Phenotypic and symbiotic characterization of rhizobia isolated from Medicago ciliaris L . growing in Zerizer from Algeria

Phenotypic characteristics of 37 rhizobia strains isolated from root nodules of Medicago ciliaris L. growing in soils collected from Zerizer (North Eastern Algeria) were studied. Tolerance to salinity, high temperatures, acid and alkaline pHs, drought and to antibiotics as well as symbiotic and cultural characteristics allowed the description of a wide physiological diversity among tested isolates. Thirteen (13) isolates from the total could grow at 45°C. Only six isolates grew at 4% NaCl. Furthermore, the isolates which showed tolerance to salinity stress also showed tolerance to water stress, indicating direct relationships between these two physiological pathways. High salt and water stress tolerant strains were isolated and tested for their ability to biological nitrogen fixation. However, seven isolates were categorized into Agrobacterium.


INTRODUCTION
Many species of the legume genus Medicago are native to the Mediterranean basin (Lesins and Lesins, 1979) and are important as agricultural crops (Irwin et al., 2001).Compared to Medicago sativa, the most important species for cultivation, and Medicago truncatula, the model chosen for studies in nitrogen fixation (Cook, 1999;de Billy et al., 2001;Ben Amor et al., 2003), investigations with Medicago ciliaris as the focus have been very limited (Laouar and Abdelguerfi, 2000) .The species M. ciliaris is an annual plant that is tolerat to salt stress (Ben Salah et al., 2009) and may show promise for cultivation in salt-affected soils.Due to the reason that this species grows in soils that are heavy with clay it has application as a cover crop, in pastures, or for producing forage (Laouar and Abdelguerfi, 2000).
Many species of this genus Medicago have significant and wide-ranging agricultural and environmental applications, such as the perennial species M. sativa L, alfalfa (Irwin et al., 2001).Alfalfa is one of the most important forage crops in the world because of its high nutritive quality, yield, drought-resistance and good adaptation to various climatic and soil conditions and, therefore, is reputed to be the "queen" of the forages, although it has been reported to be water use inefficient (Li et al., 2007).Furthermore, the annual species, collectively known as "medics", are naturally distributed over a very wide range of environmental conditions in the Mediterranean basin, and the great importance in pastures in the Mediterranean and known to establish a nitrogen-fixing symbiosis with soil bacteria of the genus Ensifer (formerly Sinorhizobium) (Bena et al., 1998;Lesins and Lesins, 1979;Badri et al., 2008).
Nitrogen is a major limiting factor for plant productivity despite the inexhaustible reserve of atmosphere (78% N 2 ) (Foth, 1990).Biological fixation of molecular nitrogen (N 2 ) from the atmosphere is one of the main sources of nitrogen pool enhancement in agricultural soils (Bradic et al., 2003).The ability of legume species to establish nitrogen-fixing symbiosis with rhizobia makes them excellent candidates for use in sustainable agricultural systems (Howieson et al., 2000) .
Although most Medicago species form symbioses with the two species Sinorhizobium meliloti and S. medicae (Brunel et al., 1996;Rome et al., 1996), it is becoming evident that several different species of Medicago may have dissimilar affinities for infection by these 2 rhizobial species.For example, Garau et al. (2005) demonstrated that S. medicae frequently nodulated Medicago species that are adapted to acid soils, while S. meliloti formed symbioses with those growing in more alkaline to neutral soils.Bena et al. (2005) indicated that the geographic distribution of these rhizobial species appeared related to the incidence of the species of Medicago resulting from the characteristics of the soils.
In Algeria, diminution of pasture areas and deficit of forage production are major problems for development and extention of ovine and bovine breeding.Annual medics are grown as forage legume and regenerating pasture in the agro-pastoral Mediterranean systems or Australian ley-farming systems.However, the commercial Australian medic cultivars are not well adapted to most of agroecologicalzanes encountered in North Africa.Therefore, selection of better-adapted medics in association with appropriate symbiotic bacterial partners is agronomically important.Zerizer is an area of Algeria where most of M. ciliaris is represented, but less is known about their associated rhizobia.The advantages of this species are increased by the fact that, like most legumes, Medicagociliarisis able to form a symbiotic association with rhizobia and thus fix atmospheric nitrogen, which enriches the soil.The use of atmospheric nitrogen makes M. ciliaris a pioneer species capable of colonizing nitrogen poor soils.
This study is a preliminary step contributed to research efforts designed to uncover the biodiversity of rhizobia and, at the same time, select promising strains for the production of inoculants to improve M. ciliaris nitrogen fixation ability.We characterized 37 efficient rhizobia isolated from M. ciliaris L. collected from Zerizer area.The phenotypic characterization of these strains was conducted to evaluate their capacity to grow under abiotic stress such as severe temperatures, salinity, drought and high pH.Finally, the symbiotic properties of the representative strains were evaluated in terms of nodule numbers.

Sampling zone
Thirty nodulated M. ciliaris plants were collected from from Zerizer (North Eastern Algeria).Healthy plants were uprooted carefully and those plants prossessing healthy nodules with pink colour were selected to isolate rhizobia.

Soil sample and isolation of rhizobia from Medicago ciliaris
Soil sample was collected from area Zerizer in 2010 and used for M. ciliaris cultivation as trap hosts (Soil samples were collected from the Zerizer area in 2010 and used for trapping rhizobia).Strains were isolated from naturally occurring root nodules collected on M. ciliaris.Nodules were washed several times with tap water and rinsed with sterile distilled water.They were surface sterilized by immersion for 30 s in ethanol (96% v/v), 3 min in 3℅ sodium hypochlorite and then washed ten times with sterile distilled water.
A single surface-sterilized nodule was placed into a Petri dish and crushed with a sterile glass rod in the presence of a sterile solution of sterile distilled water.A loopful of the resulting suspension was then streaked on Tryptone Agar medium surface containing 25 μg/ml Congo red (TA) in a Petri dish and incubated at 28°C.
Bromothymol blue (BTB) agar medium was used for differentiating of the isolates.The cultures were streaked on BTB agar plates.BTB agar was made by adding 5 ml of (0.5% BTB in ethanol) to 1 L of YEMA medium.The plates were incubated at 28°C for 4 days.The change in color of medium was observed.The isolates were classified as slow growers (medium turns blue) or fast growers (medium turns yellow) on their reaction on YEMA supplemented with BTB (Table 1) (Somasegaran and Hoben, 1994).Isolates were purified by repeated streaking of a single colony on TA medium and were checked for purity by light microscopic examination of living cells and Gram staining (Vincent, 1970).They were then stored at 4°C on TA slants and at -16°C in Tryptone yeast extract (TY) liquid culture aliquots in the presence of 20 and 50% glycerol (v/v).

Purification of isolates a)-Growth on congo red medium
Rhizobia colonies appeared white, translucent, gummy, glistening elevated and comparatively small withenremargine were selected in contrast to of Agrobacterium on congo red medium which were red in color.

b)-Gram staining
Gram staining was done to ensure purity and freedom from Gram +ve bacteria.Gram-staining reaction was carried out by using a loopful of pure culture grown on Tryptone agar and stained as per the standard Gram's procedure (Somasegaran and Hoben, 1994).to a 10 ml solution containing 17.3 g of CuS0 4 .5H 2 O, and the mixture is diluted to 100 ml).The presence of 3-ketolactose in the medium is indicated by the formation of a yellow ring around the growth of a positive strain (Table 1, Figure 3).Maximum intensity of the yellow ring (2-3 cm in diameter) of cuprous oxide around the bacterial spot.Around 3-ketolactose positive strains is reached in about 1-2 h after flooding with Benedict's reagent.Biovar 1 strains have the unique ability to oxidize lactose into 3-ketolactose (Bernaerts and Deley, 1963).

Morphological studies
The thick bacterial smear of all the isolates was Gram stained and morphological characterization was done on the basis of colony morphology including shape,color and surface margin (Table 2).

Biochemical studies
Biochemical characterization was done on the basis of oxidase, catalase.

Stress tolerance screening
The isolates characterized in this study were examined for growth under different stress conditions of high temperature, high salinity, alkaline pH and extreme drought.In the case of temperature tolerance, isolates were kept at 28 (as a control), 37, 40, 42 or 45°C on YMA plates for four to five days.To check the ability of isolates to grow under different concentrations of NaCl, the medium was supplemented with 0 (control), 1, 2, 3, or 4% NaCl.To test the tolerance to acid and alkaline pH, the pH of the medium was adjusted with 0.5 M HCl or 0.5 M NaOH to 4.5, 5.5, 6.8 (as a control), 8, and 9.
The salinity and pH test were performed on YMA plates kept at 28°C for 4-5 days.To test drought resistance, different concentrations of polyethylene glycol (PEG 4000) were applied to the sterile distilled water at 10, 15, 20, 25℅.In this experiment, isolates were first grown in TA medium for 3 days at 28°C and the resulting bacterial suspensions containing approxi-mately 10 9 cells ml -1 , were transferred to YMA plates as previously indicated.The screening for stress tolerance was performed in Petri dishes divided into equal squares.Each square was spot inoculated with 10 µL of the cell suspensions at 10 9 cell ml -1 grown in Tryptone agar at an exponential phase.After incubation under different stressful conditions, the growth of isolates was estimated in comparison with that following control treatment, as follows: −, no growth; + weak growth (10-30% in relation to the control); ++, good growth (30-80% in relation to the control); and +++, very good growth (similar to the control).

Antibiotic susceptibility
Antibiotic resistance tests were performed by measuring the diameters of inhibition zones on YEM agar plates containing the following antibiotic discs: streptomycine (10 μg), tetracycline (30 μg), chloramphenicol (30 μg), nalidicic acid (30 μg); kanamycin (30 μg); ampicillin (10 μg).The antibiotic resistance was detected by an inhibition zone measured over a seven day period for each disc.Determination of intrinsic antibiotic resistance was evaluated in plates of YEM with different concentrations of antibiotics (rifampicine, erythromycin and neomycin).Filter-sterilized aliquots of each antibiotic were added aseptically to sterile YEM medium at 50°C to give the final concentrations.Control plates contained no antibiotic (Van Berkum et al., 1998).

Plant test
To assess their abilities to generate root nodules on their original hosts, the isolates were grown on Tryptone agar for 3 days at 28°C and the resulting bacterial suspensions containing approximately 10 9 cells mL −1 were inoculated on aseptic M. ciliaris seeds.Seeds were surface sterilized in 3℅ sodium hypochlorite for 10 min, rinsed with sterile distilled water, and then scarified.These seeds were germinated for 72 h on water agar (0.7 w/v) and planted at the rate of two seedlings in plastic pots containing sterilized sand.As controls, two pots (T0) with non inoculated seedlings were tested.Plants were sprayed with sterile distilled water every two days, in addition to being provided once a week with a nitrogen-free nutrient solution.
Plants were inoculated with 1 ml of early stationary phaserhizobial culture (10 8 -10 9 cells mL -1 ) cultivated at 28°C in TA medium.Two replications were carried out for both inoculated and non-inoculated plants (negative controls).Plantlets were harvested six weeks growth.Nodulation was recorded by the existence of nodules and the efficiency was estimated by the presence of red coloring (leghemoglobin) inside the nodules (Vincent, 1970).Shoot weight and root nodule numbers in each plant were also determined.For each isolate, the inoculation effect was estimated by determining the relative index of dry weight increase according to the following formula: relative index of dry weight increase = (inoculated plant dry weight) / (control plant dry weight).

Stress tolerance
The 37 isolates obtained from M. ciliaris and 6 reference strains of laboratory collection were first screened for resistance to high salinity, alkaline pH, high drought and high temperature conditions.According to this preliminary characterization, high diversity in stress resistance was observed.The data in Figure 1 show that M. ciliaris rhizobia exhibited a wide diversity in their salt tolerance.The salt inhibitory concentrations varied among strains and salt nature.Indeed, tolerance to sodium chloride (NaCl) was found since than 100% of the tested rhizobia continued to grow with 1% NaCl.However, at higher concentrations, the percentage of tolerant strains decreased rapidly and only two isolates (4.65% of all isolates) were able to grow at 4% NaCl, while at 3% NaCl.All the isolates were sensitive to the high salinity level of 3 and 4% NaCl, however, they showed relatively  good growth a 2% NaCl.Regarding high temperature resistance, response to extreme temperatures was positive for all the strains.Optimum temperature range for growth of culture is 28-30 0 C. 100 % of the isolates were able to grow at 37°C.Above those values, the percentage of isolates that grew decreased to reach 69.76 % at 42°C and 27.90% at 45°C.
In the case of the pH test (Table 3, Figure 2), showed a wide diversity in their pH tolerance.From 93.02 to 100% of the isolates grew in lightly acid and neutral pH.At low pH, some isolates exhibited anacido-tolerant character.Above pH 8, 100% of the isolates grew in alkaline pHs.Osmotolerance of all isolates was tested in minimal medium YMA supplemented with increasing concentrations PEG 4000 (drought), as described in the Materials and methods.All strains were able to grow in YMA with PEG added.Nearly all isolates survived at 25% PEG 4000 with the exception of Medp01.

Distinguishing test between Rhizobium and Agrobacterium
Agrobacterium is common in soil and in the plant rhizosphere, but was never described inside root nodules.Distinguished Agrobacterium from Rhizobia by 3-ketolactose test, whereas the Agrobacterium produced yellow ring of precipitate of CuO 2 around the colonies of the bacterium when plates were flooded with Benedict's reagent.In the present study six isolates showed positive results for 3-ketolactose test (Table 1 and Figure 4).

Antibiotic susceptibility
Intrinsic resistance to antibiotics showed a general resis-tance to erythromycin, and 69.67% of the strains were also resistant to 10 μg ml -1 ampicillin.18.60% of strains were scored resistant to 30 μg m -1 chloramphe-nicol and 55.81% were also resistant to 30 μg ml -1 nalidixic acid; 51.16 and 93.02% were also resistant to 10 and 30 μg ml - 1 of streptomycin and neomycin, respectively.Nevertheless, all strains were highly sensitive to tetracycline and kanamycin; rifampicin the concen-tration of 30 μg ml - 1 (Table 5 and Figure 5).

Plant tests
The 37 isolates and reference strains of laboratory collection were tested for their capacity to form root nodules on their original host plants under controlled laboratory conditions of temperature and relative humidity.All the isolates induced root nodules to form on their original hosts, and the uninoculated plants used as negative controls were not nodulated -root nodule numbers in the original host plants.
The mean number varied from 2.25 in Agrobacterium tumefacians (reference strain)to 19.50 in S. meliloti (reference strain).Seven isolates MedS14, MedS24, MedS25, Meds26, Meds29, and Meds31 and Medp07 have the enzymatic ability to aerobically convert lactose to 3-ketolactose, Strains Medp09 and A. tumefaciens were the less infective, with a respective average of 2 and 0 formed per plant.While strains MedS32 and S. meliloti were the most infective with 22 and 28 nodules formed per plant respectively.Relative indexes expressed as a shoot dry weight of the inoculated plants compared to the positive control plants, was largely variable (Table 4 and Figure 7).The most infective strains were also the most effective (Table 4 and Figure 6).

DISCUSSION
Although phenotypic and genotypic approaches provided very different information on the M. ciliaris rhizobia strains, they were similarly sensitive in demonstrating the large diversity found amongst these bacteria.The phenotypic characterization of the sampled 37 isolates and six reference strains of laboratory collection for above  Shamseldin et al. (2009) reported that E. meliloti strains from faba bean root nodules survived at 3% NaCl.Payakapong et al. (2006) also reported that an Ensifer strain of BL3 obtained from root nodules of Phaseolus lathyroides could survive at 3.5% NaCl.Shamseldin et al. (2006) reported the proteomic characterization of Rhizobium etli at 4% NaCl.In the case of high temperature resistance, out of 37 isolates, and six of reference strains thirteen isolates and two of reference strains survived at 45°C as shown in Table 3. Concerning high temperature resistance, some tolerant rhizobial isolates have been described.For instance, Fall et al. (2008) isolated rhizobia from Acacia Senegal that showed good growth at 45°C.Furthermore, Ge-Hong et al. ( 2008) reported a temperature tolerant strain of Mesorhizobium at 35°C.However, there is no report about E. meliloti surviving at 45°C.Moreover, two isolates, Meds13 and Meds14 survived at both 45°C and 2% NaCl.Nonetheless, we could not find an isolate able to survive abilities at 45°C and 3% NaCl.Out of these 37 isolates, only 2 (5.40% of the total) could survive at both 45°C and 2% NaCl as shown in Figure 1 and Table 3.This shows that the frequency of isolates having both high temperature resistance and high salt resistance is low.For salinity tolerance, we observed a wide variability for tolerance (0-4%).The isolates showed variation for NaCl tolerance, indicating that the rhizobia nodulating Medicago spp.are more tolerant compared to other rhizobia species (Struffi et al., 1998;Zahran, 1999).However, as suggested by El Sheikh and Wood (1989) and Odee et al. (1997), we found that fast growing strains were generally more tolerant to high salt concentrations than slow-growing strains.Salinity imposes both ionic and osmotic stresses.Indeed, the imposition of any stress to rhizobia results in adaptive responses, which lead to changes in the regular metabolic processes that are then reflected in protein profiles.The tolerant rhizobia to osmotic stress accumulate the osmolytes, and changes their morphology and dehydration of cells (Buss and Bottomley, 1989).There was a good correlation between the tolerances to both stressors; strains that were halotolerant were in general also tolerant to PEG, suggesting that common osmoadaptation mechanisms were operating.Hypersaline stress, most bacteria synthesize and accumulate small organic molecules called compatible solutes, as they compensate for hyperosmotic stress without interfering with cellular metabolism.As drought also imposes osmotic stress, it is plausible that at least part of the cell response to drought involves the synthesis and accumulation of compatible solutes.Our results showed that all the isolates grew at 25% of PEG 4000.This result showed that Zerizer soils contained rhizobial strains well adapted to dry conditions.For the most rhizobia, optimum temperature range for growth of culture is 28-31°C, and many cannot grow even at 37°C (Graham, 1992).At 28, 37 and 40°C, the isolates grew well like most Sinorhizobium species (Lindstrom and Lehtomaki, 1988;De Lajudie et al., 1994), could grow above 40°C.There was a varied response of the isolates tested to pH.All the isolates tested grew in alkaline pH (pH 8 and 9).At very low pH (pH 4.5), isolates grew normally with the exception of Meds09, Medp01, Medp04 and Medp05 S. fredii.According to Jordan (1984), slow-growing strains appear to be more tolerant to low pH than fast-growing strains.Some fast-growing strains such as Rhizobium tropici and Mesorhizobium loti can grow at a pH as low as 4 (Cooper, 1982;Cunningham and Munns, 1984;Graham, 1992;Gao et al., 1994).It has been reported that S. melilotig rows at a pH range of 5.0-9.5 (Jordan, 1984) and is tolerant to 2.0% (w/v) NaCl.Another Medicago-nodulating species, S. medicae, can grow at pH 5.0-10.0 and is resistant to 2.0% NaCl (Rome et al., 1996).Growth at pH 5.0 has been recorded for only some strains (Jordan, 1984) from acidic soils.Regarding the intrinsic resistance to antibiotics, it has been reported that fast-growing strains are more sensitive to antibiotics (Jordan, 1984) than slow-growing rhizobia.The evaluation of intrinsic resistance to antibiotics showed that most tested isolates had high resistance to erythromycin and nalidixic acid, chloramphenicol, and streptomycin.However, the degree of resistance to antibiotics was higher than in other species of rhizobia (Wei et al., 2003), indicating that S. meliloti and S. medicae had higher levels of tolerance to these antibiotics.

Percent of isolates
All tested strains were able to infect their host plant and to fix atmospheric nitrogen leading to plant shoot production above the noninoculated controls.Strains Meds10,Meds14,Meds15,Meds28,Meds32 developed 15.75,16,15.25,16.50 and 19 mean number of root nodules respectively, while S. meliloti developed 19.50 mean number of root nodules with the same host plant.This result shows that the root nodule forming ability of the symbiotic Agrobacterium is significantly lower than that of the reference strain S. meliloti.The root nodule number inoculated with trap host isolates is clearly lower than that of host plants inoculated with the remaining isolates.Sullivan and Ronson (1998) reported that a symbiotic element of M. loti was transferred into three non symbiotic species.Bailly et al. (2007) reported that several interspecific horizontal gene transfers occurred during the diversification of Medicago symbionts.Similarly, Wong and Golding (2003) reported that a large portion of pSym B genes in E. meliloti are most closely related to genes in A. tumefaciens linear chromosomes.These reports support the existence of symbiotic Agrobacterium isolates produced by horizontal transfer of symbiotic genes.
We conclude that rhizobia strains isolated from M. ciliaris nodules in Algerian soil are both phenotypically diverse.To verify this suggestion, we need to complete

Figure 1 .
Figure 1.Dendrogram showing effect of different concentrations of Nacl and temperature on growth of Medicago ciliaris rhizobia.

Figure 2 .Figure 3 .
Figure 2. Effect of different concentrations of pH on growth of Medicago ciliaris rhizobia.

Figure 5 .
Figure 5.Effect of different antibiotics on growth of Medicago ciliaris rhizobia.

Antibiotic Figure 6. .
Infectivity of M Medicago ciliaris rhizobia of nod ulation test in p plastic pot.

Table 4 .
Symbiotic properties of isolates on M. ciliaris nodulation test in plastic pots.

of nodules/ plant * nodulation test in plastic pot Plant dry weight (mg) Relative indexes
a Average ± standard deviation.