The syntaxin 31-induced gene, LESION SIMULATING DISEASE1 (LSD1), functions in Glycine max defense to the root parasite Heterodera glycines

Abstract

Experiments show the membrane fusion genes α soluble NSF attachment protein (α-SNAP) and syntaxin 31 (Gm-SYP38) contribute to the ability of Glycine max to defend itself from infection by the plant parasitic nematode Heterodera glycines. Accompanying their expression is the transcriptional activation of the defense genes ENHANCED DISEASE SUSCEPTIBILITY1 (EDS1) and NONEXPRESSOR OF PR1 (NPR1) that function in salicylic acid (SA) signaling. These results implicate the added involvement of the antiapoptotic, environmental response gene LESION SIMULATING DISEASE1 (LSD1) in defense. Roots engineered to overexpress the G. max defense genes Gm-α-SNAP, SYP38, EDS1, NPR1, BOTRYTIS INDUCED KINASE1 (BIK1) and xyloglucan endotransglycosylase/hydrolase (XTH) in the susceptible genotype G. max_{[Williams 82/PI 518671]} have induced Gm-LSD1 (Gm-LSD1–2) transcriptional activity. In reciprocal experiments, roots engineered to overexpress Gm-LSD1–2 in the susceptible genotype G. max_{[Williams 82/PI 518671]} have induced levels of SYP38, EDS1, NPR1, BIK1 and XTH, but not α-SNAP prior to infection. In tests examining the role of Gm-LSD1–2 in defense, its overexpression results in ∼52 to 68% reduction in nematode parasitism. In contrast, RNA interference (RNAi) of Gm-LSD1–2 in the resistant genotype G. max_{[Peking/PI 548402]} results in an 3.24–10.42 fold increased ability of H. glycines to parasitize. The results identify that Gm-LSD1–2 functions in the defense response of G. max to H. glycines parasitism. It is proposed that LSD1, as an antiapoptotic protein, may establish an environment whereby the protected, living plant cell could secrete materials in the vicinity of the parasitizing nematode to disarm it. After the targeted incapacitation of the nematode the parasitized cell succumbs to its targeted demise as the infected root region is becoming fortified.

Keywords:

• Golgi
• LESION SIMULATING DISEASE1 (LSD1)
• membrane fusion
• pathogen resistance
• salicylic acid
• signaling
• syntaxin 31
• vesicle

Abbreviations

EDS1	=	enhanced disease susceptibility1
PR1	=	pathogenesis-related 1
NPR1	=	nonexpressor of PR1
SA	=	salicylic acid
LSD1	=	lesion simulating disease1
α-SNAP	=	alpha soluble N-ethylmaleimide-sensitive factor attachment protein
BIK1	=	botrytis induced kinase1
XTH	=	xyloglucan endotransglycosylase/hydrolase
RNAi	=	RNA interference
Sed5p	=	suppressors of the erd2-deletion 5
sec	=	secretion
JA	=	jasmonic acid
PAD4	=	phytoalexin deficient 4
SID2	=	salicylic-acid-induction deficient2
LOL1	=	LSD1-like
qPCR	=	quantitative polymerase chain reaction
INA	=	2,6-dichloroisonicotinic acid
SAR	=	systemic acquired resistance
O2−	=	superoxide
ROI	=	reactive oxygen intermediates
CuSOD	=	copper superoxide dismutase
SHMT	=	serine hydroxymethyltransferase
GOI	=	gene of interest
ER	=	endoplasmic reticulum
MATE	=	multidrug and toxin extrusion
PCD	=	programmed cell death

Introduction

Knowledge of the ability of biological membranes to fuse, resulting in the delivery of vesicle contents to different cellular destinations, is longstanding.¹ Genetic experiments and screens in model organisms have identified the proteins that function in the process and ordered the events that lead to material delivery in the form of secretion.^2-4 Subsequent work in other systems has demonstrated that the core protein machinery involved in membrane fusion is highly conserved, found in all eukaryotes (Reviewed in ⁵). The process of membrane fusion requires fidelity and protective measures are taken by the cell to ensure it happens properly.⁶

Through recent studies, a link between membrane fusion at the cell membrane and also the cis face of the Golgi apparatus with SA signaling has been made in plants.^7-9 Genetic work in the plant genetic model, Arabidopsis thaliana has also identified essential roles for proteins involved in membrane fusion.¹⁰ The essential nature of these membrane fusion proteins makes them difficult to study since their mutants are lethal or cause highly detrimental developmental anomalies.^2,3,10,11 However, it is possible to study these proteins under certain circumstances. For example, a genetic screen employed by Mayer et al.¹⁰ has determined the role of vesicles in embryo cytokinesis. This approach has succeeded because the biosynthesis of the phragmoplast which relies on vesicles occurs early during embryo development. Subsequent identification of one of the A. thaliana genes involved in cytokinesis (KNOLLE [At-SYP111]) has determined it to be related to a Saccharomyces cerevisiae membrane associated protein known as suppressors of the erd2-deletion 5 (Sed5p) which is structurally homologous to syntaxin.^12-14 Syntaxin is a protein involved in secretion, functioning in the fusion of membranes.^12,13 Syntaxins perform membrane fusion through their interaction with a number of other proteins (Reviewed in ⁵). One of these proteins is α-SNAP whose relation to plant defense has been demonstrated.^8,12,15,16 Since these discoveries, membrane fusion and vesicle transport have been well documented in plants, with many of the related genes having orthologs in yeast and other systems.^14,17,18

The roles that these core membrane fusion proteins perform in eukaryotes is extensive, ranging from signaling, cell growth, mitosis, the endocytic cycle, exocytosis, hormonal release, neurotransmission, fertilization, embryogenesis, development, sporulation and cell death.^{2,3,13-15,17-33} A variety of studies show membrane fusion to be important to the defense process that plants have toward pathogens as well as different types of defense responses.^9,11,34-42 While the list of functions that the membrane fusion and vesicle transport proteins have is large, it is less clear whether the proteins also are engaged in other, but related functions.

Recent experiments in G. max have demonstrated that α-SNAP contributes to the resistance of G. max to the plant parasitic nematode, Heterodera glycines.^8,43 The α-SNAP gene was first identified in S. cerevisiae as Sec17p in a genetic screen for temperature sensitive secretion (sec) mutants.³ Subsequent research has demonstrated Sec17p is required for vesicle transport from the endoplasmic reticulum (ER) to the Golgi apparatus as mutants accumulated 50 nm vesicles.^4,44 The results presented by Matsye et al.⁸ identified the existence of a role for α-SNAP that went beyond membrane fusion. Matsye et al.⁸ examined the effect that the overexpression of an α-SNAP gene had on genes associated with different types of hormonal signaling that have known defense functions. While not comprehensive, these genes included an analysis of the SA-regulated cysteine rich secretory protein gene, pathogenesis-related 1 (PR1).⁴⁵ Furthermore, the study examined the transcriptional activity of other genes whose protein products are secreted. These genes included the ethylene responsive β-1,3-glucanase, PR2,⁴⁶ the ethylene and jasmonic acid (JA) responsive chitinase gene, PR3⁴⁷ and the SA-responsive thaumatin, PR5.⁴⁸ In those experiments, Matsye et al.⁸ demonstrated α-SNAP overexpression causes induced expression of PR1, PR2 and PR5. Thus, the induced expression of components of the membrane fusion and vesicular transport machinery (α-SNAP) appears to influence the expression of genes that are vesicle cargo. To expand on this concept further, related experiments have been performed analyzing the effect that the overexpression of the α-SNAP binding partner, syntaxin 31 has on transcription.⁹ In these experiments, the overexpression of α-SNAP or SYP38 also results in the transcriptional induction of the SA signaling genes EDS1 and NPR1.⁹ In A. thaliana, SA biosynthesis and signaling occurs through a well-understood pathway including the EDS1 protein binding to the lipase PHYTOALEXIN DEFICIENT 4 (PAD4).^49-51 This heterodimer functions upstream of SALICYLIC-ACID-INDUCTION DEFICIENT2 (SID2), a putative chloroplast-localized isochorismate synthase, its allelic EDS16, along with the multidrug and toxin extrusion (MATE) efflux transporter EDS5 to activate SA biosynthesis.^52-54 Downstream, a complex composed of SA, the SA hormone receptor protein NPR1, copper ions and the transcription factor TGA2 forms.^55,56 The complex binds to a DNA promoter sequence composed of TGACG which results in the induction of PR1 transcription.^55,61 Another gene that relates to SA signaling in A. thaliana is LESION SIMULATING DISEASE1 (LSD1).⁶² In A. thaliana, the LSD1 gene is a negative regulator of programmed cell death (PCD) and its activity is antagonized by a related positive regulator of cell death gene called LSD1-like (LOL1).^63-68 Currently, it is unknown whether the G. max LSD1 functions in defense. However, its involvement in establishing a tight boundary between cells targeted and not targeted for apoptosis makes it an intriguing candidate.

In the analysis presented here, the relationship between the G. max α-SNAP, Gm-SYP38 and SA signaling is examined further, adding to information generated in prior experiments.⁹ Gene expression experiments have identified induced levels of Gm-LSD1 (Gm-LSD1–2) in roots engineered to overexpress α-SNAP or SYP38. These results further strengthen a link between vesicle transport and SA signaling. Genetic engineering experiments reveal that the overexpression of Gm-LSD1–2 results in engineered resistance. In contrast, RNAi of Gm-LSD1–2 in a G. max genotype that is normally resistant to H. glycines infection results in roots that permit parasitism at a higher frequency. It is shown the Gm-LSD1–2 overexpression positively influences the transcriptional activity of G. max SYP38, EDS1, NPR1 and BIK1. Furthermore, the overexpression of Gm-LSD1–2 also results in the induction of the expression of the hemicellulose-modifying, vesicle-cargo gene XTH43. In contrast, their expression is suppressed in roots expressing an LSD1–2 RNAi construct. The experiments presented here identify an antiapoptotic aspect of defense in the G. max-H. glycines pathosystem.

Results

Gm-LSD1 is expressed in roots overexpressing α-SNAP, SYP38 and genes relating to SA signaling

Deep sequencing experiments show that the overexpression of the G. max Gm-SYP38, results in the induction of 5 α-SNAP paralogs, including the rhg1 component Glyma18g02590 and Glyma11g35820 (Table 1). This result strengthened prior observations of the importance of α-SNAP to the process of defense.⁸ Furthermore, Pant et al.⁹ has demonstrated that along with the involvement of Gm-SYP38 during the defense of G. max to H. glycines, its overexpression also results in induced levels of the SA signaling gene EDS1. The demonstration that SA signaling genes function in the defense of G. max to H. glycines has led to an analysis showing that Gm-LSD1 (Gm-LSD1–2) is induced in roots overexpressing Gm-SYP38 (Table 2). During parasitism, a well demarcated boundary is established between parasitized and nonparasitized cells in the G. max-H. glycines pathosystem (Fig. 1). To understand the nature of Gm-LSD1–2 in relation to resistance (Fig. 2), qPCR experiments have been performed using cDNA template from genetically engineered G. max roots that acquired the ability to defend itself from H. glycines parasitism. Roots genetically engineered to overexpress G. max α-SNAP, SYP38, NPR1, EDS1, BIK1 or XTH43 exhibit induced levels of LSD1–2 (Table 2). The association of Gm-LSD1–2 expression in roots undergoing defense indicates that it may be performing an important role in the process. To test this hypothesis, the susceptible G. max_{[Williams 82/PI 518671]} has been engineered to overexpress Gm-LSD1–2 (Fig. 3). No statistically significant effect is observed in root growth (Fig. S1). In experiments presented here, the overexpression of the Gm-LSD1–2 results in a significant reduction in parasitism (Fig. 4). To examine the specificity of the overexpression experiments, the expression of an RNAi cassette for Gm-LSD1–2 in the normally resistant genotype G. max_{[Peking/PI 548402]} was done (Fig. 3). No statistically significant effect is observed in root growth (Fig. S1). The expression of an RNAi cassette for Gm-LSD1–2 in the normally resistant genotype G. max_{[Peking/PtdIns 548402]} results in an increased capability of H. glycines to parasitize the resistant G. max_{[Peking/PI 548402]} (Fig. 5).

Table 1. Deep sequencing of mRNA isolated from uninfected Gm-SYP38 overexpressing roots reveals altered transcriptional activity of the rhg1 resistance gene, α-SNAP (Glyma18g02590) and paralogs of α-SNAP

a-SNAP	log2(fold_change)	test_stat	P_value	q_value	significant
Glyma18g02590	0.396298	1.49148	0.0204	0.03961	yes
Glyma11g35820	0.39959	1.51642	0.0192	0.0375849	yes
Glyma14g05920	0.936435	2.72327	5.00E-05	0.000175486	yes
Glyma02g42820	2.64661	1.74839	0.00365	0.00866875	yes
Glyma09g41590	1.31903	3.53869	5.00E-05	0.000175486	yes

Table 2. qPCR of G. max roots overexpressing defense-related genes. qPCR was performed using primers designed specifically against LSD1–2 The experiments used the ribosomal S21 gene⁸ as a control to standardize the experiments

qPCR using LSD1 primers
Transgenic line	0 dpi
EDS1-OE	47.679
NPR1-OE	82.061
SYP38-OE	335.571
a-SNAP-OE	228.011
BIK1-OE	89.195
XTH43-OE	190.915

Figure 1. A 3 dpi image of H. glycines successfully parasitizing a root of G. max_{[Williams 82/PI 518671]}. Black arrow, nematode; red arrows, boundary of the nurse cell (syncytium). Bar = 100 μm.

Figure 2. The Golgi apparatus serves a central role in resistance as a defense engine, processing proteins for their eventual transport. The overexpression of α-SNAP resulted in engineered resistance.⁸ Furthermore, α-SNAP overexpression results in the induction of Gm-SYP38 transcription.⁹ In reciprocal experiments, Gm-SYP38 overexpression results in the transcriptional activation of α-SNAP and its paralogs (Table 1). The overexpression of Gm-SYP38 results in the transcriptional activation of EDS1 which functions upstream of SA biosynthesis (dashed lines). The overexpression of Gm-SYP38 also results in the transcriptional activation of the SA receptor, NPR1, the DNA binding β-ZIP transcription factor TGA2 and the GATA-like transcription factor LSD1. The binding of SA to NPR1 results in its translocation to the nucleus. NPR1 and TGA2 are directly involved in the transcriptional activation of PR1 and PR5. For presentation purposes, on the right side of the Golgi apparatus are shown vesicles undergoing anterograde transport while those on the left are undergoing retrograde transport. Vesicles are shown released from the trans-Golgi network, moving toward the endosome. Ultimately, secretory vesicles fuse with the plasma membrane to deliver receptor components and secrete contents into the apoplast. Some of these secreted contents, like Gm-XTH43, play important roles in defense.⁹ In contrast, vesicles emerge from the plasma membrane and fuse with the endosome, recycling contents. Not shown, Gm-SYP38 and α-SNAP overexpression results in induced expression of the cytoplasmic receptor-like kinase BIK1 that is important for defense ⁹

Figure 3. Representative control and transgenic LSD1–2 overexpressing and LSD1–2 RNAi G. max plants. (A) Control susceptible G. max_{[Williams 82/PtdIns 518671]} plant. (B) Genetically engineered G. max_{[Williams 82/PI 518671]} overexpressing Gm-LSD1–2. (C) Control resistant G. max_{[Peking/PtdIns 548402]} plant. (D) A resistant G. max_{[Peking/PI 548402]} plant genetically engineered to express an LSD1–2 RNAi construct. Scale provided on left of each image.

Figure 4. The female index for transgenic G. max plants genetically engineered to overexpress Gm-LSD1–2 and infected with H. glycines. Replicate 1 (R1) control plants had 28.39 cysts per gram (12 plants); LSD1–2-R1-overexpressing plants (LSD1–2-R1: oe) had 13.66 cysts per gram (12 plants). The FI = 47.92; P-value = 0.0216541 which is statistically significant (P < 0.05). R2 control plants (replicate 2) had 30.40 cysts per gram (16 plants); LSD1–2-R2-overexpressing plants (LSD1–2-R2: oe) had 9.85 cysts per gram (12 plants). The FI = 32.4; P-value = 0.000059234 which is statistically significant (P < 0.05). R3 control plants had 32.98 cysts per gram (20 plants); LSD1–2-R3 overexpressing plants (LSD1–2-R3: oe) had 14.07 cysts per gram (18 plants). The FI = 42.662; P-value = 3.36219e-06 which is statistically significant (P < 0.05).

Figure 5. G. max plants genetically engineered for RNAi of Gm-LSD1–2 and infected with H. glycines have an increased capability, shown as fold change, for parasitism. Replicate 1 (R1) control plants (resistant G. max_{[Peking/PtdIns 548402]}) had 1.98 cysts per gram (10 plants). LSD1–2-RNAi-R1 (LSD1–2-R1: RNAi) in resistant G. max_{[Peking/PI 548402]}) had 6.41 cysts per gram (11 plants). The results were statistically significant (p = 0.00255251). Replicate 2 (R2) control plants (resistant G. max_{[Peking/PtdIns 548402]}) had 0.79 cysts per gram (12 plants). LSD1–2-RNAi-R2 (LSD1–2-R2: RNAi) in resistant G. max_{[Peking/PI 548402]}) had 8.63 cysts per gram (5 plants). The results were statistically significant (p = 0.0117053). Replicate 3 (R3) control plants (resistant G. max_{[Peking/PtdIns 548402]}) had 2.51 cysts per gram (10 plants). LSD1–2-RNAi-R3 (LSD1–2-R3: RNAi) in resistant G. max_{[Peking/PI 548402]}) had 11.7 cysts per gram (7 plants). The results were statistically significant (p = 0.0120138).

Gm-LSD1–2 overexpression induces the expression of genes relating to membrane fusion and SA signaling

To understand the relationship between Gm-LSD1–2 and resistance, a series of qPCR analyses have been performed using cDNA synthesized from RNA isolated from roots overexpressing Gm-LSD1–2 (Table 3). The gene expression analysis demonstrates that Gm-LSD1–2 overexpression results in induced mRNA levels of LSD1–2 as well as EDS1–2, NPR1–2, BIK1–6, XTH43 and SYP38. In contrast, Gm-LSD1–2 overexpression results in suppressed levels of α-SNAP prior to infection. This result is not surprising since recent experiments have shown that α-SNAP becomes highly induced later during the resistant reaction.⁹ In reciprocal experiments, the expression of an RNAi cassette for Gm-LSD1–2 in the normally resistant genotype G. max_{[Peking/PtdIns 548402]} results in suppressed transcriptional activity for LSD1–2 as well as EDS1–2, NPR1–2, BIK1–6, XTH43 and SYP38 (Table 3). Expression of α-SNAP was not detected under the experimental conditions. The results confirm and provide further context for the existence of a link between the membrane fusion gene SYP38 and SA signaling.

Table 3. qPCR of G. max roots either overexpressing LSD1-2 or genetically engineered with a RNAi construct targeting LSD1–2 The experiments used the ribosomal S21 gene as a control to standardize the experiments. *expression not detected

qPCR	LSD1–2 OE	LSD1–2 RNAi
Gene tested	0 dpi	0 dpi
LSD1	293.784	−1.851
EDS1	40.129	−2.094
NPR1	145.11	−2.346
SYP38	581.545	−1.889
a-SNAP	−3.104	N/A*
BIK1	161.048	−1.359
XTH43	37.536	−1.223
PR1	4.276	3.192
PR2	159.282	−3.222
PR3	3.206	1.388
PR5	−2.005	1.2

Discussion

LSD1 was first discovered in A. thaliana in a forward genetic screen designed to identify spontaneous lesion simulating mutants.⁶² The 5 identified lsd mutants have been divided into 2 classes. One class forms spontaneous necrotic lesions that are determinate in nature.⁶² In this class, the expansion of necrosis into adjacent tissue is limited.⁶² Furthermore, lesion formation is not influenced by pathogens or chemicals such as SA and the non SA-inducing 2,6-dichloroisonicotinic acid (INA) that induce the onset of systemic acquired resistance (SAR).^62,70 The second class of lsd mutants, defined by LSD1, is described as a feedback or propagation mutant.⁶² The lsd1 mutant forms spontaneous lesions under long day growth conditions.⁶² In contrast, lesion formation is suppressed under short days.⁶² These characteristics indicate that light influences the process at some level. The lsd1 mutant is characterized by indeterminate lesions that eventually consume the whole leaf or plant.⁶² Another characteristic of lsd1 mutants is that plants grown under permissive short day conditions develop lesions that eventually consume the whole plant when switched to long day.⁶² Furthermore, the lsd1 mutant initiates lesion formation by fungal or bacterial pathogens and inducers of SAR, including SA and INA.⁶² Related experiments using lsd1 mutants demonstrate that superoxide (O₂−) accumulates in the cells adjacent to the cells undergoing cell death.⁶³ This result demonstrates that O₂⁻ is both necessary and sufficient to initiate lesion formation and promote its spreading into adjacent cells.⁶³ This result also identifies a link between photorespiration and lesion development.

It is clear from these studies that lsd1 mutants are impaired in their ability to establish a boundary beyond which the neighboring cells are not consumed in the wave of cell death. Sequence analysis of LSD1 demonstrates it to be a novel zinc finger, GATA-type transcription factor.⁶⁴ In this regard, the data presented here provides an example of a GATA-type transcription factor involved in G. max defense against H. glycines. From observations made in A. thaliana it has been hypothesized that LSD1 is responsible either to negatively regulate a pro-death pathway or activate a repressor of cell death.⁶⁴ As a regulator, LSD1 would function very early in the process. In A. thaliana, LSD1 has since been shown to function in relation to genes composing the SA signaling pathway, including EDS1, PAD4 and NPR1 as well as the signaling molecule SA.^65,71,72 Notably, LSD1 as an antiapoptotic gene, functions in the cells adjacent to the infected cell that is undergoing cell death.^65,71,72 Experiments have shown that runaway cell death was dependent on SA and NPR1 in lsd1 mutants.⁷² In contrast, LSD1 has been shown to negatively regulate SA and NPR1-independent basal disease resistance.⁷² From these studies, it has been proposed that SA and NPR1 function in runaway cell death in the lsd1 mutant through their participation in a signal amplification loop that promotes apoptosis.^71,72 It has been shown that an important component of runaway cell death is the generation of reactive oxygen intermediates (ROI) such as O₂⁻.^63,65,71,72 Additional studies further link the lsd1 mutant to impaired photorespiration, leading to the accumulation of excess excitation energy and subsequent cell death.⁷³ In contrast, cell death is prevented in the lsd1 mutants by impeding conditions that lead to photorespiration.⁷³ These results explain the link between the lsd1 and photo-oxidative damage. Thus, it has been proposed that the LSD1 protein functions like a rheostat whereby above a ROI threshold, the cell would undergo cell death.^{63,64,65,71,72,73} In contrast, below a certain threshold, the cell would survive. From this work, a signal potentiation loop has been coined to describe how in the absence of LSD1 protein, the accumulation of signaling components leads to runaway apoptosis.⁷²

These experiments focused in on the above ground portions of A. thaliana. Subsequently, a number of experiments examining LSD1 have studied specific aspects of root biology. Under certain adaptive environmental circumstances (i.e. water saturated conditions and low oxygen [hypoxia]), root cells become targeted for apoptosis through a process called lysigeny. As a consequence of this process, the roots develop aerenchyma which increases the ability of roots to maintain higher O₂ levels. Experiments in A. thaliana have shown that lysigeny is under the control of LSD1.⁷⁴ Under conditions of hypoxia, LSD1, EDS1 and PAD4 function upstream of H₂O₂ production and ethylene signaling events that lead to lysigeny.⁷⁴ Under normal conditions in A. thaliana, LSD1 functions as a negative regulator of the apoptosis-promoting EDS1 and PAD4. In contrast, under hypoxia, LSD1 is negatively regulated, permitting EDS1 and PAD4 to promote cell death in A. thaliana.⁷⁴ To understand how H₂O₂ production could be regulated in the roots, earlier experiments performed on aerial portions of A. thaliana demonstrated that LSD1 controls H₂O₂ production through SA-regulated transcription of CuSOD.⁶⁵ This is an important finding since plants can produce the highly toxic O₂− during plant defense by the activities of NADPH oxidase.⁷⁵ Recent findings performed in A. thaliana have shown a direct link between NADPH oxidase and BIK1.⁷⁶ In those experiments, BIK1 directly phosphorylates NADPH oxidase to produce O₂− and activate defense pathways. Plants then detoxify O₂⁻ to H₂O₂ through major antioxidant enzymes like CuSOD. Thus, certain aspects of LSD1 function in A. thaliana are similar between the shoot and root. Furthermore, recent findings in A. thaliana have also revealed LSD1 has many functions with regard to basic aspects of plant growth, development and its ability to function under different environmental conditions and stresses.⁷⁷ These observations place some context into the observation that Gm-BIK1 functions in defense in the G. max-H. glycines pathosystem.⁹

LSD1 transcription is induced in G. max roots overexpressing the membrane fusion gene α-SNAP

Two major H. glycines resistance loci have been identified from screening ecological collections of G. max.^78,79 These loci, the recessive rhg1 and the dominant Rhg4, have been mapped and cloned through traditional means and aided further by transcriptomics and candidate gene approaches.^{43,78,79,80-83} Genetic crosses of rhg1 and Rhg4-containing genotypes leads to progeny with further-enhanced, nearly full resistance. The additive effect that these loci have, regarding H. glycines resistance, indicate that the genes function in different genetic pathways that converge on the same outcome (resistance). The rhg1 locus, depending on the resistant genotype examined, is composed of multiple tandem repeated copies of 3 or 4 genes. These genes include an amino acid transporter, α-SNAP, a wound inducible protein and in some genotypes, a gene known as placenta-specific gene 8 protein (PLAC8).^43,81,82,84 Among these genes, the overexpression of α-SNAP has been shown to yield a resistant reaction when overexpressed on its own. As part of the secretory pathway, α-SNAP would function in many essential cellular processes.³ The other resistance gene, Rhg4, gene is a SHMT which plays a role in photorespiration. In overexpression studies, SHMT suppresses the ability of H. glycines to parasitize G. max.^69,83

The overexpression of α-SNAP leads to an increase in expression of its binding partner, syntaxin 31 (Gm-SYP38). Syntaxin 31 functions at the cis face of the Golgi apparatus to facilitate the fusion of transport vesicles transported from the endoplasmic reticulum.^3,4,44,85-88 In G. max, the overexpression of α-SNAP and Gm-SYP38 results in induced levels of the SA signaling genes EDS1, NPR1 and PR1.⁹ While the observation of an influence of vesicle transport on SA signaling is not a new concept,⁷ the results of Pant et al.⁹ indicates that SA signaling may be important to the process of defense in the G. max-H. glycines pathosystem. To test this hypothesis, the overexpression of Gm-EDS1 and NPR1 has been shown to lead to resistance.⁹ In related experiments, the overexpression of EDS1 and NPR1 in G. max leads to induced levels of SHMT prior to infection.⁹ Furthermore, the overexpression of G. max syntaxin 31 leads to slightly induced levels of EDS1 and SHMT during infection.⁹ While these experiments were not comprehensive, they indicate that genes composing the rhg1 locus can influence the expression of Rhg4.

The observation in G. max that EDS1 and NPR1 function in resistance to H. glycines indicated other genes relating to them may also function in the process. An obvious candidate is Gm-LSD1. In qPCR experiments examining G. max roots overexpressing α-SNAP, it is shown that Gm-LSD1–2 transcription is induced. Complimentary experiments presented here show that Gm-LSD1–2 is also induced in roots engineered to overexpress Gm-SYP38. Furthermore, Gm-LSD1–2 transcription is also induced in roots overexpressing BIK1, EDS1, NPR1 or XTH. The strong association of Gm-LSD1–2 with engineered forms of resistance led to the idea that it may perform a direct role in the process. Since A. thaliana LSD1 is known to play roles in establishing and maintaining a tight boundary around the cells and tissues involved in pathogen infection, it is possible that the expression of Gm-LSD1–2 could be performing an important role in regulating the expansion and/or initial survival of parasitized cells. The H. glycines-parasitized root cells undergo a slow process taking days to conclude that ultimately leads to resistance.⁸⁹ During this time, the parasitized root cell would have time to synthesize and secrete molecules in the vicinity of the nematode to neutralize its activities while fortifying the parasitized area. One such enzyme is Gm-XTH43. Notably, XTH contains a signal peptide and is transported through the vesicle transport machinery to the apoplast where it modifies hemicellulose.^9,90 Furthermore, the parasitized cell may produce O₂− whose subsequent metabolism to H₂O₂ has been shown in A. thaliana to be under regulation by LSD1.^63,65,70-74 In the analysis presented here, the overexpression of Gm-LSD1–2 in G. max_{[Williams 82/PI 518671]} roots that are otherwise susceptible to H. glycines parasitism, resulted in ∼52 to 68% reduction in nematode parasitism. Roots overexpressing Gm-LSD1–2, when tested for the expression of markers of resistance (i.e., XTH43, SYP38, NPR1, EDS1 and BIK1) show that each is induced in its expression prior to H. glycines infection. In examining molecular markers of different signaling processes, highly induced levels of PR2 were observed in Gm-LSD1–2 overexpressing roots prior to their infection by H. glycines. The induction of PR2 transcription indicates ethylene may also be a component of in Gm-LSD1–2-mediated resistance. The contribution of PR2 to resistance has been demonstrated, linking ethylene to the process.⁶⁹ In contrast, RNAi of Gm-LSD1–2 in the resistant genotype G. max_{[Peking/PI 548402]} demonstrates specificity. In these experiments, the normally resistant G. max_{[Peking/PI 548402]} roots engineered with the Gm-LSD1–2 RNAi cassette lacked the induction of LSD1–2 expression and exhibited an increase in parasitism capability. These results provide direct evidence that Gm-LSD1–2 plays an important role in the ability of G. max to prevent parasitism by H. glycines, contrasting with recent heterologous expression studies.⁹¹ In examining this discrepancy between the heterologous expression of A. thaliana LSD1 and Gm-LSD1–2 further, the conceptually translated At-LSD1 gene studied in Matthews et al.⁹¹ is 66.5% identical to the tested G. max LSD1–2 protein (Glyma08g13630) presented here. Thus, part of the difference observed between the capability of At-LSD1 and Gm-LSD1–2 proteins to function in G. max may arise from gene sequence variation. To reinforce our observation that Gm-LSD1–2 functioned in resistance, we present through a double-blind analysis experimental and biological replicates in both the Gm-LSD1–2 overexpression and RNAi experiments.

Spatial and temporal aspects regarding LSD1

The demonstration that Gm-LSD1–2 is important to the defense process clarifies the paradox that parasitized G. max root cells tolerate the establishment and maintenance of the attacked cell early during H. glycines parasitism prior to the commitment of the parasitized cell for demise. The association of LSD1 with the antiapoptotic activities of photorespiration in A. thaliana links its function to G. max Rhg4-mediated defense.^63,65,70-73 The demonstration that induced levels of Gm-LSD1–2 transcription in roots overexpressing the rhg1 gene α-SNAP and SYP38 links LSD1 to the process of vesicle transport at some level. At this point, many details remain concerning the genetic program responsible for the establishment and maintenance parasitized cell and surrounding root cells. From these observations, it is plausible that Gm-LSD1–2 functions initially in both the parasitized cell and surrounding cells to prevent cell death and establish a boundary. The demonstration that Gm-BIK1 is important to resistance implicates NADPH oxidase performing a role in the process.^9,76 NADPH oxidase would provide the O₂⁻ that could antagonize H. glycines. During this time, as the cell is protected from apoptosis, the vesicle transport machinery including the rhg1 gene α-SNAP would function to deliver antimicrobials, cell wall modifying enzymes and other substances to the site of parasitism. However, the process of resistance is not limited to this framework.

Methods

Gene cloning

The candidate gene overexpression study presented here has been done according to our published procedures using the pRAP15 and pRAP17 vectors.^8,9 The primers used to clone Gm-LSD1–2 (Glyma08g13630) are provided (Table S1). The nature of the hairy root system is that each transgenic root system functions as an independent transformant line.^9,92 Amplicons, representing the gene of interest (GOI) generated by PCR have been gel purified in 1.0% agarose using the Qiagen® gel purification kit, ligated into the directional pENTR/D-TOPO® vector and transformed into chemically competent E. coli strain One Shot TOP10. Chemical selection has been done on LB-kanamycin (50 μg/ml) according to protocol (Invitrogen®). Amplicons have been confirmed by sequencing and comparing the sequence to its original Genbank accession. The G. max amplicon has been shuttled into the pRAP15 or pRAP17 destination vector using LR clonase (Invitrogen®). The engineered pRAP15 or pRAP17 vector have been transformed into chemically competent A. rhizogenes strain K599 (K599)⁹³ using the freeze-thaw method⁹⁴ on LB-tetracycline (5 μg/ml).

The infection of G. max by H. glycines

Genetic transformation overexpression experiments have been performed according to Pant et al.⁹ in the functionally hypomorphic rhg1^−/− genetic background of G. max_{[Williams 82/PtdIns 518671]}, lacking a defense response to H. glycines parasitism. In contrast, RNAi studies have been performed in the rhg1^+/+ genetic background of G. max_{[Peking/PI 548402]} according to Pant et al.⁹ Female H. glycines_{[NL1-Rhg/HG-type 7/race 3]} have been purified by sucrose flotation.^95,96 Each root has been inoculated with one ml of nematodes at a concentration of 2,000 second stage juveniles (J2s)/ml per root system (per plant), infected for 30 d and confirmed by acid fuchsin staining.⁹⁷ At the end of the experiment, the cysts (fully matured females) have been collected over nested 20 and 100-mesh sieves.⁹ Furthermore, the soil has been washed several times and the rinse water sieved to assure collection of all cysts.⁹ The accepted assay to accurately reflect if a condition exerts an influence on H. glycines development is the female index (FI).⁹⁸ The FI has been calculated in a double blind analysis as FI = (Nx/Ns) X 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.⁹⁸ Nx is the pRAP15-transformed line that had the engineered GOI. Ns is the pRAP15 control in their G. max_{[Williams 82/PtdIns 518671]}. The effect of the overexpressed gene on parasitism has been tested statistically using the Mann–Whitney–Wilcoxon (MWW) Rank-Sum Test, P < 0.05.⁹

RNA-seq

Exon sequencing (RNA seq) has been performed according to our original published work with modifications.⁸¹ RNA has been extracted from G. max roots using the UltraClean® Plant RNA Isolation Kit (Mo Bio Laboratories®, Inc.; Carlsbad, CA) and treated with DNase I to remove genomic DNA.^8,9 RNA-seq analyses have been performed using the Illumina® HighSeq 2500® platform (Eurofins MWG Operon; Huntsville, Alabama). The RNAseq procedures that identified transcript (tag) counts and chromosomal coordinates of the G. max genome⁸⁴ along with the associated gene ontology (GO) annotations⁹⁹ are outlined here, subsequently. The qualities of raw reads have been checked using program FASTQC. The updated genome sequence and annotation of G. max⁸⁴ have been obtained from Phytozome v9.0 (dated: Nov 27, 2011). The abundance of transcripts across all samples has been measured and compared¹⁰⁰ and default setting of the programs used unless specified. Briefly, the raw reads for each sample have been mapped on G. max genome using TopHat v2.0.6.¹⁰¹ Then, Cufflinks v2.0.2¹⁰² program have been used to assemble the mapped reads into transcripts. The FPKM values were calculated for all genes in all samples and their differential transcript expression (log base 2) computed using program Cuffdiff.¹⁰²

Quantitative real-time PCR (qPCR)

The qPCR experiments examining LSD-1–2 overexpression have been performed according to Pant et al.⁹ The same root mRNA used in Pant et al.⁹ has been used here for the qPCR analyses of roots overexpressing G. max SYP38, α-SNAP, EDS1–2, NPR1–2, XTH43, BIK1–6. Primers used in qPCR gene expression experiments are provided (Table S2). The experiments, presented as log base 2, use the ribosomal protein gene S21 as a control.^8,9 Gene expression has been tested in relation to several different classes of pathogenesis related (PR) genes, and defense genes (Table S1). The qPCR experiments have used Taqman® 6-carboxyfluorescein (6-FAM) probes and Black Hole Quencher (BHQ1) (MWG Operon; Birmingham, AL). The qPCR differential expression tests have been performed according to Livak and Schmittgen.¹⁰³ The qPCR reaction conditions have been prepared according to Pant et al.⁹ and includes a 20 μl Taqman Gene Expression Master Mix (Applied Biosystems; Foster City, CA), 0.9 μl of μM forward primer, 0.9 μl of 100 μM reverse primer, 2 μl of 2.5 μM 6-FAM (MWG Operon®) probe and 9.0 μl of template DNA. The qPCR reactions have been executed on an ABI 7300 (Applied Biosystems®). The qPCR conditions include a preincubation of 50° C for 2 min, followed by 95°C for 10 min. This step has been followed by alternating 95°C for 15 sec followed by 60°C for 1 min for 40 cycles.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank George Hopper, Reuben Moore and Wes Burger (MAFES) whose support has made the research possible. Greenhouse space to support the research has been provided by the Department of Biochemistry, Molecular Biology, Entomology and Plant Pathology (BMEP) and the Department of Plant and Soil Sciences (PSS) at Mississippi State University. The authors thank Dr. Giselle Thibaudeau and Amanda Lawrence, Institute for Imaging and Analytical Technologies, Mississippi State University for imaging expertise and technical suggestions during the course of the research. Postdoctoral research support has been provided to Aparna Krishnavajhala through a competitive Special Research Initiative grant awarded by MAFES. Yixiu Pinnix (BMEP) has provided technical support. The efforts of Keshav Sharma, Jian Jiang, Prakash Nirula and Jillian Harris are acknowledged. Much of the research was supported by undergraduate students including Ashley Dowdy, Nishi Sunthwal, John Clune, Hannah Burson, Chase Robinson, Meghan Calhoun and Austin Martindale and Annedrea McMillan (DBS) and by James McKibben, Cody Roman and Micah Schneider (BMEP). The Shackouls Honors College is acknowledged. VPK acknowledges Dr. Ben Matthews and Dr. Perry Cregan at the USDA-ARS (Beltsville, MD) for support throughout the process.

Funding

The research has been supported by the start-up package provided by Mississippi State University and the Department of Biological Sciences (DBS). The authors thank the Mississippi Soybean Promotion Board for support. The research is supported jointly between the College of Arts and Sciences (DBS) and the Mississippi Agricultural and Forestry Experimental Station (MAFES). The authors acknowledge the Office of the Graduate School at Mississippi State University for providing competitive Summer Research Program for Undergraduate Students research award to Tineka Burkhead.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.