Introduction

The ‘Poblano’ pepper (Capsicum annum L.) is Mexico’s major native pepper cultivar. The ‘Poblano’ pepper has high commercial value as a hot spice and is in strong demand in international commodity markets (FAOSTAT 2015). Pepper growers face a number of production problems that, if mitigated, would greatly increase their commercial returns. Statistical data show that although sown areas have increased by approximately 10% (SIAP 2017), the reduction in the percentage of harvested area has not changed since 2006 (36,200 ha), 2010 (34,144 ha), and 2016 (32,627 ha) (FAOSTAT 2017).

Many factors have been considered to contribute to yield losses, such as the incidence of biotic diseases, among them, vascular wilt caused by Fusarium spp. This disease occurs when susceptible cultivars are grown in infected native soil using a continuous monoculture. In this case, the ‘state of alert’ has been observed to be a novel mechanism in plants that is involved in the interaction of plants with soil microbiota and drives plant responses (Chialva et al. 2018). In addition, high temperature and moisture stimulate fungal development and contribute to the spread of this disease over large planted areas (Di Pietro et al. 2003; Gordon 2017).

Species of Fusarium are considered to cause a soilborne fungal disease (Yadeta and Thomma 2013); plants are infected when vegetative or propagative structures enter through the roots, causing yellowing and wilting, which often appear later during the growing season and are first observed on the lower leaves (Goodman et al. 1986; Beckman 1987). These symptoms progress up the plant as the fungus spreads within its host; a brown discoloration of the conducting vascular system just beneath the epidermis is often observed. The discoloration can extend from the roots up the stem through the branches and into the leafstalks of the plant (Michielse and Rep 2009; Sundaramoorthy et al. 2012). As the disease progresses, younger leaves are also affected and the plant prematurely drops its fruit or eventually dies, significantly reducing fruit production (Perez-Hernandez et al. 2014).

The severe outbreaks of most of the Fusarium vascular wilt pathogens are caused by F. oxysporum Schlechtendal, which contains pathogenic and non-pathogenic strains that are morphologically indistinguishable (Booth 1971; Lievens et al. 2008). Tropical race 4 has been mentioned in the literature as historically economically damaging to the banana industry (Dale et al. 2017), and races 1 and 4 have been mentioned as being damaging to other commercially important plants, including basil (Rekah et al. 2000) and cotton (Wang et al. 2018). On the other hand, F. solani (Martius) Saccardo has also been implicated as a causal agent of chilli wilt disease and as a pathogen of a large number of solanaceous plants (Zhang et al. 2007; Sundaramoorthy et al. 2012).

During the production of many solanaceous plants, seed germination is conducted in trays that are maintained in a greenhouse for later transplanting. During this period, several affectations are observed in seedling development that lead to rapid mortality. Therefore, there is little information regarding which Fusarium spp. result in damping-off symptoms in pepper seedlings under greenhouse production and whether the root tissue or the conducting vessels are colonised by species-specific or other phytopathogens. The purpose of this study was to identify the causal agents of damping-off in pepper seedlings in greenhouse and establish their relationship with those that cause crown rot and necrotic symptoms in the stem base of ‘Poblano’ adult pepper plants. Additionally, we sought to determine whether damage occurs in seedlings at a histological level.

Materials and methods

Plant material and fungal isolates

During spring/summer from 2007 to 2009, ‘Poblano’ mature pepper plants with wilt symptoms were recovered from commercial fields in the Puebla State of southeastern Mexico. Stems and roots with necrotic tissues were separated and washed in running water. Afterwards, the exterior portion was disinfested by immersion for 1 min in a 1.5% w/v aqueous sodium hypochlorite solution and rinsed three times with sterile distilled water. Pieces of the crown rot and stem base were excised and dipped in 25% aqueous ethyl alcohol (v/v) for 30 s and finally rinsed twice in sterile distilled water. The samples were then surface dried in a sterile flow chamber at room temperature. Small samples (1–3 cm) that displayed a reddish brown, rusty looking discolouration in the interior of the crown and stem were excised using a sterile scalpel and placed in Petri dishes containing potato dextrose agar (PDA) (BD Becton Dickinson, NJ, USA). When mycelium growth and macro and microconidia were evident, the conidia were removed by adding 10 mL of sterile distilled water to each Petri dish. This volume was placed into a test tube from which 1 mL was transferred to a new tube containing 9 mL of sterile distilled water to create a 10-fold serial dilution. From this, 10−5 to 10−7 conidia per millilitre were plated onto water agar (WA) medium. After 3 days, single conidia that had germinated were transferred to a Petri dish containing PDA amended with kanamycin (Sigma-Aldrich, MO, USA) at 50 mg L−1 to obtain pure cultures. Petri dishes were maintained for 10 days for pathogenicity tests and genomic DNA extraction (Table 1).

Table 1 Species of Fusarium isolated from mature plants of ‘Poblano’ pepper showing wilt symptoms from 2007 to 2009 in Puebla State, Mexico
Full size table

DNA extraction

The mycelium was removed using a spatula from the surface of the Petri dish and crushed in a mortar with 1 mL of lysis solution (10-mM TRIS base HCL pH 8.0, Triton × 100 0.5%) and 10 μL of proteinase K (Sigma-Aldrich, MO, USA) (10 mg/mL). Then, 600 μL of each sample was transferred to a 2.0-mL Eppendorf tube for DNA extraction in accordance with the protocol of Doyle and Doyle (1990) with very few modifications. DNA from each tube was suspended in 50 μL of TE buffer (10-mM Tris-HCl, pH 8, 1-mM EDTA), and the concentration was quantified using spectrophotometry in a Nanodrop 2000C (Thermo Scientific, MA, USA). DNA quality was estimated by the ratio of A280/260 and A230/260, and then, DNA was diluted to 20 ng μL−1 for PCR.

PCR amplifications of the ITS region of rDNA and a partial sequence of the TEF-1α gene

PCR using the ITS5 (5′-GGAAGTAAAAGTCGTAACAAGG-3′) (White et al. 1990) and NL4 (5′-GGTCCGTGTTTCAAGACGG-3′) (Guadet et al. 1989) set of primers was used to amplify a partial sequence of 18S rDNA, the complete sequence of the 5.8S rRNA gene, and the internally transcribed spacer 1 and 2 regions plus domains D1 and D2 of the nuclear large subunit (LSU) of a partial sequence of the rDNA gene. A fragment of 1100 base pairs (bp) was expected, including 500 bp corresponding to the end of the 18S rRNA gene, entire ITS region, and approximately 600 bp of the conserved domains in the 28S rRNA gene. In the case of the translation elongation factor 1-α (TEF1) partial sequence, the EF1-a (5′-ATGGGTAAGGARGACAAGAC-3′) and EF2-a (5′-GGARGTACCAGTSATCATGTT-3′) primers were used (O’Donnell et al. 1998), and an ~ 700-bp product was expected.

PCR was prepared in 0.2-mL Eppendorf tubes with a final volume of 15 μL containing 5× Go Taq DNA Reaction buffer, 2 μM of each dNTPs, 10 pmol of each primer, and 2 units of Go Taq DNA (Promega, WI, USA). Finally, 100 ng of DNA was added to each tube. PCR was performed using an initial denaturing hold at 95 °C for 4 min, 35 denaturing cycles at 95 °C for 1 min, annealing at 56 °C (ITS-28S rDNA) and 53 °C (TEF-1 α) for 1 min, extension at 72 °C for 2 min, followed by a single final extension cycle at 72 °C for 10 min. PCR was performed in a Peltier Thermal Cycler DNA Engine PTC-200 (BioRad, CA, USA), and the PCR products were verified by loading 5 μL on a 1.5% agarose electrophoresis gel (Seakem, CA, USA), which was stained as previously described. The remaining PCR amplified products were purified using the QIAquick PCR purification kit (Qiagen, CA, USA) following the manufacturer’s instructions. To confirm that there were no misreads in any of the strands, forward and reverse primers were used in individual wells for sequencing with the Big-Dye Terminator version 3.1 Cycle Sequencing kit in an Applied Biosystems model 3130 automated DNA analyser (Applied BioSystems, CA, USA).

Phylogenetic analyses

Sequences corresponding to both strands of the ITS entire region, 28S rDNA gene and TEF1 (partial sequence) were assembled independently and edited using BioEdit software version 7.0.5 (Hall 1999). A consensus sequence of each isolate was created for each gene or rDNA region and partial TEF1 and later submitted to the BLASTN 2.2.19 (Altschul et al. 1997) and Fusarium ID databases (Geiser et al. 2004).

Phylogenetic reconstruction for all of the sequences was performed with Bayesian inference using MrBayes v3.2.2 (Huelsenbeck and Ronquist 2001), and a 1-M generation post-stationarity was run in concatenated analyses. Markov Chain Monte Carlo (MCMC) analysis used four chains, one cold and three heated chains, and started from a random tree topology. The sample frequency was set at 1 in 1000 for the combined analysis. The consensus tree and subsequent Bayesian inference probabilities for the most complex model of evolution available for nucleotides (GTR + gamma distribution + invariant positions) were obtained (Ronquist and Huelsenbeck 2003).

The accession numbers of the F. oxysporum and F. solani mentioned in MycoBank (http://www.mycobank.org/) and deposited in the NCBI-GenBank database (Benson et al. 2012) were downloaded and included for comparison along with the sequences obtained in this study (Table 1). For construction of the phylogenetic tree, the sequence of Phoma herbarum (accession number EU082106) was designated as the out-group (Fig. 3).

Additionally, to determine to which Fusarium species complex the isolates obtained in this study belonged, the sequences described by O’Donnell and collaborators (2016) were used to construct a new phylogenetic tree based on the amplification of the TEF1 partial gene (Table 2). A phylogenetic tree was created in the same manner as previously described (Fig. 4).

Table 2 The Fusarium species complex and their accession numbers corresponding to partial sequences of translation elongation factor 1-α
Full size table

Morphological and cultural characteristics

Small portions of ~ 5 mm2 plugs of agar containing mycelium from each species of Fusarium were transferred onto Petri dishes containing PDA medium. To study the morphological features of these species, each isolate was grown on Carnation Leaf Agar (CLA) medium (Fisher et al. 1982), and the dishes were incubated at 24 °C with 12 h of white light and 12 h of UV light (365-nm General Electric 40W F40SL, CT, USA) for 15 days. Fifty cells were examined for each isolate to identify fungi at the genus level, and the descriptions of Nelson et al. (1983), Burgess et al. (1994) and Leslie and Summerell (2006) were used.

Inoculum preparation for pathogenicity tests

Isolates were increased using the oat glume method (Llop et al. 2000). Oat glumes were sterilised in a flask (twice for 45 min). Petri dishes with mycelium from each isolate grown on PDA were incubated under previously described conditions. The mycelium from each isolate were cut into pieces that were 1 cm2 in area and mixed in flasks of suspended oat glumes. The flasks were maintained at room temperature (c. 25 °C) for 3 weeks until colonisation of whole glumes was observed. Conidia were harvested by washing the glumes with rubbing in 250 mL of sterile distilled water.

Pathogenicity tests in the greenhouse

One-hundred seeds of the ‘Poblano’ pepper cv. San Luis were germinated in peat moss for 10 days and then transplanted during the two-leaf stage to a sterilised potting mix (3 peat:1 vermiculite, by volume) in a 12-cm-diameter × 15-cm-deep plastic pot. They were maintained in the greenhouse during the experiments, with temperatures ranging from 20 to 25 °C for 30 days. The seedlings were then inoculated with a preparation of 105 conidia mL−1 according to the protocol of Punja and Parker (2000). Four groups of 25 seedlings each were inoculated with (a) F. oxysporum (FO) M3 isolate, (b) F. solani (FS) M5 isolate, (c) FO + FS jointly, or (d) sterile distilled water (control). Seedlings were distributed at random in the greenhouse. To identify the most aggressive isolates for the pathogenicity test, M3 and M5 were selected in a preliminary study performed for the same pepper cultivar (data not shown). Disease severity was scored on a scale as follows: S0 = healthy plants, S1 = canker at the base of stem, S2 = base lesions joined into a single canker, and S3 = dead plants.

Evaluation was initiated at 5 days after the first symptoms of damping-off were observed on the stem base and ended after 35 days when the plants died. Fusarium spp. were re-isolated from inoculated seedlings to fulfil Koch’s postulates, and their identities were confirmed through molecular and morphological approaches. Pathogenicity tests were conducted three times.

Statistical analysis

Based on data of severity (scale 0 to 3) obtained from four replicates of 25 plants each, a Kruskal-Wallis test was performed providing the nature of the data (ratings) to determine whether there were significant differences in the severity of the inoculated pathogens. In addition, the incidence (proportion of symptomatic plants in relation to the total plants) and seedling mortality, i.e. those that were in category 3 at the end of the experiment, were determined. Control plants were not inoculated.

Histopathological test of the stem base and crown rot

The transition regions of the crown rot were selected for this study. The seedlings were analysed during stage 1 to maintain the integrity of symptomatic vegetal tissue during sample processing. Although cutting was attempted in tissue during the posterior stages, a clear observation of the regions was not possible. Three treatments were utilised (FO-M3, FS-M5, or FO-M3 + FS-M5 isolates) in addition to a control, and the crown rot areas with necrotic symptoms were fixed in FAA solution (ethanol:glacial acetic acid:formaldehyde:water in a 50:5:10:35 proportion) for 48 h. Each sample was gradually dehydrated in ethanol and ethylene and then processed for paraffin infiltration according to Zavaleta-Mancera and Engleman (1994). Cross sections that were 20 μm in thickness were obtained using a rotary microtome Spencer 820 (American Optical Company, USA) and stained with safranin ‘O’ (0.05% safranin ‘O’, 13% NH2SO4, 0.01% phenol in water) and Fast Green FCF (0.12% FCF Green in 95% ethanol) (Ruzin 1999). The sections were observed at × 40 magnification using a compound light microscope (Zeiss Axiosiop plus, Oberkochen, Germany) and were photographed using an Axiocam MRc 5 digital camera (Zeiss, Oberkochen, Germany).

Results

Isolated fungi

Eleven isolates recovered from the stem base and crown rots of infected ‘Poblano’ pepper adult plants with typical symptoms of yellowing or wilting were selected for this study. Three colonies grown in PDA medium showed similarity to F. oxysporum; they had grown to less than 6 cm in diameter after 8 days and had abundant aerial mycelium forming a convex shape, and the isolates were pale pink to deep purple and magenta in colour with a cottony texture. The macroconidia had an average of one to five septa with conidial dimensions of 23–54 × 3–4.5 μm; they were half-moon shaped and slightly curved, with delicate, fine walls. Their apical cells were sharp, and their basal cells had an upright form, although sometimes both extremities were sharp (Fig. 1a).

Fig. 1
figure 1

Scanning electron microscope images of a Fusarium oxysporum and b F. solani. mac: macroconidia with two or three divisions. c F. oxysporum had colonies less than 6 cm in diameter after 8 days. d F. solani grew more slowly and was less than 6 cm in diameter after 10 days

Full size image

In addition, eight colonies grown in the same medium resembled F. solani; they had grown to approximately 6 cm in diameter after 10 days and had aerial mycelium with a concave shape, smooth texture, and white to pale yellow colour. The colour of the colony beneath the surface layer varied from cream to light blue or clear coffee. The macroconidia had three to four septa with dimensions of 32–55 × 3–5.0 μm and were slightly curved, and in many, the terminal cell was slightly crushed (Fig. 1b).

Sequence analyses

As expected, ITS5/NL4 primer amplification yielded a fragment of 1100 base pairs (bp) (Fig. 2a) and the EF1-a/EF2-a yielded a fragment of ~ 700 bp (Fig. 2b) for all isolates. The results from BLASTN (https://blast.ncbi.nlm.nih.gov/Blast) and Fusarium ID (http://www.fusariumdb.org/) confirmed that three of the isolates that caused wilt symptoms in ‘Poblano’ mature pepper plants were F. oxysporum and eight isolates were F. solani (Table 1).

Fig. 2
figure 2

PCR amplifications of a the internal transcribed spacer (ITS) complete region including large subunit rDNA partial sequence, and b translation elongation 1-α partial sequences. MM, molecular marker 1Kb; bp, base pair

Full size image

The ITS rDNA regions and nLSU rDNA revealed that the nucleotide sequence identities ranged from 99 to 100% for both species. The sequences of F. oxysporum and F. solani obtained in this study were deposited in GenBank at NCBI under the accession numbers EU082095-EU082105 (Table 1). The amplified sequences of TEF1 partial sequences are listed in Table 2.

Phylogenetic reconstruction

For phylogenetic reconstruction sequences, the F. solani and F. oxysporum that were downloaded from Genbank formed two different clades. In the first clade, the sequences identified in this study as F. solani were grouped with accession no. EU029589, which corresponds to the S-0900 isolate from sorghum in the USA, and had a maximum identity of 100%. Other subgroups were observed in the same clade and belonged to sequences with a maximum identity of 99% from humans in Brazil, India, and the USA; in all cases, they were reported as a causal agent of rot in vegetables. The second clade group of sequences belonged to F. oxysporum; they were clustered with sequence KM030315, which corresponds to the NRRL 66023 isolate from the stem of a pre-silk corn plant in the USA. The sequence of Phoma herbarum included in this study was clearly clustered in a different clade and considered to be an out-group (Fig. 3).

Fig. 3
figure 3

Phylogenetic tree constructed with concatenated Bayesian inference using sequences belonging to the internal transcribed spacer (ITS) complete region, large subunit rDNA partial sequence, and translation elongation 1-α partial sequences. Amplifications corresponding to Fusarium oxysporum and F. solani were recovered from ‘Poblano’ pepper cv. San Luis plants that were decaying, yellowing, or wilting. The Bayesian inference posterior probability (> 0.95) is indicated below the internodes. The tree is rooted on a sequence of Phoma herbarum EU082106

Full size image

The Fusarium species complex of each isolate obtained in the present work was clearly determined by grouping and was consistent with those mentioned by O’Donnell et al. (2016), who studied Fusarium recovered from veterinary sources in the USA. Isolates of F. oxysporum and F. solani corresponded to the Fusarium oxysporum species complex (FOSC) and the Fusarium solani species complex (FSSC) (Fig. 4), respectively.

Fig. 4
figure 4

Phylogenetic tree constructed with Bayesian inference using translation elongation factor 1-α partial sequences of Fusarium isolates belonging to four Fusarium species complexes. Plant pathogenic isolates were clustered with Fusarium solani species complex (FSSC) and Fusarium oxysporum species complex (FOSC). The bar scale represents the number of substitutions per site

Full size image

Once the identity of the isolates was confirmed, they were deposited in the ‘Colección of microorganismos from Centro Nacional de Recursos Genéticos’ with numbers CM-CNRG 0424 and CM-CNRG 0425, respectively. This collection is a member of the World Data Collection for Microorganism (WDCM) under number WDCM 1006.

Pathogenicity test

Typical wilt symptoms were observed in ‘Poblano’ pepper seedlings inoculated with F. oxysporum, F. solani, or both. Necrotic tissues in the stem and crown rot began as small spots with a brown dark colour that were 2 mm in diameter, and later, they extended to the whole stem base to form a single black canker until the plant died 35 days after inoculation (dai). Necrotic spots were first observed five dai. The control seedlings did not develop symptoms. All inoculated seedlings (100%) were infected. No obvious differences in symptom severity were observed for F. oxysporum or F. solani at the cankers in the stem base and crown rot (Table 3).

Table 3 Severity of disease in ‘Poblano’ pepper seedlings inoculated in a greenhouse with Fusarium oxysporum, F. solani, and their mixture after 35-day post-inoculation
Full size table

The results suggest that F. oxysporum and F. solani are both pathogens of ‘Poblano’ pepper seedlings, causing both necrosis of the stem base and crown rot.

Seedling mortality caused by Fusarium spp. in inoculated pepper seedlings

For the mortality variable, the data indicated that while the percentage of mortality in seedlings inoculated with the combination of F. oxysporum and F. solani (FOS) was slightly less than those of FO and FS 49, 54, and 56%, respectively, there were no significant differences between the three combinations, but there was a significant difference with the control (F = 135.4, P < 0.001).

Considering this evaluated variable, the data indicate that both FO, FS, and the combined inoculation of both resulted in a similar level of mortality, whether alone or in a mixed inoculation pathogenicity.

Histopathological observations

In control seedlings, the parenchyma of the cortex, epidermal, endodermal tissue, and vessels of the vascular cylinder were intact (Fig. 5a, b). Light micrograph cross sections (Fig. 5) of the crown rot of ‘Poblano’ pepper seedlings inoculated with F. oxysporum showed two microconidia around the xylem vessels; one had penetrated a vascular vessel of the xylem but was impeded by the thickening of this structure (Fig. 5c) due to the defence response of the plant to the pathogen. In another sample with an annular disturbance, intercellular growth of the hyphae and thickening of the Casparian strips (Fig. 5d) were observed. Seedlings inoculated with F. solani had necrosis of the epidermal cells, the presence of macroconidium (mac, Fig. 5f) and collapse of the subepidermal parenchyma (Fig. 5e). We also observed necrosis of the epidermis and external layer of the cortex (Fig. 5f), intercellular growing hyphae, and partial destruction of the Casparian strips. Inoculated seedlings with both pathogens had a vascular cylinder without damage due to the presence of pathogens, although they had localised damage to the endodermal tissue of the Casparian strips and necrosis at the epidermis (Fig. 5g). Plants inoculated with both pathogens F. oxysporum and F. solani (Fig. 5h) had an epidermal layer with necrotic tissue (NT) and abnormal development of parenchymatic cells.

Fig. 5
figure 5

Severity symptoms of ‘Poblano’ pepper seedlings inoculated with Fusarium spp. a Stage 1, canker at the base of the stem and yellowing; b Stage 2, whole necrotic crown; c Stage 3, dead plants; and d control plant, no symptoms

Full size image

Discussion

Vascular wilt is among the most destructive plant diseases that occurs in annual crops (Michielse and Rep 2009; Ferniah et al. 2014). Cultural, chemical, and biological measures to control vascular wilt are required to reduce its devastating effects, as the causative agents of this disease are considered to be soilborne pathogen fungi (Yadeta and Thomma 2013). The precise identities of the causal agents of wilting that affect ‘Poblano’ mature pepper plants in field production, as well as damping-off in pepper seedlings in greenhouses in Mexico, are not known. Different fungal species have been reported in related cultivars, including Fusarium spp. (Schumann and D’Arcy 2009), Fusarium lateritium (Vasquez-Lopez et al. 2009), F. oxysporum (Lievens et al. 2008; Perez-Hernandez et al. 2014), and F. solani–F. oxysporum together (Fayzalla et al. 2008), but not F. solani alone (Figs. 6 and 7).

Fig. 6
figure 6

Seedling mortality percentage of Capsicum annuum cv ‘Poblano’ inoculated with Fusarium oxysporum (FO), F. solani (FS), and F. oxysporum and F. solani jointly (FOS). There were no significant differences between the two isolates, but there was a significant difference with the control (F = 135.4, P < 0.001). No synergistic effect was observed with joint inoculation

Full size image
Fig. 7
figure 7

Histology of Fusarium-infected crown rot. a–b Controls. c–d Crown rot of inoculated plant with Fusarium oxysporum displaying microconidium around the xylem vessels, necrotic cells in the epidermis, and disturbance of the Casparian strips. e–f Plants inoculated with Fusarium solani species complex. g–h Plants inoculated with both pathogens F. oxysporum and F. solani. ICs, increased Casparian strips; GIH, grow intercellular hyphae; AD, abnormal development; PD, peripheral destruction of the endodermal and Casparian strip tissue; XVO, xylem vessel obstruction; NT, necrotic tissue; Mac, macroconidium; Mic, microconidium

Full size image

In the present research, phylogenetic reconstruction based on amplification of conserved ribosomal operon genes and translation elongation factor 1-α found only two species of the Fusarium genus involved in these symptoms. F. oxysporum was isolated from crown rot and F. solani was recovered from rotten tissues at the stem base. They clustered in clades that are phylogenetically different, but together form part of a larger monophyletic group with strong statistical support (Geiser et al. 2013). In both cases, there was general plant decay, including wilt and root rot. These symptoms in mature plants were attributed to a reduction in water flow from the root to the aerial parts of the plants because of the obstruction or blockage of the xylem vessel with fungal hyphae. Fungi colonise cortical cells from which the hyphae migrate intercellularly toward vascular parenchyma cells and invade the xylem vessels, where conidiospores are produced and disseminated (Di Pietro et al. 2003; Schumann and D’Arcy 2009). The results of this study indicated that the isolates recovered from the mature plants were capable of causing damping-off in seedlings under greenhouse conditions. In this tissue, penetrating fungal hyphae and their differentiation cause damage mainly in the external cell layers in the root cortex and in internal layers comprising the endodermis (Genre et al. 2008), which is characterised by broken Casparian strips (Fig. 5). Endodermal cell walls perform an important function as apoplastic barriers, like the suberin lamellae in the roots of vascular plants (Schreiber and Franke 2011; Robbins II et al. 2014). The Casparian strips are a tangential deposition of a complex mixture of biopolymers in the cell wall of the endodermis, including carbohydrates, lignin, suberin, and structural proteins. These cell wall structures offer not only a significant barrier to apoplastic water flow but also a physical barrier to the movement of microorganisms, and it is well recognised that cell walls are degraded by enzymes secreted by Fusarium spp. and other pathogens during the infection process (Jones et al. 1972). Polygalacturonases have been isolated from Fusarium plant pathogens; the enzyme endoPGs macerates plant tissue by depolymerising homogalacturonan, a major component of the plant cell wall (Collmer and Keen 1986). Recent work on hormone signalling, propagation of calcium waves, and plant-fungal symbiosis has provided evidence that supports the hypothesis that the endodermis acts as a signalling centre (Robbins II et al. 2014).

It is known that F. oxysporum secretes fusaric acid (Singh et al. 2017), polygalacturonase, cellulase, α- and β-galactosidases, α- l-arabinofuranosidase, and β-xylosidase (Jones et al. 1972). F. solani is able to oxidise α, β-unsaturated alcohols, such as coniferyl alcohol, to the corresponding aldehyde, as well as the non-esterified side chain of dehydroconiferyl alcohol. F. solani is also able to degrade lignin compounds (Iwahara et al. 1980). However, degradation of chemically distinct lignified cell walls is not easy (Schreiber and Franke 2011). The primary function of the Casparian strip is to force water through a symplastic pathway as it enters the stele. In this manner, the Casparian strip controls the ion movement into the vascular system. In consideration of this feature, the function of the conductive elements is the most important feature for maintaining equilibrium in the vascular bundles and the union between adjacent cells. Rupture of this strip causes an alteration in the regular process of water movement and ions from the exterior to the interior of the plant and also alters the interior concentration of oxygen, creating significant gaps in the diffusion barrier and causing cells remain separated from each other (Robbins II et al. 2014). In addition, the plant cell wall represents a major barrier to the penetration and spread of microorganisms. It has been observed that pathogens produce cell-wall-degrading enzymes during the development of vascular wilt disease, penetrate different layers of the root cortex to the vascular system, and colonise the host by spreading upward through xylem vessels (Beckman 1987).

Light micrographs of seedlings inoculated with either F. oxysporum-M3 or F. solani-M5 show damage to the epidermis and parenchyma cells. They also show significant damage to the Casparian strips. Similarly, seedlings inoculated with both of these pathogens had vascular and epidermal damage, including to the surrounding epidermal cells. In all cases, we clearly observed damage to the aforementioned structures. These results were unexpected in that there are no reports of which we are aware that show the breakdown of Casparian strips in pepper seedlings infected by Fusarium species. By contrast, light micrographs of the stem base of control seedlings (Fig. 5a) confirmed that there were healthy cells in the epidermis, parenchyma, endodermis, and xylem.

During the initial stages of the interaction, fungal pathogens must sense stimuli from the plant and respond with appropriate morphogenetic and biochemical changes. The general conditions of the host genotypes, environmental factors, and virulence of the pathogen can determine the speed and severity at which the symptoms develop (Roncero et al. 2003), as was observed in seedlings inoculated with F. oxysporum and F. solani both together and independently in the present study.

The fact that the fungus identified on rockwool cubes of the transplants in the greenhouse requires confirmation to substantiate this result indicates that maintaining a clean transplant production area is important. No fungicides were labelled for Fusarium stem rot in the greenhouse. Growers of ‘Poblano’ peppers are very concerned about the increasing incidence of fungal diseases that result in severe yield and economic losses due to a reduction in cultivated area. Lastly, observation of damage to the Casparian strips provides an opportunity for further research to determine if this is the mechanism that is responsible for the gradual death and ripening of ‘Poblano’ pepper plants and the prompt death of their seedlings. Additionally, we should be able to ascertain whether, with better management of fungicide applications and biological or cultural strategies, producers will be able to control these two Fusarium species that cause harm in ‘Poblano’ pepper production in Mexico. Successful management of wilt diseases will be very important to ensure future economic viability of pepper production seedlings and pepper fruits for international trade.

Conclusions

Fusarium oxysporum and F. solani are two species of Fusarium that are responsible for wilting in ‘Poblano’ pepper mature plants and damping-off in seedlings during their growth in greenhouses. In mature plants, they only colonise the crown or base stem. Histological studies performed in inoculated seedlings with the isolates both separately and jointly showed necrotic cells of the external cortex, collapse of epidermal cells, necrotic endodermis, Casparian strip rupture, and changes in the cell size and wall thickness of the xylem during stage 1 of infection when mycelial growth was absent inside the vascular tissue of the pepper seedlings. Moreover, Casparian strip rupture caused by F. oxysporum and F. solani resulted in permeability loss and regulatory activity to maintain cellular equilibrium. To address these issues, pepper growers producing seedlings must take special care during seedling production in greenhouses because the aggressiveness of these isolates leads to rapid seedling death.