APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Dec. 1978, p. 915-919
0099-2240/78/0036-0915$02.00/0 Copyright © 1978 American Society for Microbiology
Vol. 36, No. 6 Printed in U.S.A.
Genetically Marked Rhizobium Identifiable as Inoculum Strain in Nodules of Soybean Plants Grown in Fields Populated with Rhizobiumjaponicum L. DAVID KUYKENDALL* AND DEANE F. WEBER Cell Culture and Nitrogen Fixation Laboratory, Plant Physiology Institute, Federal Research, Science and Education Administration, U. S. Department ofAgriculture, Agricultural Research Center, Beltsville, Maryland 20705
Received for publication 21 August 1978
The fate of an inoculum strain of Rhizobium japonicum was studied using a genetically marked strain I-110 subline carrying resistance markers for azide, rifampin, and streptomycin (1-110 ARS). At the time of planting into a field populated with R. japonicum, seeds of soybean cultivars Kent and Peking were inoculated with varying cell densities of strain 1-110 ARS. At various times during the growing season, surface-sterilized root nodules were examined for the presence of the inoculum strain by plating onto selective media. The recovery of the inoculum strain was unambiguous, varying, in the case of Kent cultivar, from about 5% with plants (sampled at 51 days) that had been inoculated with 3 x 108 cells per cm of row to about 20% with plants (sampled at 90 days) that had been inoculated with 3 x 109 cells per cm. The symbiotically incompatible interaction of Peking and strain 110 in Rhizobium-populated field soil was confirmed by the finding that, at 60 days after planting, only one nodule in 360 sampled contained strain 1-110 ARS. The use of genetically marked Rhizobium bacteria was found to provide for precise identification of the inoculum strain in nodules of fieldgrown soybeans.
Agronomic benefits due to Rhizobium inoculation are readily observed when soybeans are grown in soils having no significant populations of soybean rhizobia. However, when grown in soils containing well-established populations of Rhizobium japonicum, soybeans often do not respond to Rhizobium inoculations with increased seed yields (2). Failure of new inoculum strains placed in populated soils to effect increased yields has been attributed to failure to "compete" successfully with the established population of root nodule bacteria (5). Studies (2, 5) have shown that only 5 to 10% of the nodules may be occupied by the inoculum strain even when supplied at numbers manyfold higher than usual. To displace less-beneficial populations of Rhizobium already established in most soybeangrowing soils with those genotypes of Rhizobium that fix more nitrogen, new methods of either introducing the inoculum or of selecting highly competitive strains must be developed. At present, the development of population displacement techniques is limited by our lack of knowledge about the environmental and genetic factors limiting the introduction of new strains. Serological methods of strain evaluation and
identification have been very useful (9) for field studies, but the use of genetically marked bacteria, as first described by Obaton (10) for the study of Rhizobium ecology in the field, has not yet been exploited, although their facility for competition tests with aseptically grown plants inoculated with pure cultures has been demonstrated (6). The objectives of this investigation were (i) to evaluate the use of genetic markers for the identification of the inoculum strain in nodules of soybeans grown in populated fields and (ii) to determine whether population displacement among closely related soybean rhizobia can be measured. In this report we present the use of genetic markers for the precise identification of the inoculum strain in nodules of field-grown soybeans. The advantages of genetically marked R. japonicum strains over serological techniques are many. The use of genetic markers for simple and direct identification of soybean rhizobia may permit the elucidation of factors limiting the introduction of more beneficial strains into populated soils. MATERIALS AND METHODS The R. japonicum wild-type parent was strain I915
916
KUYKENDALL AND WEBER
110, which was cloned (7) from strain 3I1b11O of the Beltsville Collection. The origin and characteristics of strain 1-110 have been described previously (7, 8). Spontaneous antimetabolite-resistant mutants of strain I-110 were selected by plating 25 x 108 cells on AIE HM agar plates (7) containing concentrations of inhibitory agent that completely inhibited the parent strain. Some of the few resistant mutant clones that grew were first purified by successive clonal isolations on medium of the same composition and then used in successive selections for other resistances to obtain strains with multiple markers. The strain used in this study (I-110 ARS) carries three genetic markers: azide resistance (azi), rifampin resistance (rif), and streptomycin resistance (str). The resistance levels of this strain compared with the ancestral strains are given in Table 1. The high-level resistances to rifampin and streptomycin are responsible for making this strain easily identifiable on the basis of examining for growth under vigorous selective conditions. For use in field inoculation, the bacteria were grown at room temperature from a 10% inoculum in 10-liter quantities of AIE HM medium (HM salts [3], with 2 g of arabinose and 1 g of yeast extract [Difco] per liter) in 20-liter glass carboys, with aeration by sparging and magnetic stirring. The cells were harvested by continuous-flow centrifugation and were then resuspended in 1.5 liters of HM salts at pH 6.6 and stored at 4.5°C before use. Just before use in the field, dilutions in water were performed to give 24 liters each of cell suspensions containing 1.7 x 109, 3.4 x 108, or 1.7 x 108 cells per ml. Seeds were inoculated immediately after placement in the row with 1 liter of liquid inoculum per row, and then the rows were closed. Control plots received 1 liter of water per row. Each inoculum density/cultivar combination was replicated in blocks four times (three-row plots, 20 feet [ca. 6.1 m] long). Soybean cultivar Kent was selected because of symbiotic compatibility with strain I-110, whereas soybean cultivar Peking was selected because it accepts nodulation by strain 3I1b11O when inoculated with a pure culture and aseptically grown in the greenhouse, but only rarely when grown in soils containing other R. japonicum strains (1). This is a unique host-microsymbiont (genotype-genotype) interaction. Nodules were sampled at various times from 10 plants from each replication. Nodules sampled for plating onto selective media only were taken at random regardless of size, but nodule samplings taken at about 90 days for both serology and screening for genetic markers were biased for large nodules suitable for tube agglutination serological tests. The nodules were surface sterilized in 3% H202 for 30 min, followed by two complete rinses with sterile water over a sterile membrane filter (Millipore). The nodules to be examined by serology as well as for the genetically marked strain were individually placed in vials containing 1.0 ml of sterile 0.9% NaCl and then were squashed with a sterile glass rod. Samples of the turbid suspensions were streaked in small patches (either 5 or 10 per plate) on agar plates containing AIE HM with or without supplementation with either 1 mg of streptomycin per ml or 500 yg of rifampin per ml or both. After it was established that the surface-sterili-
APPL. ENVIRON. MICROBIOL.
TABLE 1. Antimetabolite and antibiotic resistance levels of parent and mutant R. japonicum strains Minimum inhibitory concn (,Uglml)b
Straina
Azide
Rifampin
Streptomycin
1-110 10 100 40 I-llO azi-5 rif-17 >10 >500 40 I-110 azi-5 rif- 17 >10 >500 >1,000 str- 74 aStrain 1-110 azi-5 rif- 17 str- 74 is designated 1-110 ARS in the text. bApproximated as the lowest concentration that totally inhibits growth when 108 to 109 bacteria are plated on AlE HM agar.
zation procedure did not result in a loss of viability of bacteria in the nodules greater than 1 nodule per 100, the plating onto nonselective medium was omitted. In nodule samplings of 25 or 30 per replicate, only the first 10 were plated on both rifampin medium and streptomycin medium, and the remaining were plated only on streptomycin. Plates were incubated for 5 days prior to scoring for the presence of the marked strain according to growth or its absence in the patch (Fig. 1). Serology was performed on the same suspensions examined for genetic markers essentially as described by Means et al. (9).
RESULTS A preliminary sample of 15 nodules from each treatment of Kent soybeans was taken 43 days after planting. None of the nodules from the uninoculated plot contained a strain that was capable of growth after 5 days on medium supplemented with either 500 ,ug of rifampin or 1 mg of streptomycin per ml. Of plants inoculated with either 3 x 108 or 3 x 109 cells per cm of row, about one-third of the nodules (4/15 and 5/15 nodules) contained a strain that grew on media supplemented with 500 jig of rifampin per ml, with 1 mg of streptomycin per ml, and with both. The results of sampling Kent and Peking soybean plants 51 days after planting for the presence of the marked strain in the nodules are shown in Table 2. None of the nodules from control plants contained a strain phenotypically expressing both streptomycin and rifampin resistances. As expected, only about 4% of the nodules from Peking plants inoculated with strain I-110 ARS contained strain 110. Although only 5% or less of the nodules from Kent plants
inoculated at 3 x 108 or 6 x 108/cm contained strain ARS, approximately 16% of the nodules from plants inoculated at the rate of 3 X 109/cm contained the genetically marked inoculum strain. Nodules of Kent soybean plants were exam-
VOL. 36, 1978
GENETICALLY MARKED SOYBEAN SYMBIONT
917
FIG. 1. Examination of soybean nodules for the presence of strain I-110 ARS. The growth medium supplemented with streptomycin and rifampin (see Materials and Methods). As shown by patches of growth, nodules no. 1, 5, 6, and 7 contained I-110 ARS.
was
TABLE 2. Proportion of Kent and Peking soybean nodules containing the genetically marked strain at 51 days after planting Inoculum Variy (celUs/cm)
Kent
Peking
TABLE 3. Presence of strain I-110 ARS in nodules of field-grown Kent soybeansa
No. of nod- Percent ules con- containdules exanisned taining Str', ing inocuRif' strain lum strain
No.
Inoculum
of
3 x 109 6 x 105 3 x 10 0
140 40 40 90
23
3 x 109 6 x 108 3 x 108 0
40 40 27 40
0 2 1 0
1 2 0
16.4 2.5
(cells/cm)
Replicate
3 x 109
1 2 3 4
4/30 2/30 2/30 4/30
0/40 4/40 4/30 5/40
6 x 108
1 2 3 4
3/30 3/30 2/30 2/30
5/40 3/40 4/40 3/40
3 x 108
1 2 3 4
4/30 3/30 5/30 6/30
1/40 10/40 1/40 4/40
5.0 0.0
0.0 5.0 3.7 0.0
ined for the presence of the inoculum strain 58 days after planting (Table 3). A total of 40 of the 360 nodules (11%) examined from the inoculated plots 58 days after planting were found to contain the inoculum strain I-110 ARS. Of 470 nodules examined from inoculated plots at 72 days after planting, a total of 44 or about 10% of them contained the Strr, Rifr inoculum strain. These data showed that regardless of the inoculum densities that had been applied, only about 10% of the nodules contained the inoculum strain I110 ARS. At 60 days after planting, 30 nodules from
No. of nodules containing I110 ARS/no. of nodules examined at (days): 58 72
0
1 0/30 0/40 2 0/30 0/40 3 0/30 0/40 4 0/30 0/40 At 58 and 72 days from planting, 10 plants from a border row of each plot were dug up, and the nodules were removed from the roots, pooled, surface sterilized, and sampled at random for the presence of the genetically marked inoculum strain as described in the text. a
918
APPL. ENVIRON. MICROBIOL.
KUYKENDALL AND WEBER
each replication of the Peking treatments were examined (not shown) for the presence of 1-110 ARS. Only one of the 360 nodules examined from the inoculated plots contained the genetically marked strain derived from strain 3I1b110. These data strongly confirm the observations of Caldwell and Vest (1) for the symbiotically incompatible genotype x genotype interaction of Peking and strain 110 found in the field. Also, the data show that the failure to detect strain 110 by serology in Peking nodules was not due to relatively small strain 110 populations in mixed nodules where intranodule population displacement might have been occurring. Obviously this incompatibility is in a recognition step, where the Peking soybean plant preferentially chooses other strains for nodulation. Data on the serology of each relatively large, mature nodule assayed at 90 days after planting were obtained for comparison with those obtained by examining for genetic markers (Table 4). Thirty-four (34/99) percent of the larger nodules from uninoculated plants contained strains identifiable as serogroup 110 by tube agglutination with anti-3IlbllO rabbit serum. In comparison, larger nodules from inoculated plants showed about the same proportion of 110 serogroup with an overall increase, if any, due to inoculation of about 6%. In the examination for genetic markers, none of the 199 nodules examined from uninoculated plants contained a strain that was either streptomycin resistant or rifampin resistant. Samplings of inoculated plants showed that 15 to 28% contained the inoculum strain I110 ARS; the overall proportion of nodules from inoculated plants containing I-110 ARS was about 20%. Given a field population composed partially of 110 serogroup strains, serological methods underestimated and did not permit the observation of the successful introduction of the inoculum strain. Between 25 and 40% of the nodules containing serogroup 110 contained strain I-110 ARS (Table 4); the overall average was 33%. Between 20 and 50% of the nodules containing strain I-110 ARS were negative for 110 by serology (Table 4); the overall proportion was 30%. This may reflect the presence of another strain predominating in addition to I-110 ARS in as many as 30% of the nodules containing 1-110 ARS.
DISCUSSION The use of genetically marked soybean microsymbionts provides for precise identification of the inoculum strain in nodules of field-grown soybeans. This is the first time that this utility has been demonstrated unequivocally. Highlevel resistances to rifampin and to streptomycin
TABLE 4. Comparison of serological data with genetic marker data from Kent soybeans sampled at about 90 days after plantinga Proportion of Inocution of nodules lum Sam- nodules contain(cells/ pling contain- ing Str', cm) ing 110 Rif' strain serogroup I-110 ARS 3 x 109 1 41/98 15/98
Propor-
6 x 108
52/100 26/98
2 1 2 1 2 1 2
28/100 19/99
Nodules, containing 110 by serology, showing I-110 ARSh
Nodules, containing 1-110 ARS, but negative for 110 by serology
10/41 22/52 9/26
4/15 6/28 9/18C 9/26 3/14
38/99 26/100 17/38 44/98 14/98 11/44 40/100 14/85 11/40 3/14 0 34/99 0/100 0/34 34/99 0/99 0/34 Data pooled from four replicates. ' Individual nodules were examined by both methods. ' One nodule, positive for Rif', Strr, was not availa3x
I0W
ble for serology.
were sufficient markers to distinguish the inoculum strain from the naturally occurring population of R. japonicum strains in soil at Beltsville. The use of genetically marked strains as tools for studying the factors limiting the introduction of more efficient inoculum strains into populated soils appears to greatly surpass the utility of the standard serological methods. For instance, the results clearly indicate that it is feasible to examine population fluctuations among closely related strains. This is important to consider, since studies (4) have shown that some soils contain a predominant serogroup occupying 50 to 80% of the nodules, with predominant groups differing among soils. The utilization of genetically marked R. japonicum strains, as compared to serological techniques, eliminates the need to assume that increases in recovery of identical serotypes above the control level are from applied strains. These results may lead to the development of techniques for displacing indigenous strains with strains that are more symbiotically competent. ACKNOWLEDGMENT1S We gratefully acknowledge stimulating discussions with Lowell D. Owens. We thank Lester A. Tayman for skilled technical assistance in both the field work and laboratory work. Also, we appreciated the help of Dorothy Jessop and Jeffrey Powers in the preparation of media and the examination of nodules. We thank Charles Sloger for taking the
photograph. LITERATURE CITED 1. Caldwell, B. E., and G. Vest. 1968. Nodulation interac-
VOL. 36, 1978
GENETICALLY MARKED SOYBEAN SYMBIONT
tions between soybean genotypes and serogroups of Rhizobium japonicum. Crop Sci. 8:680-682. 2. Caldwell, B. E., and G. Vest. 1970. Effects of Rhizobium japonicum strains on soybean yields. Crop Sci. 10: 19-21. 3. Cole, M. A., and G. H. Elkan. 1973. Transmissible resistance to penicillin G, neomycin, and chloramphenicol. Antimicrob. Agents Chemother. 4:248-253. 4. Johnson, H. W., and U. M. Means. 1963. Serological groups of Rhizobium japonicum recovered from nodules of soybeans (Glycine max) in field soil. Agron. J. 55:269-271. 5. Johnson, H. W., U. M. Means, and C. R. Weber. 1965. Competition for nodule sites between strains of Rhizobium japonicum applied as inoculum and strains in the soil. Agron. J. 57:179-185.
919
6. Johnston, A. W. B., and J. E. Beringer. 1975. Identification of the Rhizobium strains in pea root nodules using genetic markers. J. Gen. Microbiol. 87:343-350. 7. Kuykendall, L D., and G. H. Elkan. 1976. Rhizobium japonicum derivatives differing in nitrogen-fixing efficiency and carbohydrate utilization. Appl. Environ. Microbiol. 32:511-519. 8. Kuykendall, L, D., and G. H. Elkan. 1977. Some features of mannitol metabolism in Rhizobiumjaponicum. J. Gen. Microbiol. 98:291-295. 9. Means, U. M., H. W. Johnson, and R. A. Date. 1964. Quick serological method of classifying strains of Rhizobiumjaponicum in nodules. J. Bacteriol. 87:547-553. 10. Obaton, M. 1973. The use of spontaneous antibiotic resistant mutants for the ecological study of Rhizobium. Bull. Ecol. Res. Comm. (Stockholm) 17:170-171.
0099-2240/78/0036-0915$02.00/0 Copyright © 1978 American Society for Microbiology
Vol. 36, No. 6 Printed in U.S.A.
Genetically Marked Rhizobium Identifiable as Inoculum Strain in Nodules of Soybean Plants Grown in Fields Populated with Rhizobiumjaponicum L. DAVID KUYKENDALL* AND DEANE F. WEBER Cell Culture and Nitrogen Fixation Laboratory, Plant Physiology Institute, Federal Research, Science and Education Administration, U. S. Department ofAgriculture, Agricultural Research Center, Beltsville, Maryland 20705
Received for publication 21 August 1978
The fate of an inoculum strain of Rhizobium japonicum was studied using a genetically marked strain I-110 subline carrying resistance markers for azide, rifampin, and streptomycin (1-110 ARS). At the time of planting into a field populated with R. japonicum, seeds of soybean cultivars Kent and Peking were inoculated with varying cell densities of strain 1-110 ARS. At various times during the growing season, surface-sterilized root nodules were examined for the presence of the inoculum strain by plating onto selective media. The recovery of the inoculum strain was unambiguous, varying, in the case of Kent cultivar, from about 5% with plants (sampled at 51 days) that had been inoculated with 3 x 108 cells per cm of row to about 20% with plants (sampled at 90 days) that had been inoculated with 3 x 109 cells per cm. The symbiotically incompatible interaction of Peking and strain 110 in Rhizobium-populated field soil was confirmed by the finding that, at 60 days after planting, only one nodule in 360 sampled contained strain 1-110 ARS. The use of genetically marked Rhizobium bacteria was found to provide for precise identification of the inoculum strain in nodules of fieldgrown soybeans.
Agronomic benefits due to Rhizobium inoculation are readily observed when soybeans are grown in soils having no significant populations of soybean rhizobia. However, when grown in soils containing well-established populations of Rhizobium japonicum, soybeans often do not respond to Rhizobium inoculations with increased seed yields (2). Failure of new inoculum strains placed in populated soils to effect increased yields has been attributed to failure to "compete" successfully with the established population of root nodule bacteria (5). Studies (2, 5) have shown that only 5 to 10% of the nodules may be occupied by the inoculum strain even when supplied at numbers manyfold higher than usual. To displace less-beneficial populations of Rhizobium already established in most soybeangrowing soils with those genotypes of Rhizobium that fix more nitrogen, new methods of either introducing the inoculum or of selecting highly competitive strains must be developed. At present, the development of population displacement techniques is limited by our lack of knowledge about the environmental and genetic factors limiting the introduction of new strains. Serological methods of strain evaluation and
identification have been very useful (9) for field studies, but the use of genetically marked bacteria, as first described by Obaton (10) for the study of Rhizobium ecology in the field, has not yet been exploited, although their facility for competition tests with aseptically grown plants inoculated with pure cultures has been demonstrated (6). The objectives of this investigation were (i) to evaluate the use of genetic markers for the identification of the inoculum strain in nodules of soybeans grown in populated fields and (ii) to determine whether population displacement among closely related soybean rhizobia can be measured. In this report we present the use of genetic markers for the precise identification of the inoculum strain in nodules of field-grown soybeans. The advantages of genetically marked R. japonicum strains over serological techniques are many. The use of genetic markers for simple and direct identification of soybean rhizobia may permit the elucidation of factors limiting the introduction of more beneficial strains into populated soils. MATERIALS AND METHODS The R. japonicum wild-type parent was strain I915
916
KUYKENDALL AND WEBER
110, which was cloned (7) from strain 3I1b11O of the Beltsville Collection. The origin and characteristics of strain 1-110 have been described previously (7, 8). Spontaneous antimetabolite-resistant mutants of strain I-110 were selected by plating 25 x 108 cells on AIE HM agar plates (7) containing concentrations of inhibitory agent that completely inhibited the parent strain. Some of the few resistant mutant clones that grew were first purified by successive clonal isolations on medium of the same composition and then used in successive selections for other resistances to obtain strains with multiple markers. The strain used in this study (I-110 ARS) carries three genetic markers: azide resistance (azi), rifampin resistance (rif), and streptomycin resistance (str). The resistance levels of this strain compared with the ancestral strains are given in Table 1. The high-level resistances to rifampin and streptomycin are responsible for making this strain easily identifiable on the basis of examining for growth under vigorous selective conditions. For use in field inoculation, the bacteria were grown at room temperature from a 10% inoculum in 10-liter quantities of AIE HM medium (HM salts [3], with 2 g of arabinose and 1 g of yeast extract [Difco] per liter) in 20-liter glass carboys, with aeration by sparging and magnetic stirring. The cells were harvested by continuous-flow centrifugation and were then resuspended in 1.5 liters of HM salts at pH 6.6 and stored at 4.5°C before use. Just before use in the field, dilutions in water were performed to give 24 liters each of cell suspensions containing 1.7 x 109, 3.4 x 108, or 1.7 x 108 cells per ml. Seeds were inoculated immediately after placement in the row with 1 liter of liquid inoculum per row, and then the rows were closed. Control plots received 1 liter of water per row. Each inoculum density/cultivar combination was replicated in blocks four times (three-row plots, 20 feet [ca. 6.1 m] long). Soybean cultivar Kent was selected because of symbiotic compatibility with strain I-110, whereas soybean cultivar Peking was selected because it accepts nodulation by strain 3I1b11O when inoculated with a pure culture and aseptically grown in the greenhouse, but only rarely when grown in soils containing other R. japonicum strains (1). This is a unique host-microsymbiont (genotype-genotype) interaction. Nodules were sampled at various times from 10 plants from each replication. Nodules sampled for plating onto selective media only were taken at random regardless of size, but nodule samplings taken at about 90 days for both serology and screening for genetic markers were biased for large nodules suitable for tube agglutination serological tests. The nodules were surface sterilized in 3% H202 for 30 min, followed by two complete rinses with sterile water over a sterile membrane filter (Millipore). The nodules to be examined by serology as well as for the genetically marked strain were individually placed in vials containing 1.0 ml of sterile 0.9% NaCl and then were squashed with a sterile glass rod. Samples of the turbid suspensions were streaked in small patches (either 5 or 10 per plate) on agar plates containing AIE HM with or without supplementation with either 1 mg of streptomycin per ml or 500 yg of rifampin per ml or both. After it was established that the surface-sterili-
APPL. ENVIRON. MICROBIOL.
TABLE 1. Antimetabolite and antibiotic resistance levels of parent and mutant R. japonicum strains Minimum inhibitory concn (,Uglml)b
Straina
Azide
Rifampin
Streptomycin
1-110 10 100 40 I-llO azi-5 rif-17 >10 >500 40 I-110 azi-5 rif- 17 >10 >500 >1,000 str- 74 aStrain 1-110 azi-5 rif- 17 str- 74 is designated 1-110 ARS in the text. bApproximated as the lowest concentration that totally inhibits growth when 108 to 109 bacteria are plated on AlE HM agar.
zation procedure did not result in a loss of viability of bacteria in the nodules greater than 1 nodule per 100, the plating onto nonselective medium was omitted. In nodule samplings of 25 or 30 per replicate, only the first 10 were plated on both rifampin medium and streptomycin medium, and the remaining were plated only on streptomycin. Plates were incubated for 5 days prior to scoring for the presence of the marked strain according to growth or its absence in the patch (Fig. 1). Serology was performed on the same suspensions examined for genetic markers essentially as described by Means et al. (9).
RESULTS A preliminary sample of 15 nodules from each treatment of Kent soybeans was taken 43 days after planting. None of the nodules from the uninoculated plot contained a strain that was capable of growth after 5 days on medium supplemented with either 500 ,ug of rifampin or 1 mg of streptomycin per ml. Of plants inoculated with either 3 x 108 or 3 x 109 cells per cm of row, about one-third of the nodules (4/15 and 5/15 nodules) contained a strain that grew on media supplemented with 500 jig of rifampin per ml, with 1 mg of streptomycin per ml, and with both. The results of sampling Kent and Peking soybean plants 51 days after planting for the presence of the marked strain in the nodules are shown in Table 2. None of the nodules from control plants contained a strain phenotypically expressing both streptomycin and rifampin resistances. As expected, only about 4% of the nodules from Peking plants inoculated with strain I-110 ARS contained strain 110. Although only 5% or less of the nodules from Kent plants
inoculated at 3 x 108 or 6 x 108/cm contained strain ARS, approximately 16% of the nodules from plants inoculated at the rate of 3 X 109/cm contained the genetically marked inoculum strain. Nodules of Kent soybean plants were exam-
VOL. 36, 1978
GENETICALLY MARKED SOYBEAN SYMBIONT
917
FIG. 1. Examination of soybean nodules for the presence of strain I-110 ARS. The growth medium supplemented with streptomycin and rifampin (see Materials and Methods). As shown by patches of growth, nodules no. 1, 5, 6, and 7 contained I-110 ARS.
was
TABLE 2. Proportion of Kent and Peking soybean nodules containing the genetically marked strain at 51 days after planting Inoculum Variy (celUs/cm)
Kent
Peking
TABLE 3. Presence of strain I-110 ARS in nodules of field-grown Kent soybeansa
No. of nod- Percent ules con- containdules exanisned taining Str', ing inocuRif' strain lum strain
No.
Inoculum
of
3 x 109 6 x 105 3 x 10 0
140 40 40 90
23
3 x 109 6 x 108 3 x 108 0
40 40 27 40
0 2 1 0
1 2 0
16.4 2.5
(cells/cm)
Replicate
3 x 109
1 2 3 4
4/30 2/30 2/30 4/30
0/40 4/40 4/30 5/40
6 x 108
1 2 3 4
3/30 3/30 2/30 2/30
5/40 3/40 4/40 3/40
3 x 108
1 2 3 4
4/30 3/30 5/30 6/30
1/40 10/40 1/40 4/40
5.0 0.0
0.0 5.0 3.7 0.0
ined for the presence of the inoculum strain 58 days after planting (Table 3). A total of 40 of the 360 nodules (11%) examined from the inoculated plots 58 days after planting were found to contain the inoculum strain I-110 ARS. Of 470 nodules examined from inoculated plots at 72 days after planting, a total of 44 or about 10% of them contained the Strr, Rifr inoculum strain. These data showed that regardless of the inoculum densities that had been applied, only about 10% of the nodules contained the inoculum strain I110 ARS. At 60 days after planting, 30 nodules from
No. of nodules containing I110 ARS/no. of nodules examined at (days): 58 72
0
1 0/30 0/40 2 0/30 0/40 3 0/30 0/40 4 0/30 0/40 At 58 and 72 days from planting, 10 plants from a border row of each plot were dug up, and the nodules were removed from the roots, pooled, surface sterilized, and sampled at random for the presence of the genetically marked inoculum strain as described in the text. a
918
APPL. ENVIRON. MICROBIOL.
KUYKENDALL AND WEBER
each replication of the Peking treatments were examined (not shown) for the presence of 1-110 ARS. Only one of the 360 nodules examined from the inoculated plots contained the genetically marked strain derived from strain 3I1b110. These data strongly confirm the observations of Caldwell and Vest (1) for the symbiotically incompatible genotype x genotype interaction of Peking and strain 110 found in the field. Also, the data show that the failure to detect strain 110 by serology in Peking nodules was not due to relatively small strain 110 populations in mixed nodules where intranodule population displacement might have been occurring. Obviously this incompatibility is in a recognition step, where the Peking soybean plant preferentially chooses other strains for nodulation. Data on the serology of each relatively large, mature nodule assayed at 90 days after planting were obtained for comparison with those obtained by examining for genetic markers (Table 4). Thirty-four (34/99) percent of the larger nodules from uninoculated plants contained strains identifiable as serogroup 110 by tube agglutination with anti-3IlbllO rabbit serum. In comparison, larger nodules from inoculated plants showed about the same proportion of 110 serogroup with an overall increase, if any, due to inoculation of about 6%. In the examination for genetic markers, none of the 199 nodules examined from uninoculated plants contained a strain that was either streptomycin resistant or rifampin resistant. Samplings of inoculated plants showed that 15 to 28% contained the inoculum strain I110 ARS; the overall proportion of nodules from inoculated plants containing I-110 ARS was about 20%. Given a field population composed partially of 110 serogroup strains, serological methods underestimated and did not permit the observation of the successful introduction of the inoculum strain. Between 25 and 40% of the nodules containing serogroup 110 contained strain I-110 ARS (Table 4); the overall average was 33%. Between 20 and 50% of the nodules containing strain I-110 ARS were negative for 110 by serology (Table 4); the overall proportion was 30%. This may reflect the presence of another strain predominating in addition to I-110 ARS in as many as 30% of the nodules containing 1-110 ARS.
DISCUSSION The use of genetically marked soybean microsymbionts provides for precise identification of the inoculum strain in nodules of field-grown soybeans. This is the first time that this utility has been demonstrated unequivocally. Highlevel resistances to rifampin and to streptomycin
TABLE 4. Comparison of serological data with genetic marker data from Kent soybeans sampled at about 90 days after plantinga Proportion of Inocution of nodules lum Sam- nodules contain(cells/ pling contain- ing Str', cm) ing 110 Rif' strain serogroup I-110 ARS 3 x 109 1 41/98 15/98
Propor-
6 x 108
52/100 26/98
2 1 2 1 2 1 2
28/100 19/99
Nodules, containing 110 by serology, showing I-110 ARSh
Nodules, containing 1-110 ARS, but negative for 110 by serology
10/41 22/52 9/26
4/15 6/28 9/18C 9/26 3/14
38/99 26/100 17/38 44/98 14/98 11/44 40/100 14/85 11/40 3/14 0 34/99 0/100 0/34 34/99 0/99 0/34 Data pooled from four replicates. ' Individual nodules were examined by both methods. ' One nodule, positive for Rif', Strr, was not availa3x
I0W
ble for serology.
were sufficient markers to distinguish the inoculum strain from the naturally occurring population of R. japonicum strains in soil at Beltsville. The use of genetically marked strains as tools for studying the factors limiting the introduction of more efficient inoculum strains into populated soils appears to greatly surpass the utility of the standard serological methods. For instance, the results clearly indicate that it is feasible to examine population fluctuations among closely related strains. This is important to consider, since studies (4) have shown that some soils contain a predominant serogroup occupying 50 to 80% of the nodules, with predominant groups differing among soils. The utilization of genetically marked R. japonicum strains, as compared to serological techniques, eliminates the need to assume that increases in recovery of identical serotypes above the control level are from applied strains. These results may lead to the development of techniques for displacing indigenous strains with strains that are more symbiotically competent. ACKNOWLEDGMENT1S We gratefully acknowledge stimulating discussions with Lowell D. Owens. We thank Lester A. Tayman for skilled technical assistance in both the field work and laboratory work. Also, we appreciated the help of Dorothy Jessop and Jeffrey Powers in the preparation of media and the examination of nodules. We thank Charles Sloger for taking the
photograph. LITERATURE CITED 1. Caldwell, B. E., and G. Vest. 1968. Nodulation interac-
VOL. 36, 1978
GENETICALLY MARKED SOYBEAN SYMBIONT
tions between soybean genotypes and serogroups of Rhizobium japonicum. Crop Sci. 8:680-682. 2. Caldwell, B. E., and G. Vest. 1970. Effects of Rhizobium japonicum strains on soybean yields. Crop Sci. 10: 19-21. 3. Cole, M. A., and G. H. Elkan. 1973. Transmissible resistance to penicillin G, neomycin, and chloramphenicol. Antimicrob. Agents Chemother. 4:248-253. 4. Johnson, H. W., and U. M. Means. 1963. Serological groups of Rhizobium japonicum recovered from nodules of soybeans (Glycine max) in field soil. Agron. J. 55:269-271. 5. Johnson, H. W., U. M. Means, and C. R. Weber. 1965. Competition for nodule sites between strains of Rhizobium japonicum applied as inoculum and strains in the soil. Agron. J. 57:179-185.
919
6. Johnston, A. W. B., and J. E. Beringer. 1975. Identification of the Rhizobium strains in pea root nodules using genetic markers. J. Gen. Microbiol. 87:343-350. 7. Kuykendall, L D., and G. H. Elkan. 1976. Rhizobium japonicum derivatives differing in nitrogen-fixing efficiency and carbohydrate utilization. Appl. Environ. Microbiol. 32:511-519. 8. Kuykendall, L, D., and G. H. Elkan. 1977. Some features of mannitol metabolism in Rhizobiumjaponicum. J. Gen. Microbiol. 98:291-295. 9. Means, U. M., H. W. Johnson, and R. A. Date. 1964. Quick serological method of classifying strains of Rhizobiumjaponicum in nodules. J. Bacteriol. 87:547-553. 10. Obaton, M. 1973. The use of spontaneous antibiotic resistant mutants for the ecological study of Rhizobium. Bull. Ecol. Res. Comm. (Stockholm) 17:170-171.
