Overexpression of DUF538 from Wild Arachis Enhances Plant Resistance to Meloidogyne spp.

: DUF538 proteins belong to a large group of uncharacterized protein families sharing the highly conserved Domain of Unknown Function (DUF). Attention has been given to DUF538 domain-containing proteins due to changes in their gene expression behavior and protein abundance during plant development and responses to stress. Putative roles attributed to DUF538 in plants under abiotic and biotic constraints include involvement in cell redox balance, chlorophyll breakdown and pectin degradation. Our previous transcriptome studies suggested that DUF538 is also involved in the resistance responses of wild Arachis species against the highly hazardous root-knot nematodes (RKNs). To clarify the role of the AsDUF538 gene from the wild peanut relative Arachis stenosperma in this interaction, we analyzed the effect of its overexpression on RKN infection in peanut and soybean hairy roots and Arabidopsis transgenic plants. AsDUF538 overexpression signiﬁcantly reduced the infection in all three heterologous plant systems against their respective RKN counterparts. The distribution of AsDUF538 transcripts in RKN-infected Arachis roots and the effects of AsDUF538 overexpression on hormonal pathways and redox system in transgenic Arabidopsis were also evaluated. This is the ﬁrst time that a DUF538 gene is functionally validated in transgenic plants and the earliest report on its role in plant defense against RKNs.


Introduction
The DUF (Domain of Unknown Function) family comprises proteins with one or more conserved domains with no functional annotation. In the Pfam database, more than 4000 DUF families are described, encompassing more than 22% of the entire database (Pfam 33.2, June 2020, 18,259 entries). Despite DUFs having as yet no clarified function, recent studies suggest that they might play an important role in plant stress responses [1][2][3]. Within this large family, DUF538 superfamily members are among the most studied proteins, found in more than 40 plant species, almost exclusively in Embryophyta [4]. Changes in DUF538 gene expression behavior and protein abundance were observed during different plant developmental and reproductive stages [5][6][7] and in response to various abiotic stresses, including drought [8]; high temperature [9]; carbohydrate deprivation [10]; and nitrogen toxicity [11]. DUF538 protein was also reported to be involved in intracellular transport of cuticular waxes, affecting leaf surface permeability and water loss control [12].
Regarding biotic stresses, some reports identified DUF538 transcripts or proteins in response to infection by fungal pathogens [13][14][15][16], where they seem to be important activators of the plant antioxidative system and a common component in the responses of
The 414 bp consensus coding sequence of AsDUF538 ( Figure 1A) was synthesized and cloned by Epoch Life Science Inc. (Missouri City, TX, USA), under the control of the Arabidopsis actin 2 promoter and Agrobacterium nopaline synthase terminator, in the binary vector pPZP_201BK_EGFP [36]. The obtained pPZP-AsDUF538 vector comprises two additional cassettes for constitutive expression of the enhanced green fluorescent protein (eGFP) reporter gene and the hygromycin phosphotransferase (hpt) gene coding for hygromycin resistance as the selectable marker [36]. cession V10309) was identified based on alignments of EH045758.1 (Genbank; [21]) and putative DUF538-domain coding sequences from Arachis species available from NCBI (http://www.ncbi.nlm.nih.gov/ and PeanutBase; accessed on 10 October 2020) and Pea-nutBase (https://peanutbase.org/; accessed on 10 October 2020), as well as from our previous transcriptome surveys (NCBI BioProject PRJNA284674). Transcript sequences were aligned using the default parameters of the MultAlin tool (http://multalin.toulouse.inra.fr/multalin/; accessed on 3 November 2020). The isoelectric point and molecular weight of the AsDUF538-deduced protein were predicted using software available at ExPASy (http://expasy.org/; accessed on 10 October 2020). The amino acids alignment and maximum likelihood phylogenetic tree were produced by MEGA X [35].
The 414 bp consensus coding sequence of AsDUF538 ( Figure 1A) was synthesized and cloned by Epoch Life Science Inc. (Missouri City, TX, USA), under the control of the Arabidopsis actin 2 promoter and Agrobacterium nopaline synthase terminator, in the binary vector pPZP_201BK_EGFP [36]. The obtained pPZP-AsDUF538 vector comprises two additional cassettes for constitutive expression of the enhanced green fluorescent protein (eGFP) reporter gene and the hygromycin phosphotransferase (hpt) gene coding for hygromycin resistance as the selectable marker [36]. The pPZP-AsDUF538 and the corresponding empty pPZP_201BK_EGFP (hereafter called pPZP-empty) binary vectors were then introduced into the Agrobacterium rhizogenes pathogenic strain 'K599' and into the disarmed A. tumefaciens strain 'GV3101' using a standard electroporation protocol. Transformed colonies were selected by PCR using specific primer pairs (5′-3′) from AsDUF538 (GAGCACAGGGTACATGTG- The pPZP-AsDUF538 and the corresponding empty pPZP_201BK_EGFP (hereafter called pPZP-empty) binary vectors were then introduced into the Agrobacterium rhizogenes pathogenic strain 'K599' and into the disarmed A. tumefaciens strain 'GV3101' using a standard electroporation protocol. Transformed colonies were selected by PCR using specific primer pairs (5 -3 ) from AsDUF538 (GAGCACAGGGTACATGTGGA/CATAAGCTCCTTGG-CCTTCA) or eGFP (CGTCGCCGTCCAGCTCGACCAG/CATGGTCCTGCTGGAGTTCGTG) sequences, according to [23,37], respectively.

Nematode Inoculum
Nematodes were multiplied for three months on greenhouse-grown tomato (Solanum lycopersicum 'Santa Clara') plants. Eggs of Meloidogyne spp. were extracted from roots using 1% (v/v) NaOCl and second-stage juveniles (J2) collected from the root and egg suspension, as previously described [29]. Eggs or freshly hatched J2s were collected by centrifugation and quantified under a light microscope using Peters' slides.

AsDUF538 Transcripts Distribution in Arachis spp. Roots Inoculated with Meloidogyne arenaria
For in situ hybridization analysis, the AsDUF538 coding sequence flanked by T7 and SP6 RNA polymerase promoters was cloned into the expression vector pGEM-T-Easy (Promega, Madison, WI, USA) and used as a template for in vitro transcription using the DIG RNA Labeling Kit SP6/T7 (Roche, Indianapolis, Indianapolis, USA) to produce both, the antisense (SP6) and sense (T7) digoxigenin-labeled RNA probes. Sections of leaves and roots from eight-week-old plants of the RKN-resistant A. stenosperma and susceptible peanut 'Runner IAC-866' were inoculated with 10,000 J2 of M. arenaria. Control plants were mock inoculated. Root samples were collected at three and six DAI (Days After Inoculation) and used to determine the temporal and spatial distribution of AsDUF538 transcripts. Roots were fixed, dehydrated and embedded in BMM (butyl acrylate-methyl methacrylate polymer) plastic resin. Semi-thin sections were obtained and hybridized either with antisense or sense probe for 16 h at 42 • C. Hybridization sites were immunocytochemically detected, as previously described [23]. Root sections, not counterstained, were observed in the AxioPhot epifluorescence microscope (Zeiss, Oberkochen, Germany) and images captured by the AxioCam MRc digital camera and the corresponding Axiovision Rel. 4.8 software (Zeiss, Oberkochen, Germany).

Peanut and Soybean Hairy Roots Inoculation with Meloidogyne arenaria and Meloidogyne javanica
A. rhizogenes strain 'K599' harboring pPZP-AsDUF538 and pPZP-empty vectors were used to generate hairy roots from both detached leaves of the RKN-susceptible peanut 'Runner IAC-866' and composite plants of the RKN-susceptible soybean 'Williams 82'.
Peanut hairy roots were produced by the ex vitro detached leaf method previously established by our group [37,38], which uses the youngest expanded quadrifoliate leaves harvested from one-month-old peanut plants grown under growth chamber conditions (25 ± 2 • C; 12 h photoperiod; 120 µmols·m −2 ·s −1 light intensity). Detached leaves were inoculated with A. rhizogenes bacterial paste and maintained under moist conditions for approximately 20 days until the emergence of the first hairy roots, essentially as described by [37,38]. Transformed hairy roots were individually screened by the presence of GFP fluorescence using a fluorescence stereomicroscope M205 (Leica Microsystems, Wetzlar, Germany) with the GFP1 filter, and roots lacking fluorescence were removed. Detached leaves containing eGFP-positive roots were kept in the growth chamber for 10 additional days and re-evaluated for GFP fluorescence. The eGFP-negative roots were removed and only the detached leaves with eGFP-positive roots, achieving up to 5 cm in length, as previously established by [38], were used for nematode inoculation. Detached leaves containing only eGFP-positive hairy roots were then transferred to wet vermiculite (grade 3), challenged with 1000 J2 of M. arenaria, and maintained in the growth chamber.
Composite soybean plants were generated by the inoculation of a fresh A. rhizogenes paste at the hypocotyl region of four-day-old seedlings, essentially as described by [39]. Hairy roots were selected one week after A. rhizogenes transformation through GFP fluorescence, as described above, and the eGFP-negative roots were removed. Composite plants containing only eGFP-positive hairy roots were transferred to a sand: plant gel mixture (3:1; v:v) and acclimated to greenhouse conditions. Hairy roots from three-week-old soybean composite plants transformed with pPZP-AsDUF538 or pPZP-empty were then challenged with 1000 J2 of M. javanica.
At 60 DAI, hairy roots from peanut detached leaves and composite soybean plants were collected, weighted and reevaluated for GFP fluorescence. The nematode infection in hairy roots transformed with both pPZP-AsDUF538 and pPZP-empty vectors was determined by counting the number of galls and egg masses and statistically tested according to [37].

Arabidopsis Transformation and OE Lines Selection
Wild type (WT) Arabidopsis thaliana plants (ecotype Col-0) were transformed with A. tumefaciens 'GV3101' harboring the pPZP-AsDUF538 vector via the floral dip method [40]. Plants were maintained in a controlled growth chamber (21 ± 2 • C; 12 h photoperiod; 120 µmols·m −2 ·s −1 light intensity). T0 transformants were selected on MS agar medium containing 20 mg/L hygromycin and screened by the presence of GFP fluorescence, as described above. Hygromycin-resistant and eGFP-positive T0 plants were transplanted to soil-containing pots and maintained in the growth chamber to produce transgenic AsDUF538 plants at T1 generation. Transformants were then screened repeatedly for hygromycin resistance, as previously described [41] to obtain homozygous AsDUF538 overexpressing (OE) lines at T2 generation for further assays and analyses.

Morphological Analysis of Arabidopsis OE Lines
To verify the occurrence of morphological alterations in root development and leaf surfaces of transgenic plants overexpressing AsDUF538, samples were collected from four-week-old Arabidopsis WT plants and two OE lines (OE4 and 8). Root and leaf morphology were observed using the fluorescence mode bright and dark field in the microscope AxioPhot, and images captured as described above.

Arabidopsis OE Lines Inoculation with Meloidogyne incognita
Ten Arabidopsis WT plants and three independent OE lines (OE4; 8; and 17) grown on a sand:substrate mixture (2:1; v:v) for three weeks were inoculated with approximately 2000 J2 of M. incognita, as previously described by [42]. Roots were then collected at 60 DAI, individually weighted and stained with acid fuchsin. The nematode infection was assessed by counting the number of females in each plant root system in the fluorescence stereomicroscope. The number of nematode females per gram of roots in WT plants and OE lines was determined and statistically compared using the t-test (p < 0.05), according to [42].

Gene Expression Analysis of Arabidopsis OE Lines
Total RNA was extracted from T2 seedlings of Arabidopsis OE lines and WT plants by RNeasy Plant Mini kit (Qiagen, Hilden, Germany) and reverse transcription was conducted for cDNA synthesis, as previously described [23]. RT-PCR analysis of 24 stress-related Arabidopsis marker genes (Supplementary Table S1) was performed according to [23].
The relative expression of the AsDUF538 transgene and five stress-related Arabidopsis marker genes was determined in OE lines and WT plants through qRT-PCR analysis using specific primers (Supplementary Table S2), as previously described [42]. The qRT-PCR reactions were conducted on a StepOne Plus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA), as previously described [23], using biological and technical triplicates for WT plants and each OE line. No template control (NTC) samples were included as negative controls. qRT-PCR cycling consisted of four steps: 50 • C for 2 min; 95 • C for 10 min; followed by 40 cycles of 95 • C for 15 s and 60 • C for 1 min, and a final dissociation curve step of at 95 • C for 15 s, 60 • C for 60 min and 95 • C for 15 s. The online real-time PCR Miner tool [43] was used to estimate the primer efficiency and optimal cycle of quantification (Cq) values.
The mean normalized expression (MNE) of AsDUF538 transgene in OE lines, relative to the AtACT2 reference gene (Supplementary Table S2), was calculated using the Q-GENE software (http://www.gene-quantification.de/download.html; accessed on 25 November 2020) [44]. The quantification of mRNA levels of the five Arabidopsis marker genes in OE lines, relative to WT plants were calculated and statistically tested via REST 2009 v. 2.0.13 software [45], using AtACT2 and AtEF-1α as the reference genes (Supplementary Table S2).

Molecular Characterization and Cloning of AsDUF538
The consensus coding sequence of the DUF538 gene from A. stenosperma (AsDUF538) was determined by the alignment of eight Arachis sequences putatively coding for DUF538 proteins, using as reference an A. stenosperma sequence (Genbank EH045758.1) previously identified as upregulated in RKN-infected roots [21] ( Figure 1A). Five aligned sequences were obtained from public databases, such as NCBI (XM_025756132; XM_016080842; and XM_016329625) and PeanutBase (Araip.IZQ8N and Arahy.E28A1S) and two (259_c3148 and large_rep_c660) from SRA NCBI database (Bio Project PRJNA284674). They comprised transcript sequences from distinct Arachis wild species, A. stenosperma (Araste), A. duranensis (Aradu), and A. ipaënsis (Araip), and from the cultivated species A. hypogaea (Arahy). The 414 bp consensus coding sequence of AsDUF538 produced from the alignment of these eight sequences ( Figure 1A) showed high nucleotide conservation, with the in-frame stop codon (TAA) at position 414, relative to the start codon (ATG) at position 1 ( Figure 1A). Only two single nucleotide polymorphisms (SNPs) were identified at positions 273 (T:A) and 285 (G:A).
AsDUF538 encodes a putative protein of 137 amino acid residues, with a unique DUF538 conserved domain (PF04398), and a theoretical pI of 9.12 and molecular weight of 15.18 kDa ( Figure 1B). No signal peptide was identified in this amino acid sequence. To show the high conservation among the DUF538 superfamily among legume species, we performed a phylogenetic analysis using the plant species at Uniprot (https://www. uniprot.org/; accessed on 5 August 2020), with the highest level of identity with AsDUF538 ( Figure 1C).

AsDUF538 Transcripts in Arachis Roots Inoculated with Meloidogyne arenaria
The temporal and spatial in situ distribution of AsDUF538 transcripts was determined on semi-thin sections of roots and leaves of A. stenosperma, the RKN-resistant species and A. hypogaea, the susceptible species. Roots of A. stenosperma collected at 3 DAI with M. arenaria showed hybridization signals in few radicular cells (Figure 2A,B), mainly concentrated in cells of the vascular cylinder (xylem, phloem, and intervening procambial or cambial cells). These same cells showed stronger hybridization in A. stenosperma roots at 6 DAI ( Figure 2C). On the contrary, root sections of A. hypogaea collected at 3 DAI lacked detectable signals ( Figure 2D,E), but with some signals detected in the vascular cylinder cells at 6 DAI ( Figure 2F). These results indicated an earlier and stronger in situ hybridization of the AsDUF538 probe in A. stenosperma than in A. hypogaea, suggesting that the accumulation of transcripts occurs differently in the susceptible and resistant Arachis species.
Interestingly, mock-inoculated roots of both Arachis species hybridized with the antisense probe (SP6) lacked hybridization signals in their roots collected six days after the start of the bioassay, as demonstrated by the root section of A. stenosperma ( Figure 2G). Likewise, roots collected at 6 DAI, from both Arachis species that were hybridized with the sense probe (T7) lacked signals, as demonstrated by the root section of A. hypogaea ( Figure 2H).
Leaves from inoculated A. stenosperma plants collected at 6 DAI showed evident hybridization signals in the vascular cylinder cells ( Figure

Meloidogyne arenaria Infection in Peanut Hairy Roots
After 20 days of A. rhizogenes transformation, roots emerging from the petiole-wounding site of peanut detached leaves ( Figure 3A) were analyzed for the presence of GFP fluorescence ( Figure 3B). From the 20 detached leaves transformed with the pPZP-empty vector, 18 (90%) developed eGFP-positive hairy roots, with an average of six hairy roots per leaf. Likewise, from the 20 detached leaves transformed with pPZP-AsDUF538, 16 (80%) were positively transformed, with the same average number of six hairy roots per leaf. All eGFP-positive roots transformed with both vectors exhibited typical hairy root phenotype, including vigorous growth and extensive lateral branches ( Figure 3A). These results indicated that the efficiency of A. rhizogenes transformation and induction of hairy roots per detached leaf were not affected by AsDUF538 overexpression.

Meloidogyne javanica Infection in Soybean Hairy Roots
Seven days after A. rhizogenes transformation, from the 40 soybean plantlets transformed with pPZP-empty vector and 40 with pPZP-AsDUF538, 28 (70%) and 25 (62%), respectively, developed typical hairy roots at the hypocotyl wounding site. The subsequent screening revealed that around 60% of the induced hairy roots per composite plant exhibited GFP fluorescence. As observed for peanut hairy roots, these results indicated Sixty days after M. arenaria inoculation, 10 detached leaves transformed with the pPZP-empty and 10 with pPZP-AsDUF538 vectors (55 and 62%, respectively), displaying GFP fluorescence, were considered for nematode infection assessment. Since each detached leaf produced an average of six hairy roots, approximately 60 roots were thus evaluated per treatment (pPZP-empty and pPZP-AsDUF538 transformation). eGFP-positive hairy roots showed average biomass of 232 and 276 mg per detached leaf transformed with pPZP-AsDUF538 and pPZP-empty vectors, respectively. The root biomass for the above treatments did not differ significantly (p < 0.01; t-test) and was similar to our previous studies with hairy roots transformed with pPZP-empty vector or overexpressing distinct candidate genes [33,37]. The M. arenaria infection was assessed in eGFP-positive hairy roots derived from pPZP-empty vector transformation by gall development (average of 19.03 galls per root gram) and production of egg masses (average 13.17 egg masses per root gram) ( Figure 3C,E). This confirms the ability of M. arenaria to infect hairy roots derived from detached peanut leaves and successfully complete its life cycle inside these roots, as previously demonstrated [33,37]. The number of galls (average of 2.44 galls per root gram) and egg mass (average of 2.15 egg masses per root gram) at 60 DAI was drastically reduced in hairy roots derived from the pPZP-AsDUF538 vector in comparison to the control roots that did not express the transgene ( Figure 3D,E). Therefore, the overexpression of AsDUF538 promoted a significant (p < 0.05; t-test) reduction of 87.18% in M. arenaria galls and 83.66% in the egg masses in peanut hairy roots.

Meloidogyne javanica Infection in Soybean Hairy Roots
Seven days after A. rhizogenes transformation, from the 40 soybean plantlets transformed with pPZP-empty vector and 40 with pPZP-AsDUF538, 28 (70%) and 25 (62%), respectively, developed typical hairy roots at the hypocotyl wounding site. The subsequent screening revealed that around 60% of the induced hairy roots per composite plant exhibited GFP fluorescence. As observed for peanut hairy roots, these results indicated that the efficiency of A. rhizogenes transformation and induction of hairy roots in composite soybean plants were not affected by the overexpression of AsDUF538.
At 60 days after M. javanica inoculation, soybean showed hairy roots ( Figure 4A) with GFP fluorescence (Figure 4A,B). From the 28 and 25 composite plants expressing pPZP-empty and pPZP-AsDUF538 vectors, 10 (55%) and six (24%), respectively, still showed eGFP-positive hairy roots at 60 DAI. Only roots showing GFP fluorescence were considered for nematode infection assessment. Regardless of the binary vector used for A. rhizogenes-mediated transformation, composite plants maintained an average of 60% eGFP-positive roots at 60 DAI and similar root biomasses, with an average of 2.75 g and 2.63 g for pPZP-empty and pPZP-AsDUF538 vectors, respectively, which did not differ significantly (p < 0.01; t-test). The average number of developed galls and egg masses per gram of pPZP-empty hairy roots (16.01 and 17.58, respectively) ( Figure 4C-E) indicated that M. javanica could successfully complete its life cycle inside soybean hairy roots, as previously demonstrated [46]. In contrast, hairy roots transformed with pPZP-AsDUF538 vector showed a strong reduction in nematode infection, with an average of 2.13 galls and 2.48 egg masses per root gram ( Figure 4A,B,E). Therefore, a significant (p < 0.05; t-test) reduction of 86.66% in M. javanica galls and 85.91% in egg mass was observed in soybean hairy roots overexpressing AsDUF538. These percentage reductions are remarkably like those for peanut hairy roots and clearly reinforces that AsDUF538 overexpression increases the resistance to RKN infection in legume species. Agronomy 2021, 11, x FOR PEER REVIEW 11 of 20  Statistical analyses were performed using the t-test and asterisks indicate significant differences (p < 0.05). Bars = 5 mm.

AsDUF538 Overexpression in Transgenic Arabidopsis
In order to investigate the effects of AsDUF538 overexpression on M. incognita resistance, we further stably transformed Arabidopsis plants. A total of 15 independent homozygous OE lines at T2 generation was successfully obtained, and the overexpression of AsDUF538 transgene was subsequently confirmed in all OE lines by qRT-PCR analysis (Supplementary Figure S1). The AsDUF538 transcript abundance relative to the reference AtACT2 gene differed among individual OE line (Supplementary Figure S1), whilst its expression was not detected in WT plants. We selected three OE lines (OE4, 8, and 17) that displayed the highest AsDUF538 expression levels for further analysis.
Morphological analysis showed that developing roots and leaves of OE lines 4 and 8 had similar morphological characteristics compared to WT roots ( Figure 5). OE lines displayed consistent cell and tissue morphology and organization in the different tissues of the roots and leaves, indicating that AsDUF538 overexpression did not affect the normal development and organization of these organs in Arabidopsis transgenic plants ( Figure 5).

Expression of Stress Related-Marker Genes in Arabidopsis OE Lines
The expression of 24 stress-related genes in non-challenged Arabidopsis OE lines (OE4, 8, and 17) and WT plants was analyzed by RT-PCR (Supplementary Table S1). As expected, the great majority of the stress-related genes analyzed, including an endogenous Arabidopsis DUF538 gene, were expressed in the OE lines and WT plants. In fact, expression of only two genes (AtAPX and AtHSFA2) was exclusively detected in trans-

Expression of Stress Related-Marker Genes in Arabidopsis OE Lines
The expression of 24 stress-related genes in non-challenged Arabidopsis OE lines (OE4, 8, and 17) and WT plants was analyzed by RT-PCR (Supplementary Table S1). As expected, the great majority of the stress-related genes analyzed, including an endogenous Arabidopsis DUF538 gene, were expressed in the OE lines and WT plants. In fact, expression of only two genes (AtAPX and AtHSFA2) was exclusively detected in transgenic plants, and one gene (AtSEB2.2) exclusively in WT plants. Together with the OE lines root and leaf morphological analysis ( Figure 5), these results support the assumption that no pleiotropic effects are observed in transgenic AsDUF538 OE lines.
Considering the importance of jasmonic acid (JA) and ethylene (ET) signaling and ROS activation on defense responses against RKN infection, we investigated the expression behavior in the three OE lines of five Arabidopsis marker genes involved with these phytohormone pathways and the antioxidant defense system. This subset comprised genes coding for ascorbate peroxidase (AtAPX); transcription factor EIN3 (AtEIN3), jasmonate ZIM domain (AtJAZ1), allene oxide cyclase (AtAOC2), and helix-loop-helix DNA-binding family protein (AtMYC2) (Supplementary Table S2).
The AsDUF538 overexpression induced the expression of the five marker genes, given that their expression was higher in the OE lines than WT plants ( Figure 6). As expected, the relative transcript abundance of each marker gene varied according to the transgenic line, regardless of its AsDUF538 transcript abundance (Supplementary Figure S1). A remarkable transcriptional induction of AtAPX and AtAOC2 genes was observed due to AsDUF538 overexpression, showing relative expression values above 56-fold for all OE lines ( Figure 6A,B). Likewise, but to a lesser extent, AtJAZ1 was highly upregulated (>23-fold) in all OE lines ( Figure 6C). The overexpression of AsDUF538 also lead to an increase in transcript levels of AtEIN3 (>8-fold) and AtMYC2 (>1.9-fold) ( Figure 6D,E). These findings suggest that AsDUF538 may be involved in the regulation of JA and ET pathways, probably modulating the transcriptional dynamics of genes involved in their biosynthesis (AtAOC2), signaling (AtEIN3 and AtJAZ1), and cross-talk (AtMYC2), besides interplaying with the antioxidant defense system mediated by ROS scavengers (AtAPX).

Arabidopsis OE Lines Infected with Meloidogyne incognita
To evaluate the potential of AsDUF538 overexpression to confer RKN resistance, 10 individuals of each selected OE lines (OE 4; 8and 17) and 10 WT plants were challenged with M. incognita. Nematode infection was confirmed in Arabidopsis WT roots by the presence of typical RKN-susceptible symptoms, such as root galling and egg production. Conversely, the three OE lines were much less affected by M. incognita infection, showing significantly lower numbers of females per gram of roots (average of 617.8) when compared to WT roots (average of 1224.5) (Figure 7).
Overall, all three Arabidopsis OE lines showed a significant reduction in the number of RKN females per root gram at 60 DAI compared to WT plants (p < 0.05; t-test). Whilst OE4 showed an average reduction of 42.4% and OE8 of 45.7%, the OE17 displayed the largest reduction of 50.1%. These results suggested that AsDUF538 overexpression increases the ability of transgenic plants to mitigate damages caused by nematodes.
No visible differences in root and gall morphologies between WT and OE lines were observed at 60 DAI. Likewise, as observed for peanut and soybean hairy roots, roots from Arabidopsis WT plants and OE lines displayed similar biomass, ranging from 237.1 to 197.6 mg of roots per plant, which did not differ significantly (p < 0.01; t-test). These results corroborated our previous observation that AsDUF538 overexpression does not induce pleiotropic effects in transgenic plants.

Discussion
Novel proteins classified as DUF have been identified over the past years, particularly in the last decade, due to the addition of numerous transcriptome and proteome datasets, with more than 3961 proteins described in the Pfam database [47]. Within this group of hypothetical proteins, DUF538 have been identified as stress-responsive proteins in a wide range of mono and dicotyledonous plants, but little is known about their functional properties [28]. To our knowledge, this is the first time that DUF538 is functionally validated in planta and the first report of its overexpression conferring resistance to RKNs.
To date, DUF538-domain proteins have been associated to plant cell walls, thus being involved in cell elongation, division and modification and in defense responses to biotic and abiotic stress [3,17]. Additionally, studies on the structural/functional characterization of C. cristata protein containing a DUF538 domain suggest that it has a lipolytic enzymatic role [17]; contains a hydrolytic domain of the esterase type [27] and methylesterase activity towards pectin molecules [28]). However, despite various studies showing significant changes in DUF538 protein abundance and gene expression modulation during plant development and in response to different stresses [48], there is still a lack of in planta functional validation.
Our previous studies in wild Arachis species showed that AsDUF538 is typically induced in RKN-resistant A. stenosperma during its incompatible interaction with M. arenaria [21][22][23], suggesting a putative role for this gene in RKN defense responses. In the present study, in situ hybridization analysis using AsDUF538 as an RNA-labeled probe showed accumulation of AsDUF538 transcripts in roots of two Arachis species inoculated with M. arenaria, whilst hybridization was lacking in roots and leaves of non-inoculated plants. Therefore, these findings suggest that DUF538 expression in roots of A. steno-

Discussion
Novel proteins classified as DUF have been identified over the past years, particularly in the last decade, due to the addition of numerous transcriptome and proteome datasets, with more than 3961 proteins described in the Pfam database [47]. Within this group of hypothetical proteins, DUF538 have been identified as stress-responsive proteins in a wide range of mono and dicotyledonous plants, but little is known about their functional properties [28]. To our knowledge, this is the first time that DUF538 is functionally validated in planta and the first report of its overexpression conferring resistance to RKNs.
To date, DUF538-domain proteins have been associated to plant cell walls, thus being involved in cell elongation, division and modification and in defense responses to biotic and abiotic stress [3,17]. Additionally, studies on the structural/functional characterization of C. cristata protein containing a DUF538 domain suggest that it has a lipolytic enzymatic role [17]; contains a hydrolytic domain of the esterase type [27] and methylesterase activity towards pectin molecules [28]). However, despite various studies showing significant changes in DUF538 protein abundance and gene expression modulation during plant development and in response to different stresses [48], there is still a lack of in planta functional validation.
Our previous studies in wild Arachis species showed that AsDUF538 is typically induced in RKN-resistant A. stenosperma during its incompatible interaction with M. arenaria [21][22][23], suggesting a putative role for this gene in RKN defense responses. In the present study, in situ hybridization analysis using AsDUF538 as an RNA-labeled probe showed accumulation of AsDUF538 transcripts in roots of two Arachis species inoculated with M. arenaria, whilst hybridization was lacking in roots and leaves of non-inoculated plants. Therefore, these findings suggest that DUF538 expression in roots of A. stenosperma and A. hypogaea may depend on a stress stimulus, such as the presence of the RKN pathogen. This is in accordance with the A. hypogaea developmental transcriptome map [49], which indicated that DUF538 expression in plants showing normal development could only be detected at two different seed developmental stages (lateral stem leaf and reproductive shoot tip), with transcripts undetected in roots and leaves.
We also found that DUF538 transcripts observed in the vascular cylinder cells of the RKN-resistant species A. stenosperma accumulated more and at an earlier stage (3 DAI) than in the roots of the susceptible peanut (6 DAI). This earlier AsDUF538 transcript accumulation after RKN infection in the resistant, in comparison to the susceptible species, suggests that this defense response's timing is crucial, as observed for those involving ROS and hormonal signaling pathways [50]. It is known that signaling by different molecules can occur via xylem (transporter cells), as for those produced during fungi infection [51]; root bacteria symbiosis [52]; environmental stress [53]; abiotic stress [54]; ROS and electrical signals [55]. These transported molecules may play important functions to rapidly prompt initiation of defense responses and signal transduction to activate downstream transcriptional regulation. Therefore, the evident and earlier AsDUF538 expression in xylem cells of the resistant A. stenosperma is another indication of its involvement in plant defense responses against RKN infection.
In this work, we functionally validated AsDUF538 via its overexpression in RKNsusceptible genotypes of three different plant species (peanut, soybean, and Arabidopsis), with each transgenic system showing significantly improved resistance against its pathogen counterpart (M. arenaria, M. javanica, and M. incognita, respectively). This broad type of resistance suggests that AsDUF538 overexpression triggers early, non-specific, defense responses, such as those initiated by PTI (pathogen triggered responses), involving hormone signaling and molecules related to the maintenance of plant redox state in stressed plants, which in turn, regulates not only local but, also systemic plant defense responses (SAR) against pathogens [56].
This seems to be the case here, as the qRT-PCR expression analysis of stress-markers genes in Arabidopsis OE lines showed that an enzyme involved with JA biosynthesis, such as allene oxide cyclase (AtAOC2) [57] and the signaling protein jasmonate-ZIM domain (AtJAZ1) [58] were both strongly induced in the AsDUF538 OE lines. In addition, the transcription factor AtMYC2, which positively regulates JA-mediated resistance genes [59] was upregulated in the OE lines. Likewise, the ET signaling pathway seems to be induced in the OE lines due to a marked upregulation of the ET marker gene AtEIN3 (ethyleneinsensitive) [60,61]. The involvement of ET in this defense response is supported by previous reports showing that the overexpression of ET in Arabidopsis activates plant defense responses against the RKN M. graminicola in rice [62], whilst Arabidopsis plants overexpressing ET are less attractive to M. hapla juveniles [63].
In addition to hormone signaling, we also found that a gene coding for ascorbate peroxidase (AtAPX) was strongly upregulated in the Arabidopsis OE lines, suggesting that AsDUF538 could be linked to maintenance of the redox balance in stressed plants, as previously proposed by [18,64]. Therefore, our findings corroborate studies that suggest a role for DUF538 in a series of physiological and genetic adaptive responses, including balancing the uptake of ions in stress-challenged plants [18,65]. This could be due to its phosphorylation activity working as a potent activator of the redox system of plant cells, which in turn can lead to improved resistance against pathogen attack.

Conclusions
Plant hormones interact in complex networks to balance the response to developmental and environmental hazards to limit defense-associated fitness costs. Our results suggest that the overexpression of AsDUF538 in Arabidopsis leads to diverse defensive states in the plant, including the induction of hormone signaling pathways (JA and ET) and control of excessive ROS accumulation by antioxidant enzymes, which contribute to the improved RKN resistance. However, further integrated molecular, physiological, and agronomical studies are needed to assess the fitness costs associated with increased RKN resistance mediated by AsDUF538 overexpression in target plants. This is the first functional validation of a DUF538 gene and report of a DUF538 family member conferring resistance to Meloidogyne spp. This study also brings new insights into how DUF538 genes can trigger plant defense responses and provides a putative novel candidate gene for genetic engineering towards improved and more sustainable RKN resistant crops.
Supplementary Materials: The following are available online at https://www.mdpi.com/2073-439 5/11/3/559/s1, Figure S1. Relative quantification of AsDUF538 expression in Arabidopsis OE lines. Relative quantification of AsDUF538 expression in 15 Arabidopsis OE lines using AtACT2 reference gene as normalizer. Table S1. RT-PCR analysis. Table S2. Primers used for qRT-PCR analysis. Data Availability Statement: All the data used to support the findings of this study are included within this manuscript.