Plant Cell Reports

, Volume 32, Issue 2, pp 249–262

In vitro generation of somaclonal variant plants of sugarcane for tolerance to Fusarium sacchari

Authors

    • South African Sugarcane Research Institute
    • School of Life SciencesUniversity of KwaZulu-Natal
  • R. Stuart Rutherford
    • South African Sugarcane Research Institute
    • School of Life SciencesUniversity of KwaZulu-Natal
  • Sandy J. Snyman
    • South African Sugarcane Research Institute
    • School of Life SciencesUniversity of KwaZulu-Natal
  • M. Paula Watt
    • School of Life SciencesUniversity of KwaZulu-Natal
Original Paper

DOI: 10.1007/s00299-012-1359-0

Cite this article as:
Mahlanza, T., Rutherford, R.S., Snyman, S.J. et al. Plant Cell Rep (2013) 32: 249. doi:10.1007/s00299-012-1359-0

Abstract

Key message

A combination of in vitro culture and mutagenesis using ethyl methanesulfonate (EMS) followed by culture filtrate-mediated selection produced variant sugarcane plants tolerant and resistant toFusarium sacchari.

Abstract

Eldana saccharina is a destructive pest of the sugarcane crop in South Africa. Fusarium sacchari PNG40 (a fungal strain harmful to E. saccharina) has the potential to be an endophytic biological control agent of the stalk borer. However, the fungus causes Fusarium stalk rot in sugarcane. In the current study, sugarcane plants tolerant and resistant to F. sacchari PNG40 were produced by exposing embryogenic calli to the chemical mutagen ethyl methanesulfonate (EMS), followed by in vitro selection during somatic embryogenesis and plantlet regeneration on media containing F. sacchari culture filtrates (CF). The incorporation of 100 ppm CF in the culture media at the embryo maturation stage, at germination, or at both, resulted in callus necrosis and consequent reduced plantlet yield. Subsequent trimming of the roots of regenerated plants and their exposure to 1,500 ppm CF served as a further selection treatment. Plants produced from EMS-treated calli displayed improved root re-growth in the presence of CF pressure compared with those from non-treated calli. The tolerance of CF-selected plants was confirmed in greenhouse tests by inoculation with F. sacchari PNG40, re-isolation of Fusarium spp. from undamaged tissue of asymptomatic plants and establishment of the identity of fungal isolates as PNG40 using molecular analysis. The restriction of PNG40 presence to the inoculation lesion in some plants suggested their resistance to the fungus. Genotypes exhibiting symptomless endophytic colonization by PNG40 were identified and will be utilised for testing biological control strategies against E. saccharina.

Keywords

Culture filtrate (CF)Eldana saccharinaEthyl methanesulfonate (EMS)Fusarium sacchariIn vitro selectionSomatic embryogenesis

Introduction

Sugarcane (Saccharum spp. hybrids) is negatively affected by pests and diseases largely due to its vegetative propagation by stem sections (setts), perennial use of ratoons and monoculture in large adjacent fields, all of which result in the easy spread of microbial diseases and pests (Bailey 2004). The most important pest in the South African sugarcane industry is the stalk borer Eldana saccharina Walker (Lepidoptera: Pyralidae), the cause of extensive damage and economic losses since the 1970s (Mokhele et al. 2009). Among other control measures, the use of resistant varieties is the cornerstone of Integrated Pest Management strategies employed against it (Rutherford and Conlong 2010).

It takes 11–15 years to release a new improved cultivar of sugarcane through conventional breeding (Butterfield et al. 2001). Potential new commercial cultivars undergo a five-stage selection programme during which clonal material of selected genotypes is increased and clones are chosen for their sucrose content and resistance to pests and diseases (Parfitt 2005). Although clones combining high sucrose content and high sugar yield can be identified, they are often not released as they may lack resistance to one or more pests or diseases (Butterfield and Thomas 1996). Due to difficulties in conducting E. saccharina-inoculated screening trials for resistance, only those relatively few clones remaining in the later stages of the selection programme can be tested (Nuss 1991). Consequently, there is a considerable negative impact of pests and diseases on the selection programme, with 40 % of the best yielding clones at each stage being rejected due to susceptibilities (Butterfield and Thomas 1996).

In the case of E. saccharina, the combination of conventional breeding strategies with genetic engineering has been considered as a means to overcome the susceptibility of high sugar yielding clones to the insect (Meyer et al. 2000). However, the use of this technology in sugarcane has limitations, such as transgene silencing (Manners and Casu 2011), the cost and length of time required to prepare biosafety regulatory dossiers (Cheavegatti-Gianotto et al. 2011) and intellectual property restrictions.

Eldana saccharina has been found to be associated with Fusarium species in maize (Schulthess et al. 2002) and sugarcane (McFarlane et al. 2009). In maize, endophytic colonisation by F. verticilliodes was correlated with higher E. saccharina infestation and damage compared with plants treated with fungicide, suggesting a beneficial relationship between the fungus and the insect (Schulthess et al. 2002). In sugarcane, E. saccharina damage is associated with infection by Fusarium spp. where larval borings facilitate access of the fungus to the inner stalk resulting in rot of tissue surrounding the insect borings (McFarlane et al. 2009). Usually, Fusarium stalk rot in sugarcane is minimal without such borer-inflicted wounds to the stalk rind as the fungus cannot overcome this barrier unaided. McFarlane et al. (2009) also showed that certain Fusarium isolates were beneficial to the stalk borer’s survival and growth rate, whilst other isolates (e.g. F. sacchari isolate PNG40, which was isolated from a larval boring containing a moribund insect) were detrimental. Developing resistance or tolerance to these Fusarium strains in sugarcane can assist in controlling Fusarium stalk rot. In addition, the production of sugarcane genotypes exhibiting resistance to the Fusarium strains beneficial to E. saccharina may result in collateral increase in resistance to the stalk borer. Furthermore, selecting genotypes tolerant to strains antagonistic to E. saccharina, e.g. F. sacchari PNG40, may enable use of the fungus as an endophytic biological control agent against the stalk borer.

In vitro culture systems are known to be a potential source of variant plants from which those with desired traits may be selected. This approach has been widely applied in the improvement of a number of agricultural species (Van den Bulk 1991; Jain 2001; Lakshmanan et al. 2005; Snyman et al. 2011). In many such studies, physical (Ali et al. 2007; Sharma et al. 2010) or chemical (Imelda et al. 2000; Shah et al. 2009) mutagens, together with in vitro culture, have been used to increase mutation frequency to obtain disease-tolerant regenerated plants, e.g. bunchy top virus in banana (Imelda et al. 2000), Fusarium wilt in abaca (Purwati and Sudarsono 2007) and Black sigatoka and Fusarium wilt in banana (Jain 2010). In sugarcane, in vitro strategies that incorporate specific selection agents have been used to produce plants with improved tolerance to red rot disease (Ali et al. 2007), salt (Patade et al. 2008) and herbicide (Koch et al. 2012).

Purified toxins, or pathogen culture filtrates containing toxins, involved in pathogenesis, have been shown to be suitable selection agents for use in vitro (Daub 1986; Svabova and Lebeda 2005). For example, Fusarium produces fumonisin B1, which triggers the salicyclic acid pathway and encourages necrotrophic colonisation (De la Torre-Hernandez et al. 2010). Reducing the plant’s susceptibility to such toxins may lead to a decrease in damage by disease. This may be achieved by employing these compounds in selection of plants that permit fungal growth without toxin-inflicted damage (tolerance) or those that inhibit proliferation of the fungus (resistance) (Roy and Kirchner 2000). Studies in maize (Gengenbach et al. 1977), rice (Ling et al. 1985), barley, wheat (Wenzel and Foroughi-Wehr 1990) and sugarcane (Sengar et al. 2009) have confirmed a correlation between tolerance of plants to toxins, or culture filtrates, and that obtained to the pathogen. Nevertheless, it is important to expose such plants selected in vitro to the pathogen to confirm tolerance (Thakur et al. 2002; Sengar et al. 2009).

The objective of the present study was, therefore, to establish the feasibility of in vitro-induced mutagenesis, followed by fungal toxin-mediated screening, selection of somaclonal variant sugarcane cells and subsequent plantlet regeneration to achieve tolerance to F. sacchari PNG40. This is the first part of a larger study aimed towards employing F. sacchari PNG40 in control strategies against E. saccharina. Tolerance to F. sacchari will reduce Fusarium stalk rot, usually associated with E. saccharina damage, and also permit endophytic colonisation in sugarcane plants by PNG40. This will facilitate exploitation of the fungus as a biological control agent against the stalk borer.

Materials and methods

Fusarium sacchari culture and filtrate preparation

Fusarium isolate PNG40 was originally obtained from aborted E. saccharina borings in sugarcane stalks, identified as Fusarium sacchari and reported to be toxic to the stalk borer (McFarlane et al. 2009). To prepare cultures for storage, colonies were grown on potato-dextrose-agar (PDA) (Biolab, Wadeville, RSA) for 5 days, after which 5 × 5 mm mycelial squares were transferred to 15 % (v/v) glycerol (Merck, Wadeville, RSA) and stored at −80 °C. Starter cultures were prepared by placing a thawed mycelial square on PDA for 3 days. Thereafter, a mycelial square was cut from the leading edge of the F. sacchari PNG40 colony, transferred to 250 ml of potato-dextrose-broth (PDB) (Fluka, St Louis, USA) and agitated at 145 rpm at 28–30 °C. After 7 days, the culture was centrifuged at 12,000 rpm for 5 min. The supernatant was filtered through sterile muslin cloth to remove the mycelia and its dry mass (80 °C for 24 h) was determined. The filtrate was sequentially filtered through Whatman No. 1 filter paper, 0.45 and 0.2 μm filters (Millipore, Ireland). This culture filtrate (CF) was stored at 4 °C for a maximum of 24 h. The concentration of each batch of CF was expressed as fungal dry mass/volume of PDB used in the Fusarium liquid culture. To ensure batch-to-batch consistency, culture conditions with regards to duration, media and temperature were identical. The culture filtrate was diluted according to mycelial dry mass.

Indirect somatic embryogenesis and plantlet acclimation

The cultivar NCo376 was field grown at the South African Sugarcane Research Institute (SASRI), Mount Edgecombe, Durban. Leaf roll decontamination, explant preparation (30 × 2 mm-thick leaf discs/stalk) and culture conditions were as described by Snyman (2004), and all media (at pH 5.8) are listed in Table 1. Ten leaf discs/90 mm Petridish were initiated on 30 ml Embryo Initiation Medium (EIM) and incubated in the dark at 25–27 °C, sub-culturing on to fresh medium every 2 weeks. After 6–8 weeks, 0.2 g of embryogenic calli were transferred to 30 ml Embryo Maturation Medium (EMM) incubated in the dark at 25–27 °C for 3 weeks. After embryo maturation, the calli were transferred to 30 ml Embryo Germination Medium (EGM1) in 90 mm Petridishes and incubated in 16 h light (200 μm/m2/s photon flux density)/8 h dark photoperiod, at 26–30 °C for 4–8 weeks. Individual rooted plantlets (>20 mm in height) were transferred to Sterivent® vessels (110 × 100 × 80 mm) (Duchfa, Belgium) with EGM1 (20 plants per vessel). For acclimation, plantlets (70–100 mm in height) were planted in polystyrene seedling trays (67 × 33 cm, 98 cells) containing a 1:1 peatmoss–vermiculite medium (v/v), supplemented with dolmitic lime (Calmasil®, Middleburg, RSA) (5 g/10 kg substrate). The trays were transferred to a poly-carbonate tunnel, watered for 5 min (600 ml/min) twice a day and fertilised every 2 weeks (NPK 5:1:5, Profert, Noordsberg, RSA) for 2 months.
Table 1

Media composition for embryo initiation (EIM), maturation (EMM), germination (EGM1) and plantlet establishment (EGM2) stages

Constituents (g/L)

Medium

EIM

EMM

EGM1

EGM2

MS salts and vitamins

(Murashige and Skoog 1962)

Full-strength

Full-strength

Full-strength

Half-strength

Casein hydrolysate

0.5

0.5

0.5

Sucrose

20

20

20

5

2,4-Dichloro-phenoxyacetic acid

3

1

Agar–agar

8

8

8

8

Establishment of culture filtrate-selection treatments

To determine the effect of CF during embryo maturation, embryogenic calli were established for 6–8 weeks on EIM. The calli (0.2 g per replicate, n = 12–21) were placed on EMM containing CF and incubated in the dark at 25–27 °C, sub-culturing weekly. Different concentrations of CF (0, 20, 50 or 100 ppm) were used to determine a suitable concentration for screening and selection during embryo maturation. Callus necrosis (defined as at least 50 % of a callus piece being brown in appearance), fresh and dry mass (70 °C for 48 h) were recorded after 3 weeks. Embryo germination and plantlet growth were subsequently carried out on media without CF and, for each treatment, the total number of plants produced per 0.2 g callus was recorded over 12 weeks. For the CF effect on embryo germination, embryo initiation (6–8 weeks) and maturation (3 weeks) were both carried out on media without CF. Subsequently, embryogenic calli were transferred to EGM1 supplemented with 0, 20, 50 and 100 ppm CF, and maintained in the photoperiod growth room for 4–12 weeks, sub-culturing weekly. Callus necrosis and number of plants per replicate were recorded after 4 and 12 weeks, respectively.

To investigate the effect of CF on established plantlets (70–100 mm in height), green leaves were pruned just above the ligule and roots were trimmed to <1 mm. The trimmed plantlets were transferred to Magenta® vessels (5 plants/vessel) containing EGM2 with 0, 750 or 1,500 ppm CF to establish an appropriate concentration for selecting plantlets tolerant to CF. After 3 weeks of incubation in the photoperiod growth room, root re-growth was determined by measuring root length.

Production of variant plants and CF-mediated selection

The EMS preparation and callus treatment were according to the methods of Koch et al. (2012). Embryogenic calli (0.2 g) were placed in 32 mM EMS solutions for 4 h, after which they were rinsed three times with liquid EMM; liquid EMM was used for the controls.

After EMS exposure, eight CF treatments (Table 2) were tested for selection. Calli were transferred to EMM for 3 weeks followed by EGM1 for 4–8 weeks. Culture filtrate at 100 ppm (the most toxic of the tested concentrations) was incorporated in either EMM or EGM1 and in both media (Table 2). Fresh and dry mass and number of plants per 0.2 g of callus were recorded after embryo maturation and germination, respectively; callus necrosis was recorded at each stage. Plantlets (20 mm in height) produced from each treatment were transferred to Sterivent® vessels until they were 70–100 mm tall. The roots and leaves of the plantlets were then trimmed before being transferred to Magenta® vessels containing 80 ml EGM2 supplemented with 1,500 ppm CF, the concentration deemed in preliminary studies to be suitable for selection of CF-tolerant plants. Root length was recorded after 3 weeks.
Table 2

Treatments used to select ethyl methanesulfonate (EMS)-treated calli and controls tolerant to CF. Embryogenic calli (0.2 g) were exposed to EMS (32 mM) for 4 h. Culture filtrate was included at embryo maturation stage, at germination, or both

Treatment

EMS

Culture stages

CF (100 ppm)

Embryo maturation

(3 weeks)

Embryo germination and plantlet establishment (4–8 weeks)a

1

2

+

3

+

4

+

+

5

+

6

+

+

7

+

+

8

+

+

+

aUntil plants reached a height of 20 mm

Ex vitro selection studies using Fusarium sacchari

After screening on medium containing 1,500 CF for 3 weeks, plantlets with roots that re-grew to at least 10 mm in length were used for ex vitro investigations. Such plants were acclimated in the glasshouse at 20/34 °C (night/day temperature), watered for 5 min (600 ml/min) twice a day and fertilised every fortnight with NPK (5:1:5). When they had 1–2 stem internodes they were transferred to pots (200 mm diameter; 170 mm height), placed in troughs and not watered for a week. They were then inoculated by stabbing the stem 2–3 cm above the soil surface with PNG40-colonised toothpicks (Gilbertson 1985). Controls were stabbed with either uncolonised or colonised toothpicks. Secondary infection was avoided by swabbing the stem with 70 % ethanol prior to stabbing and the protruding toothpick was cut and the wound wrapped with parafilm.

Seven to 8 weeks after toothpick stab inoculation, the stems of dead and live plants (some with necrotic, chlorotic and crinkled leaves) were surface sterilised by submerging them sequentially in 95 % (v/v) ethanol for 2 min, 10 % (v/v) sodium hypochlorite for 5 min and sterile water (twice) to rinse, after which they were dried in the laminar air flow cabinet for 5 min. Removal of surface contaminants was confirmed by pressing the leaves and stems on PDA. The leaves were cut across the blade and placed on Nash and Snyder (1962) medium. The stems were split into longitudinal sections and inoculation lesion severity for individual stems was visually rated as: 0, no lesions; 1, mild; 2, moderate; 3, severe. One of the two longitudinal sections was used for re-isolation of the fungus from the lesion and from the undamaged area above it, and the other section was used for staining. Fungal re-isolation from the inoculation lesion, and the undamaged area 20–30 mm above it, was carried out by cutting longitudinal stem sections and placing the segments on Nash and Snyder (1962) medium and incubating at 28–30 °C for 5 days. Isolates were visually compared with PNG40 and were subjected to molecular analysis by inter-simple sequence repeats (ISSRs). Stem tissue was stained with lactophenol cotton blue (Sigma, St. Louis, USA) and wet mounts prepared for compound light microscope detection of fungal colonisation, according to the method of De Mars and Boerner (1995).

Molecular analysis of isolated fungus

Fungal DNA extraction was conducted according to the PrepMan® Ultra Sample protocol (Applied Biosystems, California, USA). The DNA concentration was determined using a spectrophotometer (NanoDrop Technologies, Delaware, USA) and adjusted to 200–250 ng/μl with elution buffer.

Regions between microsatellites were amplified using ISSR 1, ISSR 2, ISSR 4 and ISSR 8 primers (McFarlane et al. 2009) (Table 3). The efficacy of these primers to discriminate Fusarium species and isolates was tested by performing ISSR-PCR using six isolates belonging to F. andiyazi, F. proliferatum and F. sacchari (two isolates/species) obtained from the SASRI culture collection. PCR was carried out for each primer using a PCR kit (Kapa Biosystems, Massachusetts, USA) in a final volume of 19.5 μl composed of 13.42 μl PCR water (Ambion, Texas, USA), 2 μl Taq buffer with MgCI2 (15 mM), 0.4 μl dNTPs (10 μM), 1.6 μl ISSR primer (10 μM), 0.5 U Taq polymerase and 2 μl DNA template. Cycling conditions for PCR amplification were 95 °C for 2 min, followed by 32 cycles of denaturation at 94 °C for 30 s, primer annealing (temperature in Table 3) for 30 s, extension at 72 °C for 30 s, and a single final extension at 72 °C for 5 min. PCR products were stained with GelRed™ nucleic acid stain and separated by gel electrophoresis with a 100 bp O’Gene™ Ruler DNA Ladder Mix (Fermentas, Maryland, USA) as the molecular weight marker. Banding patterns of the Fusarium isolates were visually analysed and polymorphic bands were noted. PCR was repeated three times with each primer. DNA extracted from re-isolated fungi from dead and live plants was subjected to ISSR-PCR using primers ISSR 1, ISSR 4 and ISSR 8, and the banding patterns were compared with those of PNG40 to confirm similarity between the inoculated F. sacchari PNG40 and the re-isolated fungus.
Table 3

ISSR primer sequences and annealing temperatures used in discriminating Fusarium species and isolates

Primer

Sequence

Annealing temperature (°C)

ISSR 1

CCCGCATCC(CA)9

57

ISSR 2

CCCGGATCC(GA)9

55

ISSR 4

(AG)9G

51

ISSR 8

(CCA)5RY

45

R purine, Y pyrimidine

Statistical analyses

The data were analysed using Genstat statistical package 13th edition (VSN International, UK. 2010). The data were initially tested for normality using the Shapiro–Wilk test and for homogeneity using the Bartlett test. The Restricted maximum likelihood (REML) method was used to estimate random and fixed effects and significant differences amongst treatments were detected using the Holm-Sidak test.

Results

Establishment of callus and in vitro plantlet screening conditions

Exposure of embryogenic calli to 0–100 ppm F. sacchari CF at either embryo maturation or germination stages resulted in callus necrosis, i.e. at least 50 % of a callus piece being brown in appearance, which intensified with increasing CF levels in the media (Fig. 1a). At the maturation stage, some necrotic calli, especially those exposed to 100 ppm CF, turned mucilaginous and developed root hairs after 3 weeks. Despite these results, no significant differences in fresh and dry mass were detected amongst the CF treatments during maturation (results not shown). When the embryogenic calli were subjected to CF stress only during embryo germination (Fig. 1a), necrosis set in during the first week of exposure and, by 4 weeks, nearly 100 % of calli on all tested CF treatments were necrotic.
https://static-content.springer.com/image/art%3A10.1007%2Fs00299-012-1359-0/MediaObjects/299_2012_1359_Fig1_HTML.gif
Fig. 1

The effect of culture filtrate concentration in either the embryo maturation or germination medium on a percentage callus necrosis, and b plantlet yield. The number above the bar indicates percentage abnormal plantlets. Lines above the bars represent standard error of the mean, n = 12–21

Incorporation of CF in the embryo maturation media and subsequent embryo germination on EGM1 without CF resulted in decreased plant yield (Fig. 1b). Embryos from the control and 20 ppm treatments at maturation germinated normally. However, some non-necrotic calli only produced roots and those calli that were brown eventually turned black. No significant differences in plantlet yield were observed between the control and 20 ppm, and the 50 and 100 ppm CF treatments (Fig. 1b). However, the latter produced significantly fewer plants than the control (P < 0.001). Although there were no statistically significant differences with respect to plantlet yield amongst the tested CF concentrations (20, 50 and 100 ppm), there was a trend indicating an inhibiting effect of the CF (Fig. 1b). Although application of the CF stress at the embryo germination stage resulted in most calli being recorded as necrotic, the non-necrotic areas within each callus started greening during the first week on EGM1 + CF and plantlet yield was recorded by week 12 (Fig. 1b). Plant yield was significantly inhibited (P < 0.001) by increasing concentrations of CF with the least number of plants being produced at 100 ppm CF (Fig. 1b).

As expected, there was an inverse relationship between percentage callus necrosis and plantlet yield with the treatments recording the highest necrosis producing the lowest number of plants (Fig. 1). Also, callus exposure to CF at the embryo germination stage resulted in more severe necrosis than at the embryo maturation stage, with the highest percentage necrosis recorded at each stage being 95.5 ± 0.9 and 61.6 ± 3.9 %, respectively. Consequently, fewer plants were produced from the former than from the latter. The number of abnormal plants from each treatment after incorporation of CF during either maturation or germination was recorded and never exceeded 2 % (Fig. 1b).

Culture of plantlets with trimmed roots and leaves on EGM2 + CF resulted in stunted root growth and discolouration at the base of the stem after 3 weeks (Fig. 2a). Significant inhibition (P < 0.001) of root re-growth was observed in the 750 ppm and 1,500 ppm treatments with plant root lengths of 17.8 ± 1.7 and 8.5 ± 2 mm, respectively, compared with 39.4 ± 2.1 mm in the control plants (Fig. 2b). The 1,500-ppm CF was the most root re-growth inhibiting treatment after 3 weeks and was, therefore, adopted for selection of CF-tolerant plants. Since trimmed roots of 50 % of the plantlets cultured on medium containing 1,500 ppm CF re-grew to at least 10 mm in length, this value was set as the criterion for selection of CF-tolerant plants. The leaves re-grew to approximately the original length in all the treatments, although they were pale-green in the 750 and 1,500 ppm CF treatments compared with those of the control.
https://static-content.springer.com/image/art%3A10.1007%2Fs00299-012-1359-0/MediaObjects/299_2012_1359_Fig2_HTML.gif
Fig. 2

The effect of Fusarium sacchari culture filtrate (CF) on a the visual appearance of roots, and b root re-growth from plantlets which had their roots and leaves trimmed before being placed on media with 0–1,500 ppm CF for 3 weeks. Lines above the bars represent standard error of the mean, n = 10

Selection of calli and plants putatively tolerant to F. sacchari

The effect of CF incorporated in either embryo maturation or embryo germination medium, and in both media on EMS-treated calli was investigated (Table 2). The EMS treatment did not result in callus necrosis, as no significant differences were recorded between the EMS-treated (treatments 4 and 8) and the non-treated calli (treatments 3 and 7) (Fig. 3a). Only calli cultured on EMM + CF (treatments 3, 4, 7 and 8) were found to be necrotic (Fig. 3a).
https://static-content.springer.com/image/art%3A10.1007%2Fs00299-012-1359-0/MediaObjects/299_2012_1359_Fig3_HTML.gif
Fig. 3

The effect of culture filtrate on a percentage necrosis, and b plantlet yield of EMS-treated calli. The culture filtrate was incorporated at the embryo maturation stage or at germination or both stages. The number above the bar indicates percentage abnormal plants. Lines above the bars represent standard error of the mean, n = 7–12

After embryo maturation, calli from treatments 1–4 were transferred to EGM1 − CF and those from treatments 5–8 to EGM1 + CF, until plantlets developed (maximum 12 weeks). As expected, after 4 weeks in germination medium, no callus necrosis was detected in calli from treatments 1 and 2 (no CF at both maturation and germination), and necrosis in calli from treatments 3 and 4 (CF stress at maturation only) remained at the end of the maturation stage (Fig. 3a). Also, as previously determined, necroses in calli from treatments 5 and 6 (CF stress at germination only) increased significantly (P < 0.001) during the germination stage. Further, they reached significantly higher values than those recorded when CF stress was imposed at maturation (treatments 3 and 4) (Fig. 3a). The calli from treatments 7 and 8 were exposed to CF during both maturation and germination stages. For treatment 7, the percentage callus necrosis increased significantly from the maturation to the germination stage, but no such difference was obtained for treatment 8. Again, there was no apparent effect of EMS on callus necrosis (treatments 3 vs. 4, 5 vs. 6, 7 vs. 8).

The results from all of the EMS and CF treatments, indicate that plantlet yield was not affected by EMS (treatments 1 and 2) but generally decreased with the severity of the CF-imposed stress (Fig. 3b). As in the previous investigation (Fig. 1b), the number of plants produced from calli exposed to CF during the germination stage (treatments 5, 6, 7 and 8) was significantly fewer (P < 0.001) than those from calli cultured on medium with CF during maturation (treatments 3 and 4). Despite treatments 7 and 8 having CF pressure in both stages, and treatments 5 and 6 in germination only, no significant differences in plantlet yield were observed (Fig. 3b). However, treatment 8 produced significantly more plants than treatment 6. The number of plants produced from the EMS + CF at maturation (treatment 4) was significantly higher (P < 0.001) than from the other EMS treatments (6 and 8). A relationship was again observed between number of plants and percentage callus necrosis, with treatments that exhibited high percentage callus necrosis producing low plant yields (treatments 5–8). All treatments produced a relatively small percentage of abnormal plants, which included albino and chimeric individuals (Fig. 3b).

Since Fusarium CF inhibits root growth (Chen and Swart 2002; Khan et al. 2004), all of the plants that survived the manipulations described above (Fig. 3) were further screened for tolerance to CF based on root re-growth in the presence of 1,500 ppm CF for 3 weeks (Fig. 4). The root lengths for plants from most EMS treatments (treatments 2, 4 and 8) displayed wider interquartile ranges (distance between the 25th percentile and 75th percentile) than their corresponding controls [1 (+CF), 3 and 7, respectively] (Fig. 4). In addition, the EMS treatments 2, 6 and 8 had wider total ranges than their respective controls (1, 5 and 7) (Fig. 4). This reflects a greater variation in root length in the plants regenerated from the EMS-treated calli than in those from the non-EMS treatments. Furthermore, the EMS treatments resulted in more plants with at least 10 mm in root length than the non-EMS treatments (with the exception of treatment 6) (Fig. 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00299-012-1359-0/MediaObjects/299_2012_1359_Fig4_HTML.gif
Fig. 4

The effect of EMS on root re-growth after 3 weeks in the presence of 1,500 ppm culture filtrate. The boxes represent the interquartile range and the lower, middle and upper limits of each box represent the 25th percentile, median and 75th percentile, respectively. The lower and upper bars represent lowest and highest values recorded, respectively. The numbers above the bars indicate the percentage number of plants (outside the brackets) and actual number of plants (in brackets), with a root length above the 10 mm threshold (dashed line), n = 6–45

All plants with root length of 10 mm or more were inoculated with F. sacchari PNG40 by stabbing with F. sacchari-colonised toothpicks. Inoculated controls exhibited leaf crinkling, chlorosis and necrosis 3–4 weeks after stabbing the stems with F. sacchari-colonised toothpicks, and their shoot growing point dried and died after 7–8 weeks. Wilting leaves and dead shoot growing points were also observed in some plants from treatment 8 and one plant from treatment 3 became chlorotic and then necrotic. Longitudinal sections of stems from dead and live plants, from all treatments, revealed lesions of varying severity progressing from the stabbed area. Dead plants were observed only in the inoculated control and in treatment 8 (Table 4). Lesion severity rating for symptomless plants ranged from 1 to 2, and from 2 to 3 in the symptomatic ones (Table 4).
Table 4

A summary of disease response and tissue colonisation by Fusarium sacchari PNG40 in plants (treatments 1–8) 2 months after toothpick stab inoculation with the fungus. Plants selected from the root re-growth test were inoculated and re-isolation was carried out from the lesion and area above it to confirm the presence of the fungus in the plant tissues. Inter simple sequence repeats (ISSR)-PCR analyses were used to confirm identity of isolates as PNG40

Treatment

Plant no.

External symptomsa

Severity of lesionb

Dead (D) or alive (A)

Re-isolation on NS agar from lesionc

ISSR comparison of isolate with PNG40d

Re-isolation on NS agar from undamaged area above lesione

ISSR comparison of isolate with PNG40f

Resistant (R), tolerant (T) or susceptible (S)

T1 (sterile toothpick)

1a

0

0

A

N

nd

N

nd

ni

1b

0

0

A

N

nd

N

nd

ni

1c

0

0

A

N

nd

N

nd

ni

T1 (inoculated with PNG40)

1d

1,2,3,4,5

3

D

Y

+

nd

nd

S

1e

2,3,4,5

3

D

Y

+

nd

nd

S

1f

2,3

2

A

Y

+

N

nd

S

1g

0

2

A

Y

+

Y

+

T

1h

3,4

3

A

Y

+

N

nd

S

1i

3,4

3

A

Y

+

N

nd

S

1j

3,4,5

3

D

Y

+

nd

nd

S

T2

2a

0

1

A

Y

+

Y

+

T

2b

0

2

A

Y

+

Y

+

T

2c

0

2

A

Y

+

N

nd

R

2d

0

2

A

Y

+

N

nd

R

T3

3c

0

1

A

Y

+

Y

+

T

3g

0

2

A

Y

+

Y

+

T

3j

2,3

2

A

Y

+

N

nd

S

T4

4a

0

1

A

Y

+

N

nd

R

4h

0

1

A

Y

+

Y

+

T

4l

0

2

A

Y

+

N

nd

R

T5

5a

0

2

A

Y

+

N

nd

R

5b

0

2

A

Y

+

N

nd

R

5c

0

3

A

Y

+

N

nd

S

5e

0

2

A

Y

+

N

nd

R

5f

0

1

A

Y

+

N

nd

R

T6

6b

0

2

A

Y

+

Y

+

T

6d

0

1

A

Y

+

N

nd

R

T7

7a

0

1

A

Y

+

Y

+

T

7b

0

1

A

Y

+

N

nd

R

T8

8a

0

2

A

Y

+

N

nd

R

8f

4,5

3

D

Y

+

nd

nd

S

8g

0

1

A

Y

+

Y

+

T

8i

4,5

3

D

Y

+

nd

nd

S

8j

0

2

A

Y

+

N

nd

R

8k

0

2

A

Y

+

N

nd

R

8n

0

1

A

Y

+

N

nd

R

8r

4,5

3

D

Y

+

nd

nd

S

8s

0

1

A

Y

+

N

nd

R

8t

4,5

3

D

Y

+

nd

nd

S

ni not inoculated

aExternal symptoms: 0, no symptoms; 1, leaves crinkling; 2, leaves chlorotic; 3, leaves necrotic; 4, dead growing point; 5, wilting

bLesion severity rating: 0, no lesions; 1, mild; 2, moderate; 3, severe

cY, fungus re-isolated; N, fungus not re-isolated

dIsolate retrieved from lesion: +, isolate similar to PNG40; −, isolate not similar to PNG40; nd, not determined (no fungus retrieved)

eY, fungus re-isolated; N, fungus not re-isolated; nd, not determined (no undamaged tissue)

fIsolate retrieved from undamaged area above lesion: +, isolate similar to PNG40; −, isolate not similar to PNG40; nd, not determined (no fungus retrieved)

Detection, re-isolation and confirmation of identity of F. sacchari PNG40 from putative-tolerant plants

Asymptomatic plants supporting endophytic colonisation were distinguished from the negative control plants by visualisation of hyphae (stained using lactophenol blue) in between plant cells above the inoculation lesion (results not shown). The presence of Fusarium was confirmed in stems of symptomatic, asymptomatic and dead plants, and symptomatic leaves by surface sterilising and placing transverse sections of these organs on selective Nash and Snyder (1962) medium. The test for microbial growth after pressing stems and leaves on PDA was negative for all samples, thus confirming the effectiveness of the surface sterilisation. Fusarium-like colonies grew from the lesions of the dead, symptomatic and asymptomatic plants and the symptomatic leaf sections. No growth was observed from the stem sections from non-inoculated plants. Fusarium could not be re-isolated from undamaged tissue above the lesion in 15 of the 24 asymptomatic plants from all treatments, but the fungus was retrieved from the other nine plants (Table 4). Visualisation of the fungus in undamaged tissue was also achieved in three asymptomatic plants. However, the retrieval of Fusarium from undamaged tissue of nine plants, including the three from which the fungus was visualised, asserted re-isolation as a better approach for determining endophytic colonisation than microscopic observation.

Molecular analysis was carried out on DNA from the fungal isolates retrieved from the plants. The efficacy of primers ISSR 1, ISSR 2, ISSR 4 and ISSR 8 in distinguishing between different Fusarium isolates and species was tested by subjecting six isolates of F. andiyazi, F. proliferatum and F. sacchari (two isolates/species) to ISSR-PCR. The number of monomorphic, polymorphic bands and unique banding patterns obtained from the Fusarium isolates using each primer indicated that a combination of primers ISSR 1, ISSR 4 and ISSR 8 was able to separate each of the six different Fusarium isolates (results not shown). Using these primers, banding patterns generated from isolates retrieved from the lesions and the undamaged area above them were the same as for PNG40 (Table 4).

F. sacchari PNG40-tolerant and -resistant plants

The inoculation with F. sacchari PNG40 confirmed the tolerance or resistance of CF-selected plants as follows: (1) Six of the seven plants from treatment 1 inoculated with PNG40-colonised toothpicks (positive controls) displayed symptoms, lesion severity ratings (LSR) of 2–3 and three of them died after 8 weeks. A single control was symptomless and showed an LSR of 2. Stabbing with sterile toothpicks had no adverse effects on the plants (negative controls) (Table 4). (2) Twenty-three out of the 29 inoculated plants from all the other treatments were symptomless and exhibited a LSR of 1–2, 8 weeks after toothpick stab inoculation (Table 4). Re-isolation of Fusarium from undamaged tissue and ISSR analyses of the isolates from eight of these plants indicated that they permitted potential endophytic colonisation by F. sacchari PNG40 and were, therefore, tolerant to the fungus (Table 4). The single asymptomatic inoculated control was also considered tolerant. In the other 15 asymptomatic plants, the fungus was retrieved from the inoculation lesions only. These plants were classified as resistant to the fungus as they appeared to limit PNG40 growth and proliferation. (3) Six plants were regarded as susceptible as one plant showed symptoms only (treatment 3), another exhibited an LSR of 3 (treatment 5) and the other four displayed symptoms, a LSR of 3 and died (treatment 8) (Table 4).

Discussion

Callus and plantlet response to fungal culture filtrate

Fungi produce toxic metabolites that are involved in plant pathogenesis (Yoder 1980). In vitro plant cultures of susceptible and tolerant genotypes vary in their response to these compounds, thus making such phytotoxins useful in in vitro selection strategies for disease tolerance (Binarova et al. 1990). In this regard, establishment of an appropriate phytotoxin concentration that negatively affects cells, tissues, organs and whole plants increases the probability of obtaining tolerant lines (Daub 1986). Fusarium spp. culture filtrates and purified toxins have been used widely in callus (Toyoda et al. 1984; Hidalgo et al. 1998; Thakur et al. 2002), root growth (Baker et al. 1981; Chen and Swart 2002; Khan et al. 2004) and leaf necrosis (Hidalgo et al. 1998) tests to select tolerant genotypes. Similarly, in the present study, inclusion of different concentrations of F. sacchari PNG40 CF in the culture media resulted in deleterious effects on both embryogenic callus and plantlets. The fungal filtrate inflicted dose-dependent callus necrosis (Fig. 1a) and consequent decrease in plant yield (Fig. 1b), as well as inhibition of root re-growth in regenerants (Fig. 2). Furthermore, necrosis was found to be a better indicator of the effect of CF on calli during embryo maturation than callus mass as no significant differences in fresh and dry mass were obtained amongst the CF treatments. However, the effect of the CF was more severe when incorporated at the embryo germination stage than at maturation, as evidenced by the higher levels of necrosis at the former than the latter (Fig. 1a). Further, there were no significant differences in necrosis at the germination stage after inclusion of CF in both maturation and germination media and when it was added in the germination medium only (Fig. 3a). This is significant as no reports on incorporation of CF at the germination stage of embryogenesis were found in the literature. In most studies, for example in alfalfa (Binarova et al. 1990), gladiolus (Remotti et al. 1997), carnation (Thakur et al. 2002), and rose geranium (Saxena et al. 2008), CF was supplied during embryogenic callus induction and surviving calli were then transferred to germination medium without CF. The observed response of CF during embryo germination may be due to light-dependant toxin activity (Asai et al. 2000) or differential effects of the CF on the various biochemical and physiological processes occurring at embryo germination or maturation stages. Nevertheless, a few necrotic calli produced plants suggesting that a callus piece could consist of both tolerant and susceptible cells. However, inadequate exposure of susceptible embryos to the CF may also have resulted in production of plants, i.e. ‘escapes’. In a study to select tomato plants tolerant to Fusarium CF, Toyoda et al. (1984) pointed out that it is critical to ensure infiltration of the CF to deeper callus cell layers to avoid escapes. In that study, adequate exposure to F. oxysporum CF was achieved by submerging tomato calli placed on semi-solid medium incorporated with CF, and in liquid medium also containing the CF (liquid-on-agar method). In another approach, Jin et al. (1996) incorporated F. solani CF in soya bean embryogenic suspension cultures to select CF-tolerant embryos. Incorporation of CF in liquid media during culture using methods such as suspension or temporary immersion cultures, together with use of smaller callus pieces can, therefore, ensure adequate exposure of callus cells to the CF. However, work in our laboratory (unpublished) has shown that sugarcane calli do not survive such immersion treatments.

In the current study, ‘escapes’ that may have resulted as a consequence of inadequate exposure of cells to CF stress at the embryo maturation and germination stages were discarded by assessing root re-growth in regenerated plants on CF-containing medium (Fig. 4). In addition, as tolerance expressed by single cells may vary from that of the whole plant (Van den Bulk 1991), this test was employed to screen the putative tolerant lines at the whole plant level. The inhibition of root growth by Fusarium CF reported in other studies (Chen and Swart 2002; Khan et al. 2004) and that observed in the current one using F. sacchari CF, proved that assessing root re-growth is a suitable approach for this purpose. Hence, the observed responses of calli and regenerated plants to the CF treatment indicated its suitability as an in vitro selecting strategy for screening plants tolerant to this toxin at various stages of morphogenesis. In this regard, the established CF concentration for selection during embryogenesis was 100 and 1,500 ppm for root re-growth.

Production of EMS-induced variants and in vitro and ex vitro selection for tolerance to F. sacchari

The ability of EMS to induce mutations in plants is well documented (Jabeen and Mirza 2002; Lee et al. 2003; Hoffmann et al. 2004). Mutagenesis can enhance in vitro-induced variation, thereby improving the probability of selecting plants with desired attributes. Although in vivo mutagenesis using EMS has been undertaken in a number of plant species (Khairwal et al. 1984; Jabeen and Mirza 2002; Sharma et al. 2010), the use of in vitro cultures allows for the treatment of single cells, thus preventing formation of chimeras, and screening and selection under controlled conditions. In this regard, there are several reports on EMS treatment of calli, followed by appropriate screening protocols, for production of plants tolerant to herbicides (Jander et al. 2003; Koch et al. 2012), salt (Luan et al. 2007) and diseases (Imelda et al. 2000; Purwati and Sudarsono 2007; Matsumoto et al. 2010). An attribute of EMS that makes it a popular mutagen is its ability to induce high point mutation frequencies without causing lethal abnormalities to the chromosomes (Waugh et al. 2006). In the present study, there were no significant differences in callus mass and necrosis (Fig. 3a) between EMS-treated and non-treated calli. As stress in calli due to EMS treatment, i.e. growth inhibition (Svetleva and Crino, 2005) and necrosis (Koch et al. 2012) has been reported to be directly related to the mutagen dose, these observations indicated the suitability of the employed mutagenic treatment. However, whilst it is important to minimise negative effects during a mutagenic treatment, it is also critical to attain high mutation frequencies, i.e. variation. Sadat and Hoveize (2012) reported that treatment of sugarcane calli with 32 mM EMS for 4 h induced variation whilst causing minimal negative effects. Similarly, in the current study, although the EMS treatment did not induce negative effects in calli, variability was achieved as evidenced by plants with greater contrast in root re-growth in the presence of CF pressure than the controls (Fig. 4). This observation suggested that treatment of calli with EMS improves mutation frequency compared with in vitro culture without the mutagen.

Fusarium sacchari CF-tolerant plants were obtained from the variants generated from the EMS treatment by selection of treated embryogenic calli and subsequently regenerated plants with the desired mutation(s) using CF at the established concentrations. More plants (2–5 times) with improved root re-growth were obtained from EMS treatments (Fig. 4, treatments 2, 4 and 8) than their respective controls (Fig. 4, treatments 1(-CF), 3 and 7), suggesting that the mutagenic treatment resulted in enhanced ability to overcome CF-induced root re-growth inhibition, i.e. increasing the production of CF-tolerant plants. It is possible that the generation of a mutation(s) that deters the action of root-growth inhibiting compounds produced by Fusarium spp., e.g. fusarubin, javanicin, anhydrofusarubin (Baker et al. 1981) and the peptide Nep1 (Bae et al. 2006), may have led to this result. Challenging the plants exhibiting improved root re-growth with F. sacchari PNG40 verified their tolerance or resistance to the fungus (Table 4). This ability of EMS to induce tolerance to Fusarium spp. has been described in other studies (Purwati and Sudarsono 2007; Shah et al. 2009; Matsumoto et al. 2010). Sharma et al. (2010) reported that in vivo EMS treatment of seeds was more effective in producing genotypes tolerant to Fusarium wilt in garden pea than selection of in vitro culture-induced variants using CF. However, a combination of in vitro mutagenesis using EMS and selection via CF pressure allows for better penetration and uniform exposure of callus cells to both the mutagen and the CF, thereby improving chances of selecting the desired mutation(s). In addition, as observed in the present study, in vitro mutagenesis and selection allow for screening at different stages of morphogenesis, i.e. embryo maturation, germination and plantlet establishment. The results indicated that screening at the embryo germination stage is more stringent than at maturation. Further, escapes can be discarded and putative-tolerant lines selected from the surviving regenerants by assessing root re-growth in whole plants exposed to CF pressure.

Based on studies in which tolerance to CF displayed by calli was uncorrelated to that exhibited to the toxin-producing pathogen (Vardi et al. 1986; Rowe and Stortz-Lintz 1993), there is a possibility of plants achieving tolerance to other compounds present in the CF and not the putative toxin (Van den Bulk 1991). Inoculation of CF-selected plants with the toxin-producing fungus is, therefore, necessary to confirm tolerance. Afolabi (2008) reported that toothpick stab inoculation of maize with Fusarium spp. induced lesions in the stem and severity was associated with the level of tolerance of the genotypes. In the current study, after toothpick stab inoculation with PNG40, most of the CF-selected plants exhibited no disease symptoms, whilst those from the controls (not exposed to CF or EMS) (treatment 1) were symptomatic or dead and displayed lesion severity ratings of 2–3 (Table 4). Re-isolation of the fungus from dead, symptomatic and asymptomatic plants and confirmation of the identity of the retrieved fungus as PNG40 using molecular analysis (Table 4), supported the tolerance or resistance of the symptomless plants to F. sacchari PNG40. Nine of the twenty-four asymptomatic plants were regarded tolerant as they allowed PNG40 growth without causing damage to the occupied tissue (endophytic colonisation), presumably because they were unaffected by toxins produced by the fungus. Studies in maize indicate that symptomless endophytic colonisation of young plants by Fusarium spp. resulting from artificial inoculation of seeds or seedlings persists up to their maturity (Munkvold and Carlton 1997; Schulthess et al. 2002). Similarly, endophytic colonisation of young plants by F. sacchari was determined in the current study and it is also possible that the fungus’ occupation of plant tissue may perpetuate over long periods. However, a 10-month investigation has been initiated to confirm such long-term colonisation in these plants. The other 15 symptomless plants appeared resistant to F. sacchari as they limited the presence of PNG40 to the inoculation lesion, possibly due to biochemical or physiological mechanisms that inhibited fungal growth and proliferation in the plant tissues (Roy and Kirchner 2000).

These observations support a correlation between tolerance to the CF in vitro and that displayed to the organism (F. sacchari), thereby confirming the usefulness of CF as an in vitro selection agent. Such a validation has been reported in several crop species (Gengenbach et al. 1977; Ling et al. 1985; Wenzel and Foroughi-Wehr 1990; Sengar et al. 2009). Fusarium spp. produce phtyotoxins, such as fumonisin B1, moniliformin (Van Asch et al. 1992) and fusaric acid (Diniz and Oliviera 2009), known to elicit disease development. As illustrated in the present study, it can be expected that developing tolerance to such compounds (purified or in culture filtrate) may also lead to tolerance or resistance to Fusarium spp. However, Fusarium CF are also known to contain mutagenic compounds, e.g. Fusarin C (Lu and Jeffrey 1993). Successive exposures to the CF during screening may, therefore, lead to undesirable mutations which result in plants exhibiting susceptibilities to biotic and abiotic factors (Matsumoto et al. 2010). This may explain the high mortality observed in treatment 8 in the present study.

The reported findings, therefore, indicate that plants both tolerant and resistant to F. sacchari can be produced by treating calli with 32 mM EMS for 4 h, exposing them to 100 ppm CF at the embryo germination stage and assessing root re-growth in regenerated plantlets in the presence of 1,500 ppm CF. Whilst phenotypic analysis of these EMS-induced variants and tests to establish long-term presence of endophytic F. sacchari in tolerant plants are necessary in ongoing work, the present study indicated the feasibility of the described approach in obtaining tolerance to the fungus. This protocol has value in developing tolerance and resistance to F. sacchari in commercially important cultivars and possibly even to other sugarcane pathogens. Further, genotypes expressing these traits may be used as parents to provide sources of tolerance and resistance genes in sexual crosses for development of new commercial cultivars. The tolerance to F. sacchari PNG40 and associated endophytic colonisation by the fungus obtained using the approach described in this study, are essential prerequisites towards the exploitation of PNG40 in biological control strategies against the E. saccharina. Future investigations will employ these tolerant genotypes to test the effect of the fungus on larval survival and growth of the stalk borer.

Acknowledgments

Many thanks to the National Research Foundation (NRF)-South Africa and the South African Sugarcane Research Institute (SASRI) for funding to carry out this study.

Copyright information

© Springer-Verlag Berlin Heidelberg 2012