Subcellular localization and differential expression provide insights into the putative function of the nematode resistance gene Hs4

TL;DR

This study explores the Hs4 gene in beet plants and related species to understand its role in resistance to nematodes.

Contribution

The study identifies Hs4 homologs in Beta and Patellifolia species and reveals differences in expression and sequence.

Findings

Patellifolia species have highly similar Hs4 homologs with unique exonic variations.

BvHs4 homologs in Beta vulgaris are expressed mainly in leaves, unlike Hs4 in roots.

No significant response to nematode inoculation was observed in Hs4 expression patterns.

Abstract

Plant-parasitic nematodes are economically important threats to global crop production. The beet cyst nematode (Heterodera schachtii) is a crucial pest in sugar beet (Beta vulgaris ssp. vulgaris). While all species of the genus Beta are highly susceptible, the three species of the beet wild relative genus Patellifolia are entirely resistant. Recently, we cloned the Hs4 gene from P. procumbens, which confers complete resistance. In this study, we aimed to determine whether putative Hs4 orthologs exist in Beta and Patellifolia species. The Hs4 gene consisted of 4999 bp, with six exons and five introns. Patellifolia species contain highly similar Hs4 homologs. Single nucleotide polymorphisms and insertions/deletions between accessions and species could be detected. We found an exonic integration of three bases, resulting in the addition of one amino acid. Interestingly, this variant was present in single accessions of all three Patellifolia species. Beta vulgaris ssp. vulgaris contains an Hs4 homolog (BvHs4) with 60 % protein identity to Hs4. BvHs4 homologs were present in all Beta species analyzed. Further, we examined the expression patterns of Hs4 and BvHs4 homologs. While Hs4 homologs from Patellifolia species are strongly expressed in roots, BvHs4 homologs are expressed mainly in leaves. When the spatio-temporal expression of Hs4 was examined, no response to nematode inoculation was observed. These results are highly relevant for searching for functional Hs4 alleles and breeding nematode-resistant varieties.

The online version contains supplementary material available at 10.1038/s41598-026-40666-5.

Figures

Structure of *Hs4* (2022 bp) and its *Beta* homolog *BvHs4* (4342 bp) and the intron-less open reading frame (ORF) derived from them. Exons are shown in green (*Hs4*) and orange (*BvHs4*). The primer positions to amplify a region of both genes simultaneously (F7/R7) or the whole genes (*Hs4*: F5/R6; *BvHs4*: F1/R0) are depicted by arrowheads. The figure was generated using the Illustrator for Biological Sequences software^[49](#CR49)^.Fig. 2PCR amplicons of *Hs4*- and its *BvHs4*-homologous genes separated by agarose gel electrophoresis. Using the same set of primers, a single 1217 bp fragment

A phylogenetic tree was created with polypeptide sequences from different *Patellifolia* and *Beta* accessions. The tree was constructed using the maximum likelihood method with MEGA11^[38](#CR38)^. Bootstrapping was conducted with 500 replicates. Quinoa (*C. quinoa*), spinach (*S. oleracea*), and mung bean (*V. radiata*) were used as outgroups.

Expression levels of *Hs4* and its homologs in leaves (L) and roots (R) of resistant *Patellifolia* (A) and susceptible *Beta* (B) accessions as determined by RT-qPCR (primer: AK_F7/AK_R5 for *Hs4* and AS_BvHs4_F1/AS_BvHs4_R7 for *BvHs4*; reference gene: *GAPDH*). (A) *Hs4* expression in fifteen *Patellifolia* accessions and three resistant translocation lines. In all samples, the expression was higher in the roots. (B) Expression of *BvHs4* orthologs across five different *Beta* species. Due to a lack of biological replicates, the expression data of two different *B. macrorhiza* accessions we

Expression levels of *Hs4* in *P. procumbens* (primer: AK_F7/AK_R5) and *BvHs4* in *B. vulgaris* ssp. *vulgaris* (primer: AS_BvHs4_F1/AS_BvHs4_R7) determined by RT-qPCR. *GAPDH* was used as a reference. i: inoculated; ni: non-inoculated; R: root; H: hypocotyl; C: cotyledon; L: leaf. A Kruskal-Wallis test (p < 0.05) was performed for the *P. procumbens* samples, and significant differences between groups were calculated using the Hochberg-corrected Dunn post-hoc test (p < 0.05). Statistical significances between *B. vulgaris* ssp. *vulgaris* samples were calculated using a one-way ANOVA (p < 0.

Full text

Introduction

Sugar beet, fodder beet, garden beet, and chard are diploid species (B. vulgaris ssp. vulgaris, 2n=2x=18) belonging to the first section of the genus Beta, together with the wild species B. macrocarpa, B. patula, B. vulgaris ssp. adanensis, and B. vulgaris ssp. maritima. The second section, Corollinae, comprises the Eastern Mediterranean and Southwestern Asian species B. corolliflora, B. intermedia, B. lomatogona, B. macrorhiza, B. nana, and B. trigyna^1^. The genus Patellifolia was formerly described as Beta section Procumbentes Ulbr. but was revised to constitute a separate genus^2,3^. It diverged from the genus Beta around 25.3 million years ago^1^. The genus forms the third gene pool and comprises the closely related diploid species P. procumbens and P. webbiana, as well as the tetraploid P. patellaris^4,5^.

While cultivated beets have a narrow genetic makeup, the wild Patellifolia species are genetically heterogeneous and represent an important source of resistance against diverse pathogens like the Beet Necrotic Yellow Vein Virus (vectored by the fungus Polymyxa betae) and the beet cyst nematode (BCN) Heterodera schachtii^6^. Therefore, the different species were employed extensively in crossing experiments with various Beta species. Still, the lack of chromosome homology and pronounced hybrid sterility severely hindered their use as genetic resources.

Sugar beet is highly susceptible to BCN, which attacks the roots and induces a nutrient cell called syncytium^7,8^. After passing through four larval stages, the maturing female forms a cyst filled with eggs. The syncytium is the only feeding source throughout the nematodes’ life cycle, and its development depends on its integrity. The BCN is the most important pest in beet cultivation^9,10^. There is no complete resistance within the Beta genus. In contrast, the Patellifolia species are entirely resistant to the pathogen. However, transferring the resistance genes is extremely difficult due to crossing barriers. After decades of research, only a few resistant beet lines have been selected that carry translocations from P. procumbens and P. webbiana^11,12^.

Recently, Hs4 was cloned from a Patellifolia translocation attached to the sugar beet chromosome 9^13^. Knocking out Hs4 turned the resistant sugar beet variety Nemata into a highly susceptible one. In contrast, Hs4 overexpression in a susceptible B. vulgaris ssp. vulgaris background resulted in varying resistance depending on the gene expression level. Hs4 was predicted to be an Endoplasmic Reticulum (ER)-bound rhomboid-like protease of 210 amino acids. Most plant rhomboid proteases contain seven transmembrane domains (TMDs). They are characterized by a serine protease catalytic unit that cleaves the substrates inside the membrane. Rhomboid proteins can also lack the catalytic residues, rendering them inactive proteases, and are then classified as iRhoms^14,15^. Hs4 itself is predicted to be a transmembrane protein and contains the typical catalytic residues^13^. A B. vulgaris ssp. vulgaris homolog (BvHs4) showed 60 % polypeptide similarity to Hs4. It differs from Hs4 by 102 additional amino acids at the N-terminus, and it lacks the leader sequence that directs Hs4 to the ER. Therefore, differences in subcellular localization of the two homologs are highly likely.

In this study, we aimed to determine whether functional orthologs of Hs4 exist in other Patellifolia and Beta species. We reasoned that the highly conserved structure of the Hs4-like proteins points to their function as BCN resistance genes in other species. We found Hs4 orthologs with high sequence identity in all Patellifolia species, whereas BvHs4-homologous genes in Beta species were highly divergent. Because all Beta species are susceptible, we reasoned that the lack of function could be due to the ectopic expression. Therefore, we determined the local expression of Hs4 homologs in a broad range of species across various tissues and under nematode infestation. We found that Hs4 was the highest expressed in roots, while BvHs4 expression peaked in leaves. The results suggest that Hs4 and its Patellifolia orthologs are the only functional nematode-resistance genes. These results shed new light on the function of Hs4 and have implications for breeding nematode-resistant Beta crops.

Results

Sequence variations between Hs4 and its homologs from Patellifolia and Beta species

The published Hs4 sequence comprises 5664 bp with five exons and five intron^13^. This sequence contained a large stretch (964 bp) of unknown nucleotides in the first intron. We sequenced the unknown region to 107 bp, thereby reducing the genomic region of Hs4 to 4999 bp (Supplementary Figure 1). The composition of the cDNA and the resulting protein remained unchanged. In the following, 'Hs4’ refers to the 4999 bp long sequence.

Using the primer combination AS_F7/AS_R7, we detected Hs4 amplicons in all P. procumbens, P. webbiana, and P. patellaris accessions analyzed. The P. procumbens, P. webbiana, and P. patellaris orthologs were named PpHs4, PwHs4, and PpatHs4, respectively (Supplementary Table 1). Additionally, we screened the sugar beet translocation lines TR363 and TR520, as well as the hybrid variety Nemata, using the same primer set. The primers also bind to the exons of the B. vulgaris ssp. vulgaris Hs4 homolog (Figure 1), so double bands could be observed after agarose gel electrophoresis (Figure 2A). The lower and upper bands are P. procumbens (Hs4, 1217 bp) and B. vulgaris ssp. vulgaris (BvHs4, 1499 bp) amplicons, respectively.Fig. 1. Structure of Hs4 (2022 bp) and its Beta homolog BvHs4 (4342 bp) and the intron-less open reading frame (ORF) derived from them. Exons are shown in green (Hs4) and orange (BvHs4). The primer positions to amplify a region of both genes simultaneously (F7/R7) or the whole genes (Hs4: F5/R6; BvHs4: F1/R0) are depicted by arrowheads. The figure was generated using the Illustrator for Biological Sequences software^49^.Fig. 2PCR amplicons of Hs4- and its BvHs4-homologous genes separated by agarose gel electrophoresis. Using the same set of primers, a single 1217 bp fragment was amplified in all Patellifolia accessions (A), while a 1499 bp fragment was generated in most Beta accessions (B). The resistant translocation lines (TR363 and TR520) and the resistant cultivar Nemata display both Hs4 and BvHs4 fragments. M: 1 kb ladder; N: negative control.

We expected genes identified from P. procumbens to exhibit the highest sequence identity to Hs4. Due to the close relatedness between P. procumbens and P. webbiana, we expected a high identity between PpHs4 and PwHs4 sequences, while PpatHs4 should display less identity to Hs4. As a first result, Hs4 was conserved across all Patellifolia species. Apart from single-nucleotide polymorphisms (SNPs) and short insertions-deletions (InDels) (Supplementary Table 1), the protein-coding exons 2-6 (Figure 1, Supplementary Figure 2) were highly conserved. The overall sequence identity was 95-99 % (Table 1). As expected, the Hs4 sequences from the translocation lines (Hs4TR363/TR520) and Nemata (Hs4Nemata) were 100 % identical to Hs4, thus confirming their origin. Moreover, PwHs4 sequences showed the highest identity to Hs4 (99 %). Identity values between Hs4 on one side and PpHs4 and PpatHs4 were lower (95-99%) (Table 1). These sequences exhibited insertions and deletions (InDels) extending the sequence by 29-30 bp (Table 1). Accordingly, most SNPs were shared between P. procumbens and P. patellaris accessions. Four InDels were located within intronic regions. Moreover, a three base pair insertion (position 1302, exon 5) led to the addition of one amino acid Supplementary Figure 4, Supplementary Figure 8, Supplementary Table 3). This insertion was present in four P. procumbens, two P. webbiana, and all P. patellaris accessions, as well as in the P. procumbens draft genome sequence.Table 1DNA sequence identities of different Hs4 sequences compared to the Hs4 reference gene of 2022 bp. The Hs4 orthologs were amplified from P. procumbens (Pp), P. webbiana (Pw), or P. patellaris (Ppat). The subscripted suffix denotes the seed code or other identifiers. Identities were calculated with the blastn suite^40^.**Query sequenceLengthIdentity [%]**Gaps [%]**Hs420221000PpHs4_1008102051951PpHs41008202052951PpHs41008212032970PpHs4IPK4192022990PpHs4IPK9512025990PwHs41000022022990PwHs41008322024990PwHs41008332022990PwHs4IPK5262025990PwHs4IPK9272025990PpatHs41000122052951PpatHs49300652052951PpatHs4IPK102052951PpHs4Draft2025990Hs4TR36320221000Hs4TR52020221000Hs4Nemata_20221000

BvHs4 homologs are present in a broad set of Beta species

Based on the phylogeny of the genus Beta, we expected BvHs4-like sequences in species of the genus Beta. We used the primer combination AS_F7/AS_R7 to amplify 1499 bp of BvHs4 (NCBI XM_010669575.1, LOC104884871) in 14 Beta accessions (Figure 2B). As a result, amplicons of the expected size were visible in 12 accessions (Supplementary Figure 5A). The B. macrorhiza and B. patula amplicons were substantially shorter (BmrHs4, ca. 1300 bp; BpHs4, ca. 1300 bp). B. corolliflora (BcHs4) exhibited an additional amplicon ca. 1300 bp in size (Figure 2B). Sequencing the smaller B. corolliflora fragment revealed no homology to any known gene from B. vulgaris ssp. vulgaris (data not shown), while the larger BcHs4 fragment was composed of different sequences. BcHs4A was a perfect match to BvHs4 from the reference genome sequence. In contrast, another sequence (BcHs4B) showed several SNPs, small InDels, and a prominent insertion of 97 bp (after 1522 bp of the BvHs4 RefBeet sequence) matching a BvHs4 ortholog from the tolerant B. vulgaris ssp. vulgaris genotype U2Bv. BcHs4B and BvHs4U2Bv displayed a 312 bp deletion (after 2713 bp of BvHs4) (Supplementary Figure 5B). We then used a BvHs4-specific primer combination (AS_BvHs4_F1/AS_BvHs4_R0, Figure 1) to amplify the entire BvHs4 open reading frame of 4342 bp from six Beta accessions (Supplementary Figure 6). Here, we found that the 97 bp insertion and the 312 bp deletion are also present in BpHs4 and BmrHs4. Further, they contained the same small InDels found in BcHs4B. BpHs4, BcHs4B, and BvHs4U2Bv shared single SNPs but not BmrHs4 (Supplementary Figure 4). Excluding BmrHs4, which shows exonic deviations of up to 15 bp, all sequence variations larger than three base pairs lie within intronic regions (Supplementary Figure 3B, Supplementary Figure 5B).

In summary, the BvHs4 homologs from species of the genus Beta are highly identical (sequence identity values 97-100%, Supplementary Figure 3), and the intron/exon structures are the same as in BvHs4. None of the genes had a higher identity to Hs4 than BvHs4 (Table 2). Thus, none of the Beta Hs4 homologs will likely function as a nematode resistance gene.Table 2DNA sequence identities of different BvHs4 homologs compared to BvHs4 (4342 bp) and Hs4 (2022 bp). The BvHs4 homologs’ suffix denotes the seed code or other identifiers. Identities were calculated with the blastn suite^40^.**Query sequenceLengthcompared to BvHs4compared to Hs4Identity [%]**Gaps [%]**Identity [%]**Gaps [%]**BaHs44341990774BcHs4_A_4343990764BcHs4_B4168970774BmHs44347990774BmHs4Bmar1.04369990774BmcHs44332990774BmrHs49600184166903764BpHs44166970774BpHs4Bpat-1.04300970774BvHs4EL10_24334990774BvHs4W357B43421000764BvHs4U2Bv_4165970774Hs42022764--

Phylogenetic analysis indicates distinct clades of Hs4- and BvHs4-like polypeptides

We compared the open reading frames of all Beta and Patellifolia accessions together with the rhomboid-like proteins of quinoa (Chenopodium quinoa, LOC110702170 and LOC110705623), mung bean (Vigna radiata, LOC106763922), and spinach (Spinacia oleracea, LOC110775063) as outgroups.

Apart from minor amino acid substitutions, the polypeptides from the Patellifolia species showed high conservation (Supplementary Figure 8). A single amino acid insertion (after Hs4 position 132) compared to Hs4 was present in various accessions of all three Patellifolia species (Supplementary Figure 7).

The sequence motif ‘LLRDRCPDN’ was suggested to be Hs4-specific due to its absence from BvHs4^13^. Interestingly, the complete motif comprising 13 amino acids is absent from the Beta homologs (Supplementary Figure 8). Surprisingly, the exact motif could only be found in four out of thirteen Patellifolia sequences, the translocation lines, and Nemata (Supplementary Table 4). Because all Patellifolia polypeptides carried two more amino acids at their 3’-end, which were absent from the Beta polypeptides, the Patellifolia-conserved motif could be revised to ‘CPDNKE’.

The Beta polypeptides were highly similar to each other and to the outgroup proteins (Supplementary Figure 8). The minor sequence variations of the BmrHs4 sequence may result in a non-functional polypeptide because of a stop codon at position 141 (Supplementary Figure 8).

Across the two genera, motives of up to fifteen amino acids (aa) were conserved. A prominent difference was the addition of roughly 100 aa at the N-terminus of all Beta proteins, resulting from an additional exon (Figure 1).

Then, we generated a maximum likelihood tree based on the polypeptide sequences (Figure 3). The BvHs4 homologs clustered closest to the outgroup proteins, and the Patellifolia polypeptides were grouped further away. The Beta species formed a distinct clade, although it was supported only by a low confidence level of 51%, except for B. macrorhiza. The divergence point towards the Patellifolia polypeptides was supported by a high bootstrap value of 100, indicating that this clade is distinct from the Beta and the outgroup proteins (Figure 3). Conclusively, the original Hs4 polypeptide grouped with the translocation lines and Nemata, which were all derived from the same P. procumbens source. Additionally, one PpHs4 and three PwHs4 polypeptides are part of that subclade. Interestingly, one PpHs4 and the remaining PwHs4 polypeptides formed a distinct subclade. The polypeptide derived from the P. procumbens draft genome was somewhat distantly related to Hs4. Unexpectedly, three P. procumbens accessions grouped with the P. patellaris accessions.Fig. 3A phylogenetic tree was created with polypeptide sequences from different Patellifolia and Beta accessions. The tree was constructed using the maximum likelihood method with MEGA11^38^. Bootstrapping was conducted with 500 replicates. Quinoa (C. quinoa), spinach (S. oleracea), and mung bean (V. radiata) were used as outgroups.

Protein localization predictions underpin the diversification of Hs4 homologs across species of the genus Beta

Hs4 was predicted to be localized within the ER membrane^13^ (Table 3). We analyzed the protein structure with DeepTMHMM^16^, which predicts that Hs4 has six transmembrane domains (Supplementary Table 5). The addition of valine in PpHs4_100810_ did not alter its predicted subcellular localization. Contrarily, the probability of an ER localization increased from 0.7973 for Hs4 to 0.8184 for PpHs4_100810_ (Table 3). However, the additional valine led to a slight conformation change. Five amino acids assigned to a transmembrane domain in Hs4 were now predicted to be part of a loop located in the lumen, and hence, the proximate fourth transmembrane domain was slightly decreased in size (Supplementary Table 5).Table 3. Predicted localization and membrane association for Hs4 and its homologs from Patellifolia and Beta species compared to the A. thaliana rhomboid protease AtRBL11. The probability values were calculated using DeepLoc2.1^42^.Hs4****PpHs4100810BvHs4****AtRBL11LocalizationPrediction probabilitycytoplasm0.06670.06950.07070.0796nucleus0.07610.07900.05660.0689extracellular0.19020.17120.02090.0218cell membrane0.13340.13300.05050.0623mitochondrion0.30800.27820.27480.3320plastid0.17210.20450.91580.9346endoplasmic reticulum0.79730.81840.22470.1656lysosome/vacuole0.40650.43350.05810.0521Golgi apparatus0.40890.42830.06730.0372peroxisome0.02740.03710.10040.1345Membrane associationperipheral0.15300.15800.28500.3000transmembrane0.96700.96800.95800.9610lipid anchor0.07400.07700.05900.0550soluble0.14300.14500.12200.1130

BvHs4, in turn, is predicted to be a plastid-localized transmembrane protein exhibiting six transmembrane domains (prediction value 0.9158) (Table 3). The main conformation difference lies in its long cytosolic N-stretch, which is much shorter in Hs4 and PpHs4_100810_ (Supplementary Table 5). Interestingly, using the AlphaFold Protein Structure Database ^17,18^, we found that Arabidopsis contains an Hs4 homolog with 57% and 71% polypeptide similarity to Hs4 and BvHs4, respectively. The protein is classified as a rhomboid-like protease 11 (AtRBL11, UniProt identifier: Q84MB5) and, like BvHs4, located in the plastids (Table 3). It also has six transmembrane domains (Supplementary Table 4).

Hs4 and its homologs are differentially regulated between Patellifolia and Beta species

We measured the transcriptional activities of Hs4 and its homologs in Patellifolia and Beta species with and without nematode infection. The translocation lines and the Patellifolia species displayed up to 26.6-fold higher Hs4 expression in roots than in leaves (Figure 4A). Contrastingly, across all Beta species, the BvHs4 homologs showed up to 6x higher expression in leaves (Figure 4B). The leaf-to-root expression ratio was similar across the different species.Fig. 4. Expression levels of Hs4 and its homologs in leaves (L) and roots (R) of resistant Patellifolia (A) and susceptible Beta (B) accessions as determined by RT-qPCR (primer: AK_F7/AK_R5 for Hs4 and AS_BvHs4_F1/AS_BvHs4_R7 for BvHs4; reference gene: GAPDH). (A) Hs4 expression in fifteen Patellifolia accessions and three resistant translocation lines. In all samples, the expression was higher in the roots. (B) Expression of BvHs4 orthologs across five different Beta species. Due to a lack of biological replicates, the expression data of two different B. macrorhiza accessions were pooled in this analysis. Bars represent the standard error of the mean (SEM).

Next, we analyzed the expression in the presence and absence of nematode infection. After infection with H. schachtii, no females or cysts were observed on P. procumbens roots, whereas B. vulgaris ssp. vulgaris roots were heavily infected. Female numbers ranged between 54 and 119 (mean: 83.4). P. procumbens showed the highest Hs4 expression in roots across all sampling points (Figure 5). The root/leaf ratio did not differ much between non-inoculated and inoculated plants. Upregulation after nematode inoculation was not significant (Figure 5). In the young sugar beet plants, BvHs4 expression was higher in roots than in leaves. However, shortly before inoculation, BvHs4 expression in leaves was 4.1x higher than in roots. The BvHs4 expression rates did not respond to nematode infection.Fig. 5. Expression levels of Hs4 in P. procumbens (primer: AK_F7/AK_R5) and BvHs4 in B. vulgaris ssp. vulgaris (primer: AS_BvHs4_F1/AS_BvHs4_R7) determined by RT-qPCR. GAPDH was used as a reference. i: inoculated; ni: non-inoculated; R: root; H: hypocotyl; C: cotyledon; L: leaf. A Kruskal-Wallis test (p < 0.05) was performed for the P. procumbens samples, and significant differences between groups were calculated using the Hochberg-corrected Dunn post-hoc test (p < 0.05). Statistical significances between B. vulgaris ssp. vulgaris samples were calculated using a one-way ANOVA (p < 0.05), followed by the Tukey HSD Test. Standard deviations are given as SEM. The letters above the bars indicate statistically significant differences between group means; groups sharing the same letter are not significantly different.

Discussion

In this study, we revised the genomic sequence of the nematode resistance gene Hs4^13^. We studied Hs4 homologs in various Patellifolia and Beta species, examining their expression in different tissues with and without H. schachtii infestations.

Hs4 has been described to span 5664 bp^13^. After sequencing an unresolved region of 964 bp in the first intron, we found that Hs4 comprises 4999 bp. Kumar et al. claimed the gene to consist of five exons and introns each, disregarding the untranslated regions (UTRs) they had annotated^13^ (see Supplementary Figure 1). However, as the UTR has also to be considered in exon labelling^19^, Hs4 consists of six exons and five introns, according to our studies.

The Hs4 gene was predicted to encode a rhomboid protease^13^. The smallest catalytically active unit of rhomboid proteases consists of six transmembrane domains (TMD). These basic TMDs are primarily found in bacteria and, to a lesser extent, in eukaryotes, although eukaryotes often have a seventh TMD^14^. However, many eukaryotic rhomboids belong to the secretase subfamily, which can be divided into two clades: secretase-A rhomboids exhibit seven TMDs, and secretase-B rhomboids contain only the core region of six TMDs. Furthermore, these two clades are distinguished by distinct sequences surrounding the catalytic serine S106. In the secretase-A clade, proteins contain a highly conserved GxSxGVYA, while B-class secretases show a less stringent GxSxxxF sequence^20^. When we analyzed Hs4 using DeepTMHMM, we found the protein to consist of six TMDs, which is conclusive in its classification as a rhomboid protease. Further, Hs4 (and all related proteins analyzed in the current study) harbors the GxSxxxF sequence, indicating its affiliation with the secretase-B clade.

Species of the genus Patellifolia are highly resistant to H. schachtii ^21,22^. According to our hypothesis, we detected highly similar homologs in all Patellifolia species. Nevertheless, sequence variations could be observed between and within the species. Sequences perfectly matching Hs4 were detected in the translocation lines and the hybrid variety Nemata, reassuring the origin of the Hs4 sequence. None of the Patellifolia accessions sequenced harbored a homolog with 100 % sequence identity to Hs4. However, PpHs4IPK419, PpHs4IPK951, and all PwHs4 homologs exhibited the highest similarity to Hs4 (99 %). Three base pairs were integrated into exon 5, adding valine to the Hs4 protein present in all three species. However, we found that the additional amino acid did not change either the predicted localization of Hs4 to the ER or its TMD structure.

The phylogenetic relationships among the different Patellifolia accessions have been discussed for some time. While the diploid P. procumbens and P. webbiana were previously considered two extreme ecotypes of a single species^23,24^, recent phylogenetic analyses based on plastome data have pointed to a clear separation into two distinct, albeit closely related species. The tetraploid P. patellaris is more distantly related to the two diploid Patellifolia species^5^. In our study, two PpHs4 and all PwHs4 homologs showed high sequence similarity to Hs4. Nevertheless, three PpHs4 and the PpatHs4 homologs showed several deviations from Hs4. Considering the relationship between the Patellifolia species, similarities between P. procumbens and P. patellaris appear peculiar. However, though P. procumbens and P. webbiana are more closely related, the genus itself is very diverse. Depending on the sampling locations, plant morphologies differ even within the same species^25^. This was underpinned by molecular marker studies demonstrating high genetic diversity within P. procumbens^26^.

Sequence variations between Hs4 homologs from different Patellifolia species have been recently described^27^. Analyzing short haplotype loci polymorphic between different species, they identified 0.16-3.36 variants between Hs4 orthologs. Of these loci, 20 % contained single- or multi-nucleotide polymorphisms, and the remaining 80 % PAVs. Notably, Reeves and Richards identified differences in exonic variants at several Hs4 loci across accessions within the same species^27^, which aligns with our findings. In addition, we identified a major insertion-deletion (InDel) in exon five of several Patellifolia accessions, resulting in the addition of a valine, which had not been reported in the study by Reeves and Richards^27^. Instead, they reported four short haplotype loci with differing major variants in or around exon five, which, in turn, could not be detected in our study. Nevertheless, these sequence variations do not alter the function of Hs4 as a gene conferring resistance to beet cyst nematodes.

Due to the high similarities between Hs4 and PwHs4 from P. webbiana, we speculate that Hs4 might originate from P. webbiana. A second resistance gene is located on chromosome 7 of P. procumbens and P. webbiana^28^. Using the primers matching Hs4, no other Hs4 ortho- or homologous sequence could be amplified from the Patellifolia species, suggesting no Hs4 paralog on chromosome 7.

The genera Patellifolia and Beta have diverged 25 mya^1^. The general sequence similarity between sugar beet and Patellifolia genomes is 75 %^27^. Apart from BvHs4, we detected a highly similar homolog in the wild progenitor of sugar beet, B. vulgaris ssp. maritima. As expected, the tetraploid species B. corolliflora^29^ has two BvHs4-homologous sequences that varied at several sites. While BcHs4A resembled BvHs4, BcHs4B had a prominent insertion of 97 bp and a deletion of 312 bp, both of which were also present in B. patula and B. macrorhiza. Matching these results, B. macrorhiza is believed to be a parental species of B. corolliflora^29^. The insertion is similar to the BvHs4 homolog from the nematode-tolerant sugar beet genotype U2Bv. This genotype is characterized as nematode-tolerant based on the expression of two BvNLP7 genes that are located on chromosome 5^30^ and which do not show any resemblance to the nematode resistance gene Hs4, whose Beta homolog is located on chromosome 2. Both BcHs4 variants lie in the intron, so they do not affect the resulting protein. Furthermore, these sequences did not exhibit higher similarity to Hs4 than BvHs4 and are therefore not expected to contribute to nematode resistance.

We were interested in the expression patterns of the Hs4 homologs from other species. Hs4 is highly expressed in roots and less in leaves. Notably, we detected the highest expression levels in the translocation line TR520 and the hybrid variety Nemata derived from it. A second translocation line, TR363, did not show enhanced Hs4 expression levels. TR363 harbors a much smaller translocation than TR520^13^ and important regulatory cis-elements might be lacking in its sequence, leading to differences in expression intensities. TR363 never gained importance in beet breeding due to its inferior yield and quality characteristics. Both translocation lines and the varieties derived therefrom are entirely resistant to H. schachtii. BvHs4, on the other hand, is expressed strongly in leaves. Its protein is predicted to target the chloroplast, similar to its Arabidopsis homolog AtRBL11^31^. Its strikingly different expression pattern and low sequence homology to Hs4, and the predicted subcellular location, indicate that BvHs4 and its homologs from other Beta species do not function as nematode-resistance genes.

We speculate that Hs4 has acquired a new function in evolution, coupled with its localization in the ER membrane, which provides a typical example of the neo-functionalization of a gene. In conclusion, the significant structural differences between Hs4 and its Beta homologs, as well as their distinct expression patterns, preclude a targeted modification of their function, such as by genome editing to convert them into resistance genes. Due to the poor agronomic performance of the nematode-resistant translocation lines, the expression of the Hs4 gene after transformation into sugar beet is the only realistic solution to breed resistant varieties.

Materials and methods

Plant material and growth conditions

Fourteen Beta wild beet accessions, fifteen Patellifolia accessions, three resistant sugar beet translocation lines, and a susceptible sugar beet accession were grown in a greenhouse or a climate chamber, respectively, under long-day conditions (LD; 16 h light / 8 h dark). (Supplementary Table 1).

DNA isolation and PCR

Samples of different plant tissues were taken at several points, frozen in liquid nitrogen, and stored at -70°C until further usage. Following the grinding of the tissues, genomic DNA was isolated using the NucleoSpin Plant II kit (Macherey-Nagel, Düren, Germany) following the manufacturer’s instructions or the plant DNA mini preparation protocol^32^. Before further use, the genomic DNA was checked on a 1% agarose gel (90V, 30 min or 80V, 12 min).

A non-sequenced (unresolved) region in the first Hs4 intron was amplified from the variety Nemata (Supplementary Table 1) with a Taq polymerase (Biozym Scientific GmbH, Hessisch Oldendorf, Germany) using the primer combination AK_F2/AS_R5 (Supplementary Table 2).

To amplify putative homologs of Hs4 and BvHs4 simultaneously, we designed a primer set (Figure 1) binding to a DNA region encoding a highly conserved polypeptide sequence between Hs4 and BvHs4 (Supplementary Figure 2). The primer pair AS_F7/AS_R7 was expected to amplify 1217 bp and 1499 bp of the Hs4 and BvHs4 genes (Supplementary Table 2). Taq polymerase (Biozym Scientific GmbH) or Phusion High-Fidelity polymerase (ThermoFisher Scientific, Waltham, MA, USA) was used to amplify the genomic DNA of Hs4 and its homologs from other species. The primer combination AK_F5/AK_R6 (Supplementary Table 2) was used for Hs4 gene amplification, whereas AS_BvHs4_F1/AS_BvHs4_R0 was used to amplify the Beta homolog. The PCR products were analyzed on a 1% agarose gel (90V, 30-40 min).

Plasmid cloning and sanger sequencing

Using the CloneJET PCR Cloning Kit (ThermoFisher Scientific), DNA was cloned into the pJET1.2/blunt cloning vector according to the manufacturer’s instructions and transformed into competent Escherichia coli (strain DH5α; DNA Cloning Service eK, Hamburg, Germany) via the heat-shock method^33^. Following overnight incubation at 37°C on LB plates supplemented with ampicillin (50 µg/mL), colonies were screened for the presence of the plasmid by PCR. Positive colonies were cultured overnight in 5 mL LB with ampicillin on a shaker. Plasmids were then isolated using the NucleoSpin Plasmid QuickPure Kit (Macherey-Nagel) and used for PCR in a 1:1000 dilution.

Amplified gene fragments or whole genes were Sanger sequenced on an Applied Biosystems 3730xl DNA Analyzer (ThermoScientific, via IKMB, Kiel University) using a set of different primers (Supplementary Table 2).

Phylogenetic analysis

Sanger sequences were aligned to the reference sequences using the CLC Main Workbench 20 (Qiagen, Hilden, Germany). As a reference, the Hs4 gene sequence derived from TR520 by Kumar et al.^13^ was used for the Patellifolia sequences, and the genomic RefBeet-1.2.2^34^ sequence of the rhomboid-like protein 11 from sugar beet (NCBI Reference Sequence XM_010669575.1, LOC104884871) was deployed for all Beta sequences. After the alignment, specific Hs4 or BvHs4 sequences were generated for each accession sequenced. The CLC-incorporated tool was used for protein translation, generating the corresponding polypeptide sequences.

Furthermore, we searched reference sequences from the susceptible sugar beet line EL10_2^35^, the red beet W357B v1.0^36^, and the nematode-tolerant B. vulgaris ssp. vulgaris line U2Bv^30^. Reference sequences from the B. patula accession Bpat-1.0^37^, the B. vulgaris ssp. maritima accession Bmar1.0^37^, and a first P. procumbens draft genome sequence (USDA_Ppro_WB292_v1.0, NCBI Genome Accession No. JBMGQN000000000, derived from USDA NPGS Accession Ames 4464) were included. Using MEGA11^38^, the sequences were aligned via the MUSCLE algorithm, and a maximum-likelihood phylogenetic tree (500 bootstraps, substitution model: JTT, tree inference: NNI method) was generated. Alignments were visualized using pyBoxshade v. 1.2^39^.

The blastn suite^40^ of the National Center for Biotechnology Information^41^ was deployed for DNA sequence comparisons. The localization and conformation of single polypeptides were predicted using DeepLoc2.1^42^ and DeepTMHMM^16^, respectively.

In vivo nematode infection assay

Nematode infection tests were performed essentially as described^43^. The H. schachtii strain ‘Schach 0’^44^ was propagated on susceptible sugar beet or rapeseed in sand-filled tubes in a greenhouse under LD conditions. Larvae were hatched from cysts in 3 mM ZnCl_2_ in a Baermann-funnel apparatus. Before inoculation, the J2 larvae were concentrated to the desired density of 150 per ml. Plants were grown in plastic tubes filled with quartz sand and inoculated with 300 J2 larvae. Developing females were identified four weeks later.

Expression analysis

Leaf and root samples from three biological replicates per accession were collected for the initial expression test. For the spatiotemporal expression analysis, samples of plant roots (R), hypocotyls (H), cotyledons (C), and leaves (L) of three biological replicates were taken shortly after germination (R, H, C), upon the development of the first true leaves (R, C, L), 21 days after germination (R, L), before inoculation (R, L), and 3 and 28 days post-inoculation (dpi; R, L). The samples were frozen in liquid nitrogen, and total RNA was isolated using the Universal RNA Kit (roboklon, Berlin, Germany), according to the manufacturer’s instructions. The RNA quality was checked using a NanoDrop2000 spectrophotometer (ThermoFisher Scientific) and agarose gel electrophoresis (2 % gel, 100V, 12 min).

One hundred nanograms of RNA were incubated with 1 U of DNase (ThermoFisher Scientific) at 37°C for 30 minutes. Subsequently, the DNase was inactivated by adding 1 µL EDTA (50 mM), followed by incubation at 65°C for 10 min. Then, the cDNA was generated using the First Strand cDNA Synthesis Kit (ThermoFisher Scientific), following the manufacturer’s instructions. Using the GAPDH primer combination for Beta (BvGAPDH_F/BvGAPDH_R, Supplementary Table 2) or Patellifolia (AS_BvPp_GAPDH_F/AS_BvPp_GAPDH_R, Supplementary Table 2) species, the cDNA quality was assessed by PCR.

Following the manufacturer’s instructions, 2 µL of cDNA were used for quantitative reverse transcription PCR (RT-qPCR) using the Platinum SYBR Green qPCR SuperMix-UDG (Invitrogen, ThermoFisher Scientific). The GAPDH gene was used for normalizing gene expression. For the detection of Hs4 transcripts, the primer combinations AK_F7/AK_R5 and AS_BvHs4_F1/AS_BvHs4_R7 (Supplementary Table 2) were used. The obtained Ct-values were evaluated using the Pfaffl method^45^.

Statistical analysis

Normally distributed data were statistically analyzed by a one-way analysis of variance (ANOVA) using Microsoft Excel. The non-normally distributed data were statistically analyzed using the Kruskal-Wallis rank sum test, followed by a Hochberg-corrected Dunn post-hoc test, as performed online with Astatsa^46^ and using the open-source statistical computing software R version 4.3.3^47^ with the “PMCMRplus” package^48^.

Supplementary Information

Supplementary Information.