Plant Cell Reports (1995) 14:626-629 Plant Cell Reports �9 Springer-Verlag 1995

Root exuded nod-gene inducing signals limit the nodulation capacity of different alfalfa varieties with Rhizobium meliloti

Georgina Hernfindez, Mario Ramirez, Ramtn Sufirez, and Sara Isabel Fuentes

Centro de Investigacitn sobrc Fijaci6n de Nitrtgeno, Departamento de Ecologia Molecular, Universidad Nacional Auttnoma de Mtxico, Apdo, Postal 565-A, Cuernavaca, Morelos, Mtxico

Received 10 May 1994/Revised version received 25 January 1995 - Communicated by J.M. Widholm

Summary. Different nodulation capacities were found among nine different varieties of alfalfa, cultivated in the Central region of Mexico, by Rhizobium meliloti 2011. A correlation between nodulation capacity and foliar dry weight was observed, which points to a genotype dependance on these parameters. A correlation between the nodulation capacity and the R. meliloti nod-gene inducing activity of the root exudates from the different varieties, as measured by I]-galactosidase induction in a test system consisting of a R. meliloti nodC-lacZ strain incubated with each root exudate, was established. When the root exudate from the best nodulating variety was added to the four poorest nodulating varieties, an increase in nodule formation was observed. We conclude that root exuded nod-gene inducing signals are a symbiotically-limiting component in natural populations of the poorest nodulating varieties of alfalfa.

Key words: Alfalfa/R. meliloti symbiosis - Nodulation - Symbiotic nitrogen fixation - nod gene induction- Alfalfa varieties

Introduction.

Alfalfa (Medicago sativa), a most important forage crop, can establish symbiosis with the soil bacteria Rhizobium meliloti. During symbiosis, structures known as nodules are formed within which the fixation of atmospheric nitrogen occurs. The earliest events of alfalfa nodule formation require the expression of R. meliloti nodulation (nod) genes. Several compounds, including different flavones and betaines, exuded by alfafa roots or seeds, induce the transcription of R. meliloti nod-genes (Phillips et al. 1993). The regulation of nod-genes occurs through the cooperative action of the products of the nodD genes. Presumably signals released by alfalfa convert the nodD product to an active form, and this altered NodD protein then induces the expression of the other nodulation genes.

The availability of nod-gene inducing signals in the rhizosphere may limit alfalfa seedling growth. A

Correspondence to: G. Hernfindez

significant increase in nodulation, dry matter and the amount of nitrogen derived from nitrogen fixation was observed in the Hairy Peruvian (HP) variety of alfalfa when the rhizosphere was supplemented with luteolin (Kapulnik et al. 1987). It was shown (Kapulnik et al. 1987) that the HP32 population, which was selected for an increase in total dry matter and crude protein concentration and which assimilates more carbon and nitrogen than the parental HP, produces higher levels of flavonoid nodulation signals, such as luteolin.

In this work, we evaluated the nodulation capacity of different alfalfa germplasm with R. meliloti. Nine of the alfalfa varieties studied are cultivated in the Central region of Mexico, they include introduced or foreign varieties and Mexican varieties (Fuentes at al. 1993); their nodulation capacities with R. meliloti has not been evaluated, either in laboratory conditions or in the field. The Iroquois variety, widely used for symbiotic studies in different laboratories, was included in this study for comparison. We observed different nodulation capacities and a good correlation between this parameter and the plant yield among the varieties studied. Our results indicate that the variation of nodulation capacity among different alfalfa genotypes correlates with different concentrations of R. meliloti nod-gene inducing signals present in the root exudate.

In addition, we found that the nodulation of the varieties which exuded lower levels of nod-gene inducing signals, can be improved with the addition of root exudate from the variety with the highest nodulation capacity. Therefore, we concluded that these signals are limiting for the nodulation capacity of certain genotypes studied.

Materials and methods.

Plant material. Seeds of the foreign (Moapa 69, San Joaqnfn, Hairy Peruvian and Valenciana) and the Mexican (Puebla 76, Inia 76, Bajto 76, Sintttico I and Sintttico II) alfalfa (Medicago sativa) varieties (Fuentes et al. 1993) were kindly provided by L, Castro- Acero (Centro de Investigaciones Forestales y Agropecuarias del Estado de Mtxico CIFAP). Alfalfa seeds of the Iroquois variety were obtained from Agway, Inc. (Plymouth IN, USA).

627

Plant growth conditions. Alfalfa seeds (50 seeds per variety) were surfaced sterilized (Fuentes et al. 1993) and germinated in plastic growth pouches (Seed-Pack Growth Pouch, Vaughan's Seed Company, Downers Grove, IL. USA) with ten seeds in each growth pouch. Four days after germination, the seedlings were inoculated with a suspension of Rhizobium meliloti 2011 (wild type strain; Mulligan and Long 1985) prepared from an overnight culture grown on rich medium and adjusted to 1.0 OD54 o. Ten ml of inoculum was added per growth pouch. Plants were incubated at 25~ with a 16 h photoperiod provided by cool white fluorescent lights (541xmol.m-2.s ~) and were watered with Jensen liquid medium (Vincent 1970) as necessary.

Determination of symbiotic parameters. Six weeks after inoculation, plants from each variety were harvested and combined at random into 10 groups (5 plants per group). Nodule number, foliar dry weight and nitrogenase activity were determined as the symbiotic parameters for each group of plants. Nitrogenase activity was determined by the acetylene reduction method, and it is expressed as nmol ethylene rain ~ per group of plants.

Preparation of root exudates. Alfalfa root exudates were isolated from plant cultures as follows. After surface disinfestation 150 alfalfa seedlings of each variety were placed in a stainless steel grid, so that the roots could be immersed asceptically in 50 ral sterile distilled water placed below the grid. The cultures were incubated for 5 days under the plant growth conditions described above, and the root exudate was collected and centrifuged aseptically. Each root exudate was determined for bacterial contamination and only non- contaminated exudates were used as nod-gene inducers. The presence of flavones was determined by extracting the organic compounds from the root exudates with ethyl acetate and subjecting the extract to silica gel thin layer chromatography, using acetone:dichloromethane (70:30)

as a mobile phase. After TLC the flavones were detected as orange- yellow spots after oxidation with seric sulphate in sulphuric acid.

Bioassays for nod-gene inducing activity. Rhizobium meliloti 1021 carrying the plasmid pRmM57 (kindly provided by Dr. S. Long, Stanford University), which contains a nodC-lacZ gene fusion (Mulligan and Long 1985), was grown on minimal medium to early log phase. To determine the nod-gene inducing activity of root exudates, 2.5 ml of culture was centrifuged and the cells were resuspended in 4 ml of root exudate and incubated at 30~ for 3 h. Cells resuspended in distilled water were used as controls. After the induction period, [~- galactosidase activity in each sample was determined using the modified Miller method, (Mulligan and Long 1985). Enzyme activity is expressed in Miller units. To quantify the nod-gene inducing activity from each alfalfa variety, at least two different root exudates from each variety were assayed and in each case at least four samples were used to determine ~-galactosidase activity.

Results and Discuss ion.

The nodulation capacity (nodule number) of ten different alfalfa varieties with R. melliloti 201 lwas evaluated (Fig. 1) and a range in nodule number formation was observed. The analysis of variance performed showed significant differences (F(9,81) = 7 .40 , p>0.05) between the varieties. The four best nodulating varieties were: Valenciana, Inia 76, Sintttico II and Sintttico I (Tukey dt(0.05) = 11.79) (Keppel 1982), while those that

120 - -

El v

" B o

,.,_ 8 0 - c o o

,..-, r', :D , . i - ~ o ~ Q

E a ~ 4 0 - ~

r -

�9 -~ E

o 1L

O - -

I ;0

~0

i F1

L.I 121 r

r C ~ 0

i3

n

>

z 8

Z

~ 1,0 e- b- b- ~ 0 0 O"

.13 GI, " 0 0

~. r- Of)

T ,4

i~ L~ r

0

L_

n

L-

-r

- 6

4

0

O

O k.

ca.

~6 ca. .-a

r

E o r. o r

E

>

13 ~D ID O r

o .i.-- . m

Z

Fig. 1. Symbiotic parameters of different varieties of alfalfa. Nodule number (I;~l), foliar dry weight (17".7"A) and nitrogenase activity (I-"1) were evaluated six weeks after inoculation with R. meliloti 2011. Each value represents the mean and standard error (vertical line) per group of plants, considering the 10 plant group replicates from each alfalfa variety. The results are ordered from left to right based on descending nodule number. A one way analysis of variance was performed for each parameter; further differences were detrmined by a post hoe Tukey test.

628

produced the least number of nodules were: San Joaqufn, Hairy Peruvian and Iroquois (Fig. 1). The nodulation capacity of Valenciana is, also, significantly different from that of Bajio 76.

A general correlation between the nodulation capacity and the plant yield was observed (Fig. 1). An analysis of variance was also performed to analyze the differences in plant yield. The plant yield of the four best nodulating varieties was, also, significantly different from the plant yield of the two poorest nodulating varieties (Tukey dt(0.05)=27.59) (Keppel 1982). Sint6tico II gave the highest plant yield, which is significantly different from that of the other varieties, except for the other three best nodulating ones.

Nitrogenase activity was detected in every variety, indicating the effectiveness of the symbioses (Fig. 1). Nevertheless, a variation in total nitrogenase activity was observed among the varieties studied. The varieties Sint6tico II and Sint6tico I, which gave high plant yield and nodule formation, showed very high nitrogenase activity, which is significantly different from the activity present in all the other varieties, except for San Joaqufn and Moapa 69 (Tukey dt(0.05)=l.51) (Keppel 1982). The two poorest nodulating varieties (Hairy Peruvian and Iroquois) had very low nitrogenase activities, which were significantly different from that of Sint6tico I, Sint6tico II, Moapa 69 and San Joaqufn. In other studies a large variation on nitrogenase activity was observed among alfalfa genotypes and a significant amount of the variability was unexplained by the traits measured like shoot weight, fibrous root score or nodule mass score (Barnes et al. 1984).

It has been shown that genetic factors in alfalfa can influence nitrogen fixation and growth, and there have been attempts to select for specific traits, such as nodule mass or nitrogenase activity (Phillips et al. 1982; Barnes et al. 1984; Teuber et al. 1984; Kapulnik et al. 1987). The different responses to nodulation with R. meliloti that we observed in the ten alfalfa varieties we studied are in agreement with the genotype dependance of these parameters. The correlation between nodulation capacity and plant yield and the high variation in nitrogenase activity that we found are in agreement with previous studies on different alfalfa germplasm (Phillips et al. 1982; Barnes et al. 1984; Teuber et al. 1984; Kapulnik et al. 1987).

It has been demonstrated that transcription of important nodulation genes in Rhizobium is controlled by specific compounds, like flavones and isoflavones, produced in the legume host. Fusions of E. coli lacZ gene to Rhizobium nod-genes have been used to monitor the inducing activity of various plant exudates and extracts (Mulligan and Long 1985; Rolfe 1988; Phillips et al. 1993). Rhizobium nod- gene expression is directly coupled to the production of galactosidase enzyme which can be easily measured. The transcription of the well-described transcriptional unit: the nodABC-genes from R. meliloti, is induced in the presence of alfalfa seed or root exudate, as well as in the presence of specific compounds isolated from the exudates (Mulligan and Long 1985; Kapulnik et al. 1987; Phillips et al. 1993). The products of the nodD regulatory gene

family, which presumably interact with the plant signal, are essential for the expression of nodABC-genes in R. meliloti (Mulligan and Long 1985; Phillips et al. 1993). The presence of elevated levels of nod-gene inducing flavones has been interpreted as the first biochemical explanation for the ability of genetically altered alfalfa populations to increase symbiotic nitrogen fixation capacity (Kapulnik et al. 1987). For this reason, we investigated the relationship between the nodulation capacity of the varieties of alfalfa studied (Fig. 1) and the presence of nod-gene inducing signals in their root exudates.

160 - - ( e 3 )

1~'0

._o

r

I11

" ~ 8 0

4 0

(18)

gi

7/1 :: N

.. Y/l "" 7 /1

._. ~

!) (~7)

(10)

_~ ,_o g - g ~ g ~ o

Fig. 2. Induction of R. meliloti nod-genes by root exudates from the different varieties of alfalfa. The 13-galactosidase activity present in the R. meliloti 1021 (pRmM57) strain containing a nodC-lacZ gene fusion was measured after induction with root exndates obtained from each alfalfa variety. For each varietY, at least two different exudates were obtained and the enzyme activity was determined in at least four replicates per root exudate. Each value (17"~) represents the mean and standard error (vertical line) of the replicates from each variety; the first value ( ~ ) represents the control, i.e. ~-galactosidase activity present after incubation in the absence of root exudate. Numbers above the bars indicate fold induction relative to the control.

Root exudates, isolated from the same number of plants of each alfalfa variety, were isolated and assayed for inducing 13-galactosidase activity in a R.meliloti nodC- lacZ strain. The root dry weight of each group of plants used to isolate the root exudate was similar, ranging from 14 to 40 mg, and there was no significant difference among the better and poorer nodulating varieties. The qualitative TLC assay performed revealed the presence of

flavones in every root exudate (data not shown). The two best nodulating varieties, Valenciana and Inia 76 (Fig. 1), gave the highest level of ~l-galactosidase induction (22- and 23-fold higher relative to the control), while the four poorest nodulating varieties: Bajfo 76, San Joaqufn, Hairy Peruvian and Iroquois (Fig. 1) gave only 5- to 10-fold induction (Fig. 2). The other four varieties had intermediate levels o f [~-galactosidase induction (Fig. 2). The production of flavone signals that induce nod-gene transcription may also be genotype dependant in alfalfa.

Table 1. Nodulation capacity of varieties of alfalfa inoculated with R. meliloti in the presence of different amounts of Valenciana root exudate.

Variety Valenciana root exudate lml 5ml 10 ml

San Joaquin 122% (+/-26) 115% (+/-37) 180% (+/-22) Hairy Peruvian 103% (+/-27) 193% (+/-27) 188% (+/-43) Iroquois 164% (+/-26) 161% (+/-19) 166% (+/-24) Valenciana 109% (+/-24) 94% (+/-34) 90% (+/-11)

The indicated amount of root exudate obtained from Valenciana variety was added to each growth pouch containing alfalfa seedlings from the indicated variety, at the time of inoculation in 5-10 groups of plants of each alfalfa variety. The results are expressed as the percentage of the mean (+/- SE) with respect to the control group of plants (100%) from each variety (inoculated in the absence of Valenciana root exudate).

Kapulnik et al. (1987) reported that the levels of flavone nodulation signals normally present in the rhizosphere of the Hairy Peruvian variety of alfalfa can limit root nodulation, symbiotic nitrogen fixation and seedling growth. We tested if the poorer nodulation capacity of certain varieties of alfalfa (Fig. 1), which is associated with a decrease in root exuded nod-gene inducing signals (Fig. 2), may be improved by the addition of root exudate from a good nodulating alfalfa variety. Root exudate from Valenciana, which contains a high concentration of nod- gene inducing signals (Fig. 2), was isolated and added, in different amounts, to each of the three poorest nodulating varieties (San Joaqufn, Hairy Peruvian and Iroquois) at the time of inoculation (Table 1). The number of nodules formed by the treated plants was compared to untreated controls, which produced quantities of nodules similar to those shown in Fig. 1. All the poorer nodulating varieties tested (San Joaqufn, Hairy Peruvian and Iroquois), inoculated in the presence of Valenciana root exudate, showed an increase in nodule formation that ranged from 15% to 93% above the control (Table 1). As expected, when the Valenciana root exudate was added to the Valenciana variety, nodulation did not increase, but it remained close to the control values without added root exudate (100%) (Table 1). The response in increasing nodulation capacity was saturated with 10 ml of Valenciana root exudate and larger amounts did not give a greater increase in nodulation with any variety (data not shown). The Iroquois variety showed similar increase in nodulation when different amounts of root exudate were added, which indicates that the response was saturated even with 1 ml of Valenciana root exudate (Table 1). However, the response of Hairy Peruvian or San Joaqufn varieties was saturated only with 5 ml and with 10 ml, respectively (Table 1). The Bajfo 76, which is the next

629

to the poorest nodulating varieties (Fig. 1), also showed an increase (123% (+/-11)) in the number of nodules formed when 10 ml of Valenciana root exudate was added at the time of inoculation. In the San Joaqufn variety, an increase (163% (+/-30)) in plant yield was also observed when the Valenciana root exudate was added to the rhizosphere.

We conclude that some of the nod-gene inducing compounds that have been reported for alfalfa (Phillips et al. 1993) are limiting, at different amounts, for an efficient symbiosis of certain varieties of alfalfa. It was reported that in different lines of Phaseolus vulgaris (Bolafios and Wemer 1993) and in a P. vulgaris white- seeded mutant (Hungrfa and Phillips 1993) there was a correlation between increased nodulation by R. leguminosarum by. phaseoli and a higher flavonoid concentration in root or seed exudates. These data are in agreement with those presented here and indicate that nod- gene inducing signals may be limiting factors in other Rhizobium-legume symbioses.

Our results show that, under the laboratory conditions used, a range of different degrees of nodulation capacity can be observed anlong the varieties studied and this corresponds with a variation in the level of nod-gene inducing signals produced. Nevertheless, it remains to be evaluated if the different degrees of nodulation capacity of the alfalfa Mexican varieties reported here will also be observed under field conditions and if the N2-dependent

growth in the field will also vary among the varieties studied.

Aknowledgements. This work was partially supported by a grant from Direccitn General de Asuntos del Personal Acadtmico (DGAPA, UNAM No. IN024489). We are grateful to Jaime Mora, Rafael Palacios and Michael Dunn for critically reviewing the manuscript, to Gabriela Guerrero for computing assistance, to Luis Castro-Acero (Centro de Investigaciones Forestales y Agropecuarias del Estado de Mtxico) for providing seeds and valuable information on different Mexican alfalfa varieties, and to Patricia Garcfa and Eduardo Aranda (Universidad Autonoma del Estado de Morelos) for technical advise on TLC of root exudates and statistical analysis, respectively.

R e f e r e n c e s .

Barnes DK, Heichel GH, Vance CP, Ellis WR (1984) Plant Soil 82: 303-314

Bolafios MCV, Wemer ]DR (1993) In: Palacios R, Mora J, Newton WE (eds) New horizons in nitrogen fixation, Kluwer Academic Publishers, Dordrecht, The Netherlands, p.328

Fuentes SI, Su~ez R, Villegas T, Castro-Acero L, Hernhndez G (1993) Plant Cell Tiss. Org. Cult. 34:299-302

Hungrfa M, Phillips DA (1993) Mol. Plant-Microbe Interact. 6: 418- 422

Kapulnik Y, Joseph CM, Phillips DA (1987) Plant Physiol. 84:1193- 1196

Keppel G (1982) Design and analysis: a researcher's handbook. Prentise Hail, Inc. New Jersey, U.S.A.

Mulligan JT, Long SR (1985) Proc. Natl. Acad. Sci. USA 82, 6609- 6613

Phillips DA, Teuber LR, Jue SS (1982) Crop Sci. 22:606-610 Phillips DA. Dakora FD, Le6n-Barrios M, Sande E, Joseph CM (1993)

In: Palacios R, Mora J, Newton WE (eds) New horizons in nitrogen fixation, Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 197-202

Rolfe BG (1988) Biofactors 1:3-10 Teuber LR, Levin RP, Sweeney TC, Phillips DA (1984) Crop Sci. 22:

553-558 Vincent JM (1970) A manual for the practical study of root-nodule

bacteria, IBP Handbook, no. 15, Blackwell Scientific Publications, Ltd., Oxford