A Spreading-Stem-Growth Mutation in Lolium perenne: A New Genetic Resource for Turf Phenotypes

Abstract

In Lolium perenne, a novel growth habit mutant, named VIROIZ, was recovered following colchicine treatment, and it was confirmed to maintain the diploid chromosome number (2n = 2x = 14). The mutation affected the stem morphology by inducing prolific axillary shoot formation at nodal zones, resulting in a spreading growth habit that can extend to ~70 cm in width. Inheritance analysis based on single-plant evaluations in crosses with wild-type plants (F₁, n = 285; F₂, n = 380) and in selfed progeny (S₁, n = 255) consistently showed ~40% expression of the spreading phenotype, deviating from classical Mendelian ratios and indicating complex genetic control. Phenotypic selection further distinguished divergent classes: positively selected lines (C1⁺) averaged 3.90 axillary tillers per stem, whereas negatively selected lines (C1⁻) averaged only 0.22. Partial sequencing of 11 candidate genes implicated in shoot architecture, covering 40–90% of full-length DNA, did not provide a conclusive explanation for the altered stem growth. Notably, a single point mutation was observed in CRT3 (an endoplasmic reticulum chaperone that interacts with brassinosteroid signaling) highlighting it as a primary target for future studies. Cytological analysis of meiosis in F₁ hybrids between VIROIZ and wild-type plants revealed irregular chromosome pairing with persistent univalents (2–4 per cell), supporting the presence of structural chromosomal rearrangements that may disrupt gene organization and function in VIROIZ. The non-Mendelian segregation of the spreading phenotype, together with the observed meiotic irregularities, suggests that the mutation affects regulatory genes responsive to hormonal signals controlling axillary meristem initiation. The mutant represents a valuable resource for turf-type L. perenne breeding and for studying hormonal regulation of shoot morphogenesis in Poaceae.

1. Introduction

Perennial ryegrass (Lolium perenne L.) is widely cultivated in temperate regions for both forage and turf due to its rapid growth, high nutritive value, good digestibility, and reliable seed production, making it one of the most prevalent species in temperate grasslands worldwide [1,2,3,4,5,6,7]. In turf applications, it is commonly included in mixtures for golf courses, sports fields, and lawns, where traits such as tiller density, ground coverage, and persistence are emphasized over herbage yield [8,9]. Therefore, an intensive tillering habit that promotes dense, uniform turf is highly desirable in turf breeding.

Shoot development in perennial ryegrass has been described by Terrell [10] and Langer [11]. Each tiller consists of nodes bearing leaves and axillary meristems, most of which remain suppressed due to apical dominance, giving the species its typical bunch-type habit. Under specific conditions, however, axillary meristems can become active, producing secondary shoots, also referred to as stolons. Such creeping growth is rare in L. perenne, though sporadic occurrences have been reported in pastures [12,13,14]. More recently, distinct spreading forms have been described, including a patented stoloniferous subspecies L. perenne subsp. stoloniferum [15] and creeping cultivars used in turfgrass mixtures [9,16].

Shoot architecture is strongly regulated by hormonal networks. Auxins, cytokinins, and strigolactones interact to control apical dominance and the suppression of axillary meristems, thereby shaping shoot architecture [17,18,19,20]. Branching mutants have been reported in model and crop species, including Arabidopsis thaliana [21], pea [22], rice and wheat (review in [23]). However, to date, no naturally occurring or induced mutants affecting branching or tillering have been described in perennial ryegrass, despite its agronomic importance.

Here, we describe a novel spreading-growth mutant of L. perenne, designated VIROIZ, discovered among diploid plants treated with colchicine during an attempt to induce chromosome doubling [24]. While L. perenne is naturally diploid (2n = 2x = 14), colchicine is routinely used to generate tetraploids (2n = 4x = 28) for breeding [1]. In addition to inducing polyploidy, colchicine can cause heritable mutations affecting traits such as tillering, vegetative growth, and flowering [25]. VIROIZ was initially studied in the European GRASP project, which screened candidate genes in a set of 20 L. perenne genotypes [26]. This study identified variation in LpBRI1, a brassinosteroid receptor kinase gene, that suggested potential links to altered shoot morphology. More broadly, the brassinosteroid (BR) signaling pathway is well known for its central role in regulating shoot architecture, overall plant growth, and development [27,28].

To further investigate the genetic basis of VIROIZ’s unusual growth habit, we focused on eleven candidate genes with roles in hormone signalling and proteasome-mediated protein turnover. These include auxin- and brassinosteroid-related genes (IAA1, BRI1), strigolactone pathway regulators (D27, CCD8, MAX2), the brassinosteroid-associated chaperone CRT3, proteasome- and ubiquitin-related genes (UBC4, UBC10, RHF2A, RPT2A), and the protein kinase CIPK9.

Specifically, in this study, we (i) characterized the morphology of the mutant VIROIZ and the inheritance of the spreading growth type, (ii) assessed candidate genes (partial sequencing) and examined chromosome meiotic behavior aiming to explore potential changes linked to stem growth alteration in VIROIZ and (iii) evaluated VIROIZ potential as a genetic resource for turfgrass improvement and as a model for stem branching regulation.

2. Materials and Methods

2.1. Colchicine Treatment and VIROIZ Mutant Discovery

Seeds of diploid L. perenne representing a synthetic population were obtained from N. Lemežienė (Breeding Department, Lithuanian Institute of Agriculture). Embryos were excised from surface-sterilized seeds treated with a 0.3% (w/v) aqueous colchicine solution for 4 h and established in vitro on Murashige and Skoog (MS) medium [29] following the procedure described by [30]. A distinct mutant phenotype was identified among approximately 150 colchicine treated plants, each clonally propagated in duplicate. In field evaluations, the mutant exhibited unique characteristics, including wide and bushy growth, as well as modifications in the inflorescence, reduced spikelet number and fasciation.

2.2. Plant Material and Field Evaluation of F₁, F₂, and S₁ Generations

To study the inheritance of the spreading-growth phenotype, five pairwise crosses were made between mutant and normal plants derived from the initial seed sample used for colchicine treatment. A total of 285 F₁ plants were grown at 50 × 50 cm spacing and evaluated for growth phenotype in the experimental field of the Lithuanian Institute of Agriculture (Figure S1). The F₂ generation was obtained through free crossing within five groups of F₁ plants, and 380 F₂ plants were evaluated under the same field conditions. The S₁ generations, S1-a and 1-b, were obtained by controlled self-pollination of cloned VIROIZ plants in two independent cycles, producing 113 and 142 plants, respectively. All trials were conducted under identical field conditions to ensure comparability. The soil at the experimental site was classified as Endocalcari–Epihypogleyic Cambisols (CMg-p-w-can) with pH 7.2, 159 mg kg⁻¹ P₂O₅, 170 mg kg⁻¹ K₂O, and 1.67% humus.

When the plants reached flowering, they were screened visually and classified into three categories based on their growth habit: mutant-type plants corresponding to VIROIZ, wild-type (WT) plants and intermediate-type plants that did not fully correspond to either phenotype. Intermediate plants did not attain the width of VIROIZ but were more widely spread than WT and displayed inflorescence abnormalities resembling those observed in VIROIZ. All assessments were performed by a single observer (I. Pašakinskienė) to ensure scoring consistency.

2.3. Divergent Selection for Axillary Tiller in the Greenhouse Pot Experiment

Five L. perenne genotypes (LTS03, LTS04, LTS11, LTS15, and LTS16) [31] were intercrossed in all possible pairwise combinations, including reciprocals, to form a synthetic C₀ population. Among them, LTS15 and LTS16 are ecotypes, LTS03 and LTS04 are parent genotypes of the VrA mapping population [32], and LTS11 is a mutant exhibiting enhanced axillary tillering [24]. Equal quantities of seed from each cross were sown individually in pots to establish 340 C₀ plants, which were evaluated for axillary tiller development by counting the mean number of axillary tillers per 20 primary tillers at the beginning of flowering. The 10% of plants with the highest and lowest mean axillary tiller numbers per stem in the plant (34 plants in each group) were selected to form subpopulations C₀S⁺ and C₀S⁻, respectively. These plants were intercrossed by open pollination in isolated greenhouses to produce C₁⁺ and C₁⁻ populations (Figure S1). To synchronize flowering, all plants were vernalized for 100 days at 6 °C under an 8 h photoperiod.

2.4. Candidate Gene Amplification, Sequencing, and Analysis

Candidate gene screening was performed at the Molecular Laboratory of the Botanical Garden, Vilnius University. Genomic DNA was extracted from fresh leaf tissue using the GeneJET Genomic DNA Purification Kit (Thermo Fisher Scientific, Vilnius, Lithuania) according to the manufacturer’s instructions. Gene fragments of IAA1 (Indole-3-Acetic Acid inducible 1), BRI1 (Brassinosteroid-Insensitive 1), D27 (DWARF27), CCD8 (Carotenoid Cleavage Dioxygenase 8), MAX2 (More Axillary Growth 2), CRT3 (Calreticulin 3), UBC4 (Ubiquitin-Conjugating Enzyme 4), UBC10 (Ubiquitin-Conjugating Enzyme 10), RHF2A (RING-H2 Finger Protein 2A), RPT2A (Regulatory Particle Triple-A ATPase 2A), and CIPK (CBL-Interacting Protein Kinase) were amplified using original primers. The primers were designed with Primer3Plus (https://www.primer3plus.com) using L. perenne reference sequences from https://www.ncbi.nlm.nih.gov/nuccore/NC_067250.1/ (accessed on 14 October 2022). Primers for gene construction were designed to generate amplicons of approximately 800 bp, dividing the gene into smaller overlapping fragments. Each adjacent fragment contained overlapping end regions to enable assembly. Primer melting temperatures (Tm) calculated using the Thermo Fisher Scientific “Tm Calculator.” The primers were synthesized by METABION (Planegg, Germany).

PCR reactions (20 µL) contained 30 ng genomic DNA, 2.5 µM of each primer, 2.0 µM dNTP mix and 5 U/μL DreamTaq DNA Polymerase (Thermo Fisher Scientific, Vilnius, Lithuania), and 10x Polymerase buffer. The thermal cycling profile was: 95 °C for 2 s, followed by 35 cycles of 95 °C for 30 s, annealing temperature for 30 s, 72 °C for 60 s/kb. The reaction was terminated with extension at 72 °C for 5 min.

PCR products were visualized on agarose gels, excised and purified using the GeneJET Gel Extraction Kit (Thermo Fisher Scientific, Vilnius, Lithuania). The purified amplicons were sequenced in both directions by BaseClear B.V. (Leiden, The Netherlands) using the Sanger method. The high-quality sequences were assembled into contigs, aligned, and analyzed in AliView [33]. The sequence specificity was verified by BLAST (version BLAST+ 2.13.0) alignment against L. perenne reference genes. The reading frames and start codons were confirmed relative to reference sequences.

RNA was isolated from leaves using the GeneJET Plant RNA Purification Kit (Thermo Fisher Scientific, Vilnius, Lithuania) according to the manufacturer’s instructions. cDNA synthesis was performed using the Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Fisher Scientific, Vilnius, Lithuania) according to the manufacturer’s instructions, with the incubation at 50 °C extended to 30 min.

2.5. Chromosome Pairing Analysis in Meiosis

For the analysis of chromosome pairing during meiosis, samples from VIROIZ and F₁ hybrids were collected at the young inflorescence stage and examined for the presence of meiotic metaphase I (MI). Samples were collected from 3 clonally propagated VIROIZ plants and from 10 F₁ hybrids at random. The young inflorescences were fixed in Carnoy’s solution (ethanol:acetic acid, 3:1 v/v) and stored in 70% ethanol. Anthers were squashed in 1% acetocarmine, gently heated, and covered with a coverslip to spread meiocytes. The cells at Metaphase I were examined under a light microscope (1000× magnification), and the number of univalents, rod or ring bivalents, and multivalents was recorded in at least 50 cells.

2.6. Statistical Analysis

Phenotypic category frequencies were compared using chi-square tests. Sequence alignment statistics and variant calling were performed with AliView [33].

3. Results

3.1. VIROIZ Morphotype

Among the C₀ plants of Lolium perenne regenerated from isolated embryos treated with 0.3% aqueous colchicine, a unique mutant exhibiting an unusual growth habit was identified. Cytological analysis confirmed that the plant retained the diploid chromosome number (2n = 2x = 14). Morphologically, the mutant displayed a creeping, spreading growth pattern resulting from prolific axillary shoot formation, with additional shoots produced at each nodal zone (Figure 1A,B). The mutation also altered inflorescence morphology, leading to a reduced number of spikelets and the occurrence of fasciation symptoms (Figure 1C,D). This spreading growth mutant conferred vigorous growth, producing dense, prostrate, bushy plants up to ~70 cm in diameter—approximately twice the size of typical perennial ryegrass under field conditions (Figure 1E,F). To our knowledge, this is the first report of such a mutation in L. perenne. The mutant has been named VIROIZ.

3.2. The Inheritance of the Spreading Growth Habit Evaluated in Field and Pot Experiments

The mutant was fertile and stable, and has been maintained through vegetative clonal propagation for several decades. To investigate the inheritance of the spreading growth habit, five pairwise crosses were made between the mutant and normal plants to produce F₁. In addition, F₂ populations were obtained from free-pollination of these five F₁ groups. S₁ of VIROIZ was also generated and evaluated in two separate experiments. Across all evaluations carried out in the field, the inheritance pattern of the spreading growth habit was similar: strictly mutant plants comprised 42–44% of each population, whereas plants with the wild-type habit comprised 8–14% (Figure 2). A considerable proportion of plants displayed intermediate phenotypes, ranging from 42–50%. Therefore, the deviation from the classical Mendelian 1:2:1 ratio for codominant trait segregation was evident, as confirmed by the χ² test (p < 0.0001; Table S1).

Additionally, in a pot experiment, the evaluation of the VIROIZ branching stem phenotype for axillary tiller development in two rounds of selection (positive and negative) was carried out, which revealed a clear pattern of trait divergence (Figure 3A–C). A selection at 10% intensity (34 of 340 genotypes) resulted in selected subpopulations C₀⁺ and C₀⁻ with an average number of 3.41 and 1.83 axillary tillers per stem in the plant, respectively (Figure 3D). The second round of selection was performed with the same intensity as during the first cycle (10%, 34 selected genotypes), and it resulted in two subpopulations, C₁⁺ and C₁⁻. The average number of axillary tillers per stem in the plant was 3.90 and 0.22 in C₁⁺ and C₁⁻, respectively (Figure 3D). Thus, the selection differential was higher for negative selection (D = −1.76) than for positive selection (D = 1.44).

3.3. Mutation Analysis of Candidate Genes Implicated in Shoot Morphogenesis

Partial sequencing of 11 candidate genes associated with shoot architecture was carried out, covering 40–90% of their full-length DNA coverage, aiming to explore the genetic basis of the altered stem-growth habit in VIRIOZ. The candidate genes were selected on the basis of their known roles in stem morphogenesis and the broad regulatory functions of the proteins they encode. These genes fall into the following functional groups: (i) classical hormone pathway regulators (IAA1, BRI1); (ii) strigolactone-related genes controlling shoot branching (D27, CCD8, MAX2); (iii) the CRT3 gene encoding an ER chaperone interacting with BR signaling; (iv) proteasome- and ubiquitin-related genes (UBC4, UBC10, RHF2A, RPT2A); and (v) the protein kinase CIPK9, implicated in signaling pathways that influence shoot patterning. Three genes, D27, UBC4, and CIPK9, were sequenced to 80–90% of their full-length genomic DNA, four others, IAA1I, BRI1, CRT3, MAX2, between 50–70%, and the rest at ~40% of their length (

Table 1. Summary of candidate gene (partial sequence) screening for potential involvement in VIROIZ shoot morphology changes.

Gene	Protein Function	Full-Length Genomic DNA, bp	Gene Coverage in VIROIZ, bp (%)	Indel Allelic Variants		SNP Allelic Variants
				Description	Status in VIROIZ	No.	Unique in VIROIZ
IAA1	Auxin-inducible transcriptional repressor	2552	1821 (71.4)	-	-	7	N*
BRI1	Receptor in brassinosteroids signaling	3837	1931 (50.3)	-	-	6	N
CRT3	ER chaperon for folding of BR receptors	4396	2584 (58.8)	1378_1440 indel, in3	Homozygote 62 bp (−/−) deletion	7	c.1277C>G (p.L427V)
D27 †	Strigolactone biosynthesis	2374	2138 (90.0)	g.463_467, in2; g.878_884, in3	Homozygote 5 bp insertion (+/+); Homozygote 7 bp deletion (−/−)	5	N
CCD8 †		3752	1572 (41.9)	g.3054_3072 indel, in3	Homozygote for 19 bp deletion (−/−)	8	N
MAX2 †	F-box protein in strigolactone signaling	4684	2840 (60.6)	-	-	1	N
UBC4 †	Ubiquitin-conjugating enzyme (E2)	2670	2274 (85.2)	-	-	-	N
UBC10	Ubiquitin-conjugating enzyme (E2)	3174	1361 (42.9)	-	-	2	N
RHF2A †	RING-E3 ubiquitin ligase	3408	1457 (42.8)	-	-	-	N
RPT2A	A 26S proteasome subunit	3104	1328 (42.8)	-	-	1	N
CIPK9 †	Protein kinase in signaling	4298	3905 (90.7)	-	-	-	N

Notes: N*—not unique; †—for these gene alignment views are shown in Figure S3.

). DNA sequences of these candidate genes from the VIROIZ mutant were aligned with those of wild-type (WT) plants (N = 6–17 individuals) and analysed using bioinformatic methods. Two mutant phenotype plants (SM117 and SM120) and two WT phenotype individuals (SW44 and SW84) from S1 selfing progeny were also included in the analysis (Figure S2). Partial sequencing of these genes did not provide a conclusive explanation for the altered stem-growth habit (Table 1). However, the most notable variation was detected in the CRT3 gene: a point mutation c.1277C>G in exon 3 (corresponding to c.342C>G in cDNA from RNA transcript), leading to an amino acid substitution (p.L427V) (Table 1, Figure 4A,B). This mutation consistently distinguished the VIROIZ mutant and its mutant phenotype progeny in S1 from the WT group. Several indels were recorded in CRT3 and D27 of VIROIZ, though they were not unique to VIROIZ (Table 1, Figure 4C). Additional sequence alignments for the loci of the D27, CCD8, MAX2, UBC4, RHF2A, and CIPK9 genes are shown in Figure S3. From the partial sequencing data of these genes, no unique variants were found in VIROIZ compared with WT.

3.4. Chromosome Pairing Irregularities in Meiosis

Meiosis was examined in the mutant and in F₁ hybrids obtained from crosses between the mutant and plants with the WT phenotype. In the mutant, meiosis proceeded regularly, with the formation of seven bivalents at metaphase I. In the F₁ hybrids, however, the chromosome pairing pattern in metaphase I was different. Two univalents (occasionally four) were consistently present, along with one of the bivalents appearing as a smaller, cross-shaped configuration distinct from the remaining five regular bivalents (Figure 5A–C). In tetrads, the presence of one or two micronuclei—likely originating from the univalents at earlier stages—was also observed (Figure 5D). These meiotic irregularities are indicative of chromosome translocations in VIROIZ and may contribute to its abnormal growth phenotype.

In summary, the VIROIZ mutant of L. perenne exhibits a unique spreading-growth habit characterized by prolific axillary shoot formation and prostrate and bushy growth. Field and pot evaluations confirmed the trait to be heritable, with populations consistently segregating into mutant, wild-type, and intermediate phenotypes, and the selection for axillary tiller number produced a clear divergence between selected positive and negative subpopulations. The partial sequencing of 11 candidate genes involved in hormone signaling, strigolactone-mediated branching, and proteasome function, all of which affect shoot patterning, revealed a notable point mutation in the CRT3 gene c.1277C>G (p.L427V) (c.342C>G confirmed in cDNA from RNA transcript). This mutation was specific to the mutant and its mutant phenotype progeny and might have contributed to the enhanced branching and spreading stem phenotype. The cytological analysis indicated a largely normal meiosis in VIROIZ, but F₁ hybrids displayed univalents and micronuclei, suggesting underlying chromosome translocations present. Together, these results highlight VIROIZ as a stable, heritable, and genetically distinct source of altered shoot architecture in perennial ryegrass.

4. Discussion

The VIROIZ mutant of Lolium perenne represents a novel and valuable genetic resource for exploring shoot architecture and turfgrass improvement. Morphologically, VIROIZ exhibits a distinctive spreading growth habit, characterized by prolific axillary shoot formation, prostrate and bushy growth. This growth habit results in dense plants nearly twice the size of the typical perennial ryegrass under field conditions, emphasizing the potential agronomic and aesthetic value of the mutation in turf applications. Reports of spreading or stoloniferous growth in L. perenne are rare, limited mainly to a patented L. perenne subsp. stoloniferum [15] and a few turf-type cultivars [9,16]. To our knowledge, VIROIZ represents the first well-documented, heritable mutation conferring a spreading growth habit in L. perenne, establishing it as a valuable breeding resource and model for dissecting the genetic regulation of shoot branching in perennial grasses.

The consistent expression of the VIROIZ spreading growth habit across F₁, F₂, and S₁ populations confirms its heritability and phenotypic stability. The segregation patterns suggest the involvement of one or more major loci, potentially modulated by additional factors. The presence of intermediate phenotypes is consistent with the quantitative and multifactorial nature of shoot branching, which is governed by hormonal signaling [18,23,34,35,36]. Furthermore, the observed responsiveness of VIROIZ to artificial selection, with distinct divergent subpopulations, C₁⁺ and C₁⁻, emerging after only two selection cycles, underscores the plasticity of tiller number and suggests that this mutation could be exploited to fine-tune plant architecture in breeding programs, optimizing traits such as ground coverage and sward density for turf applications.

The partial sequencing of 11 candidate genes associated with shoot architecture did not provide a conclusive explanation for the altered phenotype. However, a nonsynonymous substitution was identified in exon 3 of the CRT3 gene and confirmed in the transcript (c.342C>G; p.L427V), which was unique to the VIROIZ mutant and its mutant phenotype progeny. CRT3 encodes an endoplasmic reticulum (ER) chaperone that interacts with brassinosteroid (BR) signaling components, ensuring proper folding and maturation of BR receptors [37]. Since BRs regulate cell elongation, organogenesis, and meristem activity, the impaired CRT3 function could reduce BR receptor maturation and alter crosstalk with strigolactone, cytokinin, and auxin pathways, thereby releasing axillary buds from apical dominance [27,28,38,39,40]. In both Arabidopsis and rice, CRT3 is co-expressed with genes involved in pathogen response and signal transduction [37,41]. Tissue-specific CRT expression, as seen in Solanum lycopersicum, suggests additional regulatory roles during development [42]. Beyond its functions in calcium storage and protein folding, CRT has been associated with the structure of plasmodesmata [43]. By facilitating the cell-to-cell transport of signaling molecules, including hormones, plasmodesmata may link CRTs to developmental regulatory pathways. While the CRT3 mutation represents a promising candidate contributing to the VIROIZ phenotype, functional validation is required to confirm its causality. Nonetheless, other gene loci or structural variants cannot be ruled out, given the limitations of our partial sequencing data; therefore, future QTL mapping and genome-wide analyses will be needed to identify additional loci contributing to this trait.

In this study, we also examined canonical regulators of shoot branching. Among these, strigolactones are well-established repressors of axillary bud outgrowth across several plant species, acting in concert with auxin and BR signaling. Strigolactone synthesis and signaling mutants exhibit increased branching across multiple species—including Arabidopsis, pea, rice, petunia, and tomato [22,44,45,46]—and notably, the VIROIZ phenotype closely resembles the “branching stem” mutants observed in these model plants. In addition, many phytohormone signaling pathways rely on ubiquitin–proteasome–mediated degradation of repressors or inducers. Several of our selected candidate genes fall into these functional categories. However, partial sequencing of strigolactone-related genes (D27, CCD8, MAX2) and ubiquitin/proteasome-related genes (UBC4, UBC10, RHF2A, RPT2A) revealed no variants specific to VIROIZ. Future studies should therefore extend to full-length gene sequencing, transcriptomic profiling, and functional assays, since regulatory changes or post-translational modifications could also underlie the phenotype.

The cytological analysis has confirmed that VIROIZ maintains the diploid chromosome number (2n = 2x = 14) and undergoes normal meiosis. In contrast, F₁ hybrids between VIROIZ and wild-type plants exhibited meiotic irregularities, such as univalents, cross-shaped bivalents, and micronuclei, indicative of structural rearrangements, possibly translocations. The data supporting gross chromosomal rearrangements triggered by mutagenesis come from other species, for example, Populus [47]. The absence of abnormalities in the mutant, but their appearance in hybrids suggests heterozygosity for such rearrangements or compensatory mechanisms that stabilize meiosis in the mutant background. Structural changes of this kind can generate position effects on gene expression, raising the possibility that the VIROIZ phenotype reflects an interaction between chromosomal rearrangements and point mutation(s) such as that in CRT3. Whole-genome sequencing and functional assays will be critical to test this hypothesis.

5. Conclusions

The VIROIZ mutant represents the first documented heritable mutation conferring such a growth habit in Lolium perenne, making it suitable for breeding programs aimed at customizing canopy architecture for lawns and sports fields. From an applied perspective, the spreading, bushy growth of VIROIZ offers a substantial potential for turfgrass improvement, enhancing ground coverage and sward density. At the same time, potential trade-offs, such as effects on seed production and long-term persistence, should be evaluated in long-term field trials. In this study, molecular and cytological analyses implicate both a candidate point mutation in CRT3 and possible chromosomal rearrangements affecting the altered phenotype; however, given the limitations of our partial sequencing data, other candidate genes cannot be ruled out in future studies. Together, these findings establish VIROIZ as both a practical breeding resource and a model for investigating the genetic control of shoot architecture in perennial grasses. Due to its perennial nature and stable growth habit, this mutant can be reliably maintained and shared for research and breeding purposes.