New sources of Sym2^A allele in the pea (Pisum sativum L.) carry the unique variant of candidate LysM-RLK gene LykX

Plant material

The search for pea lines demonstrating the “Afghan” phenotype was conducted on the basis of material obtained from the VIR plant genetic resources gene bank (N. I. Vavilov All-Russian Institute of Plant Genetic Resources (VIR), St. Petersburg, Russia), kindly provided by Prof. M. A. Vishnyakova, head of Department of genetic resources of leguminous crops, and E. V. Semenova, curator of the pea collection. The choice of material was determined by the geographical origin of samples (the Middle East). The whole sample contained 122 pea lines (see Table S1).

Additionally, a number of pea cultivars and genetic lines with the confirmed “Afghan” phenotype from the collection of the laboratory of genetics of plant-microbial interactions of the All-Russia Research Institute for Agricultural Microbiology (ARRIAM; St. Petersburg, Russia) were used as control (Table S2). The line 701 of Pisum fulvum Sibth. & Smith (Israel, Valley of the Cross) was kindly provided by Sc.D. O.E. Kosterin (Novosibirsk, Russia).

Plant growth conditions

During the summer season of 2011, plants were grown in vegetation houses in conditions corresponding to the region’s climate (St. Petersburg, Leningrad Oblast, Russia). When necessary for control, plants were grown in Vötsch Industrietechnik VB 1014 growth chambers (Balingen, Germany) under the following conditions:

• day/night: 16/8 hours;

• temperature: 21 °C;

• relative air humidity: 75%; illumination: 38000 lux.

The seeds were pre-sterilized with concentrated H₂SO₄ for between 10 and 15 minutes and then washed five times with distilled water. Sterilized seeds were germinated in Petri dishes filled with sterile vermiculite for 3 days at 27 °C, then seedlings were planted in pots. For large-scale experiments in vegetation houses, sterilization and preliminary germination of seeds was omitted.

Plants were grown in the 2 l pots filled with sand. When sterility was necessary, pots together with sand were heated in a dry-heat cabinet for 2 h at 200 °C. Mineral nutrition (Table S3) was added to substrate simultaneously with planting seedlings, 150 ml of solution per pot. When the analysis of the symbiotic phenotype was required, combined nitrogen in the form of ammonium nitrate was not added to the nutrient solution.

Bacterial strains

In experiments involving analysis of nodulation, two contrast strains of Rhizobium leguminosarum bv. viciae were used:

Strain R. leguminosarum bv. viciae RCAM1026 (= CIAM1026), used in ARRIAM as a model strain (Afonin et al., 2017). This strain is unable to colonize a plant with the “Afghan” phenotype due to absence of the nodX gene (nodX⁻ strain).

Strain R. leguminosarum bv. viciae A1, first isolated in the Leningrad Oblast from nodules of the “Afghan” pea line (Chetkova & Tikhonovich, 1986), thus possessing the nodX gene (nodX⁺ strain).

For the visualization of symbiotic phenotype, the gusA-marked derivative of RCAM1026 (RCAM1026 gusA) with constitutive expression of gusA gene was used (Kirienko et al., 2018).

Bacteria cultivation

The strains were cultivated on solid medium No. 79 (Table S4) in a thermostat at a temperature of 28 °C for three days. In the case of streptomycin-resistant strain RCAM1026, streptomycin (100 g/ml) was added to the medium; in the case of strain A1, no antibiotics were added. Plants were inoculated with a suspension of bacteria: the culture (“lawn”) from one Petri dish was resuspended in 1 l of sterile distilled water (final CFU/ml : not less than 10⁶) and the resulting solution was added to the substrate while planting the seedlings, 100 ml per pot.

Plant DNA extraction

DNA was extracted from young leaves (top or second-from-top node). Extraction was conducted according to the previously described CTAB protocol (Rogers & Bendich, 1985; Sulima et al., 2017).

Sequencing of DNA fragments

To amplify the first exons of LykX, primers were used as described in Sulima et al. (2017). The PCRs were performed in 0.5 ml eppendorf-type microcentrifuge tubes on an iCycler (Bio-Rad, Hercules, CA, USA) or Dyad (Bio-Rad) thermocycler using the ScreenMix-HS kit (Evrogen, Moscow, Russia). The PCR cycling conditions were as follows: 95 °C (5 min), 35 × [95 °C (30 s), Tm (varying depending on primers) (30 s), 72 °C (1 min)], 72 °C (5 min). The PCR fragments were sequenced using the ABI Prism3500xL system (Applied Biosystems, Palo Alto, CA, USA) at the “Genomic Technologies, Proteomics, and Cell Biology” Core Center of the ARRIAM (St.Petersburg, Russia). The resulting sequences have been deposited into the NCBI database under accession numbers MN187362 –MN187364, MN200353 –MN200358.

Analysis of sequences was conducted by means of MEGA7 software (Kumar, Stecher & Tamura, 2016). A Minimum Evolution tree was built using the modified Nei-Gojobori (assumed transition/transversion bias = 2) method (Zhang, Rosenberg & Nei, 1998). A bootstrap test was performed with 500 bootstrap replications. Alignments were done in MEGA7 and EMBL-EBI Clustal Omega web service (https://www.ebi.ac.uk/Tools/msa/clustalo/).

Genotyping of F₂ population

To analyze the F₂ sample obtained from the cross between lines K-3821 and NGB2150 carrying different alleles of LykX, the CAPS (cleaved amplified polymorphic sequence) marker was designed using the dCAPS Finder 2.0 online service (Neff, Turk & Kalishman, 2002). The primers used were the same as for amplification of the 1^st exon. The digestion of SNP (single nucleotide polymorphism) site was performed by the Hpy188III restriction enzyme (New England Biolabs, Massachusetts, USA).

Microscopy

For detailed phenotype analysis, plants were grown in sand, as described above, while inoculated with suspension of RCAM1026 gusA culture. Four weeks after inoculation, 5 segments of roots 2–2.5 cm long were collected from each plant, leaving approximately two cm from the base of the root and two cm from the tips (if possible). Root segments were stained for 18 h in a solution of 50 mM Na-phosphate buffer (pH 7.0) containing 0.1% Triton, 5 mM K₃Fe(CN₆) (red blood salt), 5 mM K₄Fe(CN₆) (yellow blood salt) and 0.02 % 5-bromo-4-chloro-3-indolyl-bD-glucuronic acid (X-glc). For storage, stained segments were fixated with 4% paraformaldehyde in a TBS buffer (50 mM TrisHCl, 150mM NaCl, pH 7.5) with the addition of 0.1% Tween-20 and 0.1% Triton X-100 under a vacuum (7 min, 3–4 times) using a VacuuBrand ME 1C vacuum pump (Vacuubrand, Wertheim, Germany) and then dehydrated in graded ethanol series (from 10% to 70%). For observation, fixated samples were treated with the same ethanol series in reverted order and then with TBS buffer. Microscopic analysis was conducted using Zeiss Axiovert 35 inverted phase contrast microscope (ZEISS, Oberkochen, Germany) by 100-fold magnification. The photos were taken on the Canon EOS 4000D camera (Canon, Tokyo, Japan).

Selection of pea lines with narrow symbiotic specificity

To increase the amount of the available “Afghan” lines, a sample of 122 pea lines originating from various regions of the Middle East (Table S1) was obtained from the VIR plant genetic resources gene bank. In the summer season of 2011, the nodulation phenotype of the lines was tested by inoculation with nodX⁺ or nodX⁻ strains. Five lines were discovered which did not form nodules when inoculated with nodX⁻ strain RCAM1026, but demonstrated normal nodulation with strain A1 (i.e., the typical “Afghan” phenotype; so did plants of cv. Afghanistan (NGB2150) as well) (Table 1). From among the other lines analyzed, there were several with decreased or increased numbers of nodules formed by either the RCAM1026 (nodX⁻) strain or A1 (nodX⁺) strain, but these were not included in subsequent analysis.

Allelism test between newly discovered and previously known Sym2^A-lines

In order to determine whether the symbiotic phenotype of selected pea accessions was caused by the same gene as in the previously known “Afghan” lines (i.e., the Sym2 gene), an allelism test was conducted. F₁ seeds obtained from the cross of previously reported “Afghan” lines NGB2150, K-6883 and K-6047-2 with lines selected from VIR gene bank, along with seeds of parental lines, were planted in pots with sterile sand and inoculated with the nodX⁻ rhizobia strain RCAM1026 (see Materials and Methods). Additionally, two seeds of cv. Finale (characterized by the broad “European” symbiotic specificity and thus able to interact with the nodX⁻ strain RCAM1026) were planted in each pot to confirm the success of the inoculation.

After 4 weeks of vegetation, F₁ hybrids, like the plants of parental lines, did not form nodules with the nodX⁻ strain RCAM1026. Plants of cv. Finale in the same pots formed, on average, about 40 nodules per plant (see Table S5). The lack of nodules in F₁ plants indicates the allelism of genetic determinants of symbiotic phenotype in newly-described and previously known “Afghan” pea lines. Thus, new sources of Sym2^A alleles have been identified: the lines K-1878, K-3374, K-3821, K-4902, and K-6559.

Sequencing of the first exon of LykX in the selected lines of pea

In our previous work (Sulima et al., 2017), a new LysM-RLK gene LykX (PsLykX) was identified as a promising candidate for the elusive gene Sym2, as it demonstrated the unique allelic state found only in two pea lines with the “Afghan” phenotype which leads to substitution of three amino acids in its corresponding protein: Q44R, N45Y, and A76D, all in the first LysM module of the receptor part. Here we sequenced the first exon of LykX (corresponding to the receptor domain of the putative protein) in the newly selected “Afghan” lines from the VIR gene bank, several lines with “European” phenotype from our laboratory collection, cv. Iran characterized by the leaky temperature-dependent “Afghan” phenotype (Lie, 1971; Kozik et al., 1996a), and a wild relative of garden pea, Pisum fulvum.

It was revealed that in the majority of lines with “Afghan” phenotype the 1^st exon of LykX has the same allelic state that was discovered previously (Sulima et al., 2017). In lines K-3821 and K-3374, however, LykX demonstrated another allelic state with unique substitutions observed neither in “European” nor in other “Afghan” lines. These lines were called “Tajik”, according to the place of origin of line K-3821, which was identified first. Taken this new find into account, we suspected that the results of our initial allelism test might be ambiguous, since the similar non-nodulation phenotype in F₁ would be observed if the block of nodulation in the “Tajik” lines is caused by genetically dominant determinant independent of Sym2. In order to test that new hypothesis, the nodulation trait in F₂ progeny derived from the cross between “Tajik” line K-3821 and the typical “Afghan” line NGB2150 was analyzed in the inoculation experiment similar to the one described above. Of 119 F₂ plants, none demonstrated normal nodulation with RCAM1026 strain regardless of the allelic state of LykX (homozygous “Afghan” allele, homozygous “Tajik” allele, heterozygote), while control plants of cv. Finale formed, on average, 65.3 ± 20 (m ±SD, n = 28) nodules per plant (see Table S6 and Fig. S1). This suggests that our initial interpretation of allelism test results was correct.

The nucleotide substitutions unique for “Afghan” lines lead to the replacement of three amino acids in the protein in comparison to the “European” lines (Table 2). “Tajik” lines seem to be much more similar to the “European” ones, with only one unique amino acid substitution in the protein: R75P. It is important to note that all these characteristic substitutions are concentrated in the first LysM module (LysM1) of the receptor domain, which could indirectly indicate the importance of this particular module for the functioning of the LykX protein (Fig. 1).

The plants of cv. Iran were found to carry the same allele of LykX as the cv. Afghanistan (NGB2150), which corresponds with the idea that their phenotype is determined by the same gene (Kozik et al., 1995; Kozik et al., 1996a). Interestingly, the “European” allele of LykX was found in Pisum fulvum line 701, the wild species closely related to Pisum sativum.

Phenotype confirmation

According to previous descriptions (Geurts et al., 1997; Zhukov et al., 2008), “Afghan” peas are unable to form normally developed infection threads while inoculated with the nodX⁻ strain of rhizobia. Bearing this in mind, we performed the microscopic analysis of roots of “Tajik” lines in order to characterize their symbiotic phenotype. Seedlings of the “Tajik” line K-3821 and the “Afghan” line NGB2150, along with the “European” cv. Caméor, were inoculated with gusA-marked strain CIAM1026. 4 weeks after inoculation, the root fragments were harvested, treated with staining buffer and observed under a light microscope. Both lines demonstrated normal root hair curling and formation of microcolonies, but infection was aborted at this stage so no infection threads were observed for neither “Afghan” nor “Tajik” lines on 25 cm of roots (Figs. 2B, 2C). Control plants of cv. Caméor developed normal infection threads (Fig. 2A). Accordingly, plants of cv. Caméor formed 81.6 ± 15.2 (m ±SD, n = 10) nodules, while NGB2150 and K-3821 formed 0.2 ± 0.6 and 0.6 ± 1.3 nodules, respectively. Thus, we confirmed that “Afghan” and “Tajik” pea lines have an identical symbiotic phenotype, which is consistent with the alterations in the first exon sequences of LykX gene as compared to the “European” variant.