additional Xanthomonas type strains. It also successfully detected two Xhv isolates from a 2018 bacterial leaf spot outbreak in lettuce in Pennsylvania and did not detect the two Pseudomonas isolates from symptomatic tissue from the same outbreak. It is unclear why the reactions with DNA from Xhv race 2 BS3126 did not produce the target amplicon using this protocol—this result requires further investigation to determine whether this strain encodes nucleotide variations in the target gene cluster that inhibit primer binding. However, based on its 99% success rate in detecting the collection of 96 known Xhv strains and its ability to correctly detect Xhv from environmental isolates, we can recommend the use of our touchdown PCR protocol for Xhv-specific detection from colony suspensions or DNA extracts. This could replace the less specific conventional PCR method that uses the B162 primers. Further research is necessary to evaluate the efficacy of our protocol for other sample types, such as tissue extracts and seed wash. Additionally, we designed the amplified targets to be within the size range for efficient Taqman RealTime PCR amplification. We are currently evaluating these genomic regions and incorporating additional Xhv genomes for the development of a Taqman RealTime PCR for this pathovar. Both its selectivity and sensitivity will be quantified in these assays.

Using the same touchdown PCR method, we evaluated primers designed for Xhv race-specific detection; our results appear to highlight the need for long-read sequencing for a variety of Xhv strains. The pangenome alignment of our short-read-based assemblies revealed race-specific gene clusters; however, these in silico analyses, along with applying BLAST to search these strains for DNA sequences homologous with these clusters, did not correspond to our laboratory results. GC4021, GC4981, and GC4890 were three gene clusters we identified from the pangenome alignment that were specific to Xhv races 1, 2, and 3, respectively. GC4021-112 did not detect any race 2 or 3 strains, but it also only detected 59% of the Xhv race 1 strains. GC4981-178 did not detect any race 3 strains, but it did detect 2% (2 out of 82) of the race 1 strains and failed to detect 20% (2 out of 10) of the race 2 strains. GC4980 did not detect any race 2 strains and detected all four race 3 strains, but it also detected 1% (1 out of 82) of the Xhv race 1 strains. Despite these drawbacks, we are now using these PCR protocols along with an MLSA to hypothesize about the race of isolates from symptomatic lettuce tissue from novel outbreaks. We hypothesize that our short-read-based sequence assemblies did not capture the full variation that exists in the Xhv strain genomes and thus expect that long-read sequences will help to improve our methods. These results also indicate strains for which HR screening will be valuable and may reveal whether any of the strains predicted to belong to a particular race based on MLSA data belong to a different race from that expected. Although the clonal nature of Xhv populations suggests that races may correspond to phylogenetic relationships [12,52], the race of a strain is defined by the lettuce cultivar or accession in which it elicits an HR. Differential germplasms for race determination have been described—L. sativa cv. Little Gem, L. serriola PI491114, and L. serriola ARM-09-161-10, for races 1, 2, and 3, respectively [11]—and are being used in subsequent studies.

We expect that our pathovar- and race-specific detection methods will be useful to clinicians, researchers, and seed companies. They will allow for the rapid detection of infection in suspect samples before they result in outbreaks, shown previously to be capable of causing 100% crop loss in severe cases. They may also become useful for seed companies to use to test isolates from seed lots prior to sale or for researchers to use to track the distribution and spread of the pathogen and its races. The race-specific methods, once optimized, could become useful for clinicians who want to provide recommendations to growers for cultivars resistant to the specific Xhv race detected from their prior plantings, as the pathogen has been shown to be harbored by weeds and crop debris and cause new outbreaks in subsequent seasons [7].

4. Materials and Methods

4.1. Isolation, Culturing, and Storage of the Bacterial Strains

Table 2 lists the strains used in this study. The bacterial strains were routinely grown at 28 °C on nutrient agar (NA) and in nutrient broth (NB). Long-term storage was achieved by adding bacteria from pure cultures to 50% nutrient broth and 50% glycerol, vortexing the combination to mix it, and storing it at −80 °C. All of the colony suspensions used in PCRs were created by isolating pure colonies from long-term storage on NA, incubating the plates for three to five days at room temperature, inoculating 30 µL of sterile water in a 1.5 mL Eppendorf tube with a single colony, and vortexing it to mix it.

Following a suspected bacterial leaf spot outbreak in Pennsylvania field-grown lettuce, green leaf and red leaf lettuce samples were submitted to the lab for pathogen isolation and identification. Symptoms appeared as chlorosis and necrotic spots on the leaves, and symptomatic leaves from each lettuce type were used to isolate bacteria using the following procedure. A small tissue fragment was cut out using a sterile scalpel at the margin of a necrotic spot and placed in a sterile 1.5 mL Eppendorf tube. To the tube, 1 mL of 10% sodium hypochlorite was added, and the tube was briefly vortexed. After 1 min of incubation at room temperature, the tube was again briefly vortexed, the sodium hypochlorite was discarded, and the leaf fragment was washed thrice with 1 mL of sterile water. After the final wash was discarded, 30 µL of sterile water was added to the tube, and the tissue was crushed with a sterile pestle. Aliquots (10 µL each) of the resulting plant extract were spread onto two NA plates and incubated for at least two days. Yellow, mucoid, round colonies (typical of Xanthomonas strains) that grew were sub-cultured until we achieved purified strains, which were then stored long term at −80 °C in 1.5 mL cryovials that contained 500 µL of sterile NB and 500 µL of sterile glycerol.

4.2. Evaluation of the B162 Primer Set

Colony suspensions were prepared as described above for three Xhv strains representing the three known races, including Xhv BS0347 (race 1), ICMP 1408 (race 2), and BP5181 (race 3), as well as for Xanthomonas hortorum pv. gardneri (CFBP 8163^PT), X. hortorum pv. cynarae (CFBP 4188^PT), X. hortorum pv. hederae (CFBP 4925^T), X. hortorum pv. taraxaci (CFBP 410^PT), X. hortorum pv. pelargonii (CFBP 2533^PT), X. hortorum from radicchio (BP5178), and X. hortorum pv. carotae (CFBP 7900). These suspensions were used as templates in conventional PCR using the primer set (B162) and the reaction conditions described in Barak et al. [7]. The reactions were run using an MJ Research PTC-100 thermocycler (Montarville, QC, Canada). The PCR products were analyzed using gel electrophoresis; a total of 5 µL of product mixed with 1 µL of loading dye (VWR Life Science EZ-Vision One Dye-as-Loading-Buffer; Radnor, PA, USA) was added to wells of 1% agarose gel, which was then subjected to 84 V for 45 min. Sterile water was loaded in place of the PCR product as the experimental control, and a 1 KB ladder was used as a size reference (New England Biolabs; Ipswich, MA, USA). The amplification bands were visualized using a Bio-Rad Gel Doc XR Imaging System (Hercules, CA, USA).

4.3. Genome Alignments and Target Selection

Whole genome sequences for thirteen Xhv race 1, four Xhv race 2, and two Xhv race 2 strains, along with strains from eight closely related Xanthomonas hortorum pathovars and one Xanthomonas hydrangea strain [12,28], were aligned with a Xanthomonas hortorum pv. carotae (M081) reference strain via the anvi’o pangenome pipeline [53]. Gene clusters unique to Xhv and each of the Xhv races were identified and extracted for further analysis. NCBI’s conserved domain tool, BLASTx (database = nr), and BLASTn (database = wgs; organism = ‘Xanthomonas’) were used to search for genes with known functions within these clusters and within the B162 amplicon sequence. Matches found in the non-redundant database were considered to be those with at least 90% query coverage, 90% identity, and an e-value threshold of 1 × 10⁻⁵. The NCBI BLAST versions used in our in silico analyses were 2.12.0 through 2.15.0.

4.4. DNA Isolation and Sequencing

Strains from our collection were streaked in quadrants to isolate single colonies, and then individual colonies were used to inoculate 50 µL of sterile NB into 100 mL sterile tubes. The liquid cultures were incubated overnight at 28 °C and shaken at 200 rpm. Genomic DNA was extracted from these cultures using the Qiagen DNeasy UltraClean Microbial Kit (Valencia, CA, USA) according to the manufacturer’s instructions. The DNA concentrations following the final elution were measured using the Thermo Fischer Qubit Fluorometer 3.0 and the Invitrogen dsDNA Broad Range Assay Kit (Waltham, MA, USA). The Genomics Core Facility at the Pennsylvania State University completed all of the Sanger sequencing.

4.5. Primer Development

NCBI’s Primer-BLAST was used to design the primers for each of the Xhv- or race-specific gene clusters (database = nr; organism = ‘Xanthomonas’). Primers were selected from the output that had low or no off-target amplification, low complementarity scores, similar melting temperatures near 60 °C, and a similar GC content near 50%. Custom DNA oligos were ordered from Thermo Fisher Scientific (Waltham, MA, USA). In total, 26 primer sets were evaluated using 19 Xhv strains, including strains of all three known races; selected from these primer sets were those that provided the desired pathovar or race specificity (Table 1). All of the other primer sets that were tested are shown in Supplemental Table S2. The detection protocol using the B162 primer set [7] was also evaluated.

4.6. Touchdown PCR Development

Initial testing using conventional PCR, involving cycling 34 times under the same conditions as those given in

Table 3. Touchdown PCR cycling conditions.

Step	Instruction	Purpose
1	95 °C for 1 min	Taq polymerase activation
2 *	95 °C for 30 s	Denaturation
3 *	68 °C for 30 s, −1 °C every cycle	Annealing
4 *	72 °C for 30 s	Extension
5 *	GOTO Step 2, 10 times	Cycling
6	95 °C for 30 s	Denaturation
7	58 °C for 30 s	Annealing
8	72 °C for 30 s	Extension
9	GOTO Step 6, 23 times	Cycling
10	72 °C for 5 min	Final extension

* The touchdown phase of our protocol was when the annealing temperature was kept higher than that in typical PCR for 11 cycles to reduce non-specific primer binding. The remaining cycling temperatures were typical of general PCR.

but with a consistent annealing temperature of 58 °C during each cycle, did not provide the desired specificity for our detection reactions. Using the touchdown PCR method, using higher annealing temperatures in earlier PCR cycles to reduce non-specific primer binding [54], did provide the appropriate specificity for pathovar- and race-level detection. Each reaction was completed with the following reaction formulation: 9 µL of sterile water, 1.25 µL of each 10 µM primer, 12.5 µL of 2X Bioline Immomix containing IMMOLASE DNA polymerase, and 1 µL of the template, either DNA extract (30 ng/mL) or colony suspension. The colony suspensions were created by taking single colonies from the NA plates, inoculating them into 30 µL of sterile water, and vortexing the combination to mix it. As an experimental control, 1 µL of sterile water was substituted for the template. The reaction conditions are listed in Table 3, and the reactions were run in an MJ Research PTC-100 thermocycler (Montarville, QC, Canada).

Amplified products were separated using gel electrophoresis, as described above in Section 4.2, except that the gels were run with 84 V applied for 2 h given the larger size of these gels. The same 1 KB ladder described above was used as a reference for the amplicon size. The amplification bands were again visualized using the Bio-Rad Gel Doc XR Imaging System (Hercules, CA, USA). A collection of 96 Xhv strains, along with the pathotype strains for each of the other X. hortorum pathovars, 1 additional X. hortorum strain isolated from radicchio [15], and the type strain of X. hydrangeae (another close relative of X. hortorum), were used to evaluate the specificity of our selected primers. These experiments were conducted twice using colony suspensions from the same strain collection.

4.7. The Multilocus Sequence Analysis of the PA Isolates

Single colony suspensions of the strains isolated from the PA lettuce served as the templates for the PCR reactions to separately amplify four Xanthomonas housekeeping genes: rpoD (693 bp), fyuA (762 bp), gyrB (522 bp), and gap1 (774 bp) [10,55]. Bands of the correct sizes were confirmed using gel electrophoresis, the PCR products were cleaned using EXOSAP-IT (Thermo Fisher Scientific, Waltham, MA, USA), and the cleaned products were submitted to the Genetic Core Facility at Penn State University for Sanger sequencing. Using Qiagen’s CLC Genomic Workbench version 21 (Valencia, CA, USA), the forward and reverse sequences for each gene were assembled to the corresponding reference gene sequence of X. hortorum pv. hederae CFBP 4925^T. The consensus sequences from each gene were merged into one file for each strain, in the order listed above, and each strain’s merged sequence was aligned with similarly made files but for a collection of Xhv strains of known races. The model testing tool, with the default options, was used to determine the appropriate nucleic acid substitution model. As recommended, the general time-reversible model with four categories of rate variation, estimated topology, and 1000 bootstraps was used to generate the maximum likelihood phylogeny.