Gene silencing to investigate the roles of receptor-like proteins in Arabidopsis

Abstract

Receptor-like proteins (RLPs) are cell surface receptors that play important roles in various processes. In several plant species RLPs have been found to play a role in disease resistance, including the tomato Cf and Ve proteins and the apple HcrVf proteins that mediate resistance against the fungal pathogens Cladosporium fulvum, Verticillium spp., and Venturia inaequalis, respectively. The Arabidopsis genome contains 57 AtRLP genes. Two of these, CLV2 (AtRLP10) and TMM (AtRLP17), have well-characterized functions in meristem and stomatal development, respectively, while AtRLP52 is required for defense against powdery mildew. We recently reported the assembly of a genome-wide collection of T-DNA insertion lines for the Arabidopsis AtRLP genes. This collection was functionally analyzed with respect to plant growth, development and sensitivity to various stress responses including pathogen susceptibility. Only few new phenotypes were discovered; while AtRLP41 was found to mediate abscisic acid sensitivity, AtRLP30 (and possibly AtRLP18) was found to be required for full non-host resistance to a bacterial pathogen. Possibly, identification of novel phenotypes is obscured by functional redundancy. Therefore, RNA interference (RNAi) to target the expression of multiple AtRLP genes simultaneously was employed followed by functional analysis of the RNAi lines.

Receptor-like proteins (RLPs) are cell surface receptors that typically consist of an extracellular leucine-rich repeat (eLRR) domain, a single-pass transmembrane domain and a short cytoplasmatic tail that lacks obvious motifs for intracellular signaling except for the putative endocytosis motif found in some members.^1–3 In several plant species RLPs play important roles in development and pathogen defense. Arabidopsis CLAVATA2 (CLV2; AtRLP10) and its maize ortholog FASCINATED EAR2 are required for maintaining the meristematic stem cell population in shoot apical meristems, while Arabidopsis TOO MANY MOUTHS (TMM; AtRLP17) controls the initiation of stomatal precursor cells.^4–7 The RLP disease resistance gene family comprises the tomato Cf and Ve genes that provide resistance against Cladosporium fulvum and Verticillium spp., respectively,^8–10 LeEIX genes that encode receptors for the ethylene inducible xylanase produced by Trichoderma biocontrol fungi,¹¹ apple HcrVf genes that confer resistance to the scab fungus Venturia inaequalis,¹² and an Arabidopsis RLP gene (AtRLP52) that provides resistance against the powdery mildew pathogen Erysiphe cichoracearum.¹³ We recently reported the assembly of a genome-wide collection of T-DNA insertion lines for the 57 Arabidopsis RLP genes (AtRLP) in the Arabidopsis genome.¹⁴ This collection was functionally analyzed with respect to plant growth, development and sensitivity to various stress responses including pathogen susceptibility. Only few novel phenotypes were discovered; while AtRLP41 was found to mediate abscisic acid sensitivity, AtRLP30 (and possibly AtRLP18) was found to influence non-host resistance towards Pseudomonas syringae pv. phaseolicola.¹⁴

The lack of identification of biological functions for the majority of the AtRLP genes may be caused by functional redundancy. Here, we describe a reverse genetics strategy by employing RNA interference (RNAi) to target the expression of multiple AtRLP genes simultaneously, and thus possibly overcome functional redundancy among AtRLP genes. To select suitable fragments for RNAi silencing, the AtRLP genes were aligned and sequence stretches of a few hundred base pairs (bp) containing minimum one 21 bp stretch with 100% identity to minimum one other AtRLP gene were identified. Specificity of the selected fragments was verified with BLAST searches against the Arabidopsis genome.¹⁵ Seven AtRLP gene fragments, varying in length between 238 and 407 bp, were PCR-amplified such that the PCR products contained a 5' BamHI or HindIII site and a 3′ EcoRI and NotI site (Table 1; Fig. 1A) and cloned into the pGEM-T Easy vector (Promega, Leiden, The Netherlands). The resulting plasmids were digested in two separate reactions with HindIII (or BamHI for RNAi constructs 2 and 5) in combination with NotI and in combination with EcoRI. Both inserts were cleaned from gel using the QIAquick gel extraction kit (Qiagen, Venlo, NL) and subsequently ligated with a NotI- and EcoRI-digested 129 bp spacer segment from the Pichia pastoris Aox-1 gene into the HindIII-digested (or BamHI for RNAi constructs 2 and 5) pGreen plasmid¹⁶ to obtain inverted repeat constructs driven by the CaMV 35S promoter that target expression of multiple AtRLP genes (Fig. 1B). The resulting seven plasmids were introduced into Agrobacterium tumefaciens strain GV3101 by electroporation, transformed to Arabidopsis,¹⁷ and multiple homozygous single-insertion T₃ lines were selected on MS plates supplemented with 100 µg/mL kanamycin that were used for functional analysis.

Table 1.

RNAi constructs to target homologous AtRLP genes

RNAi construct	Tared genea)	Primer name	Restriction site	PCR product	Primer sequence (5′-3′)b)	TF No.c)		Homology tod)
1	AtRLP8	At154480F	HindIII	301 bp	AAGCTT-GGTTATCCCAGCAGAGC	5	AtRLP14	(At1g74180)	21 bp
	(At1g54480)	At1g54480R	EcoRI/NotI		GAATTC-GCGGCCGC-ATTGGTCGTGGTCCAC		AtRLP21	(At2g25470)	28 bp
2	AtRLP53	At5g27060F	BamHI	407 bp	GGATCC-AAAGGTGTAGCGATGGAGCTGG	8	AtRLP19	(At2g15080)	20 + 33 bp
	(At5g27060)	At5g27060R	EcoRI/NotI		GAATTC-GCGGCCGC-GCTGGCGTGTGAATATCTCTGC		AtRLP34	(At3g11010)	22 + 24 + 26 + 28 + 44 + 59 bp
						AtRLP35	(At3g11080)	21 + 43 + 55 + 102 bp
							AtRLP43	(At3g28890)	22 + 24 + 25 + 27 + 29 bp
3	AtRLP36	At13G23010F	HinDIII	336 bP	AAGCTT-CCGATTCTCCGGACATATCCCT	7	AtRLP37	(At3g23110)	25 bp
	(At3g23010)	At3g23010R	EcoRI/NotI		GAATTC-GCGGCCGC-GGCACATGATGGCTTTCTCCAC		AtRLP38	(At3G23120)	28 bp
4	AtRLP15	At1g74190F	HindIII	289 bp	AAGCTT-CCAGACACATTGCTTGC	4	AtRLP13	(At1g74170)	22 bp
	(At1g74190)	At1g74190R	EcoRI/NotI		GAATTC-GCGGCCGC-CATCAGAAGGGAAAGAAATGC
5	AtRLP41	At3g25010F	BamHI	312 bp	GGATCC-CCGAAATTGCAAGTCCTTCTCC	9	AtRLP23	(At2g32680)	24 bp
	(At3g25010)	At3g25010R	EcoRI/NotI		GAATTC-GCGGCCGC-GGCTGAGGAAGTAAGAACC		AtRLP39	(At3g24900)	22 + 24 + 54 + 56 bp
							AtRLP40	(At3g24954)	25 bp
							AtRLP42	(At3g25020)	24 + 24 + 26 + 27 + 32 + 36 bp
							PGIP	(At3g24982)	25 bp
6	AtRLP47	At4g13810F	HindIII	269 bP	AAGCTT-CCTCTCTGGTATTTTTCCAG	5	AtRLP48	(At4g13880)	22 + 24 + 27 bp
	(At4g13810)	At4g13810R	EcoRI/NotI		GAATTC-GCGGCCGC-TTCGCAACCTGGAGAAACTTAAAG		AtRLP49	(At4g13900)	33 + 36 + 59 + 63 bp
							AtRLP50	(At4g13920)	23 + 41 bp
							PGIP	(At4g13820)	21 bp
7	AtRLP2	At1g17240F	HindIII	238 bp	AAGCTT-TACCAGTCGAAGTTGGCCAG	8	AtRLP3	(At1g17250)	28 bp
	(At1g17240)	At1g17240R	EcoRI/NotI		GAATTC-GCGGCCGC-TCAAACTGACCTTCACTGGG

^a)Target gene used as template for PCR amplification.

^b)Restriction sites are underlined.

^c)Number of homozygous single-insert lines tested.

^d)Stretches of base pair identities (>20 bp) of the AtRLP fragment in the RNAi construct with the most homologous AtRLP genes indicated.

Figure 1.

Cloning strategy for RNAi constructs. (A) PCR fragments of specific AtRLP fragments are generated with 5′HindIII (or BamHI for RNAi construct 2 and 5) and 3′ NotI and EcoRI restriction sites. (B) Inverted repeat constructs are generated by ligating HindIII (or BamHI for RNAi constructs 2 and 5) and NotI digested PCR fragment and HindIII (or BamHI for RNAi constructs 2 and 5) and EcoRI digested PCR fragment together with a NotI- and EcoRI-digested 129 bp spacer segment from the Pichia pastoris Aox-1 gene into the HindIII-digested (or BamHI for RNAi constructs 2 and 5) pGreen backbone. The fragments are not drawn to scale.

Similar as the individual AtRLP insertion lines,¹⁴ also the RNAi lines were analyzed with respect to plant development and sensitivity to various abiotic and biotic stress factors. Development of roots, rosettes, leaf cuticle and flowers as well as stomatal patterning were examined, but no developmental anomalies were observed. In addition, the RNAi lines were assayed for altered sensitivity to plant hormones and abiotic stress factors. The only consistently altered phenotype was observed for lines containing RNAi construct 5 upon exogenous application of the plant hormone abscisic acid (ABA), as leaves of the RNAi lines bleached while wild-type leaves remained green (Fig. 2). Since RNAi construct 5 is predicted to target AtRLP41 of which a knock-out has been shown to result in enhanced ABA susceptibility¹⁴ this phenotype was expected. Moreover, this observation confirms that RNAi-mediated gene silencing can be used as a mechanism to investigate the function of RLP receptors. To determine whether AtRLP genes play a role in recognition of plant pathogens, similar as the individual AtRLP insertion lines¹⁴ the collection of AtRLP RNAi lines was assessed for altered phenotypic responses upon challenge with a range of diverse host-adapted and non-adapted necrotrophic and biotrophic pathogens.^14,18 In addition to the previously used pathogens,¹⁴ we included Fusarium oxysporum f. sp. raphani strain 815, the Verticillium dahliae strains St12.01, St17.01 and JR2, as well as the two oomycete strains Phytophthora brassicae HH/CBS782.97 and CBS686.95 in our analysis. Interestingly, no significant differences were identified when the responses of the RNAi lines were compared to those of the parental Col-0 line upon inoculation with any of the pathogens used.

Figure 2.

RNAi construct 5 triggers ABA-induced chlorosis. Comparison of the leaf phenotype of six independent transgenic lines containing RNAi construct 5 (a to f) with the parental line Col-0 three days after application of 100 mM abscisic acid (ABA).

The Arabidopsis genome harbors 24 loci containing a single AtRLP gene and 13 loci comprising multiple, between two and five, AtRLP genes.^14,19 Often, the most homologous AtRLP genes reside at the same locus,^14,19 and therefore crossing individual T-DNA insertion lines to obtain knock-out lines for multiple AtRLP genes is nearly impossible. RNAi-mediated gene silencing currently is the most suitable strategy to target expression of several highly homologous genes simultaneously. Based on the sequence comparison between Arabidopsis and rice RLP genes, and building on the hypothesis that developmental genes are less likely to be duplicated and undergo diversifying selection than are disease resistance genes,²⁰ most AtRLP genes were proposed to be candidate disease resistance genes.¹⁹ Remarkably, despite an extensive list of pathogens tested, including adapted and non-adapted pathogens of Arabidopsis, we have been able to identify only one AtRLP gene with a role in basal non-host resistance against the non-adapted bacterial pathogen Pseudomonas syringae pv phaseolicola when screening a genome-wide collection of T-DNA insertion lines in the AtRLP genes.¹⁴ It was hypothesized that the lack of identification of biological functions for AtRLP genes may be explained by functional redundancy.¹⁴ In the experiments presented in this manuscript we employed RNA interference to interfere with the expression of multiple AtRLP genes at the same time to overcome functional redundancy among AtRLP genes. Nevertheless, no biological functions could be assigned to additional AtRLP genes. Obviously, the targeted AtRLP genes might function in defense against pathogens that have not yet been assayed. As suggested previously,¹⁴ if AtRLP genes are active in non-host resistance or basal defense, the array of potential microbial targets may be significantly increased and the response to more microbes or even insects and nematodes should be tested.²¹ Furthermore, it may be questioned whether the knock-down established by RNAi is sufficiently strong to compromise RLP receptor activity, although gene silencing has been successfully used to compromise the activity of RLP-type disease resistance genes in tomato.²² Also, the observation that transformants expressing RNAi construct 5 phenocopies the AtRLP41 T-DNA insertion allele with respect to ABA responsiveness argues against this possibility. Possibly, however, the RNAi constructs do not silence all redundant AtRLP homologs as efficiently or target all the redundant AtRLP homologs. For instance, RNAi construct 4 that is derived from AtRLP15 is predicted to silence expression of AtRLP13, but not of AtRLP16 which is also close homologue of AtRLP15. Finally, redundant AtRLP genes are not necessarily those with the highest overall homology, since ligand specificity may be determined by only a small sequence stretch, making it difficult to design the most potent RNAi constructs. Therefore, a more extensive analysis using many more RNAi constructs is needed to exclude the possibility that the lack of phenotypes can be explained by a high degree of functional redundancy among the AtRLP genes. Overall, the RNAi lines developed in our studies provide a useful tool for further investigation into roles of the AtRLP genes.

Addendum to: Wang G, Ellendorff U, Kemp B, Mansfield JW, Forsyth A, Mitchell K, Bastas K, Liu CM, Woods-Tör A, Zipfel C, de Wit PJGM, Jones JDG, Tör M, Thomma BPHJ. A genome-wide functional investigation into the roles of receptor-like proteins in Arabidopsis. Plant Physiol. 2008;147:503–517. doi: 10.1104/pp.108.119487.