Subcuticular–Intracellular Hemibiotrophy of Colletotrichum lupini in Lupinus mutabilis

Anthracnose caused by Colletotrichum lupini is the most important disease affecting lupin cultivation worldwide. Lupinus mutabilis has been widely studied due to its high protein and oil content. However, it has proved to be sensitive to anthracnose, which limits the expansion of its cultivation. In this work, we seek to unveil the strategy that is used by C. lupini to infect and colonize L. mutabilis tissues using light and transmission electron microscopy (TEM). On petioles, pathogen penetration occurred from melanized appressoria, subcuticular intramural hyphae were seen 2 days after inoculation (dai), and the adjacent host cells remained intact. The switch to necrotrophy was observed 3 dai. At this time, the hyphae extended their colonization to the epidermal, cortex, and vascular cells. Wall degradation was more evident in the epidermal cells. TEM observations also revealed a loss of plasma membrane integrity and different levels of cytoplasm disorganization in the infected epidermal cells and in those of the first layers of the cortex. The disintegration of organelles occurred and was particularly visible in the chloroplasts. The necrotrophic phase culminated with the development of acervuli 6 dai. C. lupini used the same infection strategy on stems, but there was a delay in the penetration of host tissues and the appearance of the first symptoms.


Introduction
Lupins are legumes that belong to a large and diverse genus, Lupinus, comprising approximately 300 species, but only four species (L. albus, L. angustifolius, L. luteus, and L. mutabilis) present high agricultural importance [1][2][3]. Their seeds have one of the highest levels of protein among legumes, they offer potential health benefits, and they can contribute to the sustainability of cropping systems due to their net nitrogen input into the soil, which arises from their symbiotic interactions with rhizobia [1]. L. mutabilis (tarwi) is a species native to the Andes and, nowadays, is mainly cultivated throughout the Andean region, namely in Peru, Ecuador, and Bolivia. However, there is an increasing demand for protein crops that are suitable for Europe, and this species is a potential candidate [3]. L. mutabilis is characterized by the highest grain value of all cultivated lupins. It has a high protein content, an oil content that is similar to soybean, and is adapted to low input farming in temperate climates [4]. L. mutabilis cultivation is severely affected by anthracnose due to Colletotrichum lupini (Bondar) Damm, P.F. Cannon & Crous, which causes the twisting of stems, petioles, and pods and leads to the collapse of these tissues. In severe cases, progress of the disease causes plant death, compromising the crop [5]. The genus Colletotrichum was ranked in the top 10 fungal plant pathogens because of its broad host range, its ability to devastate essential crops, and its importance as a postharvest pathogen [6]. The pre-penetration phase is similar in all the pathogens of this genus, which usually begins with germination of the conidia and the formation of a melanized appressorium, a structure that is responsible for penetrating the host tissues [7][8][9]. After penetration, Colletotrichum species can use different strategies to infect and colonize host tissues, namely intracellular hemibiotrophy and subcuticular intramural necrotrophy, or even both [9][10][11][12]. The intracellular hemibiotrophic strategy is one of the most common and combines two phases: biotrophy and necrotrophy. After penetration, an infection vesicle and primary hyphae are formed. These structures do not kill the host's invaded cells (biotrophic phase). Later in the infection process, necrotrophic secondary hyphae spread within host cells and kill the host tissue, culminating in the appearance of disease symptoms [9,13]. The extension of the biotrophic phase and the switch from biotrophy to necrotrophy can differ among Colletotrichum species depending on the host species, organ, and maturity stage [9]. The hemibiotrophic strategy has been widely studied and reported in different plants, including Colletotrichum spp. interactions, such as Phaseolus vulgaris with C. lindemuthianum, Zea mays with C. graminícola [7,14], Sorghastrum nutans with C. caudatum [15], Tanacetum cinerariifolium with C. tanaceti [16], Arabidopsis thaliana with C. higginsianum [17], and Coffea arabica with Colletotrichum kahawae [18,19]. In the subcuticular intramural necrotrophic strategy, the appressorium produces a penetration hypha that penetrates the host through the cuticle. The fungus grows under the cuticle within the periclinal and anticlinal walls of the epidermal cells without penetrating the protoplasts. Before entering the necrotrophic phase, the fungus causes swelling and dissolution of the epidermal cell walls and, afterwards, quickly grows both inter-and intracellularly throughout the tissue, causing cell disruption and death. There is no primary or secondary hyphae development [7,20]. This type of infection strategy has been observed in several interactions, such as Gossypium hirsutum with C. truncatum (as C. capsici) [7], Capsicum annuum and Carica papaya with Colletotrichum truncatum [21,22], and Malus domestica with C. fructicola [23]. In some particular cases, including the infection of almond fruits and leaves by C. acutatum, after subcuticular fungal penetration, there is a biotrophic phase where the fungus develops intracellularly before the necrotrophic hyphae start colonizing the host tissues, leading to cellular destruction and the development of symptoms [7,20,24]. Interestingly, the same Colletotrichum species can adopt different infection strategies in the same host depending on the organs to be colonized [24] and references therein. Thus, Colletotrichum species have dynamic interactions with their host and can implement different lifestyles. Understanding the pathogenic interactions of fungi with their hosts has important implications for disease control [7].
A recent study conducted on Lupinus albus infected with C. lupini that used transcriptomic and proteomic approaches revealed that the dynamics of symptoms, gene expression, and protein synthesis were similar to those of hemibiotrophic pathogens [25]. However, no studies have detailed the strategy that is used by C. lupini to infect and colonize lupin tissues. To overcome this lack of knowledge, in the present investigation a susceptible genotype of L. mutabilis was selected and inoculated with a C. lupini isolate, and the infection strategy was characterized using light and transmission electron microscopy. Once it was confirmed that the infection strategy applied by C. lupini was the same in both petioles and stems, we focused our thorough analysis on the behavior of the pathogen when colonizing lupin petioles. Petioles seem to be a more sensible tissue that displays symptoms earlier, as well as being a much easier organ to manage in histopathological procedures.

Conidial Germination and Appressorium Formation
The light microscopic analysis revealed that 1 day after inoculation (dai), about 50% of the conidia germinated and formed melanized appressoria on the surface of the petioles and stems. Around 30% of the spores germinated, but this was without appressoria formation. In the germinated conidia, close to 100% of the appressoria were sessile (Figure 1a).

Conidial Germination and Appressorium Formation
The light microscopic analysis revealed that 1 day after inoculation (dai), about 50% of the conidia germinated and formed melanized appressoria on the surface of the petioles and stems. Around 30% of the spores germinated, but this was without appressoria formation. In the germinated conidia, close to 100% of the appressoria were sessile (Figure 1a).

Post-Penetration Stages
• Petioles In the petioles 1 dai, fungal penetration was evident by the appearance of internal light spots (Figure 1a) in more than 50% of melanized appressoria, as shown by the light microscopic observations. The internal light spot (ILS) corresponds to appressorium pore and penetration hypha development when viewed from above. Following cuticular penetration, the pathogen develops beneath the cuticle within the walls of epidermal cells and the apparent adjacent host cells appeared intact, as was observed at 2 dai ( Figure 1b). No visible macroscopic symptoms were detected from 1-2 dai (Figure 1c). However, at 3 dai, necrotic lesions first appeared ( Figure 2a). Light microscopy showed that hyphal growth within the cuticle was followed by the penetration of epidermal cells and further inter-and intracellular colonization of the cortex and less frequently of the vascular cells (Figure 2b,c).

Post-Penetration Stages
• Petioles In the petioles 1 dai, fungal penetration was evident by the appearance of internal light spots (Figure 1a) in more than 50% of melanized appressoria, as shown by the light microscopic observations. The internal light spot (ILS) corresponds to appressorium pore and penetration hypha development when viewed from above. Following cuticular penetration, the pathogen develops beneath the cuticle within the walls of epidermal cells and the apparent adjacent host cells appeared intact, as was observed at 2 dai ( Figure 1b). No visible macroscopic symptoms were detected from 1-2 dai (Figure 1c). However, at 3 dai, necrotic lesions first appeared ( Figure 2a). Light microscopy showed that hyphal growth within the cuticle was followed by the penetration of epidermal cells and further inter-and intracellular colonization of the cortex and less frequently of the vascular cells (Figure 2b,c).

Conidial Germination and Appressorium Formation
The light microscopic analysis revealed that 1 day after inoculation (dai), about 50% of the conidia germinated and formed melanized appressoria on the surface of the petioles and stems. Around 30% of the spores germinated, but this was without appressoria formation. In the germinated conidia, close to 100% of the appressoria were sessile (Figure 1a).

Post-Penetration Stages
• Petioles In the petioles 1 dai, fungal penetration was evident by the appearance of internal light spots (Figure 1a) in more than 50% of melanized appressoria, as shown by the light microscopic observations. The internal light spot (ILS) corresponds to appressorium pore and penetration hypha development when viewed from above. Following cuticular penetration, the pathogen develops beneath the cuticle within the walls of epidermal cells and the apparent adjacent host cells appeared intact, as was observed at 2 dai ( Figure 1b   cytoplasm disorganization: (i) discontinuities appeared in the plasma membrane, which retracted away from the host cell wall; and (ii) cells showed fragments of the host plasma membrane and other remaining cellular debris. In the epidermal and cortex cells, it was frequently observed that chloroplast membranes were losing their integrity ( Figure 3).
No distinct primary and secondary hyphae seem to have developed. The dissolution of cell walls was observed, particularly in the epidermal cells (Figure 2b,c). Disorganization of the cytoplasmic contents of some of the host's infected cells was also detected (Figure 2b,c). At this time point, transmission electron microscopy demonstrated the intense degradation of epidermal cell walls and the different levels of plasma membrane and cytoplasm disorganization: (i) discontinuities appeared in the plasma membrane, which retracted away from the host cell wall; and (ii) cells showed fragments of the host plasma membrane and other remaining cellular debris. In the epidermal and cortex cells, it was frequently observed that chloroplast membranes were losing their integrity ( Figure 3).  At 4 dai, the intense degradation of the walls of epidermal cells and those of the first layers of the cortex were visible by light microscopy. The cytoplasm of these cells collapsed or was disrupted, with only fragments remaining (Figure 4a,b). At 4 dai, the intense degradation of the walls of epidermal cells and those of the first layers of the cortex were visible by light microscopy. The cytoplasm of these cells collapsed or was disrupted, with only fragments remaining (Figure 4a,b). At 6 dai, acervuli had been formed. Acervuli were initiated by the formation of a dense hyphal stroma in the epidermis that gave rise to conidiophores, which disrupted the cuticle with the subsequent release of conidia and the appearance of dark sunken lesions (Figure 5a,b).  At 6 dai, acervuli had been formed. Acervuli were initiated by the formation of a dense hyphal stroma in the epidermis that gave rise to conidiophores, which disrupted the cuticle with the subsequent release of conidia and the appearance of dark sunken lesions (Figure 5a,b). At 4 dai, the intense degradation of the walls of epidermal cells and those of the first layers of the cortex were visible by light microscopy. The cytoplasm of these cells collapsed or was disrupted, with only fragments remaining (Figure 4a,b). At 6 dai, acervuli had been formed. Acervuli were initiated by the formation of a dense hyphal stroma in the epidermis that gave rise to conidiophores, which disrupted the cuticle with the subsequent release of conidia and the appearance of dark sunken lesions (Figure 5a,b).   The strategy applied by C. lupini to infect stem tissues was the same as that which was applied to the petioles. The initial subcuticular-intramural penetration and biotrophic intracellular growth were followed by necrotrophic fungal growth and the development of symptoms ( Figure 6). However, there was a delay in the fungal penetration and colonization of stem tissues. The fungal penetration was not observed until 3 dai (Figure 6a), and the first symptoms only appeared from 6 dai. The strategy applied by C. lupini to infect stem tissues was the same as that whic was applied to the petioles. The initial subcuticular-intramural penetration and bi trophic intracellular growth were followed by necrotrophic fungal growth and the deve opment of symptoms ( Figure 6). However, there was a delay in the fungal penetratio and colonization of stem tissues. The fungal penetration was not observed until 3 dai (Fig  ure 6a), and the first symptoms only appeared from 6 dai.  A diagram of the infection process that is used by Colletotrichum lupini to colonize Lupinus mutabillis petioles and stems is represented in Figure 7.

Discussion
To the best of our knowledge, this is the first histological and ultrastructural study concerning the infection process of Colletotrichum lupini in Lupinus tissues. As in other Colletotrichum-plant pathosystems [7,18,[20][21][22], the early stages of fungal development involved conidia germination and the differentiation of melanized appressoria. Highly specialized, melanized appressoria have extremely high turgor pressure, which is crucial for cuticle penetration [26]. At 1 dai, the cuticles of the petioles were already penetrated, which was confirmed by the presence of appressorial internal light spots. According to Diéguez-Uribeondo et al. (2003) [27], the internal light spot corresponds to the penetration pore from which the infection peg emerges, and host colonization begins. These are considered indicators of successful penetration. After cuticle penetration, the fungus grew beneath the cuticle and the epidermal cell wall, which is similar to what has been described in interactions between Capsicum annuum, Carica papaya and Colletotrichum truncatum [21,22], Malus domestica and C. fructicola [23], and Fragaria × ananassa and C. acutatum [28]. The host epidermal cells that were adjacent to these subcuticular intramural hyphae appeared intact, suggesting the existence of a biotrophic fungal growth phase. Identical biotrophic development was also observed during the subcuticular intramural colonization of apple leaves by C. fructicola [23]. In L. mutabilis petioles, hyphal growth within the cuticle was followed by the penetration of epidermal cells and by further inter-and intracellular colonization of the cortex and vascular cells. The ultrastructural modifications observed 3 dai suggested that this time point corresponds to the switch for necrotrophy. Besides the intense degradation of epidermal cell walls, different levels of plasma membrane and cytoplasm disorganization were reported. Some cells showed discontinuities in the plasma membrane which retracted away from the host cell wall, other cells exhibited fragments of the host plasma membrane and other remaining cellular debris, and in other cells, it was possible to observe the loss of chloroplast membrane integrity. Analogous cellular alterations were described in the interactions between coffee and C. kahawae

Discussion
To the best of our knowledge, this is the first histological and ultrastructural study concerning the infection process of Colletotrichum lupini in Lupinus tissues. As in other Colletotrichum-plant pathosystems [7,18,[20][21][22], the early stages of fungal development involved conidia germination and the differentiation of melanized appressoria. Highly specialized, melanized appressoria have extremely high turgor pressure, which is crucial for cuticle penetration [26]. At 1 dai, the cuticles of the petioles were already penetrated, which was confirmed by the presence of appressorial internal light spots. According to Diéguez-Uribeondo et al. (2003) [27], the internal light spot corresponds to the penetration pore from which the infection peg emerges, and host colonization begins. These are considered indicators of successful penetration. After cuticle penetration, the fungus grew beneath the cuticle and the epidermal cell wall, which is similar to what has been described in interactions between Capsicum annuum, Carica papaya and Colletotrichum truncatum [21,22], Malus domestica and C. fructicola [23], and Fragaria × ananassa and C. acutatum [28]. The host epidermal cells that were adjacent to these subcuticular intramural hyphae appeared intact, suggesting the existence of a biotrophic fungal growth phase. Identical biotrophic development was also observed during the subcuticular intramural colonization of apple leaves by C. fructicola [23]. In L. mutabilis petioles, hyphal growth within the cuticle was followed by the penetration of epidermal cells and by further inter-and intracellular colonization of the cortex and vascular cells. The ultrastructural modifications observed 3 dai suggested that this time point corresponds to the switch for necrotrophy. Besides the intense degradation of epidermal cell walls, different levels of plasma membrane and cytoplasm disorganization were reported. Some cells showed discontinuities in the plasma membrane which retracted away from the host cell wall, other cells exhibited fragments of the host plasma membrane and other remaining cellular debris, and in other cells, it was possible to observe the loss of chloroplast membrane integrity. Analogous Plants 2022, 11, 3028 8 of 10 cellular alterations were described in the interactions between coffee and C. kahawae [18] and between apple and C. fructicola [23] when the necrotrophic phase was established. The growth of C. lupini in L. mutabilis during the necrotrophic phase was associated with intense cell wall degradation and the death of the host's protoplasts, and culminated in the production of acervuli. As observed in the L. mutabilis petioles, C. lupini used the subcuticular intracellular hemibiotrophic strategy to infect and colonize the stems, as was revealed by the histological studies. Most Colletotrichum species are hemibiotrophic [9]. Interestingly, the transcriptomic and proteomic data obtained by Dubrulle et al. (2020) [25] also support the existence of a hemibiotrophic phase in L. albus and C. lupini interactions.
Lupin anthracnose is notorious for affecting stems. Since lupins bloom apically, the effect of stem infection hampers fruit set and, therefore, can be disastrous. Lupin anthracnose symptoms are peculiar due to the twisting of the affected organs (petioles, stems, and pods), a situation seldom found in anthracnose pathosystems (e.g., onion twister disease or celery anthracnose) [5]. The collapse of host cells in the infected area, as documented in the present study, could explain the twisting of lupin organs. Although L. mutabilis is generally regarded as susceptible, differences in disease responses have been noted [29] and seem to be influenced by anthocyanin pigmentation [30]. In the present work, a highly susceptible genotype was chosen, enabling the detailed characterization of infection mechanisms. We believe that this information in Lupinus will allow plant breeders to develop better control strategies against this disease and will enable better analysis of these interactions with less susceptible genotypes.

Biological Material
For the present work, a Colletotrichum lupini isolate (RB221, isolated from L. albus in France [29,31]) was used. L. mutabilis LM231 accession was selected as vegetal material due to its high susceptibility to anthracnose [29]. Plant and fungus growth conditions were the same as those described by Guilengue et al. (2020) [29].

Inoculation and Incubation
Whole plants (3-4 weeks old, containing 6-8 leaves) were placed horizontally on plastic trays on a wet nylon sponge and then the stem and petioles were inoculated with a 10µL drop of a conidia suspension (2 × 10 6 spores/mL). In order to ensure the adhesion of spores to the plant, the spore suspension contained 1% gelatin [29]. The covered trays were incubated for the first 24 h in a dark moist chamber at 22 • C and were then kept for a photoperiod of 12 h at the same temperature.
The plants that were used as controls were treated exactly like the inoculated ones, but instead of conidial suspension, distilled water was used.

Sampling and Microscopy
Conidial germination and appressoria differentiation were evaluated on the petioles and stem pieces (≈5 cm 2 ) 1 day after inoculation (dai), as previously described [32]. The samples were painted with transparent nail polish on the inoculated surface in order to recreate a tissue surface replica [33]. Once dried, the nail polish was removed, stained, and mounted in lactophenol cotton blue. For each experiment, a minimum of six microscope fields, each containing 100 conidia and/or differentiated appressoria on the surface of the tissues, were used.
To visualize the progression of the C. lupini infection under a bright-field, the petiole and stem samples were collected (at 1, 2, 3, 4, 6, and 9 dai) and were fixed using a technique adapted from Fernandes et al. (2021) [34]. The sample pieces were immersed in formaldehyde (37 to 38%)/glacial acetic acid/ethanol (70%) in a 1:1:18 proportion for a minimum period of 24 h. Before sectioning, the tissues were rehydrated with distilled water for 10 min. Sectioning (20-25 µm thick) was performed using a freezing microtome (Leica CM1850; Zurich, Switzerland).
To analyze fungal post penetration stages, the cross-sections were stained and mounted in cotton blue lactophenol [18]. In order to optimize the observation of the fungal infection process, strips from healthy and infected tissues (at the same time-points used to collect fresh tissues: 1, 2, 3, 4, 6, and 9 dai) were prepared according to a technique previously described [32]. The tissues were cut and fixed in a 2.5% (v/v) solution of glutaraldehyde in 0.1M sodium cacodylate buffer at pH 7.1. After 2 h, the tissues were washed (3 × 20 min) in sodium cacodylate buffer and were post-fixed with 1% osmium tetroxide in the same buffer for 2 h. The tissues were then washed in distilled water for 10 min. Subsequently, the tissues were dehydrated in a graded ethanol series (10-100% at 10% increments), embedded in Spurr's resin (TAAB), and polymerized overnight at 70 • C. As it was difficult to impregnate the plant material with resin, alterations to the described technique were introduced. The dehydration periods in ethanol were increased (20 min at each dehydration point up to 90% and two rinses were performed with 100% ethanol for 1 h each). The process of embedding the tissues in Spurr's resin was also more gradual (proportion resin:ethanol: 1:3, 1:2, 1:1, 2:1, and 3:1 for 12 h each), and was performed at 4 • C with agitation (ca. 10 rpm). Semi-thin sections (2 µm) that were made with an ultramicrotome Leica (Leica Ultracut R. Zurich, Switzerland) were stained with 0.5% aqueous toluidine blue O solution and were observed using light microscopy. Observations were made using a light microscope Leica DM-2500.
In order to better elucidate the infection strategy of C. lupini, polymerized blocks of healthy and infected material collected at 3 dai (corresponding to the appearance of the first necrosis) were used for transmission electron microscopic observations. Ultrathin sections (80-90 nm) of polymerized blocks were made with a diamond knife (mod. DIATOME ultra 458; DIATOME, Hatfield, PA, USA) using an ultramicrotome (Leica Ultracut R. Zurich, Switzerland). The sections were then collected on Formvar-coated nickel grids (200 mesh), stained with uranyl acetate and Reynold's lead citrate, and were examined using a 120 kV Transmission Electron Microscope (Hitachi H-7650. Tokyo, Japan) equipped with a XR41M mid mount AMT digital camera for data acquisition.