Co-regulation of the Glycine max soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptor (SNARE)-containing regulon occurs during defense to a root pathogen

ABSTRACT Genes functioning in membrane fusion were originally identified genetically in Saccharomyces cerevisiae and are found in all eukaryotes. Components of the unit, soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptor (SNARE), function in the plant genetic model Arabidopsis thaliana during its defense to shoot pathogens. Regarding defense, little is understood about SNARE in roots or its regulation. Experiments in Glycine max (soybean) have provided an opportunity to perform such studies, revealing that SNARE genes are expressed under natural conditions in root cells undergoing defense to parasitism by the nematode Heterodera glycines. Presented here, the G. max homolog of S. cerevisiae suppressor of sec1 (SSO1), identified genetically in A. thaliana as PENETRATION1 (PEN1) and named in its genomic annotation as syntaxin 121 (SYP121) functions in the resistance of G. max to H. glycines. Genetic experiments demonstrate Gm-SYP121 is co-expressed with homologs of other SNARE genes exhibiting measurable transcript levels in infected cells undergoing resistance. These genes include synaptosomal-associated protein 25, homologous to A. thaliana SNAP33 (SNAP-25/SNAP33/SEC9); mammalian uncoordinated-18 (MUNC18/SEC1); synaptotagmin/tricalbin-3 (SYT/TCB3); synaptobrevin/vesicle associated membrane protein/YKT6/SEC22 (SYB/VAMP/YKT6/SEC22); N-ethylmaleimide-sensitive fusion protein (NSF/SEC18) and alpha-soluble N-ethylmaleimide-sensitive fusion protein associated protein (α-SNAP/SEC17). Experiments show each SNARE component functions in resistance. In contrast, a coatomer zeta/retrieval3 (Cζ/RET3) homolog known to function in retrograde transport within and between the Golgi and endoplasmic reticulum does not appear to function in resistance. Experiments show that SNARE is co-regulated along with a β-glucosidase having homology to PEN2 and an ATP binding cassette transporter exhibiting homology to PEN3.


Introduction
Secretion is a central component of natural physiological processes of all eukaryotic cells (Zhou et al. 2015). The process of secretion examined genetically, beginning with studies in the model organism Saccharomyces cerevisiae (yeast), have resulted in the identification of the Secretion phenotype from which the sec mutant alleles have been determined (Novick et al. 1980(Novick et al. , 1981. The protein products of the SEC genes function in an orderly stepwise manner, mediating membrane fusion (Novick et al. 1980(Novick et al. , 1981. The functional unit responsible for membrane fusion is the soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptor (SNARE) (reviewed in Ferro- Novick & Jahn 1994;Rothman 1994;Rothman & Warren 1994;Sudof 1995;Hanson et al. 1997;Jahn & Fasshauer 2012). The proteins were classified as vesicle (v) (i.e. synaptobrevin/VAMP) or target (t) (i.e. syntaxin and SNAP-25) SNARES (Söllner et al. 1993;Rothman 1994). Because of how the vesicle fusion steps occur and that homotypic fusion happens between vesicles, an attempt was made to reclassify these proteins based on their contribution to fusion (Fasshauer et al. 1998). These proteins were subsequently reclassified based on their contribution to a field called an ionic zero layer (Fasshauer et al. 1998). With few exceptions, in this layer arginines (R) and glutamines (Q) are provided by R-and Q-SNAREs, respectively (Fasshauer et al. 1998). SNARE homologs have been identified in all eukaryotes, functioning in cellular stasis (Clary et al. 1990;Lukowitz et al. 1996;Geelen et al. 2002;Zhou et al. 2015).
Subsequent genetic analyses in A. thaliana have demonstrated the involvement of additional components functioning in defense, including the secreted signal peptide-containing βthioglucoside glucohydrolase gene PENETRATION2 (PEN2) which is part of a large family of β-glycosidases (Lipka et al. 2005;Stein et al. 2006). Plants produce a vast number of secondary compounds known as β-glycosides that are conjugated to various sugar moieties to increase solubility and inactivate the molecule for storage. The conjugated β-glycoside is part of a binary system that requires its cognate β-glycosidase to activate the compound. The presence of a signal peptide is consistent with PEN2 entering the secretion system (Lipka et al. 2005;Stein et al. 2006).
The transport of glycosides to the apoplast is mediated by the eukaryotic ATP-binding cassette (ABC) superfamily of proteins. The roles of ABC transporters in plants are diverse, including pathogen resistance, lead tolerance, resistance to antimicrobials, resistance to auxin-perturbing herbicides, volatile compound production and rhizosphere signaling. The vast majority of ABC transporters are membrane bound and have been divided into eight subfamilies (ABC A-H) (Verrier et al. 2008). In particular, the ABC-G subfamily has undergone extensive diversification in plants. Early work in A. thaliana on the ABC-G subgroup revealed a function in the secretion of cuticular wax (Pighin et al. 2004;Bird et al. 2007). Genetic and molecular analyses have shown that the PM-localized ABC-G type transporter PENETRATION3 (PEN3) resistance protein functions in the export of a toxic glucoside known as a glucosinolate to the fungal penetration site, neutralizing the barley powdery mildew Blumeria graminis f. sp hordei pathogen (Lipka et al. 2005;Stein et al. 2006). Furthermore, the PEN3 protein functions with PEN1 and PEN2 during a race specific defense reaction (Johansson et al. 2014). These studies explain the long-known involvement of a two component system functioning in legume shoots against various herbivores, identified from natural genetic variants (Armstrong et al. 1913;Ware 1925;reviewed in Hughes 1991). From these studies and the genetic analyses involving A. thaliana PEN1, PEN2 and PEN3, a cell biological framework called a regulon has been coined to describe the defense system (Humphry et al. 2010). However, the intricacies and extent of how these genes interact genetically are not well understood. Furthermore, experiments in Oryza sativa (rice) have demonstrated a role for an ABC half transporter playing essential roles in mycorrhizal arbuscule formation in Oryza sativa (rice) (Gutjahr et al. 2012). This observation indicates that ABC-G type transporters function in both symbiotic relationships in the root as well as events that aid in antagonizing plant-pathogen interactions in the shoot. Little information exists for an involvement of these genes in plant resistance to root pathogens except for the identification of a natural variant of α-SNAP functioning in some capacity in the defense of Glycine max (soybean) to its root pathogen the parasitic nematode Heterodera glycines (Matsye et al. 2012). In this pathosystem, the overexpression of the α-SNAP variant is accompanied by elevated transcript levels of syntaxin 31 which resides on the cis face of the Golgi apparatus (Hardwick & Pelham 1992;Lupashin et al. 1997;Bubeck et al. 2008;Matsye et al. 2012;Pant et al. 2014). Therefore, SNARE components function in the defense of G. max to H. glycines parasitism and are co-regulated. However, the extent of this co-regulation has yet to be demonstrated and the functionality of the other SNARE components tested.
In the analysis presented here, an examination of data from published gene expression experiments that have detected the presence of G. max transcripts in H. glycinesparasitized feeding sites known as syncytia undergoing the natural process of resistance in roots have aided in candidate gene selection (Klink et al. 2005(Klink et al. , 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010bMatsye et al. 2011). Functional experiments have examined SNARE, including G. max homologs of the PEN1 gene SYP121 (Gm-SYP121-1). These experiments have been followed by an analysis of MUNC18 (MUNC18-5), SNAP-25 (SNAP-25-3), synaptobrevin (SYB-2), synaptotagmin (SYT-3), N-ethylmaleimide-sensitive fusion protein (NSF-1) and alpha-soluble N-ethylmaleimide-sensitive fusion protein associated protein (α-SNAP-5) that is different from the natural variant described here, previously. The experiments have also examined a gene that is related to SYB, known in A. thaliana as VAMP721 which functions in defense (Gm-VAMP721-2) (Collins et al. 2003;Wang et al. 2007;Klink et al. 2010bKlink et al. , 2011Matsye et al. 2011;Kim et al. 2014). Experiments show a resistance outcome occurs when the relative transcript levels are increased for each of the candidate SNARE genes. In contrast, by decreasing the relative transcript abundance in RNAi lines for the SNARE genes, the defense reaction in the normally H. glycines-resistant G. max [Peking/PI 548402] is impaired. An examination of the relative transcript abundance found in roots engineered to overexpress or undergo RNAi of each of these SNARE genes reveals that the SNARE components appear to be expressed a unit (co-regulation). Thus, the overexpression of these genes appears to recapitulate the natural condition (Klink et al. 2010bMatsye et al. 2011). The extent of the importance of the secretion system during defense to H. glycines parasitism has been examined by identifying the contribution of G. max homologs of the A. thaliana PEN2 and PEN3 genes. Furthermore, the demonstration of callose deposition in the root cells undergoing resistance, a proven indicator of the contribution of a functional secretion system involving homologs of the PEN1, PEN2 and PEN3 genes in defense, is shown (Collins et al. 2003;Lipka et al. 2005;Kwon et al. 2008;Bednarek et al. 2009;Clay et al. 2009;Meyer et al. 2009;Ellinger et al. 2013;Caillaud et al. 2014;Johansson et al. 2014). The importance and usefulness of the functional genomics approach is shown here through these experiments, revealing that while it is likely that other yet to be tested components are also involved in an extensive, co-regulated network, the process exhibits specificity (Somerville & Somerville 1999;Silady et al. 2004

Selection of candidate genes for genetic analyses
In A. thaliana, the PEN1 SNARE protein functions in defense (Supplemental Figure 1). PEN1 functions in concert with PEN2 and PEN3. Presented here, the identified candidate genes are being studied to determine if they perform a role in defense analogous to that observed in A. thaliana ( Figure  1). G. max candidate genes examined here have been selected from published gene expression experiments analyzing the natural defense responses of G. max [Peking/PI 548402] and G. max [PI 88788] (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010bMatsye et al. 2011). The gene is considered expressed in syncytia undergoing defense if the probe set representing the gene measures probe in all six examined arrays (three arrays for G. max [Peking/PI 548402] and G. max [PI 88788] ) at a statistically significant level above background (p < .05) for a given time point (3 or 6 dpi) ( Table 1; Supplemental Table 1) (Klink et al. 2010a(Klink et al. , 2010bMatsye et al. 2011). Expression in control cells did not preclude the genes from consideration since SNARE genes have important functions in normal root cells (Table 1; Supplemental Table 1) (Arpat et al. 2012). In most cases, the gene transcript is detected in the samples collected from cells undergoing the process of defense at both time points in each genotype. In some cases the transcript is detected in the control cells. The results show the candidate genes exhibit expression under natural, unengineered conditions in syncytia that have been induced to form by H. glycines [NL1-Rhg/HG-type 7/ race 3] during defense.

Demonstration of the specificity of the pRAP vector system
The objective of using the complimentary approaches of gene overexpression and RNAi in studying a developmental process is that the combined opposite outcomes, respectfully, are hallmarks of the involvement of the gene in the process (Zhou et al. 2005;Baena-González et al. 2007;Pant et al. 2014;Sun et al. 2014;Dóczi et al. 2015). In the analysis presented here, these opposite outcomes are engineered resistance in the normally H. glycines-susceptible G. max [Williams 82/PI 518671] and engineered impairment of resistance in the normally H. glycines-resistant G. max [Peking/PI 548402] (Pant et al. 2014(Pant et al. , 2015a. A control experiment targeting coatomer zeta (Cζ) has been designed to prove the specificity of the pRAP vector system (Kuge et al. 1993;Klink et al. 2009b;Matsye et al. 2012;Matthews et al. 2013). The G. max genome is represented by three paralogous Cζ genes, Glyma07g03510 (Cζ-1), Glyma08g22580 (Cζ-2) and Glyma15g01150, (Cζ-3). Probe for Gm-Cζ-1 has not been detected at a statistically significant level within syncytia undergoing the resistant reaction while Gm-Cζ-2 and Gm-Cζ-3 probes have (Klink et al. 2010bMatsye et al. 2011) (Table 1; Supplemental Table 1). The control experiments presented here examine Gm-Cζ-2 in more detail. The overexpression or RNAi of Gm-Cζ-2 gene has no obvious observable effect on root growth ( Figure 2). An examination of the relative transcript levels of Gm-Cζ-2 in the overexpression and RNAi lines, respectively, confirm their expected effect of increasing or lessening of its relative transcript abundance ( Figure 2). In other experimental systems, α-tubulin folding cofactor B has been shown to be an effective control for gene expression experiments (Caracausi et al. 2016). The α tubulin cofactor B control (Glyma05g38210) shows no change in relative transcript abundance in comparison to the genetically engineered lines designed to overexpress or suppress Gm-Cζ-2 transcription, respectively ( Figure 2).
In order to quantify the effect that a biological treatment has on H. glycines, an analysis procedure has been performed that takes advantage of its obligate parasitic life cycle in G. max. The measurement output is called a female index (FI; Golden et al. 1970) (please refer to Methods). In these analyses, the pRAP15 control is set to a FI of 100 (100% parasitism) for comparative purposes ( Figure 2). In comparison, the FI shows that the overexpression of Gm-Cζ-2 in G. max [Williams 82/PI 518671] has no statistically significant effect on H. glycines parasitism ( Figure  2; Table 2). In the experiment examining Gm-Cζ-2 overexpression, replicate 1 (n = 10 plants) shows a FI of 91.0 (p = .287578) when examining the parasitism by H. glycines in the whole root as compared to the pRAP15 control roots (n = 12). This FI value means that parasitism has been reduced by 9.0% when measuring the total number of cysts per Gm-Cζ-2 overexpressing root mass. However, the reduction is not statistically significant. In those same roots presented in replicate 1, an analysis of the average amount of cysts per gram of Gm-Cζ-2 overexpressing root tissue has resulted in a calculated FI of 85.3 (p = .28237). This means that when the FI is calculated as a function of cysts per gram of Gm-Cζ-2 overexpressing root tissue that parasitism is reduced by 14.7%. However, the reduction is not statistically significant. The two additional replicates show no statistically significant differences in parasitism occurring between the genetically engineered Gm-Cζ-2 roots and controls ( Figure 2; Table 2). Experiments using pRAP17 to suppress the relative transcript levels of Gm-Cζ-2 in the G. max [Peking/PI 548402] genetic background have been performed, resulting in no statistically significant differences in parasitism occurring between the genetically engineered Gm-Cζ-2 roots and controls ( Figure 2; Table 3). The results presented here demonstrate that it is possible to engineer G. max roots using the pRAP15 or pRAP17 vectors to increase or decrease the relative transcript levels of a gene of interest (GOI), respectively, resulting in no effect on root growth or H. glycines parasitism. The experiments prove the specificity of the pRAP plant transformation system demonstrating the feasibility of using the system to examine other genes having measurable transcript abundance in syncytia undergoing the process of resistance.

G. max SNARE functions in defense in the root
The full length Gm-SYP121-1 has been cloned and engineered into the pRAP15 vector to drive its overexpression in the H. glycines-susceptible G. max [Williams 82/PI 518671] (Supplemental Figure 2). In complimentary studies, Gm-SYP121-1 has been engineered into the pRAP17 RNAi vector to suppress its relative transcript level in the H. glycines-resistant G. max [Peking/PI 548402] (Supplemental Figure 2). Roots engineered to undergo overexpression of Gm-SYP121-1, respectively, appear normal as compared to engineered control plants (Supplemental Figure 3). No discernable effect on growth has been observed in Gm-SYP121 RNAi roots. Gm-SYP121-1-OE and RNAi roots, respectively, have then been infected with H. glycines and examined histologically to examine the effect on parasitism (Supplemental Figure  4). The FI of Gm-SYP121-1 overexpressing roots in G. max [Williams 82/PI 518671] reveals suppressed parasitism ( Figure 3; Table 2). In complimentary studies, Gm-SYP121-1 RNAi lines exhibit an impairment of resistance in G. max [Peking/PI 548402] (Figure 3; Table 3). The analysis procedure then has been used to calculate the FI in the overexpressing and RNAi lines for the other SNARE genes, including MUNC18-5, SNAP-25-3, SYB-2, VAMP721-2, SYT-3, NSF-1 and α-SNAP-5 (Figure 3; Tables 2 and 3). To examine the specificity of the RNAi experiments, quantitative PCR (qPCR) primers targeting three additional Gm-α-SNAP paralogs (Glyma02g42820 [α-SNAP-1], Glyma11g35820 [α-SNAP-3] and Glyma14g05920 [α-SNAP-4]) have been examined in α-SNAP-5 RNAi roots. The results, using two different control genes (S21 and α-tubulin cofactor) demonstrate the specificity of the α-SNAP-5 RNAi experiment by showing the expression of α-SNAP-1, 3 and 4 is not altered (Supplemental Figure 5). Furthermore, expression of the functional control, Gm-Cζ-2, is not altered (Supplemental Figure 5). These expected results contrast with the expected decrease in measured transcript for α-SNAP-5 in the G. max [Peking/PI 548402] α-SNAP-5 RNAi lines (Supplemental Figure 5). The results presented here demonstrate that the overexpression of the candidate membrane fusion gene results in a suppressed capability for H. glycines to parasitize G. max [Williams 82/PI 518671] . In contrast, the results presented here demonstrate that the RNAi of the candidate membrane fusion gene results in an impaired capability of G. max [Peking/PI 548402] to suppress H. glycines parasitism.
A G. max homolog of PEN2 function in defense in the root A. thaliana PEN1 delivers the β-glycosidase PEN2 to the infection site of B. graminis f. sp hordei to activate resistance, demonstrating the importance of delivered cargo to resistance and that SNARE mediates the process (Stein et al. 2006). In the legume Lotus japonicus, a β-glycosidase that exhibits homology to the PEN2 gene is the root-expressed LjBGD7 which belongs to a family of genes. Two L. japonicus LjBGD7 paralogs that have been shown to be expressed in the shoot, LjBGD2 and LjBGD4, exhibit homology to α-hydroxynitrile glucosidase. Experiments have shown α-hydroxynitrile glucosidase functions effectively in defense through their role as part of a biochemical pathway resulting in the biogenesis of hydrogen cyanide (HCN) (Supplemental Figure 5). A G. max homolog related to the root-expressed LjBGD7 is Gm-βg-4 (Gly-ma11g13810), sharing 68.3% amino acid (aa) identity (Supplemental Table 2). Gm-βg-4 transcript has been detected in syncytia undergoing the resistant reaction (Table 1). The homology that Gm-βg-4 has to the secreted L. japonicus LjBGD7 indicates it may function in the defense process.
In L. japonicus, the biochemical pathway leading to the production of HCN begins upstream of α-hydroxynitrile glucosidase (Supplemental Figure 6). An analysis of the G. max genome resulted in the identification of five genes whose conceptually translated protein products share 53.7-66.9% aa identity to LjCYP79D4 (Supplemental Table 2). Of the five G. max protein homologs of LjCYP79D4, Gm-CYP79D4-3 (Glyma13g06880) is most closely related sharing 66.9% aa identity (Supplemental Table 2). An analysis of genes proven to have detectable levels of transcript within syncytia undergoing the defense response has been performed. Except for Gm-CYP79D4-1 Table 1. The genes originally identified by detection call methodology (DCM) and studied here in the functional analyses (Klink et al. 2010b;Matsye et al. 2011).
Note: For the gene to be considered expressed, the probe set for the accompanying gene had to detect probe above threshold in all three arrays in each G.
max genotype (Peking/PI 548402 and PI 88788); p < .05, Wilcoxon's rank test. M, measurable expression (red); N/M no measurable expression (blue); n/a, not applicable (gray). a Please see Supplemental Table 1 for details. The FI for Gm-Cζ-2-OE (note: there is no asterisk above the histogram bar because there is no statistically significant difference between the pRAP15-ccdB control and Gm-Cζ-2-OE lines [ Table 2]); (f) The FD for Gm-Cζ-2-RNAi (note: there is no asterisk above the histogram bar because there is no statistically significant difference between the pRAP17-ccdB control and Gm-Cζ-2-OE lines [ that has a corresponding probe set fabricated on the Affymetrix ® GeneChip, but did not measure detectable levels of transcript, the other G. max CYP79D4 paralogs lack corresponding probe sets (Supplemental Table 1). Therefore, transcript measurements could not be made for the remaining Gm-CYP79D4 paralogs (Klink et al. 2010a(Klink et al. , 2010b. The Gm-βg-4 and Gm-CYP79D4-3 genes closely related to LjBGD7 and LjCYP79D4, respectively, have been cloned and genetically engineered for overexpression in G. max [Williams 82/PI 518671] or RNAi in G. max [Peking PI 548402] . No statistically significant effect has been observed on root growth (Supplemental Figure 7). An examination of Gm-βg-4 and CYP79D4-3 overexpressing roots in G. max [Williams 82/PI 518671] identified suppressed H. glycines parasitism ( Figure 3). In contrast, Gm-βg-4 and Gm-CYP79D4-3 RNAi lines in G. max [Peking/PI 548402] exhibit an increase in H. glycines parasitism ( Figure 3). These observations have been confirmed histologically (Supplemental Figure 8). The results show that homologs of components representing enzymatic steps in the α-hydroxynitrile glucosidase metabolic pathway function effectively in resistance when engineered into the H. glycines susceptible genotype G. max [Williams 82/PI 518671] . In contrast, their RNAi results in an impaired capability of G. max [Peking/PI 548402] to suppress H. glycines parasitism. A qPCR analysis shows the expression of components of the α-hydroxynitrile biosynthetic pathway, including CYP79D4, CYP71, UDP-glucosyltransferase, βg-4, α-hydroxynitrile lyase (AHL) and β-cyanoalanine synthase (BCS) are induced in the Gm-βg-4 and Gm-CYP79D4-3-OE lines (Supplemental Figure 9).
A G. max ABC-type transporter related to PEN3 functions in defense in the root In A. thaliana, the PEN1 and PEN2 genes function in concert genetically with PEN3 to mediate defense against B. graminis f. sp hordei (Stein et al. 2006;Johansson et al. 2014). Examination of the G. max genome shows it contains 35 ABC-G-type transporters. Among them, Gm-ABC-G-26 (Gly-ma17g04360) exhibits detectable levels of transcript in syncytia undergoing the natural process of resistance to H. glycines parasitism in unengineered G. max [Peking/PI 548402] and G. max [PI 88788] (Klink et al. 2010bMatsye et al. 2011) (Table 1, Supplemental Table 1). The Gm-ABC-G-26 cDNA has been cloned and overexpressed in G. max [Williams 82/PI 518671] or engineered as RNAi lines in G. max [Peking/PI 548402] , exhibiting no statistically significant change in root growth (Supplemental Figure 10). These roots then have been examined histologically for H. glycines parasitism (Supplemental Figure 8). Gm-ABC-G-26 overexpression in G. max [Williams 82/PI 518671] roots suppresses H. glycines parasitism ( Figure 3). In contrast, Gm-ABC-G-26 RNAi lines suppress resistance in G. max [Peking/PI 548402] , resulting in increased parasitism by H. glycines (Figure 3). The results presented here show that there are G. max ABC-G type transporters that exhibit detectable levels of transcript abundance within syncytia undergoing the process of resistance. When These observations led to the hypothesis presented here that the G. max SNARE components, including its PEN1 homolog Gm-SYP121-1, glycoside metabolizing genes, including the PEN2 homolog βg-4 and the PEN3 homolog ABC-G-26 may be co-regulated during its process of defense. In the analysis presented here, qPCR has been used to examine cDNA synthesized from RNA isolated from the overexpressing lines at 0 and 6 dpi and the RNAi lines at 0 dpi. At 0 and 6 dpi, the overexpressing lines are accompanied by an increase in relative transcript levels of the remaining genes examined in this study ( Figure 4). The effect is specific since the relative transcript abundances of α-tubulin folding cofactor B and Cζ-2 control genes are not affected ( Figure  4). As expected, RNAi of the target gene is accompanied by a decrease in relative transcript abundance of the remaining genes examined in this study while the relative transcript abundances of the control genes are not affected (Figure 4).

Callose is present at the site of defense
Genetic experiments in A. thaliana show that the PEN2 and PEN3 genes mediate part of the defense response by being required for callose deposition (Clay et al. 2009). Callose is a cell wall polymer composed of 1,3-β-glucan subunits that functions in defense (Ellinger et al. 2013). Related genetic experiments in A. thaliana show the process of defense is complex, possibly involving additional syntaxins other than the PEN1 protein and/or the SYB homologs VAMP721 and VAMP722, but still resulting in callose deposition Meyer et al. 2009;Nielsen et al. 2012;Caillaud et al. 2014). The complex nature of the action of the regulon has been examined in genetic experiments in A. thaliana, showing that the PEN1, PEN2 and PEN3 genes function in race-specific resistance (Johansson et al. 2014 Figure 5). In contrast, aniline blue label is less evident in syncytia that have undergone the localized susceptible response in Gm-SYP121-1-RNAi engineered G. max [Peking/PI 548402] roots ( Figure 5). These roots appear similar to unengineered, susceptible G. max [Williams 82/PI 518671] roots that have undergone the susceptible reaction to H. glycines parasitism ( Figure 5). Controls are provided ( Figure 5). In the experiments presented here, localized callose accumulation is demonstrated at syncytia undergoing their natural resistant reaction in unengineered G. max [Peking/PI 548402] roots. From the Gm-SYP121-OE experiments, it appears that the engineered defense response is accompanied by callose accumulation. However, the results do not prove callose is responsible for resistance. A more detailed analysis is required to understand its presence at defense sites.

Discussion
Prior experiments have demonstrated the functioning of the G. max syntaxin 31 (Gm-SYP38), which is homologous to the S. cerevisiae suppressors of the erd2-deletion 5 protein (Sed5p), in the root during its resistance to H. glycines parasitism (Hardwick & Pelham 1992;Sanderfoot et al. 2000;Pant et al. 2014Pant et al. , 2015a. The results of those experiments have led to the development of a model predicting the involvement of other SNARE genes including the G. max homolog of PEN1   Pant et al. 2014). However, a functional test of Gm-SYP121-1 had not been presented. The experiments presented here have expanded that model of defense, reinforced in functional analyses of G. max homologs of the PEN1 and other SNARE components as well as homologs of PEN2 and PEN3. These functional analyses have been followed by the demonstration of co-regulation of the G. max SNARE genes and homologs of PEN2 and PEN3 during the defense process. Callose has also been observed at the sites of defense.

SNARE functions in defense in the G. max root
The experiments presented here have focused in on analyzing SNARE, employing gene overexpression and RNAi to examine its relationship to the G. max-H. glycines root pathosystem. The specificity of the plant transformation platform used here has been reported elsewhere, used in large scale genetic screens to study plant-pathogen interactions (Matthews et al. , 2014. We have demonstrated further the specificity of the experimental procedure by examining Gm-Cζ-2. Cζ has been first isolated from bovine (Bos taurus) and is related to the S. cerevisiae YCZ1 and Ret3p (Kuge et al. 1993;Cosson et al. 1996;Yamazaki et al. 1997). Cζ is part of a 600 kD heptameric coat protein complex I (COPI) that is involved in many cellular processes, functioning during retrograde trafficking between the Golgi and endoplasmic reticulum (ER), the maturation of endosomes and autophagy (Kuge et al. 1993;Cosson et al. 1996;Razi et al. 2009;Beck et al. 2009). The Gm-Cζ gene family is composed of three members (Gm-Cζ-1-3) with Cζ-2 and Cζ-3 having measurable transcript levels in syncytia undergoing defense (Klink et al. 2010bMatsye et al. 2011). In a control experiment examining Gm-Cζ-2, overexpression and   parasitism. The lack of an observable effect on H. glycines parasitism found here in experiments targeting Gm-Cζ-2 may be due to the remaining gene family members functioning redundantly. Redundancy for Cζ occurs in other biological systems (Wegmann et al. 2004;Moelleken et al. 2007;Shtutman et al. 2011). These results indicate the effect observed in the experiments presented here reflect the actual role that the tested genes perform in defense. The experiments then have focused in on understanding SNARE. The results presented here showing the involvement of Gm-SYP121-1 in G. max defense to H. glycines parasitism corroborates earlier experiments that demonstrated Gm-SYP121-1 exhibits detectable levels of transcript in syncytia in unengineered roots undergoing their natural process of resistance (Klink et al. 2007(Klink et al. , 2010bMatsye et al. 2011). The functional experiments presented here demonstrate that Gm-SYP121-1 acts in resistance, indicating that part of the defense process in the G. max-H. glycines pathosystem employs some of the same components that function in A. thaliana shoots (Collins et al. 2003). This observation is consistent with the identification of G. max homologs of A. thaliana defense genes functioning in its resistance to H. glycines Pant et al. 2014Pant et al. , 2015a. It is also likely that different syntaxins are involved in resistance as has been revealed in Nicotiana benthamiana showing the involvement of its homolog of the A. thaliana SYP132 in the secretion of the defense protein PATHOGENESIS RELATED 1 (PR-1) and other apoplastic proteins (Kalde et al. 2007). Furthermore, NbSYP132 has been shown to be involved in basal and salicylic acid (SA)associated defense (Kalde et al. 2007). This observation is in agreement with our results showing the involvement of the SA signaling proteins ENHANCED DISEASE SUSCEPTI-BILITY1 (EDS1) and NONEXPRESSOR OF PR1 (NPR1) and the expression of PR1 gene during defense in the G. max-H. glycines pathosystem (Cao et al. 1994;Falk et al. 1999;Matsye et al. 2012;Pant et al. 2014). These observations support diverse roles for plant syntaxins (Sanderfoot et al. 2000;Shirakawa et al. 2010). Furthermore, it is likely that other regulatory components of this apparatus function in defense as has been shown for the ADP ribosylation factor-GTP exchange factor, GNOM (Nielsen et al. 2012). GNOM delivers SYP121 and callose to the PM during resistance to B. graminis f.sp. hordei (Nielsen et al. 2012).
These experiments have been followed by the examination of other SNARE components including MUNC18-5, SNAP-25-3, SYB-2, SYT-3, NSF-1, α-SNAP-5 and VAMP721-2 showing they function in resistance. Specificity is demonstrated in the control experiments whereby G. max [Williams 82/ PI 518671] engineered with the pRAP15-ccdB overexpression cassette and G. max [Peking/PI 548402] engineered with the pRAP17-ccdB RNAi cassette exhibit levels of infection that are comparable to unengineered control plants (Klink et al. 2009a,b;Matsye et al. 2012;Matthews et al. 2013Matthews et al. , 2014Pant et al. 2014Pant et al. , 2015a. In A. thaliana, PEN1 protein functions in the shoot in one pathway leading to resistance by forming a complex on the PM with VAMP721/VAMP722 and SNAP33 and mediating the secretion of PR1 to the apoplast (Collins et al. 2003;Assaad et al. 2004;Kalde et al. 2007;Kwon et al. 2008;Pajonk et al. 2008;Kim et al. 2014). Those results clearly show the A. thaliana SNARE components function in secretion in the shoot during resistance. The A. thaliana SNAP33 protein is homologous to the Gm-SNAP-25-3 presented here. For comparative purposes, we have included in the analysis presented here an examination of a G. max SYB homolog of the A. thaliana VAMP721/VAMP722 gene (Gm-VAMP721-2). The results show Gm-VAMP721-2 plays a role in resistance of G. max to H. glycines parasitism. In A. thaliana, VAMP721 co-immunoprecipitates with PLAS-MODESMATA-LOCATED PROTEIN 1 (PDLP1) and regulates callose deposition at developing encasements at Hyaloperonospora arabidopsidis infection sites during defense (Caillaud et al. 2014). In A. thaliana VAMP721 protein also plays an important role in the delivery of the resistance (R) protein RESISTANCE TO POWDERY MILDEW8 (RPW8) paralog, RPW8.2, to the extrahaustorial membrane of Golovinomyces orontii (Kim et al. 2014). The RPW8.2 and VAMP721 proteins function along with PEN1 and SNAP33 during infection by G. orontii to accomplish defense (Kim et al. 2014). Therefore, as presented by Kim et al. (2014), vesicles deliver R proteins to the site of defense and this fusion of vesicle and PMs is mediated by SNARE. In this regard, the experiments presented here help in explaining our prior observations of the involvement of the membrane-bound G. max homolog of the A. thaliana BOTRYTIS INDUCED KINASE1 (BIK1) functioning in resistance (Veronese et al. 2006;Pant et al. 2014). In A. thaliana, BIK1 is a PM-tethered receptorlike cytoplasmic kinase that becomes activated by phosphorylation stimulated by bacterial flagellin (flg22) peptide (Veronese et al. 2006;Lu et al. 2010;Zhang et al. 2010). Flg22 activates FLAGELLIN SENSING PROTEIN2 (FLS2) protein and transphosphorylation of BRASSINOSTEROID ACTIVATED KINASE1 (BAK1) which then phosphorylates BIK1 to induce downstream signaling events (Chinchilla et al. 2007). In A. thaliana, the FLS2 pathway activates defense processes including, but not limited to, SA signaling and callose deposition (Boller & Felix 2009). In A. thaliana, the RPW8.2 gene has been identified along with RPW8.1 functioning to confer broad-spectrum resistance to diverse species of powdery mildew fungi (Xiao et al. 2001;Wang et al. 2007). The protein products of the RPW8.1 and RPW8.2 R genes transduce their signal through the SA signaling pathway by activating EDS1 (Falk et al. 1999;Xiao et al. 2001Xiao et al. , 2003. As stated, the G. max homologs of EDS1 and NPR1 genes have already been shown to function effectively during resistance to H. glycines parasitism (Cao et al. 1994;Matsye et al. 2012;Pant et al. 2014Pant et al. , 2015a. The activation of these signaling pathways is consistent with the observation of transcripts for hundreds of genes becoming increased in their relative abundance in syncytia undergoing the process of defense (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010bMatsye et al. 2011). Furthermore, in A. thaliana the transcription of the callose synthase AtGsl5 is induced by SA in wild type plants (Ostergaard et al. 2002). This observation is important from the standpoint that in A. thaliana, complete resistance to G. cichoracearum and B. graminis f. sp. hordei is mediated by the callose synthase gene PMR4 (GSL5) although this response was not SA-dependent (Ellinger et al. 2013). Therefore, the cellular machinery that facilitates the defense of A. thaliana against multiple shoot pathogens also appears to function at least in part in the defense of the G. max root under parasitism by H. glycines. The experiments presented here have also examined the relative changes in transcript abundance of SNARE, demonstrating that the genes appear to be co-regulated. The co-regulation of different vesicle components observed here in this system has been seen in other organisms, some of them non-plant systems, and functional genomics screens have revealed this co-regulation can be quite extensive (Shanks et al. 2012;Liberali et al. 2014;Pant et al. 2014Pant et al. , 2015aZick et al. 2015). However, very little published data is available in plants.
A homolog of PEN2 functions in defense in the G. max root The involvement of SNARE in the root during the resistance of G. max to H. glycines parasitism has led to the hypothesis that homologs of A. thaliana PEN2 gene are involved in the process since it has been demonstrated in A. thaliana that the PEN2 protein functions during an inducible pre-invasion resistance process (Lipka et al. 2005;Stein et al. 2006;Clay et al. 2009). In A. thaliana, the PEN2 genetic pathway functions in the extracellular deposition of callose, working in concert with PEN3 gene (Collins et al. 2003;Lipka et al. 2005;Kwon et al. 2008;Bednarek et al. 2009;Clay et al. 2009;Johansson et al. 2014). In contrast, a protein functioning very effectively in defense in the legume L. japonicus is a PEN2 homolog belonging to a family of glucosidases known as α-hydroxynitrile glucosidase (Morant et al. 2008;Takos et al. 2011). Bioinformatics analyses presented here show that the conceptually translated Gm-βg-4 is most closely related to the root-specific L. japonicus α-hydroxynitrile glucosidase LjBGD7, belonging to a small family of enzymes involved in the production of cyanogenic α-hydroxynitrile glucosides (Morant et al. 2008;Takos et al. 2011). While Gm-βg-4 likely functions differently than the PEN2 gene in A. thaliana, overexpression and RNAi experiments show Gm-βg-4 functions in the G. max root during defense. In L. japonicus, the production of cyanogenic α-hydroxynitrile glucosides involves CYP79, CYP71, UDP-glucosyl transferase, α-hydroxynitrile glucosidase and AHL with cyanide detoxification occurring through the activity of BCS (Gleadow & Moller 2014). Except for Gm-CYP79D4 where four of its five paralogs lack the fabrication of corresponding probe sets on the Affymetrix ® soybean GeneChip ® , each of these genes in this pathway exhibit measurable levels of transcript syncytia undergoing the process of resistance (Klink et al. 2010bMatsye et al. 2011). The experiments are further supported by overexpression and RNAi of CYP79D4-3, an enzyme which has been shown in other systems to function at the initial conversion of amino acids to oximes (Gleadow & Moller 2014). The production of the αhydroxynitrile glucosides is accomplished by specific cytochrome P450 enzymes including CYP79D3 and CYP79D4, respectively (Forslund et al. 2004;Bjarnholt et al. 2008). Morant et al. (2008) has demonstrated increased relative levels of expression of LjCYP79D3 in aerial parts of L. japonicus plants which is also where LjBGD2 and LjBGD4 are expressed. In contrast, LjCYP79D4 has been shown to have increased relative levels of expression exclusively in the roots where LjBGD7 occurs (Forslund et al. 2004). The results presented by Morant et al. (2008) have demonstrated the co-expression of α-hydroxynitrile glucoside and their cognate hydrolyzing α-hydroxynitrile glucosidase. We have presented a similar observation here for Gm-βg-4 and CYP79D4-3. Furthermore, in L. japonicus, the heterologous expression of a Manihot esculenta (cassava) CYP79D2 driven by the cauliflower mosaic virus 35S promoter results in the accumulation of cyanogenic α-hydroxynitrile glucosides (Forslund et al. 2004). Our experiments expand on that work by demonstrating that Gm-βg-4 and Gm-CYP79D4 overexpression is accompanied by the co-regulation of the rest of the enzymes composing the biochemical pathway leading to the production and detoxification of HCN. From the presented gene expression experiments of the syncytium, it is likely that other β-glucosidases and biochemical pathways requiring their activity are involved in defense and function in parallel (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010bMatsye et al. 2011).
A PEN3 homolog functions in defense in the G. max root The involvement of G. max homologs of PEN1 and PEN2 genes implicate the involvement of a G. max homolog of the A. thaliana PEN3 functioning in resistance to H. glycines. Genetic experiments in A. thaliana have shown this to be true for race-specific defense processes occurring in the shoot (Johansson et al. 2014). One of the functions of PEN3 in defense is to export toxins to the penetration site to neutralize B. graminis f. sp hordei (Stein et al. 2006;Clay et al. 2009;Meyer et al. 2009). Therefore, the hypothesis that a G. max homolog of the PEN3 gene functions in defense to H. glycines parasitism as presented here has merit. The G. max genome has 35 ABC-G transporters and some exhibit detectable levels of transcript abundance in syncytia undergoing the process of resistance (Klink et al. 2010bMatsye et al. 2011). Through overexpression and RNAi experiments, the G. max PEN3 homolog Gm-ABC-G-26 is shown to function in its root during resistance to H. glycines parasitism. The results presented here establish the involvement of full ABC-G type transporters functioning in defense in the root.

The regulation of the regulon
Based on ecological genetic variants and how PEN1, PEN2 and PEN3 genes function in A. thaliana, the cellular apparatus acting in resistance is described as a binary system composed of two parallel pathways called a regulon that converge on defense (Humphry et al. 2010;Johansson et al. 2014). In this manner, the defense apparatus identified here that acts during G. max resistance to H. glycines functions like the regulon described for A. thaliana and the ecological variants identified in other plant systems over a century ago ( Figure  1) (Armstrong et al. 1913;Ware 1925;reviewed in Hughes 1991;Humphry et al. 2010;Johansson et al. 2014). The experiments presented here provide context to the observation of the functionality of a number of membrane bound and secreted proteins, SA signaling and transcription factors in defense in the G. max-H. glycines pathosystem and relating their activities to callose biosynthesis which can indicate defense (Mangin 1895;Matsye et al. 2012;Ellinger et al. 2013;Matthews et al. 2013Matthews et al. , 2014Youssef et al. 2013;Caillaud et al. 2014;Johansson et al. 2014;Pant et al. 2014Pant et al. , 2015a. Through these experiments it is shown that it is possible to recapitulate at least part of the defense response found naturally in G. max that is utilized as it defends itself from H. glycines parasitism. Continued work, in particular phylogenetic studies aimed at to better understand the evolutionary relationships of these proteins and conservation of specific residues will certainly aid in our understanding of their function under normal metabolic conditions and as they are recruited for defense (Collins et al. 2003;Dacks et al. 2008).

Selection of candidate genes
The selection of candidate genes has been aided by mining data from published gene expression experiments (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010bMatsye et al. 2011). This procedure is an effective means to identify genes that function in G. max defense to H. glycines parasitism, proven further in independently performed genetic mutational analyses (Liu et al. 2012;Matsye et al. 2012;Matthews et al. 2013Matthews et al. , 2014Pant et al. 2014Pant et al. , 2015a. To summarize those published experimental procedures, G. max [Peking/PI 548402] and G. max [PI 88788] have been infected with H. glycines [NL1-Rhg/ HG-type 7/race 3] , resulting in a resistant reaction as proven histologically in unengineered roots which is the natural resistance response found in these G. max genotypes (Ross 1958;Endo 1965Endo , 1991Klink et al. 2007Klink et al. , 2009aKlink et al. , 2010aKlink et al. , 2010bKlink et al. , 2011. Roots have then been processed for histology and laser microdissection (LM), a procedure that has been used to collect syncytia undergoing the defense response (Klink et al. 2005(Klink et al. , 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010b. The mRNA has been isolated from the syncytia and converted to probe for hybridization onto the Affymetrix ® Soybean GeneChip ® (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010bMatsye et al. 2011). The hybridizations have been run in triplicate (arrays 1-3) using probe derived from RNA isolated from LM-collected syncytia obtained from three independent replicate experiments each run independently in the two different H. glycines-resistant genotypes (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010b. For the gene to be considered expressed at a given time point (3 or 6 days post-infection [dpi]), probe signal had to be measurable above threshold on all three arrays for both G. max [Peking/PI 548402] and G. max [PI 88788] (6 total arrays), p < .05 (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010b. The original analysis procedure has been performed as follows; the measurement for a particular probe set (gene) transcript on a single array has been determined using the Bioconductor implementation of the standard Affymetrix ® detection call methodology (DCM) (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010b. DCM consists of four steps, including (1) removal of saturated probes, (2) calculation of discrimination scores, (3) p-value calculation using the Wilcoxon's rank test, and (4) (Klink et al. 2007(Klink et al. , 2009a(Klink et al. , 2010a(Klink et al. , 2010bMatsye et al. 2011). The mined data used in the analysis is presented (Supplemental Table 1). From these data, genes used in the analysis have been selected for functional experiments and/or qPCR.

Gene cloning
G. max root mRNA has been isolated according to Matsye et al. (2012) using the UltraClean ® Plant RNA Isolation Kit according to the manufacturer's instructions (Mo Bio Laboratories ® , Inc.; Carlsbad, CA). Genomic DNA has been removed from the mRNA with DNase I according to the manufacturer's instructions (Invitrogen ® , Carlsbad, CA). The cDNA has been synthesized from mRNA using the SuperScript First Strand Synthesis System for RT-PCR (Invitrogen ® ) with oligo d(T) as the primer (Invitrogen ® ) according to the manufacturer's instructions. The accession numbers and DNA primer sequences for the genes examined in the study presented here are provided (Supplemental Table 3). Genomic DNA contamination has been assessed by PCR by using β-conglycinin primer pair that amplifies DNA across an intron, thus yielding different sized products based on the presence or absence of that intron (Klink et al. 2009b).

G. max genetic transformation
The pRAP plant transformation system used here has been designed and tested specifically for studying the interaction between G. max and H. glycines (Klink et al. 2008(Klink et al. , 2009bMatsye et al. 2012;Matthews et al. 2013Matthews et al. , 2014. The pRAP plant transformation system has been proven independently in other labs to obtain the same outcomes (resistance to H. glycines parasitism) as genetic mutational analyses and virus-induced gene silencing (VIGS) (Liu et al. 2012;Matthews et al. 2013). The pRAP vector system that has been proven to function in G. max is based off of the published Gateway ® cloning vector platform that has been developed and proven to work in other plant systems (Curtis & Grossniklaus 2003;Matsye et al. 2012;Matthews et al. 2013). The published pRAP vector platform uses an enhanced green fluorescent protein (eGFP) transgenic reporter system (Haseloff et al. 1997). The pRAP vector platform, depending on the integrated cassette, is used to activate or suppress the transcription of a targeted gene (Jefferson et al. 1987;Fire et al. 1998;Collier et al. 2005;Klink et al. 2009b;Matsye et al. 2012;Matthews et al. 2013Matthews et al. , 2014Pant et al. 2014Pant et al. , 2015a. The expression of the gene cassettes is driven by the figwort mosaic virus subgenomic transcript promoter (FMV-sgt) promoter (Bhattacharyya et al. 2002). The FMV-sgt promoter has been proven to drive gene expression in transgenic G. max roots throughout the life cycle of H. glycines (Klink et al. 2008). The activation of transcription of a targeted gene is accomplished using the pRAP15 vector which has been designed for and has been proven to result in an increase in the relative transcript levels of the GOI (Matsye et al. 2012;Matthews et al. 2013Matthews et al. , 2014Youssef et al. 2013;Pant et al. 2014Pant et al. , 2015aPant et al. , 2015b. The pRAP17 vector has been designed for and proven to result in a decrease in the relative transcript levels of the GOI (Klink et al. 2009b;Pant et al. 2014Pant et al. , 2015a. Between the left and right border of the pRAP15 and pRAP17 vectors exists the attR homologous recombination sites of the Gateway ® system (Invitrogen ® ) where the GOI integrates (Klink et al. 2009b;Matsye et al. 2012;Pant et al. 2015b). Thus, roots exhibiting the expression of the eGFP visual reporter will also possess the GOI, each with their own promoter and terminator sequences (Collier et al. 2005;Klink et al. 2009b;Matsye et al. 2012;Matthews et al. 2013;Pant et al. 2015b).
The amplicons representing the GOI have been cloned from G. max [Williams 82/PI 518671] and ligated into the directional pENTR/D-TOPO ® Gateway ® -compatible vector (Invitrogen ® ) according to the manufacturer's instructions. The reaction contents then have been transformed into chemically competent E. coli strain One Shot TOP10 ® and selected on kanamycin (50 μg/ml) according to the manufacturer's instructions (Invitrogen ® ). Gene sequences have been confirmed by matching them to the G. max [Williams 82/PI 518671] genome accession (Schmutz et al. 2010). Amplicons representing full length genes have been cloned into the pRAP15 overexpression vector (Matsye et al. 2012;Pant et al. 2015b). Alternatively, full length genes or subcloned portions of genes have been engineered into the pRAP17 RNAi vector (Klink et al. 2009b). This approach has been proven effective for RNAi studies in plants (Klink & Wolniak 2001). In the overexpression studies, the amplicons have been ligated into the pRAP15 destination vector using LR Clonase ® (Invitrogen ® ) according to the manufacturer's instructions (Matsye et al. 2012). The pRAP15-ccdB control and engineered pRAP15 vector containing the GOI have been used to transform chemically competent Agrobacterium rhizogenes K599 (K599) (Hofgen & Willmitzer 1988;Haas et al. 1995;Collier et al. 2005). The transformation mix then has been plated on LB-agar, selecting with tetracycline (5 μg/ml) according to Matsye et al. (2012). A PCR reaction using pRAP15 primers that amplify the 717 bp eGFP gene and the 690 bp A. rhizogenes root inducing (Ri) plasmid (EU186381) VirG gene (VirG) have confirmed that the K599 contains both plasmids prior to transformation. The pRAP15 vector containing the GOI then has been confirmed by PCR using primers for the respective genes and DNA sequencing. Genetic transformation experiments resulting in gene overexpression in G. max roots have been performed according to Matsye et al. (2012) in H. glycines-susceptible genetic background of G. max [Williams 82/PI 518671] (Concibido et al. 2004;Schmutz et al. 2010). Genetic transformation experiments designed to decrease the level of target gene mRNA have been performed according to Klink et al. (2009b). This procedure uses the pRAP17 RNAi vector in the functionally H. glycines-resistant genetic background of G. max [Peking/PI 548402] (Concibido et al. 2004). The procedure for making genetically engineered plants that are used in overexpression or RNAi experiments involves the co-cultivation of 7-9-day-old G. max [Williams 82/PI 518671] (overexpression experiments) or G. max [Peking/PI 548402] (RNAi experiments) with the K599 engineered to harbor the appropriate genetic construct. The roots of these plants are excised while the cut plants are immersed in Murashige and Skoog (MS) media containing the K599 harboring the engineered pRAP15-ccdB or pRAP17-ccdB controls while at the same time different plants are cut and transformed with K599 harboring the engineered pRAP15-GOI or pRAP17-GOI experimental constructs (Murashige & Skoog 1962;Klink et al. 2009b;Matsye et al. 2012;Pant et al. 2014). Due to the way K599 transfers the DNA cassettes situated between the left and right borders of the plasmid into the root cell chromosomal DNA, the subsequent growth and development of the stably transformed genetically engineered cell into a transgenic root results in the production of a plant that is a genetic mosaic called a composite plant (Collier et al. 2005). These composite, genetically mosaic plants have the entire shoot being non-transgenic and the entire root being transgenic (Haas et al. 1995;Collier et al. 2005;Klink et al. 2008Klink et al. , 2009bMatsye et al. 2012;Matthews et al. 2013;Pant et al. 2014). In these studies, therefore, each individual transgenic root system functions as an independent transformant line (Tepfer 1984;Matsye et al. 2012;Matthews et al. 2013;Pant et al. 2014Pant et al. , 2015a. qPCR has been used to confirm the relative levels of transcript abundance in the pRAP15-GOI engineered overexpressing lines or the pRAP17-GOI-engineered RNAi lines.

Quantitative PCR
The DNA sequences for the qPCR primers used in quantitative gene expression experiments are provided (Supplemental Table 3). The experiments involving G. max have used three different control genes for monitoring the relative levels of transcript abundance, (1) ribosomal protein gene S21 (S21), (2) α-tubulin folding cofactor B and (3) coatomer zeta (Cζ). The Gm-S21 gene has been tested and used as a control in prior studies (Klink et al. 2005;Matsye et al. 2012;Pant et al. 2014Pant et al. , 2015a. S21 is a highly conserved gene proven to be transcribed into mRNA and translated into protein (Morita-Yamamuro et al. 2004). With regard to assessing the relative abundance in transcript levels in qPCR experiments, prior qPCR analyses have shown that the Gm-S21 control performs in the same manner as elongation initiation factor protein 3 (Matsye et al. 2012). Therefore, Gm-S21 has been selected to serve as the control for the qPCR experiments presented here. Added gene expression controls have been performed using the G. max α-tubulin folding cofactor B, selected because in other biological systems it has been determined in genomics analyses to be an effective control gene (Caracausi et al. 2016). The α-tubulin folding cofactor B gene is transcribed and translated, but functions in the cytosol by direct protein-protein interaction during α-tubulin stasis (Radcliffe & Toda 2000;Dhonukshe et al. 2006a,b). A third control gene proven to be transcribed and translated into protein, that has also been used in functional transgenic experiments presented here, is Cζ of which there are three in the genome of G. max (Kuge et al. 1993). Cζ acts in retrograde transport, functioning in retrieval between the Golgi and ER (Kuge et al. 1993;Cosson et al. 1996;Yamazaki et al. 1997).

FI analysis
The overexpression and RNAi experiments presented here for all of the studied genes each have three independent biological replicates, respectively. In every experiment, each biological replicate has multiple experimental replicates represented by 5-20 individual plants. The communityaccepted assay that is used to determine if an experimental condition exerts an influence on H. glycines development (parasitism) is calculated and presented as the FI (Golden et al. 1970). The FI is calculated as FI = (Nx/Ns) × 100, where Nx is the average number of females on the test cultivar and Ns is the average number of females on the standard susceptible cultivar (Golden et al. 1970;Riggs & Schmitt 1988Niblack et al. 2002;Klink et al. 2009a, b;Matthews et al. 2013). Various levels of resistance have been designated, based on the calculated FI (Niblack et al. 2009). These levels include 0-9, highly resistant; 10-24, resistant; 25-39, moderately resistant; 40-60, low resistant; >60, not effectively resistant. The description presented here uses the general term resistant for FI of 0-59 with the data presented. In the experiments of Golden et al. (1970), Schmidtt (1988, 1991), Kim et al. (1998) and Niblack et al. (2002), the labs that originally developed and modified the FI, the FI is calculated from a total of 3-10 experimental and 3-10 control plants. In those studies, each individual plant serves as a replicate and biological replicates may or may not be performed (Golden et al. 1970;Riggs & Schmidtt 1988Kim et al. 1998;Niblack et al. 2002). All of the experiments presented here at least meet and in most cases exceed these published standards (Golden et al. 1970;Riggs & Schmidtt 1988Kim et al. 1998;Niblack et al. 2002). The FI assay is also the community-accepted standard analysis method used in experiments in other labs employing genetically engineered constructs in G. max, including those using K599, to examine H. glycines biology (Mazarei et al. 2007;McLean et al. 2007;Steeves et al. 2006;Li et al. 2010;Melito et al. 2010;Liu et al. 2011Liu et al. , 2012Matthews et al. 2013Matthews et al. , 2014. Following the published methods employed in those studies, in the analysis presented here Nx is the pRAP15-GOI or pRAP17-GOI-transformed line and Ns is the pRAP15-ccdB or pRAP17-ccdB control. Because the pRAP15 or pRAP17 control has the ccdB gene located in the position where, otherwise, the GOI is inserted during the LR clonase reaction, those control vectors also control for non-specific effects caused by gene overexpression or RNAi (Klink et al. 2009b;Matsye et al. 2012;Matthews et al. 2013;Pant et al. 2014Pant et al. , 2015b. Therefore, by definition, the pRAP15-ccdB or pRAP17-ccdB transformed plants serve as a control. In the analysis performed here, the FI has been calculated and presented as a function of the cysts per mass of the whole root (wr) and also cysts per gram (pg) of root. The cyst per gram analysis was done to account for any possible root growth effect that may result by the overexpression or RNAi of a GOI. The experiments have been analyzed statistically using the Mann-Whitney-Wilcoxon (MWW) Rank-Sum Test, p < .05 cutoff (Matsye et al. 2012;Pant et al. 2014). Following community-accepted, standard published methods, error bars are not calculated when using the FI analysis (Golden et al. 1970;Riggs & Schmidtt 1988Kim et al. 1998;Niblack et al. 2002). The effect that the overexpressed gene exerts on root growth has been taken from a representative experiment and determined as a function of root mass tested statistically using the MWW Rank-Sum Test, p < .05 cutoff (Matsye et al. 2012;Pant et al. 2014).

Histology
Histological observation has been performed according to Klink et al. (2005). Tissue has been fixed in Farmer's solution (FS) composed of 75% ethanol (ETOH), 25% acetic acid (Klink et al. 2005). G. max root tissue has been cut into 0.5 cm pieces and vacuum infiltrated with FS for 1 h at 4°C. Fresh FS fixative has then been added to their respective samples and subjected to an incubation step of 12 h at 4°C. Dehydration of FS-fixed tissue proceeded through a graded ETOH series (75%, 85%, 100%, 100%, 100%), 30 min each. ETOH has been replaced with 1:1 Hemo DE ® (Scientific Safety Solvents; Keller, TX): ETOH for 30 min. Subsequently, three, 100% Hemo DE ® incubations (30 min each) have been performed. The specimens in Hemo DE ® have been moved from 4°C and placed into a 58°C oven. Hemo DE ® has then been replaced by paraffin. The roots have been infiltrated sequentially in 3:1, 1:1, 1:3 Hemo DE ® :Paraplast+ ® tissue embedding medium (Tyco Healthcare Group LP ® ; Mansfield, MA) in each step for three h. Three changes of 100% Para-plast+ ® in each step for three h followed. Tissue has then been cast and subsequently mounted for sectioning. Serial sections of roots have been made on an American Optical 820 ® microtome (American Optical Co ® .; Buffalo, NY) at a section thickness of 10 μm. Sections have been stained in Safranin O (Fisher Scientific Co.; Fair Lawn, NJ) in 50% ETOH and counter-stained in Fast Green FCF (Fisher Scientific Co.) (Klink et al. 2005). For histological analyses, the microscopic sections (10 μm thick) have been permanently mounted in Permount ® (Fisher Scientific Co.) and visualized.

Callose labeling
Callose labeling has been accomplished by staining with aniline blue using roots from the three independent replicates described previously (Eschrich & Currier 1964;Yim & Bradford 1998;Levy et al. 2007;Schuette et al. 2009;Ellinger et al. 2013). Serial sections of roots (10 μm) have been made and prepared according to Klink et al. (2005) for staining with aniline blue (Sigma-Aldrich Co., St. Louis, MO). Sections have been stained using 0.05% aniline blue in 67 mM phosphate buffer, pH 8.5, for 30 min following two 0.5-min washes with the buffer. Aniline blue stain visualization in samples has been accomplished with a 405-nm diode laser with emission filtering set at 490-510 nm.