*3.1. Di*ff*erential Expression between Invasive A. philoxiroides and Native A. sessilis during Disease Development*

We found that there were interspecific differences in JA and ET-signaling, as well as in the degree of antagonistic cross-talk between SA and JA (which was stronger in the invasive species and weaker in the native species). This variation in the expression of defense hormone genes between species suggests *R. solani* may affect host plant defense hormones, specifically by inducing the major defense hormone SA (which may regulate other hormones, such as JA and ET, differently between plants for successful infection of plant tissue, Table 1). Our findings also suggest that the infection of *R. solani* was enhanced in the native *A. sessilis*, perhaps due to the weaker antagonistic cross-talk between SA and JA (which may allow faster or enhanced disease development to occur in the native species). Furthermore, our results indicate that the stronger antagonistic cross-talk between SA and JA in the invasive *A. philoxiroides* may delay disease progression. To our knowledge, this is the first study that demonstrates a necrotrophic pathogen influencing plant defense hormone pathways in an invasive clonal weed differentially compared to a native congener (Figure 10). A previous study [30] identified several genes that may be involved in the introduction success of invasive *A. artemisiifolia*. Of the 180 genes identified in this earlier study, some genes were found to be involved in the metabolism of plant hormone signaling and biosynthesis (e.g., lipoxygenases and cytokinins-zeatin o-glucosyltransferase) [30]. Gene expression differences between the invasive and native species suggested that invasive species may have evolved to stressful conditions during the invasion process [29–31].

**Figure 10.** Proposed model showing response of defense hormone genes, salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) during pathogenesis in an invasive *Alternanthera philoxeroides* (AP) and native *A. sessilis* (AS). We investigated the molecular interactions between *R. solani* compared to two species (focusing on defense hormone signaling). Two key differences in hormone gene expressions were identified between species during pathogenesis. Firstly, the JA and ET-signaling was differentially regulated between species (partially suppressed in *A. sessilis*, whereas there was a consistent reduction in expression in *A. philoxiroides*, shown as difference in height and color of JA). Reduction in JA-dependent *LOX*, *JAR1* and *PR6* expressions during disease development suggested that JA-signaling is responsible for resistance to *R. solani*. Secondly, ET-*EIN3* expression was reduced in *A. sessilis*, but was induced in *A. philoxiroides* (shown as difference in height and color of ET). SA level was induced in both species. The elevated levels of SA in both species during disease development suggest that the unknown virulence factor from pathogenic *R. solani* may potentially target SA. This in turn may affect the antagonistic effect between SA and JA/ET differentially between species (shown as a different arrow between SA and ET). Circled SA-*PAL* and JA-*JAR1* show antagonistic effects in both the local inoculated and neighboring systemic leaves. Dotted circles (*PR3*, *EIN3* and *PR6*) represent differential regulation in both inoculated and un-inoculated systemic sites.

The differences in defense hormone signaling between the invasive and native species observed in our study during pathogenesis may be the result of interspecific genetic variation. It would be interesting to study how differences in genetic variation between the invasive *A. philoxeroides* (which has previously been found to have low genetic diversity in China due to predominantly clonal vegetative propagation, [60,64]) and native *A. sessilis* (which may have higher genetic diversity in its native range, although this needs to be confirmed) may influence the resistance response of each species to pathogens. The data from our study provide an appropriate baseline for investigating this line of inquiry in the future.

#### *3.2. Changes in Defense Hormone Gene Expression during Pathogenesis in A. philoxeroides*

Our study demonstrates how a common and widespread pathogen regulates plant defense hormones (SA, JA, and ET) allowing for the successful infection of an invasive species, *A. philoxeroides* (using RT-qPCR). Our results showed that both JA-*JAR1* and *PR6* transcript expressions decreased during disease development following inoculation with *R. solani* AG4 HGI (Figure 6a,b), indicating that JA-signaling pathway is likely to be involved in plant resistance to the pathogen. There is much evidence to suggest that endogenous JA is triggered against necrotrophic pathogens [71,72]. For example, gene expression changes induced by *R. solani* AG1 IA in resistant and susceptible rice plants have been reported and it has also been shown that JA plays an important role in disease resistance [73]. In the present study, we found that the JA biosynthetic *LOX* gene was inconsistent in expression (Figure 6c). This was consistent with the findings of a previous study [74], where it was found that *LOX* expression increased approximately 6-fold in response to virulent and avirulent strains of *Pseudomonas syringae*.

In our study, *R. solani* induced SA (*PAL* and *PR3*), which may antagonize JA during pathogenesis in invasive *A. philoxeroides* (Figure 6a,b and Figure 7a,b). Earlier studies have shown cross-talk between hormone signaling pathways can greatly help the plant to regulate defense responses to a wide range of pathogens [75–77]. The predominant cross-talk observed between SA and JA is antagonistic [50,52]. For example, an exogenous application of SA was found to inhibit JA-induced proteinase inhibitor expression in tomato [54,78], whereas MeJA treatment inhibited SA-induced acidic *PR* gene expression in tobacco [55]. Like in our study, previous studies have found that phytopathogens can take advantage of the cross-talk between SA and JA allowing them to successfully infect plants [79,80]. For example, the hemi-biotrophic pathogen *P. syringae* was shown to manipulate antagonism between SA and JA by producing coronatine, which is a phytotoxin that mimics plant JA to suppress SA. This mechanism promotes disease infection in *Arabidopsis* and tomato [81]. The necrotrophic *Botrytis cinerea* produces exopolysaccharide (EPS), a virulence factor that elicits SA to activate antagonism to JA, thereby enhancing the ability of the fungus to infect tomato [56]. Our results are consistent with the findings of El Oirdi et al. [56], although the virulence factor from *R. solani* needs to be investigated further. A follow-up study by El Oirdi et al. [56] showed SA-signaling can also contribute to disease development (caused by another necrotrophic pathogen, *Alternaria solani*) in tomato [82]. It has also been found that the SA pathway might contribute to disease development (i.e., a *SAR8.2k* gene induced by SA may be involved in disease susceptibility caused by *R. solani* in tobacco [83]).

*PR* genes are often associated with specific signaling pathways and their expression can be regulated by different plant hormones [84]. In our study, at 48 hpi, the SA-inducible *PR3* showed a 10-fold higher expression to *R. solani* infection (Figure 7b), a 100-fold higher expression for SA pretreatment (Figure 7e), and a 75-fold higher expression for JA pretreatment (Figure 8b). This increase in SA-inducible *PR3* supports the hypothesis that SA enhances disease development by suppressing JA in *A. philoxeroides*.

Our results indicate that, as with SA, *EIN3* (an active form of the ET transcription factor) [77] enhances disease development in *A. philoxeroides* (Figure 7c). Although *EIN3* was induced at moderate levels, its expression was consistently induced over time during inoculation (Figure 7c,f). Previous studies have shown that ethylene is a potential modulator of plant pathogen defenses [85]. JA and ET together form an effective defense against necrotrophs [86], and can act positively or negatively with SA, depending on the specific pathogen interactions [87]. For example, ET modulates the antagonism between SA and JA pathways, and is mediated by NPR1 [53]. ET induces SA-responsive *PR1* gene expression in *Arabidopsis* [88]. In contrast, ET (*EIN3* and EIL1) represses the SA biosynthesis gene *ICS*/*SID2*, thereby reducing SA accumulation [89].

## *3.3. Hormonal Cross-Talk and Systemic Resistance in A. philoxeroides during R. solani Pathogenesis*

Invasive plants may encounter multiple pathogens with different infection strategies in their non-native range. Hormonal cross-talk provides signals to adjacent leaves from local infected tissues to resist a forthcoming infection [90–92]. This systemic defense is effective against pathogens with a similar attacking strategy [93]. For pathogens with a different infection mode, hormonal cross-talk between pathways (specifically SA and JA) plays a critical role in plant resistance [90,94]. The biotrophic pathogen inducing SA can activate antagonism to JA in infected local and systemic tissues, which in turn favors insect herbivores or necrotrophic pathogens [91].

In our study, trade-offs between the SA and JA pathways was clearly observed in infected local leaves (Figures 6 and 7). This trade-off was lower in neighboring leaves (Figure 9). Only the core signaling component of SA-*PAL* and JA-*JAR1* showed antagonism in the neighboring leaves, whereas other gene transcriptions (i.e., JA-*LOX* and *PR6*) were not suppressed (Figure 9). Our results are consistent with the finding of Spoel et al. [90], who reported that *P. syringae* suppressed JA-mediated resistance to *Alternaria brassicicola* at the infected site. However, this antagonism was at modest levels in neighboring tissues in *Arabidopisis* [90]. It was suggested that, while antagonism between SA and JA was moderate, their cross-talk expression was detected in systemic tissues [90]. However, spatial separation (local, adjacent or systemic), time (immediate or delayed), pathogen type (biotroph or necrotroph), and inoculation dosage are important factors determining systemic resistance trade-offs in plants [91].

## **4. Materials and Methods**

## *4.1. Plant Species and Pathogens*

*Alternanthera philoxeroides*(Amaranthaceae) is one of the 100 worst invasive species in the world [59]. Climate modeling suggests that many regions around the world are suitable habitat for the growth of *A. philoxeroides*, including areas in Southeast Asia, Southern Africa, and Southern Europe [95]. In China, it has successfully invaded 19 provinces since its introduction (usually on roadsides and lakeshores) [96].

*Alternanthera sessilis* (L.) DC. is native to China, and was used for comparison with *A. philoxeroides* in our study [22]. Ramets of *A. philoxeroides* and *A. sessilis* were randomly collected in August (summer) 2016 from naturally occurring sites in Fuzhou National Forest Park (119◦1701200E, 26◦9 03500N, Figure S7) and propagated in a greenhouse at Jiangsu University, Zhenjiang, China. Healthy stems with two nodes and without roots were planted in 250 g of sterilized sand in plastic pots (8 × 7 × 5 cm). Fixed volumes (30 mL) of distilled water or full-strength Hoagland liquid nutrient solution [97] were supplemented alternatively every two days to maintain optimal growth conditions.

*Rhizoctonia solani* is a ubiquitous soil-borne necrotroph that causes significant yield loss in many economically important crops globally [98]. The most common symptom is the 'damping-off' of seedlings or failure of infected seeds to germinate. The *R. solani* AG4-HGI (accession ACCC 30374) used in this study was obtained from the Agriculture Culture Collection of China (Agricultural and Microbial Culture Collection Management Center, Beijing, China). This strain was further sequenced to identify the specific anastomosis group (AG4-HGI), which was confirmed by PCR amplification of AG common and subgroup specific primers [99], and internal transcribed spacer (ITS) primers [100]. The fungus was grown and sub-cultured on potato dextrose agar (PDA) and incubated for five days at 28 ◦C.

## *4.2. Isolation of Putative Hormone-Responsive Genes for RT-qPCR*

Partial gene sequences of six hormones and responsive genes (one for ET, two for SA and three for JA) were successfully isolated from both *A. philoxiroides* and *A. sessilis*. Although there is currently a lack of hormone specific sequence information for many invasive and native plant species (including *Alternanthera* species), the species from our study are in the Amaranthaceae family (which was used for sequence isolation). Twenty-one hormones and their responsive gene sequences (i.e., eight for SA, seven for JA, and six for ET) were retrieved from species in the Amaranthaceae (Table S6). Each gene sequence from species in the Amaranthaceae was aligned using Clustal Omega [101] for primer design in the conserved region (using Primer3) [102] (for initial screening). We also performed a phylogenetic analysis of these sequences (Figure S2) to compare with other species.

Plant leaf material was ground using liquid nitrogen, and 50–100 mg of ground tissue was used for DNA extraction using the Rapid Plant Genomic DNA Isolation Kit (Sangon Biotech, Shanghai, China). Genomic DNA was further precipitated by adding sodium acetate and ethanol and purified using the method of Dellaporta et al. [103]. PCR was performed, and DNA was purified using the UNIQ-10 Column MicroDNA Gel Extraction Kit (Sangon Biotech, Shanghai, China) prior to sequencing. A total of six genes (i.e., *PAL* and *PR3* for SA; *LOX*, *JAR1* and *PR6* for JA; *EIN3* for ET) were successfully isolated based on the PCR amplification of similar amplicon lengths obtained from both *A. philoxiroides* and *A. sessilis* (Table S3). Furthermore, the coding (exonic) region of each isolated gene sequence was confirmed by cDNA PCR. We performed a NCBI nucleotide search of each of the isolated sequences and found that they showed closest homology to sequences of species in the Amaranthaceae. RT-qPCR primers (amplicon length < 200 bp) were designed using web-based Integrated DNA Technologies (IDT)-Primer Quest [104] from initially screened sequences of each gene. Primers with similar amplification efficiency in the cDNA of both invasive and native plants were used in RT-qPCR. All the primer sequences used for RT-qPCR are listed in Supplementary Table S2.

#### *4.3. Plant Inoculations with R. solani for RT-qPCR*

Pathogenicity tests consisted of detached (in vitro) and in planta leaves from four-week old stem cuttings of *A. philoxeroides* and *A. sessilis* inoculated with 5-mm agar plugs (*R. solani* culture). Both detached and in planta leaf assays were performed using the method used by El Oirdi and Bouarab [105]. Inoculated leaves of in planta inoculations with mycelium plugs were covered with clear zip-lock plastic bags to maintain high humidity. Detached and in planta disease symptoms were photographed every day, up to five days after inoculation.

In planta inoculations of *A. philoxeroides* and *A. sessilis* were used for gene expression experiments by quantitative RT-qPCR. Inoculated leaf samples were harvested at different time intervals, including 0, 6, 12, 24, 48, 72, and 96 h post-inoculation (hpi). There were three biological replicates per time interval (i.e., one plant per replicate each for *A. philoxeroides* and *A. sessilis*). The 0 hpi time interval refers to un-inoculated control samples. All inoculated plants were grown at 25 ◦C and 70% humidity with a 16 h photoperiod. Harvested leaf material at different time intervals was quickly frozen in liquid nitrogen for RNA extractions. For the systemic acquired resistance (SAR) tests, samples of un-inoculated neighboring leaves (i.e., younger leaves just above the inoculated leaf) were investigated. For SAR gene expression experiments, samples of three biological replicates were taken per time interval (0, 6, 12, 24, 48, 72, and 96 hpi). That is, one plant per replicate for each species was investigated. The experiments of *R. solani* infection and SAR were repeated twice.

The effect of each hormone (SA, JA, and ET) on resistance to *R. solani* was tested using RT-qPCR by spraying each hormone two days before inoculation with *R. solani*. Hormone pretreatments were performed on four-week old plants of *A. philoxeroides* and *A. sessilis*. There were three treatments in this experiment group for each species (SA, JA, and ET pretreatments). Hormones SA (0.5 mM, BBI Life Sciences, Shanghai, China), methyl jasmonate (MeJA, 0.1 mM, Sigma-Aldrich, St. Louis, MO, USA), and ethephon (ET, 0.5 mM, BBI Life Sciences, Shanghai, China) were dissolved in water. Hormones were sprayed directly on plant leaves (in each treatment group) until drenched (surface run-off) once

per day, for three consecutive days. All treatment plants were grown at 25 ◦C and 70% humidity with a 16 h photoperiod until they were harvested. Each of the hormone pretreated samples was harvested at different time intervals: 0, 6, 12, 24, 48, 72, and 96 hpi with *R. solani*. The hormone pre-treatment samples for SA and ET spray were unable to be collected after 72 and 96 h. This was because the ET-spray samples wilted early and could not be sampled (i.e., healthy leaves could not be obtained). The control treatment (0 hpi) consisted of only water pretreated and un-inoculated. There were three biological replicates for each time interval (i.e., one plant per replicate) for each species. All harvested samples were quickly frozen in liquid nitrogen for RNA extractions. The experiment was repeated twice.

## *4.4. RT-qPCR Analysis*

Total RNA was isolated from leaves of invasive and native plants in each treatment using the TaKaRa MiniBEST Plant RNA Extraction Kit according to the manufacturer's instructions (Takara, Shiga, Japan). First-strand cDNA was synthesized from 500 µg total RNA using PrimeScript RT Master Mix (Takara, Japan). The targets were amplified using primers that are listed in Table S2. RT-qPCR was performed using SYBR Premix Ex-Taq (Takara, Japan) on CFX96 Real-Time PCR Detection System (Bio-rad). Melt-curve analysis was performed after each PCR run to ensure specific amplification of each gene-specific primer. To identify the suitable housekeeping gene, we tested three reference genes, β*-tub* (β-tubulin), *EF1*α (Elongation factor 1-alpha), and *Act* (Actin), in *R. solani* infected samples based on Cycle threshold (Ct) difference and the coefficient of variance method [106]. Primers are listed in Table S2. *Actin* had the lowest Ct difference in *R. solani* infected samples across different time intervals (Manoharan B. et al., unpublished data) and was used for normalizing the expression level of each target gene. Relative gene expression was calculated using the comparative 2−∆∆*CT* method [107]. Three biological replicates were used for each gene and each reaction was independently replicated three times.
