1. Introduction

As a single species, Fusarium oxysporum is rated the fifth most important fungal plant pathogen in the world [1]. Within the Fusarium oxysporum species complex (FOSC), there are over 106 known formae speciales which infect more than 100 different hosts causing vascular wilts ^[1][2][1,2]. They are soilborne, can survive for long periods, are often unaffected by chemical management, and can evolve to overcome host resistance quickly ^[3][4][3,4]. As hemibiotrophic pathogens, they not only cause yield loss, but also result in total plant death and crop loss [5]. Recent research trends in several formae speciales of the FOSC have focused on the molecular analysis of pathogenicity genes, whole genome sequence analysis, and proteomics, to understand the infection process. It is the case that FOSC isolates are ubiquitous in the soil around the world, but are largely nonpathogenic ^[6][7][6,7]. The mobility and plasticity of the FOSC genomes is also of particular interest to determine if nonpathogenic strains can gain pathogenic function either by traditional mutation or through chromosomal movement and exchange ^[8][9][8,9].

In the Southeastern United States, Fusarium oxysporum forma specialis niveum (Fon) is widely distributed and causes major yield losses in watermelons, the only host of Fon ^[10][11][10,11]. There are additional FOSC formae speciales that infect watermelon but also infect other crops [12]. Fusarium wilt of watermelon, the disease caused by Fon, results in vascular clogging leading to wilting in one or two infected runners ( Figure 1 ). Over time, the infection intensifies leading to total plant wilting and eventual necrosis [13]. The cool and moist springtime climate (below 70 °F) common to the Southeastern US is favorable for disease spread and initiation and occurs during the watermelon seedling stage. Thus, Fon is responsible for significant damping off in seedlings, and outbreaks primarily occur early (February–April) in the season [14].

Figure 1. Fusarium wilt symptoms (A) in the field, and (B) in the vasculature (arrow).

Some formae speciales are further divided into races (usually) to indicate cultivar-specific pathogenicity, however, this terminology is not universally accepted and thus raises questions about appropriate nomenclature [2]. Subdivisions of race within an individual forma specialis are defined by their pathogenicity on specific and newly susceptible cultivars. Races are characterized by their ability to overcome a specific cultivar’s host resistance ^[2][15][16][2,19,20]. This reaction also can describe virulence, as how the differential interactions linked to virulence genes (R-genes) govern the susceptibility of the previously resistant cultivars. Aggressiveness is a quantitative component of pathogenicity and does not have a connection to race differentiation ^[17][21]. Resistant watermelon cultivars have been developed, but new resistant races of Fon have arisen ^[18][22]. Fon has four recognized races: 0, 1, 2, and 3. Each subsequent race is determined by its ability to cause infection on previously resistant cultivars, with race 3 having the largest range of pathogenicity. Although the specific avirulence gene responsible for overcoming cultivar resistance is not known in most races, the ability to cause infection should be characterized as virulence rather than pathogenicity as their differentiation rests on a 1–9 scale of disease rating. The aggressiveness of Fon has been thought to positively correlate with increased cultivar range, meaning race 3 isolates both have a larger range of pathogenicity (virulence) and aggressiveness than race 1 isolates. This has been argued due to multiple isolates of the same race being tested and showing variation within their aggressiveness ^[19][20][23,24].

Fon is found in most, if not all, watermelon growing regions in the world ^[21][25]. It is present in the six possible continents and is recorded in 44 countries ( Figure 23 ). The majority of countries that report the presence of Fon do not specify the race (only 13 nations report specific races) which could suggest the presence of additional races or levels of virulence of detected isolates. Distinct countries use different watermelon cultivars for race differentiation purposes as well, complicating the comparison across continents and countries ^[22][26]. This goal of this review is to discuss the available diagnostic and race-differentiation methods for Fon while also providing a brief overview of presence, distribution, and management.

Figure 2. Global Fon distribution. Countries are colored based on highest race detected: green, race 1; yellow, race 2; red, race 3; and purple, race not reported. (A) Country area proportional to watermelon yield (hg/hectare), and (B) country area proportional to watermelon production (metric tons). Images produced using an online program which scales images depending on a determined variable while maintaining the original object boundaries as computed by a flow-based algorithm ^[23].

2. Diagnostics

Forma specialis status of Fon was designated based on the specificity of Fon to be solely pathogenic on watermelon and not infecting closely related cucurbit hosts [11]. Race differentiation is also done with a bioassay, but using multiple different cultivars which have a corresponding reaction to each particular race of the pathogen ( Table 1 ) ^[24][28].

More difficult and subsequently more important for growers and breeders, is Fon race differentiation. While the traditional bioassay method has been discussed, there exist a few molecular markers that claim the ability to differentiate between several of the races present in the literature. The first, by Niu et al. (2016), identified the avirulence gene Secreted in Xylem 6 ( SIX6 ) in races 0 and 1, but was absent in race 2 ^[28][36]. Race 3 was not tested in their study, but in subsequent research, race 3 was found to contain SIX6 ^[20][24]. Mutation studies with this gene increased virulence in race 1 isolates without SIX6 and reduced virulence in race 2 isolates when given SIX6 , drawing a connection between isolate virulence and the Secreted in Xylem gene family which is discussed further below. The absence of SIX6 as detected by FONSIX6 specific markers has been used for race 2 differentiation, with races 0, 1 and 3 showing a positive reaction. In the literature, isolates with levels of pathogenicity that would normally identify the isolate as race 2 seldom use the marker for SIX6 to determine a correlation between the bioassay and the marker. When carried out, it does not appear to correlate at a high percentage ^[22][26], which suggests multiple genetic contributions for pathogenicity, of which SIX6 may be one.

Further race differentiation research focused on race 3 differentiation similar to that of race 2. Hudson et al. (2021) ^[20][24] sequenced the whole genomes of suspected race 1, 2 and 3 isolates before determining the regions that could be used for differentiation. The primer set FNR3F/R, while targeting a chromosomal region involved in pathogenicity, did not directly link the amplified region to any gene function other than coding for a hypothesized protein ( Figure 35 ). FNR3F /R does not provide race 0 identification, nor identification of nonpathogenic Fon strains. FNR3F/R amplifies a region in race 1 and race 2 isolates but does not amplify anything in race 3 isolates. Using Fon-1/Fon-2 and FONSIX6F/R primer sets alongside FNR3F/R, races 1, 2 and 3 can be differentiated. Based on a sampling of over 90 race typed Fon isolates, 89% of race 1 isolates were predicted by the marker, 80% of race 2 isolates, and only 60% with the race 3 marker. Problems with this marker include the correlation between the bioassay results and the molecular results, as well as not including nonpathogenic or race 0 isolates. In order to determine the function of the gene targeted with FNR3F/R, knockout mutants lacking the genomic region must be made for various isolates ^[28][29][30][36,37,38].

Figure 3. Molecular method of Fon race differentiation adopted from Hudson et al. (2021) ^[20].

SIX ^[31][32][33][47,48,49] proteins have been shown to be strongly correlated with disease progression and virulence in hosts infected by Fo species ^[34][50]. The exact mechanism and how they function to cause disease is unknown, however, research has highlighted their possible involvement in interfering with host signal transduction and the Jasmonic acid-mediated response to detecting the presence of PAMPS ^[35][36][51,52]. SIX proteins are characterized by containing fewer than 300 amino acids, an abundance of cysteine residue, and the inclusion of a secretion peptide signal. The exact composition of SIXs and their homologs can predict the pathogen host range including formae speciales and races and have been used to develop molecular diagnostic assays ^{[37][38][39][40]}[53,54,55,56]. Since the first suggested use of distinguishing races and formae speciales by its application as proposed by Lievens et al., research in FOL showed that the three described races were differentiated by the presence/absence of SIX1, 3, and 4 genes ^[41][42][57,58]. In another example, Czislowski et al. who showed that the SIX absence/presence profile was strongly correlated with the pathogenicity of known Foc lineages ^[43][59]. Similarly, it is hypothesized that the virulence demonstrated by various Fon isolates on differential cultivars is likely the result of specific permutations of SIX, or other, effectors. For example, Table 24 shows the distribution and identity of known SIX effectors in all publicly available Fon genomes (SAMN15791673, SAMN15791674, and SAMN15791675). While the race classification for some of these assemblies is known, most are not. BLAST searches reveal seven unique combinations which suggests either additional races or effectors with a negligible contribution to pathogenicity. Clearly, additional isolate genomes with race classification information is necessary for the application of this method to sufficiently differentiate the genetic basis underlying the variability of virulence in Fon.

3. Conclusions

Based on the limited host range of Fon, it likely has a reduced, yet specific, set of effectors that allow for setting up a disease to watermelon. The set of effectors present in Fon which permit infection in watermelons is likely shared with other Fusarium species whose host range includes watermelon ( F.o. melonis , F.o. cucumerinum , possibly even F. solani cucurbitae ). However, the exact, unique complement of effectors which overcome cultivar-specific host resistance remains uncharacterized. Because of this perspective, we suggest that, moving forward, an effort should be made to increase the number of available Fon genomes with or without race ID. Once a significant number of genomes are available from variable geographic locations, the profiling done here should be repeated with the new isolates to determine the number of distinct gene profiles. Effector function on Fon virulence affecting available watermelon cultivars should then be examined and each SIX gene’s function on virulence can be determined using the susceptible cultivars. Connections then can be made for breeders based on corresponding resistance (R) genes in watermelon and race designation can be done using these genetic interactions.

While it is true that the designation of specific races based on an effector gene profile requires more steps and greater computational resources, once determined, the most important genetic combinations can be adapted for molecular diagnostic methods ^[46][35]. Additionally, the ability for FOSC members to exchange chromosomal elements and perform horizontal gene transfer no longer interferes with molecular diagnostics because detection will be based on the specific genes which govern virulence, not simply conserved regions with other metabolic functions. Effector genes that are common to all isolates can then be analyzed compared to the other Fusarium species that infect watermelon to design new and improved molecular markers for Fon-specific detection.

It has been noted by several researchers studying Fon that the difference between races 0 and 1 may be quantitative rather than qualitative, but that race 2 is distinct ^[21][25]. These perspectives were written before the appearance of race 3, but it would still appear that multiple genes facilitate the virulence of any given race and that a particular combination of those virulence genes may dictate the ability of a Fon isolate to cause infection. A single identifiable gene required for infection on a resistant watermelon cultivar is needed to consider the resistance qualitative; FONSIX6 has been studied as that target gene to isolate race 2 Fon isolates ^[28][36]. Confirmation of FONSIX6 as the only gene responsible for infection on the resistant cultivars (i.e., qualitative disease resistance) must occur by demonstrating that the presence or absence of FONSIX6 causes no difference in aggressiveness against susceptible cultivars, but only allows for the infection to take place on the previously resistant cultivars. If, when FONSIX6 is absent, the Fon isolate is shown to increase the aggressiveness on previously susceptible cultivars and that it overcomes resistance, then the resistance is quantitative and breeders must incorporate resistance against all avirulence genes involved in the disease process.