Updated May 24, 2023:
Selection of two soybean cyst nematode effectors that we have shown to suppress plant defenses:
Among the nine previously identified effectors able to suppress plant immunity, we have selected two effectors showing the strongest Effector Triggered Immunity (ETI) suppression abilities: GLAND1 and GLAND9.
While GLAND9 is a pioneer protein without recognizable domains or similarity to other proteins in the NCBI NR databases, GLAND1 comprises three distinct domains: (i) a previously undescribed N-terminal domain; (ii) an N-acetyltransferase domain reported as being acquired from soil bacteria (Streptomyces sp.) through Horizontal Gene Transfer; and (iii) a C-terminal domain carrying six tandem repeats of unknown function. GLAND1 and GLAND9 both are soluble proteins and exhibit secretory signal peptides and no transmembrane domains, strongly suggesting that they are secreted from the nematode salivary glands. The presence of a Dorsal Gland Motif (DOG box) in their respective promoters is consistent with the elevated expression of GLAND1 and GLAND9 during plant infection as emphasized by RNA-seq data (Figure 1A) and in situ hybridization. Indeed, this DOG box constitutes a well-known signature of effectors produced in the dorsal salivary gland of cyst nematodes, the most active gland during late infection stages.
We used recently published expression data to monitor at which time point GLAND1 and GLAND9 are expressed during the life cycle of H. schachtii, another species of cyst nematode for which seven time points of its life cycle have been subjected to RNA-seq analyses. We observed that GLAND1 and GLAND9 show elevated expression very early on at 48 and 10 hours post infection (hpi), respectively (Figure 1A). This expression timing corroborates their being involved in plant defense suppression because plant defense has to be suppressed as soon as the nematodes enter the plant.
We previously showed that GLAND1 and GLAND9 are able to suppress the ETI response induced by Pseudomonas syringae in N. benthamiana leaves. GLAND1 strongly delays the hypersensitive response (HR) triggered by the recognition of P. syringae by the host, whereas GLAND9 is a more moderate suppressor of the ETI response induced by P. syringae
Functional analysis of the GLAND1 domains and their involvement in ETI suppression:
Since GLAND1 comprises three distinct domains, we produced several truncated versions of this protein:
A ?N-terminal version (G1N); a ?C-terminal version (G1C); a partial ?C-terminal version where the three last repeats have been deleted (G1Gr 3R); a partial ?C-terminal version where the five last repeats have been deleted (G1Gr 1R) and a ?N-terminal version + ?C-terminal version (AcT) (thus only the N-acetyl-transferase domain remains). Moreover, to test if the N-acetyltransferase domain is involved in ETI suppression, we first predicted the 3D structure of the acetyltransferase domain of GLAND1 (using the TrRosetta webserver) and we performed a structural alignment with other acetyl-transferase domains for which the 3D structures have been resolved by crystallography. This analysis pointed out a conserved tyrosine residue described in the literature to be essential for the acetyltransferase activity, which is conserved in the GLAND1 acetyltransferase domain. Subsequently, we used directed-site mutagenesis to introduce point mutations on the conserved catalytic Tyrosine residue (G1 Y231F) involved in the transfer of the acetyl group from Acetyl-CoA to the potential plant targets of GLAND1 (Figure 1B).
All GLAND1 mutants have been secreted into N. benthamiana leaves by P. syringae, which allowed us to test which domains are important for ETI suppression. We observed that the N-terminal and C-terminal domains of GLAND1 are essential for its ETI suppression ability (Figure 2A, 2B). Since the C-terminal domain contains multiple repeated motifs, we produced truncated versions of GLAND1 for which the three or five last repeats were deleted. The three-repeats-deleted versions still show an ability to delay the HR cause by P. syringae, but the HR is less delayed compared to the wild-type version of GLAND1, suggesting that those three last repeats are not essential but support the ETI suppression ability of GLAND1 (Figure 2A, 2B). The five-repeats-deleted version of GLAND1 could not suppress the HR as observed for the full ?C-terminal version of GLAND1, suggesting that at least two or three repeats are essential to the ETI suppression ability of GLAND1. When the acetyltransferase domain is expressed alone, we did not observe any HR suppression ability, suggesting that the N and C-terminal domain have to be present to make GLAND1 functional. We confirmed that all those full-length point-mutated or truncated versions of GLAND1 are indeed expressed and produced in P. syringae as shown by Western Blot (Figure 2C).
We reasoned that if the acetyltransferase domain is involved in the ability of GLAND1 to suppress the ETI response, then introducing a point mutation to modify the conserved tyrosine, involved in the catalytic transfer of the acetyl to its unknown plant substrate, into phenylalanine should impair the ETI suppression ability of GLAND1. We discovered that modifying the conserved tyrosine into phenylalanine does not impair the ETI suppression ability of GLAND1 (Figure 2A, 2B).
Because tyrosine and phenylalanine differ only in one hydroxyl group and in order to further scrutinize if this amino acid plays a central role in the acetyl-transferase domain and could be involved in the suppression of plant immunity, we subsequently modified this tyrosine into alanine, a much more distant amino acid. Also, since another tyrosine is located right next to the tyrosine of interest, we have produced multiple point mutations aimed at mutating one or both tyrosines into phenylalanine or alanine (Figure 3A).
We again used P. syringae to secrete those point-mutated versions of GLAND1 into N. benthamiana leaves and we observed that constructs carrying an alanine instead of the conserved tyrosine are not able to suppress ETI anymore, whereas the phenylalanine-mutated versions are still able to suppress ETI. This result demonstrates that the acetyltransferase activity of GLAND1 is important for ETI suppression (Figure 3B). One can hypothesize that GLAND1 interacts with one or more plant targets through its N-terminal and/or C-terminal domains and acetylates those targets to impair or stabilize their function involved in ETI using its acetyl-transferase domain.
To gain a better insight into how GLAND1 might function inside plant cells, we produced GFP fusions with full-length or truncated versions of GLAND1. Those constructs were transiently expressed in N. benthamiana leaves using Agrobacterium tumefaciens, and the fluorescence signals were monitored using a confocal microscope at 48 hpi (Figure 4). GLAND1 exhibits a nucleocytoplasmic subcellular localization which is excluded from the nucleolus in all the cells observed, whereas the N-terminal, full C-terminal or C-terminal repeats-truncated versions lead to a nucleolar accumulation in some cells or show the same pattern as the full-length GLAND1 in other cells from the same infiltrated leaf area. More interestingly, the acetyltransferase domain alone also localized to the nucleolus and particularly in nuclear speckles. Since histones are located in the nucleus, and knowing that actively transcribed chromatin is localized in nuclear speckles, we hypothesize that GLAND1 could target chromatin-associated proteins, such as histones, to acetylate those and regulate the expression of genes involved in the ETI pathway. A validation of these subcellular localizations will be performed in soybean roots to confirm our observations.
We set-up a protocol in the laboratory to generate composite roots in soybean (Figure 5). This protocol allows now to rapidly generate transgenic roots in a non-sterile manner, yielding high rates of SCN infection.
To go further, we have generated a dozen of GATEWAY-compatible vectors to be able to express our effectors of interest in soybean transgenic roots (e.g., soybean composite plants). We firstly modify the antibiotic resistance on pICH75322 KanaR to Tetracycline resistance to ensure compatibility with the GATEWAY cloning system. Therefore, we used the GoldenGate cloning system to assemble two intermediate vectors. Furthermore, we converted those intermediate vectors into GATEWAY-compatible vectors (Figure 6). Those vectors allow now to quickly and easily fuse any gene of interest. Furthermore, we added markers for the selection of transgenic roots: the red fluorescent protein mCherry or the non-invasive reporter gene RUBY. We have tested the functionality of those vectors by expressing subcellular markers in soybean root (Figure 7). We also demonstrated the feasibility of observing those subcellular markers in feeding sites of plant-parasitic nematodes.
Concerning the GLAND1 effector, we have conducted analysis of its expression pattern in H. glycines infecting soybean roots at 4, 7, 12 and 30 dpi using RT-qPCR, and observed an early expression peak at 4 dpi, correlating with its defense suppression activity previously demonstrated (Figure 8).
To identify the potential plant target of GLAND1, we performed two independent Yeast-two hybrid screens, one against an infected soybean library and another against an infected A. thaliana library. We have first tested that GLAND1 does not autoactivate the reporter genes in yeast by itself and confirmed its expression in yeast by Western Blot (Figure 9). Concerning the soybean library, we have tested 7 500 000 interactions that lead to the selection of two colonies. Those two genes correspond to a Hsp40-like protein, unfortunately this gene is reported to be an auto-activator in previous studies and should be ignored. The second gene corresponds to an hydroxyproline-rich glycoprotein, which is an important component of the plant cell wall. Concerning A. thaliana library, we have tested 30 million interactions that lead to the selection of 15 colonies. We started the process of sequencing those yeast colonies to find out which plant interactors of GLAND1 have been captured (Figure 9).
To find interactors in planta using the composite soybean plant system along with the new GATEWAY vector recently generated, we planned to use immunoprecipitation and proximity labelling approaches. As a proof of concept for the immunoprecipitation approach, we successfully detect by Western Blot fusions of GLAND1 with the epitope 3xHA in soybean root. We successfully detect the WT and catalytic version of GLAND1 (Figure 10). Using this material, we are now prepared to conduct immunoprecipitation followed by analysis by mass spectrometry to identify the plant targets of GLAND1 in planta. The catalytic mutant of GLAND1 will be used to compare the list of plant interactors with the WT version, which should allow the identification of the acetylation site on the plant target. Concerning proximity labelling, we detected by Western Blot the expression of the miniTurboID protein alone or fused to GFP (control) (Figure 10). More optimization will be needed to detect the miniTurboID fusions with GLAND1 to perform proximity labelling in order to identify its plant targets through this method.
This project has led to the functional characterization of the GLAND1 effector. We have selected GLAND1 as an effector candidate since it is the strongest immune suppressor identified so far by our team. We have identified domains and catalytic amino acid responsible for its acetyltransferase activity and demonstrated that those mutations impair its plant immune suppression ability. We have shown that GLAND1 is expressed early-on during infection supporting its involvement in plant defense suppression since defense suppression is expected to occur early-on by SCN. We showed that GLAND1 localized mainly to the cytoplasm and nucleus of plant cells. Using Yeast-two hybrid screens, we have identified an hydroxyproline-rich glycoprotein as a potential target in soybean. We are currently sequencing yeast colonies found to interact with GLAND1 using the A. thaliana library. In order to be able to work directly with soybean roots, we developed a composite plant system along with a set of GATEWAY-compatible vectors that will drastically accelerate research on SCN effectors. Using these tools, we will use two other promising approaches in soybean roots to capture the plant targets of GLAND1 in planta. We have prepared constructs to perform similar approaches on the second selected effector (GLAND9) using composite plant system.