EFFECT OF ARBUSCULAR MYCORRHIZAL FUNGI AND SOIL PHOSPHORUS LEVEL ON EXPRESSION OF PROTEIN AND ACTIVITY OF PEROXIDASE ON PASSION FRUIT ROOTS

The effects of mycorrhizal inoculation and increasing soil P levels on the expression of total proteins and peroxidase activity on passion fruit roots were evaluated. The experimental design was entirely at random, with four treatments of inoculation (a – control; b – Gigaspora albida; c – Scutellospora heterogama; d – mixture of G. albida, G. margarita, S. heterogama, and Glomus clarum) × three levels of soil P (4, 11, and 30 mg/dm of soil), each with three replicates. Plants were harvested 70 days after inoculation, when root colonization, shoot P level, protein content, and enzymatic activity of peroxidase (PAGE – 7%) on root extract were evaluated. Regarding protein, there was no significant difference among the treatments, except between those roots receiving mixed inoculum and 11 mg P/dm of soil. Effect of P on protein concentration, when compared with the inoculation effect was observed. For peroxidase, there was an eletrophoretic band common to all treatments (rf: 0.43) and another that was absent only in noncolonized plants, grown in soil with lower P (rf: 0.46). Mycorrhizal specific bands were not present but a small decrease of intensity of bands in noncolonized plants was observed. Conversely, the control roots presented a single band (rf: 0.33) not observed in the other extracts, that may demonstrate an inhibitory effect of AMF on some host activities. The data showed the influence of P level in soil on the protein expression of roots, suggesting the influence of this nutrient on root genetic expression as well as on the mechanisms of symbiotic control/recognition.


INTRODUCTION
Arbuscular mycorrhiza are mutualistic associations with great ecological and economical importance, formed between some fungi (Glomales) and roots of most plant species, in natural and agricultural ecosystems.Much has been speculated regarding the molecular bases that direct the functional integrity of endomycorrhiza; however, little is known about these mechanisms (Gianinazzi-Pearson & Gianinazzi, 1995).Physiological and biochemical alterations have been observed in mycorrhizal roots, with production of specific proteins (endomycorrhizins) which are probably related with aspects of recognition between the symbionts (Gianinazzi-Pearson, 1992;Lambais, 1996).The differential expression of some proteins, as well as the mycorrhizal colonization may be regulated by the plant P level (Gianinazzi et al., 1992).
There are few observations related with specific products of colonization by endomycorrhizal fungi, including the arbuscular (Gianinazzi et al., 1992).Gianinazzi-Pearson & Gianinazzi (1995) compared a few studies and observed that there are more results showing higher protein concentration in mycorrhizal plants, the majority of them with unknown functions.However, other analyses do not show this effect, which indicates that further studies should be performed for better comprehension of this phenomenon.Hepper & Sem (1986) investigated the enzymatic and proteic components of spores of AMF and mycorrhizal roots, observing similar activities for peptidases and malato dehydrogenase and others for sterases in different fungal species.These authors mentioned that the proteic pattern in the treatments was relatively constant for spores and hypha (external and internal) during mutualism, while roots (mycorrhizal or not) varied in the band pattern.Additional information regarding variability of the proteic pattern in isolated species and physiological stages of AMF spores was presented by Avio & Giovannetti (1998): Glomus mosseae and G. caledonium presented equal patterns of bands, independently of the associated host, while different isolates of G. coronatum showed the same protein pattern.It was further mentioned that the physiological state of spores may also change the quantity and quality of bands.Mohr et al. (1998) analysed in detail defense gene expression in bean roots (Phaseolus vulgaris L.) colonized by an AMF (G.mosseae) and a phytopathogenic fungus (Fusarium solani f. sp.phaseoli).
They observed that mycorrhizal colonization had no effect on the expression of plant defense genes, as opposed to the expressive increase of enzymatic activity found in the pathogenic interaction compatible with the Fusarium, which was related with production of chitinase, glucanase, and PAL (phenilalanine ammoniumliase).These data differ from results of other investigations, where in general production and activation of enzymes during plant mycothophy have shown similarities between fungal infection processes and the early stages of development of endomycorrhizas (Bonfante-Fasolo & Spanu, 1994).
The activity of peroxidase in arbuscular mycorrhiza has been mapped because this enzyme functions as a barrier to fungi invasion of plant tissues, as a characteristic marker of stress situations.Spanu & Bonfante-Fasolo (1988) evaluated peroxidase activity in cell walls in garlic plants colonized by Glomus versiforme.They observed that the higher points of activity occur at the beginning of symbiotic establishment, although in more advanced stages the activity had decreased to the same level of non-mycorrhizal plants.These data are quite different from those presented by Dehne & Schönbeck (1978) who showed increase of peroxidase activity in mycorrhizal plants, even working with enzymes from extracts, in a fixed observation time.
The data available indicate the need for investigations about protein and enzymes synthesized in mycorrhizal plants, which are valuable tools for the ecophysiological understanding of the association, and possible factors of regulation, of the mutualistic partners.In this study we evaluated the variation on protein expression and peroxidase activity in roots of passion fruit colonized by AMF species and maintained in soils with increasing phosphorus levels.

MATERIAL AND METHODS
The experiment was performed in a greenhouse in natural conditions of temperature, light, and humidity.The maximum and minimum averages for temperature and humidity were respectively: 38.6 o C and 19.6 o C, 90% and 33%.

Fungal isolates
Four

Substrate and P levels
A red yellow Podzol was used for the experiments; its characteristics are shown in Table 1.The soil was sterilized with methyl bromite, 20 days before used.Three levels of P in the soil solution were tested (4, 11, and 30 mg/dm 3 of soil) based on nutrient recommendations for this crop (Cavalcanti et al., 1998) and using superphosphate (P 2 O 5 ).Soil moisture was maintained at 60% of total pore volume (TPV).

Plant material and mycorrhizal inoculation
Seeds of passion fruit were allowed to germinate in sterilized soil, after desinfection with 20% sodium hypochlorite/2 min.and being rinsed with distilled sterilized water.
Eight days after germination, seedlings of 4.5 cm height with only one true leaf were transferred to plastic containers with 40 g of soil, and inoculated with AMF.All inoculated treatments received 100 spores from either one species (G.albida or S. heterogama) or from four species (G. albida, G. margarita, S. heterogama, and G. clarum), in this case, 25 spores from each species.In the control treatments, 2 ml of the soil filtrate solution (1:10, soil:water, v/v) was used to add natural microbiota (except AMF).Ten days after inoculation, the seedlings were transfered to plastic bags with 2 kg of soil and maintained in the greenhouse for 70 days.

Determination of P on shoots
After harvest, the shoots were separated from the roots, dried at 65 o C, ground and processed for determining concentration of phosphorus by nitroperchloric digestion and spectrophotometric reading at 880 nm.

Protein extraction
After rinsing, sub-samples of roots were dried at room temperature, placed in plastic bags, and kept at -4 o C. For protein extraction, 0.5 g of roots from each plant were placed in a mortar with liquid nitrogen and ground with 0.3 g sucrose and 0.3 g of PVP (polivinilpirrolidona).After that, 2 ml of extraction buffer were added for new trituration.The buffer consisted of a mixture (1:9) of buffer A -litium borate 0.2 M (1.2 g litium hydroxide + 11.89 g boric acid + 1,000 ml distilled water -pH 8.3), and buffer B -tris-citrate 0.2 M (1.6 citric acid + 6.2 g tris + 1,000 ml distilled water -pH 8.3) (Scandalios, 1969).The macerated material was transfered to Eppendorfs and centrifuged at 14,000 rpm/10 min.at 40 o C. The supernatant was either transfered to new tubes and immediately used for electrophoresis or frozen for later analysis.Protein concentration was estimated by the Bradford (1976) method through linear regression, using a BSA curve (bovine serum albumin) as a pattern.

Electrophoresis
For qualitative analysis, the samples were submitted to horizontal electrophoresis for native proteins in gel of 7% poliacrylamide (72 ml buffer A + 8 ml buffer B), 0.28 bis-acrylamide, 5.32 g acrylamide, 0.8 ml ammonium sulphate (10%, w/v), and 0.08 ml TEMED.Each role received 10 ml of root extract.The electrophoresis was carried out in constant current (20 mA) during 4 hours at 4 o C, using bromophenol blue as the carrier.

Densitometry
For confirmation of the electrophoretic bands, after dehydration the gel was submitted to densitometric analysis at 520 nm of wave length.

Statistical treatment
Data were submitted to analysis of variance and correlation tests using the Statistic Program (Statsoft, 1995) and means were compared through the LSD (5% of significance).Values of colonization were transformed (arcosen of square root of percentage of root colonization/100).

Root protein
The concentration of root proteins in the treatments of fungi and soil P levels were not significantly different.However, comparison of means showed significant differences among roots receiving the inoculum mixture and kept in soil with 11 mg P/dm 3 of soil and among plants associated with G. albida, and in soils with 30 mg P/ dm 3 of soil, respectively with 445 and 137 µg of protein.ml - of extract (Table 2).
The effects of each single species and of the mixture of AMF on values of root peptides were not significantly different when compared with colonized and noncolonized plants or even the inoculum treatments.
For the soil P factor, there were significant differences in protein concentrations between roots maintained in soils with 11 mg P/dm 3 and those kept in soil with 30 mg P/dm 3 of soil (Table 3).
Significative negative correlation was observed between values of protein and soil P concentration (-0.41) which was not observed in concentration of this mineral in plant tissues.

Isoenzymatic activity
Peroxidase marking in gel electrophoresis presents few differences among extracts of the treatments, specifically in relation to mycorrhizal and non-mycorrhizal plants.The samples presented one band in common, which was less intense in non-colonized plants grown in soils with 11 and 30 mg P/dm 3 of soil (lines 5 and 9, Fig. 1).However, from the protein amount, it was observed that this effect was not proportional to the concentration of total proteins, since more strongly marked samples received lesser amounts of protein than the others (lines 9 and 10, Fig. 1) with 2.32 and 1.69 µg of protein.ml - of extract, respectively.This indicates not only a quantitative but also a proportional alteration.
In most of the samples, there was a band above the main one, which was absent in non-inoculated plants grown in substrate with less P amounts (treatments 1 and 5).In the control roots, at the level of 3 mg P/dm 3 of soil (treatment 1) a band with lower molecular weight (rf: 0.46) was observed that was not present in the other extracts.Specific bands for mycorrhizal activity, which were expected in associated plants, were not seen.

DISCUSSION
Absence of significative variance among protein values in the treatments differs from earlier results where qualitative-quantitative physiological changes in the gene expression of fungi and hosts producing different families of proteins were registered.In a revision paper, Gianinazzi-Pearson & Gianinazzi (1995) observed that results showing increases in protein concentration in root extracts of mycorrhizal plants are more frequentely found.Arines et al. (1993) also noticed high content of protein in roots of Trifolium pratense L. colonized by Glomus mosseae, although noting that during development of arbuscular mycorrhiza, changes in the polypeptyde expression happens in low levels, and that purification of the extracts is necessary in order to visualize possible changes in genetic translation.In our research, the extracts were not purified and this might have interfered in the results.Statistical differences among means of root proteins when soil P effect is observed, may demonstrate the role of this mineral on regulation of symbiosis through synthesis, stimulation, or activation of proteins and enzymes (Lambais & Mehdy, 1995).Smith et al. (1985) studied the effect of mycorrhization and phosphate fertilization on the expression of glutamate synthetase (GS) and glutamate dehydrogenase (GD) in roots and shoots of T. subterraneum and Allium cepa.For T. subterraneum the authors did not observe an apparent effect of mycorrhization or fertilization on activity of GD in plant tissues.In GS the activity was low when plants developed in soil deficient in P, but increased with P fertilization and mycorrhizal colonization, which made it difficult to isolate the effect of each factor.In onion plants, similar responses were observed with increase in enzymatic activity (GS) following directly the soil P level and intensity of colonization by AMF.Conversely, Gianinazzi et al. (1992) pointed out that the activity of alcaline phosphatases is enhanced when the host plant grows in P-deficient substrates.These data agree with those of Tibbett et al. (1998) who mentioned that with species of the ectomy-Fig. 1 -Peroxidase activity (PAGE -7%) in extract of passion fruit roots, grown in soils with 4 mg P/dm 3 of soil (1, 2, 3, 4), 11 mg P/dm 3 of soil (5,6,7,8), and 30 mg P/dm 3 of soil (9,10,11,12).Lines 1, 5 and 9: non-inoculated; 2, 6 and 10: Gigaspora albida; 3, 7 and 11: Scutellospora heterogama; 4, 8 and 12: mixed inoculum (G.albida, G. margarita, S. heterogama and Glomus clarum).Big arrow: common band for all extracts.Small arrow: varied band, absent in control treatments, with lower levels of soil P. PTN = amount of proteins applied.corrhizal fungus Hebeloma there was increase of production of acid phosphatase when the microrganisms were maintained in media with low organic P.This may demonstrate the strong effect of this mineral on the physiology of all mycorrhiza, not only on those of the arbuscular type.However, Epstein (1975) emphasizes that phosphorus has a direct effect on diverse enzymatic plant processes, very often acting as a biochemical co-factor.Detailed studies regarding effects of P and fungal colonization on protein expression of roots are preliminary, and a higher number of observations are necessary for broad understanding of this process.Lambais (1996) mentioned that for some authors the association would be regulated by the P level in plant tissues and not by the soil P level.

Level of soil P (mg/dm
There was no significant interaction between amount of P on plant shoots and amount of root peptides.This can be due to problems related with the time (70 d) for harvesting the experiment.O' Keefe & Sylvia (1992) affirmed that P absorbtion may have more impact at the beginning of mycorrhizal colonization.The soil P, however, presented negative interaction with protein concentration.This may indicate an inhibitory effect on expression of some proteins, occasionally more greatly produced during the physiological stage in which the plants were harvested.These effects were mentioned before in some mycorrhizal interactions (Gianinazzi et al., 1992;Tibbett et al., 1998).Peroxidase activity has been investigated as a marker of mycorrhizal development since it works in the plant defense system.Spanu & Bonfante-Fasolo (1988) observed that the zimogram of peroxidase showed less strong bands in noninoculated plants grown in intermediate and high P concentrations, which is in accordance with results of Dehne & Schönbeck (1978).Spanu & Bonfante-Fasolo (1988) noticed high amount of peroxidases from cell walls of Allium porrum L. roots which were beginning mycorrhizal development (16 d); after that there was a decrease in activity which, reached the same level that of noncolonized plants (50 d).For the authors, this might be due to the direct effect of AMF (Glomus versiforme) with massive colonization signifying a low level of enzymatic activity and vice versa.These data support the idea that low levels of hydrolitic enzyme activity and transcription of plant defense genes would be related with delayed recognition between the partners, promoting the functional integrity of the symbioses (Morh et al., 1998;Blilou et al., 1998).However, differences found in the responses may be due to variations in the methodology used (soluble enzymes or those from cell wall) or even may reflect the result of the complexity of the mycorrhizal system regarding evaluation of peroxidase activity on colonized and noncolonized plants.
The additional band observed on the control extract may demonstrate an inhibitory effect of the mycorrhizal fungus or the effect of high concentration of soil P on translation of specific genes.García-Garrido et al. (1999) studied the effect of xyloglucans on expression of endo-and exoxyloglucanases, observing that many bands present in non mycorrhizal material were absent in the association as well as in the mycelium, which suggests qualitative inhibitions in some host activities.Although some investigations have demonstrated the effect of concentration of soil P on enzyme expression, there are no data available connected with peroxidase.Effect of this factor (P concentration) on intensities of enzymatic band was not detected in our studies.
The inoculation with different isolates (alone or in a mixture) did not affect the amount of protein in passion fruit, indicating an indifferentiated reaction of this plant in relation to mycorrhizal colonization.In some cases, purification of the root extract is necessary in order to show its quantitative alterations.On the other hand, soil P level had a more relevant effect on the genetic expression, which reinforce theories regarding defense systems and control of symbioses as being modulated by P amount in the soil.
Peroxidase activity apparently responds more strongly to the fungus effect, presenting bands with low intensity for non-mycorrhizal plants (except for treatment 1).The appearance of enzymatic bands in material from noncolonized roots may indicate an inhibitory effect of AMF colonization on the expression of some host root protein.For mycorrhiza formed between passion fruit and the AMF tested, peroxidase dye is not useful as an ecophysiological marker, since mycorrhizal specific bands were not observed.

TABLE 2 Effect of inoculation with arbuscular mycorrhizal fungi and soil P levels in concentration of passion fruit root proteins.
1 -non-inoculated; 2 -Gigaspora albida; 3 -Scutellospora heterogama; 4 -mixed inoculum (G.albida, G. margarita, S. heterogama and Glomus etunicatum).Means followed by the same letter are not significantly different from the LSD test at 5%.