Plant Physiol. (1980) 66, 1027-10310032-0889/80/66/1027/05/$00.50/0

Early Events in the Infection of Soybean (Glycine max L. Merr)by Rhizobium japonicumI. LOCALIZATION OF INFECTIBLE ROOT CELLS'

Received for publication February 19, 1980 and in revised form June 25, 1980

T. V. BHUVANESWARI, B. GILLIAN TURGEON, AND WOLFGANG D. BAUER2Charles F. Kettering Research Laboratory, Yellow Springs, Ohio 45387

ABSTRACT

The infectible cells of soybean roots appear to be located at any giventime just above the zone of root elongation and just below the position ofthe smallest emergent root hairs. The location of infectible cells on theprimary root at the time of inoculation was inferred from the position ofsubsequent nodule development, correcting for displacement of epidermalcells due to root elongation. Marks were made on the seedling growthpouches at the time of inoculation to indicate the position of the root tipand the zones of root hair development. Virtually all of the seedlingsdeveloped nodules on the primary root above the marks made at the roottips at the time of inoculation. None of the plants formed nodules on theroot where mature root hairs were present at the time of inoculation. Theseresults and profiles of nodulation frequency indicate that the location ofinfectible cells is developmentally restricted. When inoculations weredelayed for intervals of I to 4 hours after marking the positions of the roottips. progressively fewer nodules were formed above the root tip marks,and the uppermost of these nodules were formed at progressively shorterdistances above the marks. These results indicate that the infectibility ofgiven host cells is a transient property that appears and then is lost withina few hours. The results also indicate that host responses leading toinfection and nodulation are triggered or initiated in less than 2 hours afterinoculation. The extent of nodulation above the root tip mark increased inproportion to the logarithm of the number of bacteria in the inoculum.

The mechanisms by which soybean plants interact with Rhizo-biumjaponicum to establish a symbiotic association have not beenstudied extensively (3). Bieberdorf (2) reported that R. japonicumenters its host by means of infection thread structures formed inroot hair cells. This observation has been confirmed by morerecent studies (refs. 5 and 8; B. G. Turgeon and W. D. Bauer,unpublished observations).

Little is known of the early events which precede the formationof infection thread structures in soybean or other legumes. Oneimportant problem in the study of these early events is to deter-mine which root hair cells of the host are capable of respondingto the bacterial symbiont and becoming infected. Here, a conven-ient method for studying the location and development of in-fectible cells on legume roots is described.

'This work was supported in part by grant PCM 77-24930 from theNational Science Foundation. This is contribution No. 709 of the C. F.Kettering Research Laboratory.

2 To whom reprint requests should be addressed.

MATERIALS AND METHODS

Rhizobium Culture. Cultures of R. japonicum strains 31lb 138and 311b 83 were obtained from D. F. Weber, United StatesDepartment of Agriculture, Beltsville, MD. Stock cultures weremaintained on yeast extract-mannitol medium (9) modified tocontain 0.5% mannitol, 0.5% sodium gluconate and 0.5% yeastextract. Every 2 to 3 weeks, single cell colonies from the stockcultures were used to inoculate 10 ml modified B5 medium (1),adjusted to pH 6.8 to 7.0, in 25-ml culture flasks. These startercultures were maintained on a rotary shaker (100 rpm) at 25 C for7 days (early stationary growth phase). The A620 nm of the starterculture was determined and a volume of the culture equivalent to0.5 ml of an 1.0 A620 nm suspension was added to 50 ml freshmedium in a 125-ml culture flask. The remaining starter culturewas stored in the refrigerator for use the following week. Inoculumcultures were grown for 3 days in the same manner as the startercultures, harvested by centrifugation, washed once with sterilephosphate-buffered saline (pH 7.2) (1), and resuspended in one-tenth strength phosphate-buffered saline. Suspensions used forinoculation were diluted to 0.08 A620 nm (I x 108 cells/ml) withsterile, half-strength N-free Jensen's plant growth medium (9)prior to use.Growth of Seedlings. Soybean seeds of cultivars Beeson and

Williams were obtained from DeWine and Hamma Seed Co.,Yellow Springs, OH. Harris seeds were a gift from Dr. K. Nadler,Michigan State University. Seeds (cv. Williams unless otherwisespecified) were surface-sterilized with 5% NaOCI plus I drop ofTween 20 for 7 min, washed 5 to 10 times with sterile H20, andsoaked in H20 for 2 hours. Swollen seeds with unbroken coatswere germinated on yeast extract-mannitol agar plates for 3 daysin the growth chamber, protected from light with aluminum foil.Seedlings without visible microbial contamination were trans-ferred to plastic growth pouches (diSPo Seed-Pack, ScientificProducts, Evanston, IL; ref. 10). Ten ml sterile, half-strength N-free Jensen's medium were added to moisten the paper towel ofeach pouch prior to inserting the roots through holes (made withtweezers) in the bottom of the paper towel seed trough. Twoseedlings were transferred to each pouch. All of the above opera-tions were conducted under aseptic conditions. Seedlings in thegrowth chamber were not kept sterile since the pouches are openat the top. Seedlings were maintained in a Conviron growthchamber under the following conditions 50 to 70% RH; 28 C day(16 h); 24 C night; light intensity, 1.7 x 104 lux. The pouches werekept in an upright position in plastic trays with high sides andwere checked daily to determine if watering was required. Thepaper towels were kept moist with sterile, half-strength N-freeJensen's medium, but not so wet that liquid dripped out when thepouches were inverted.Marking and Inoculation. Seedlings were inoculated I day after

1027 www.plantphysiol.orgon May 15, 2018 - Published by Downloaded from Copyright © 1980 American Society of Plant Biologists. All rights reserved.

BHUVANESWARI, TURGEON, AND BAUER

transfer to the pouches, when the roots were 3 to 9 cm long. Awaterproof marking pen was used to indicate the position of theRT3 at the time of inoculation. The position of the SERH at thetime of inoculation was determined by inspection through adissecting microscope. The RT and SERH marks (Fig. IA) weremade as gently as possible on the plastic overlaying the root.

Seedlings were inoculated by one of three methods. (a) In thepouch technique, 500 tI inoculum suspension were dripped ontothe lower portion of the root. In some cases, the inoculum suspen-sion (I ml) was added on the paper towel between the roots of thetwo plants in a pouch. (b) In the droplet technique, a small pieceof Parafilm (I x 2 cm) was inserted between the paper towel ofthe growth pouch and the portion of the root nearest the tip. A 5-1il droplet of inoculum suspension then was placed on the topsurface of the root with a microliter pipette fitted with a disposableplastic tip. Droplets adhered to the root surface more tenaciouslythan to the plastic tip, so that transfer was usually quantitative.Droplets on the root normally remained in place until the upperplastic wall of the pouch was allowed to touch the root again. Theliquid then spread within a few minutes to form a thin filmcovering the root surface for a distance of about 2 cm (the lengthof the Parafilm strip). (c) In the spot technique, a 1-,ul capillarypipette was filled with a suspension of bacteria and India ink,sealed at one end, and then touched gently and quickly to thedesired point on the surface of a root. The suspension was preparedby centrifuging the India ink, resuspending the particles to theoriginal volume in H20, and mixing this volume for volume witha bacterial cell suspension (usually I x 107 cells/ml).

Scoring of Nodulation. Soybean seedlings were normally main-tained in the growth chamber for 7 to 8 days after inoculation. Bythis time, the most mature nodules on the primary root wereusually 0.5 to 2 mm in diameter and easily scored. Small noduleswere distinguished from emerging lateral roots by their spherical(rather than pointed) shape. In cases when it was necessary todistinguish between nodules on the primary root and nodulesdeveloping at the bases of lateral roots, nodules were carefullybroken off from the roots in order to determine the point ofattachment. The positions of the nodules on the primary root weredetermined to the nearest 0.5 mm with the aid of a dissectingmicroscope (X6). Nodules developed on lateral roots as well asprimary roots and developed on the primary root below, as wellas above, the RT mark made at the time of inoculation. Nodula-tion on lateral roots and on the primary root below the RT markwas not of major concern in the present studies.

Root Elongation. Changes in the position of epidermal cells onsoybean roots, relative to the RT and SERH marks on the pouches,were assessed by measuring the displacement of ion-exchangebeads (4) placed on the root surface. Soybean seedlings with roots3 to 8 cm long were marked as shown in Figure IA. The growthpouches were cut and folded back to expose the roots. Individualbeads (Bio-Rad AG I -X8, Cl-, 45-106 t,m) were placed gently withfine forceps on the side of a root at intervals of 1 to 2 mm. Thedistance of each bead from a fixed point near the root crown wasmeasured to the nearest 0.1 mm with the aid of a dissectingmicroscope. The pouches were resealed with tape and placed inthe growth chamber until the next measurement. Measurementswere made periodically over a 22- to 26-h period.

RESULTS

Pattern of Nodulation. Nodules on the primary root developedmost frequently in the region between the RT and SERH marks(Table I). No root hairs had emerged in this region of the root atthe time when the plants were inoculated. A relatively smallproportion of the test plants formed nodules above the SERH

Abbreviations: RT, root tip; SERH, smallest emergent root hairs.

Table I. Nodulation of Soybean Cultivars in Various Zones of RootNumbers in parentheses indicate percentage of plants that were nodu-

lated in the indicated zone relative to the total number of nodulated plants.Some plants nodulated in both the no root hair and developing root hairzones. Uninoculated control plants were not nodulated.

Proportion of Plants Nodulated

Zone of Root cv. Williams cv. Beeson

Poucha Droplet' Poucha Droplet"Mature root hairzone 0/36 (0) 0/29 (0) 0/42 (0) 0/19 (0)

Developing roothair zone 4/36 (11) 1/29 (3) 9/42 (21) 0/19 (0)

No root hairzone 30/36 (83) 19/29 (66) 36/42 (86) 10/19 (53)

Only on lateralroots or belowRT mark 6/36 (17) 9/29 (31) 6/42 (14) 9/19 (47)a Inoculation techniques were as described under "Materials and Meth-

ods."

mark, in the zone which contained developing root hairs at thetime of inoculation. None of the plants developed nodules in thatregion of the primary root where mature (i.e. fully elongated) roothairs were present at the time of inoculation.The location of infectible cells on the root surface at the time of

inoculation was deduced from the positions at which nodulessubsequently formed on the root, taking the effects of root elon-gation into consideration. Root elongation caused portions of theprimary root to be displaced relative to the marks made at thetime of inoculation on the plastic face of the growth pouch. Ion-exchange beads placed on the roots above the SERH mark werenot measurably displaced by elongation. Beads initially positionedin the zone without root hairs were displaced by root elongationduring the first 24 h of an experiment, but not thereafter. Rootselongated at an average rate of 2.4 ± 0.6 mm/h. The arrows andassociated numbers in Figure I indicate the distances that givenpoints on the root surface of an average plant were displaced as aresult of root elongation. The zone of significant elongation typi-cally extended 5 to 7 mm back from the root tip, ie. approximatelyhalf the total length of the "no root hair" zone. Epidermal cells inthe lower third of the no root hair zone, nearest the tip, weredisplaced by elongation to positions below the RT mark.

Profiles of nodulation frequency provided further informationregarding the location of infectible host cells on the root. Thehistogram in Figure 2 shows the frequency of nodulation as afunction of distance from the RT mark. The relative distance ofeach nodule from the RT mark was calculated as a percentage ofthe distance between the RT and SERH marks for the given plant.This was done in order to facilitate comparisons between profilesobtained at different times and under different conditions.

Nodulation frequency above the RT mark was zero or very lowat distances greater than approximately 120% of the RT-SERHinterval. The frequency increased rapidly at shorter distances fromthe RT mark. A maximum frequency of nodulation was observedin the region 20 to 80% of the RT-SERH distance above the RTmark. The frequency of nodulation then diminished substantiallyin the vicinity of the RT mark and below it. The profile illustratedin Figure 2 is representative of results obtained from several suchexperiments in these respects. Comparison of profiles from repli-cate experiments has indicated that the peak of maximum nodu-lation frequency quite consistently reached two-thirds its maxi-mum height between 80 and 90% of the RT-SERH distance abovethe RT mark. However, this peak diminished to two-thirds itsmaximum height anywhere between 20% of the RT-SERH dis-

1028 Plant Physiol. Vol. 66, 1980

www.plantphysiol.orgon May 15, 2018 - Published by Downloaded from Copyright © 1980 American Society of Plant Biologists. All rights reserved.

LOCALIZATION OF INFECTIBLE ROOT CELLS

(1) ~~~ZON*E Of DEV.MATUREROOTU ZONE OF NO

ROOT HAIRS HAIRS ROOT HAIRS

A. ROOT AT TIME nZ^l 111|111OF INOCULATION I

SERH RTKMARK MR30mm 71mm 14Bmm

B. ROOT 7 DAYS 13 11131lIHIUIIIjjpIII Il IAFTER INOCULATION U J=

SCALE in mm 124 20 16 12 8 4 0 4 8

(2)

Z .

LL

z

160 120 80 40 RT 40 80 120 160RELATIVE DISTANCE FROM RT MARK [%RT-SERH DISTANCE)

FIG. 1. Diagrammatic representation of a soybean root at the times ofmarking and determination of nodule position. A, zones of root hairdevelopment and positions of marks, made at the time of inoculation, onthe growth pouches at RT and SERH. B, the same region of the root Iweek after inoculation. The displacement of epidermal cells, relative tothe RT and SERH marks, is indicated by arrows from the root at the timeof marking (A) to the root at the time of scoring (B). Results from fourseparate experiments were averaged and normalized to a "typical" plantwith a distance of 14.6 mm between RT and SERH (see legend to Fig. 2).The distances given with the arrows indicate how far epidermal cells atparticular initial locations were displaced.

FIG. 2. Profile of nodulation frequency. The positions of all noduleson the primary roots of 150 healthy seedlings were measured relative tothe RT mark. Each seedling was inoculated by the pouch technique with500 t1d of a I x 108 cells/ml suspension of R. japonicum 31 lb 138 grown tomid-log phase on modified YEM medium (9). The distance from RT toSERH was determined for each plant at the time of inoculation, estimatedto the nearest 0.1 mm. The relative distance of each nodule on the primaryroot from the RT mark was calculated as a percentage of the RT-SERHdistance for the given plant. Of the 150 seedlings, 120 (80%1) developednodules above the RT mark. The mean distance between the RT andSERH marks for these plants was 14.6 + 5.0 mm (SD), with a range from3.0 to 25.5 mm.

tance above the RT mark and 40%o of this distance below themark. The extent to which nodulation was diminished in theregion below the RT mark was also somewhat variable, rangingroughly between a 3-fold to a 20-fold decrease with respect to themaximum frequency. Profiles similar to that shown in Figure 2were also obtained with rhizobia cultured on the modified B5medium and with seedlings inoculated by the droplet technique.However, with droplet-inoculated seedlings, the frequency ofnodulation was found to increase again in the region 30 to 70%o ofthe RT-SERH distance below the RT mark. Approximately 7% ofthe nodules in Figure 2 developed above the SERH mark. This isconsistent with the low proportion of plants nodulated above theSERH mark (Table I).The pattern of nodulation indicated by the data in Table I and

Figure 2 was not greatly affected by changes in strain, cultivar, ormethod of inoculation. Three cultivars of soybean (Beeson, Harris,and Williams) were tested with R. japonicum 311b 138, and twostrains of R. japonicum (31 lb 83 and 31 lb 138) were tested withcultivar Williams. Essentially the same pattern of nodulation was

obtained in each case. Nodules were observed to form on lateralroots in the same zones of root development as described for themain root. When inoculated by the spot technique, plants quiteconsistently formed nodules at the point of inoculation in the noroot hair zone, only occasionally formed nodules at the point ofinoculation in the developing root hair zone, and did not formnodules at the point of inoculation in the mature root hair zone.

Effect of Inoculum Dose. The extent and location of noduledevelopment on primary root above the RT mark depended onthe number of bacteria in the inoculum (Figs. 3 and 4). Theproportion of plants that developed nodules above the RT markincreased linearly with the logarithm of the inoculum dose forseedlings inoculated by the pouch technique (Fig. 3). The samelogarithmic dependence on inoculum dose was obtained withplants inoculated by the droplet technique (Fig. 4). The numberof nodules that developed above the RT mark, as well as theaverage distance of the uppermost nodule above the mark and theproportion of plants that nodulated above the mark, all varied inan approximately linear manner with the logarithm of the inocu-lum dose. Relatively few of the plants were nodulated afterinoculations with 20 to 100 R. japonicum cells/plant, and thoseplants that did develop nodules ordinarily developed only one ortwo nodules on the primary root. As a consequence, standarderrors were quite large for data points obtained at low inoculumdoses. In particular, the apparent increase in average distance ofthe uppermost nodule at the lowest inoculum dose was not repro-ducible (Fig. 4). The log-linear nature of the dose-response curves,the maximum values of the nodulation parameters reached at thehighest inoculum doses, and the apparent intercepts of the extrap-olated curves on the inoculum dose axis, however, were all quitereproducible.

Delayed Inoculation. When inoculations were delayed for var-ious intervals, from I to 4 h after marking the positions of the roottips, it was found that nodulation above the RT mark decreased

E

0:

0

.0

0

~00

C

O-C

0

0.

80-

601

40-

20k

2 3 4Log bacterial cells

5 6 7 8added / pouch

FIG. 3. Effect of inoculum size on nodule development: pouch inocula-tions. Sets of 20 to 25 soybean plants were inoculated by the pouchtechnique with I ml inoculum suspension placed on the paper towelbetween the roots of the two plants in a pouch. A 10-fold serial dilutionseries of R. japonicum 311b 138 was prepared and each set of plants was

inoculated with one concentration of rhizobia at the time of marking(marked at RT only). Bacterial cell numbers were measured by OD andverified by direct counting in a Petroff-Hauser chamber. The percentageof plants in each set which developed nodules on the primary root abovethe RT mark was determined 7 to 8 days after inoculation and marking.The results from two separate experiments, done at different times, are

shown. The curve was fitted to the data points by the least-squares method.Plants that died or were injured were not scored. No nodules were observedon uninoculated control plants.

50O 40 _ DIRECTION of ROOT GROWTH40

10 Ii

II

A/

0

A

0/ A

I I

Inn{ . . . . . .

Plant Physiol. Vol. 66, 1980 1029

!

www.plantphysiol.orgon May 15, 2018 - Published by Downloaded from Copyright © 1980 American Society of Plant Biologists. All rights reserved.

BHUVANESWARI, TURGEON, AND BAUER

a

0

.0E

0

cr

0

~00av)

0

C

0l-0

4

E

E

I-E

0S.

00

c0

0

I-I~~~~~~~~~~-0 1-21j3 4

Log number of cells added /plantFIG. 4. Effect of inoculum size on nodule development: droplet inoc-

ulations. Sets of 20 to 30 soybean plants were inoculated by the droplettechnique at the time of marking (RT mark only). Each set of plants wasinoculated with one concentration of R. japonicum 31 lb 138 from a 5-foldserial dilution series. Bacterial cell concentrations were measured byoptical density and verified by plate counting of colonies. Subsequentnodulation was scored to determine the percentage of plants with nodulesabove the RT mark (U), the average number of nodules above the RTmark per plant (0), and the average distance of the uppermost primaryroot nodule from the RT mark (A). Curves were fitted to the data pointsby the least-squares method, omitting the data point at the lowest inoculumdose for the curve giving the average distance of the uppermost nodule.Vertical bars indicate SE. Plants were scored by cutting through the rooton the upper edge of the RT mark and examining the upper and lowerportions of the root for nodule development under a dissecting microscope.Uppermost nodules that developed below the RT mark, but within 10 mmof the mark, were assigned negative values for distance. The relatively fewplants which failed to develop primary root nodules above the RT markor within 10 mm below the mark were disregarded in computing theaverage distance of nodules from the mark. Unhealthy plants were notscored. Uninoculated control plants developed no nodules.

as the time interval between marking and inoculation was in-creased (Fig. 5). The average position of the uppermost noduledecreased in a linear manner in these experiments at a rate quitesimilar to the average rate of root elongation (2-3 mm/h). Therewas relatively little change in the average number of nodulesabove the RT mark or in the percentage of plants with nodulesabove the mark during the first I or 2 h of the experiments.However, these two parameters decreased in a relatively linearfashion when inoculations were delayed between 2 and 4 h. Theygradually approached and reached zero when inoculations weredelayed by 5 to 6 h. Average values for the position of theuppermost nodule and for the number of nodules above the markwere quite consistent from point to point within an experiment orfrom week to week between experiments despite the wide range ofvalues for individual plants and the correspondingly large stand-ard errors.

DISCUSSION

At any given time, the soybean root cells infectible by Rhizobiumare located in a small, developmentally restricted zone just abovethe root tip. Nodules developed on the primary root most fre-quently in the region from 20 to 80% of the RT-SERH distanceabove RT mark (Fig. 2). At the time of inoculation, the cells in

14

0 E0.4 0- 12

E E1-lOE2310cr~~~~~~~~~~I

cr3.0-

Inera bewe akn ndiouain(

makSeso80 to 30syenpat >eemakda h ottpa

0

0

ID60 a)20-6~0 2-~c 0 0

oaR.jpncm401- 3 supnso cotiigIx1'cllsm) fe

0.

20-2 00

2 3 4Interval between marking and inoculation (h)

FIG. 5. Effects of delayed inoculation on nodulation above the root tipmark. Sets of 20 to 30 soybean plants were marked at the root tip atvarious time intervals prior to inoculation by the droplet technique (5 'Ulof a R. japonicum 31 lb 138 suspension containing I x 107 cells/ml). AfterI week, the number and position of primary root nodules above the RTmark were determined for each plant as described in the legend for Figure4. Calculations were made of the percentage of plants in each set withnodules above the RT mark (a), the average number of nodules above themark per plant (0), and the average distance of the uppermost nodulefrom the RT mark (A). Vertical bars indicate SE.

this most frequently nodulated region were located approximately50 to 80% of the RT-SERH distance above the RT mark (Fig. 1).Since the rapidly elongating portion of the root typically extendsto about 50% of the RT-SERH distance above the root tip, itseems that infections leading to nodulation are normally initiatedin epidermal cells that have finished most of their elongation buthave not reached the stage of root hair emergence. The infectibilityof cells in this region refers to their capability ofbecoming infectedso that nodules are subsequently formed. It is possible that infec-tions that do not lead to nodule formation are initiated in otherregions of the root. This point requires further study.The identity of the cells that are susceptible to infection by

rhizobia cannot be specifically determined from the present stud-ies. However, light microscopic studies of infections in soybeanhave provided evidence that rhizobia normally enter the root bymeans of infection threads formed in root hair cells (2, 5, 8).Examination of the infectible region of roots by serial sectioninghas revealed infection threads developing only in short (15-40/Lm) emergent root hairs (B. G. Turgeon and W. D. Bauer,unpublished data). Such microscopic evidence suggests that emer-gent root hairs are the infectible cells. Nonetheless, nodules formedmost frequently in that region of the root where no emergent roothairs were present at the time of inoculation (Table I, Fig. 2). Aresolution of this apparent discrepancy cannot be provided withany certainty. We speculate that developing, pre-emergent roothair cells must be activated by the bacteria, or by some substancefrom the bacteria, in order for these hair cells to develop as sitesof subsequent infection and nodulation. The development of pre-emergent hair cells as sites of subsequent infection and nodulationappears to require a period of 2 to 3 h after inoculation. This timeestimate is based on an average distance between the zone ofSERH and the zone of maximum nodulation frequency (Figs. Iand 2), and on the average rate of root elongation (2.4 mm/h).Time estimates of this sort, however, depend on the assumption

1030 Plant Physiol. Vol. 66, 1980

www.plantphysiol.orgon May 15, 2018 - Published by Downloaded from Copyright © 1980 American Society of Plant Biologists. All rights reserved.

LOCALIZATION OF INFECTIBLE ROOT CELLS

that the bacteria are ready to activate the infection process, withoutdelay, at the time of inoculation.

Delayed inoculation experiments demonstrated that nodulationabove the RT mark decreased rapidly as the interval between thetime of marking and time of inoculation was increased. This isconsistent with the notion that infections leading to nodule devel-opment cannot be initiated in root hair cells that are too old. Haircells are no longer infectible approximately 5 h after they originatefrom the root meristem. Further support for this conclusion isprovided (Fig. 2) by the substantially diminished frequencies ofnodulation at distances greater than about 12 mm above the RTmark, which is equivalent to about 5 h root growth at 2.4 mm/h.The data in Figure 5 indicate that infectibility is acquired and lostby a process of acropetal development. The transient, acropetaldevelopment of infectibility in soybean makes the position of theuppermost nodule a sensitive indicator of factors that affect therate at which infections are initiated. This should be useful insubsequent studies of the infection process.The average distance of the uppermost nodule from the RT

mark decreases significantly (to less than half its initial value)when inoculations are delayed by only 2 h after marking (Fig. 5).We infer from this that target epidermal cells of the host root areable to respond to the presence of the bacterial symbiont verysoon after inoculation, perhaps within minutes and certainly inless than 2 h. It will be of great interest to determine the nature ofthis early response and how it is elicited.From observations of infection thread formation in root hairs

of clover, Nutman (6) concluded that apical infections (i.e. infec-tions initiated at the tips of root hairs) occur only in the developingroot hair zone. Lateral infections (initiated at a branch near thebase of a hair), however, were observed in the zone of mature,fully elongated root hairs. The apical infections described byNutman (6) seem to be localized in an acropetally developingzone in clover just as infectible root epidermal cells are in soybean.However, the development of lateral infections in the mature roothair zone of clover does not appear to occur in soybean. Thispoints out the need for caution in assuming that Rhizobiuminfections proceed by the same paths in different hosts.We have consistently observed a lower frequency of nodulation

in the vicinity of the RT mark and below this mark (Fig. 2). Wespeculate that the lower frequency of nodulation in this regionmay reflect the existence of a quick-acting regulatory mechanismin the host that serves to prevent overnodulation.The logarithmic relationship between nodulation above the RT

mark and the number of bacteria in the inoculums (Figs. 3 and 4)is a noteworthy phenomenon. From a statistical point of view, theprobability that a Rhizobium cell will come in contact with asuitable site on the root and initiate an infection essentially doublesif the number of cells in the inoculum is doubled. Actually,nodulation above the RT mark is increased by much less than 2-

fold when the inoculum dose is doubled. Thus, there appears tobe an inhibitory or regulatory effect superimposed on the statisticaleffect of increasing the inoculum dose. This inhibitory or regula-tory effect is independent of the method of inoculation. Bothpouch and droplet inoculation techniques result in nodulationthat is logarithmically dependent on inoculum dose (Figs. 3 versus4). The inoculum dose must be roughly 100-fold larger for pouchinoculations than for droplet inoculations in order to achieve thesame extent of nodulation, probably because most of the bacteriain a pouch inoculation contact or adhere to the paper towel ratherthan the root, whereas essentially all the bacteria in a dropletinoculation come in contact with the root.The relationship between inoculum dose and the extent of

nodulation leads to the possibility that more than one Rhizobiumcell might be required to initiate the infection process. Extrapo-lation of the dose-response curves to the inoculum dose axis hasconsistently provided intercepts in the range of 5 to 20 bacteria/plant. Perhaps a minimum of 5 to 20 R. japonicum cells must actin concert to initiate an infection. Alternatively, perhaps only onecell in every 5 to 20 cells in the inoculum is infective. Furtherexperiments are required to explore such alternatives. Purchaseand Nutman (7), using very different methods, have previouslyreported evidence for a logarithmic dependence of total nodulenumber on inoculum size in clover, as well as evidence thatnodulation requires a minimum of about 10 Rhizobium cells. Thusthe phenomenon may prove to be quite general.

Acknowledgments-We thank Andrew Mort for his many helpful suggt.stionsduring the course of the work and preparation of the manuscript. We appreciate theexcellent and steadfast technical assistance of Greg Gottschhich and Steve Rozzo.

LITERATURE CITED

1. BHUVANESWARI TV, SG PUEPPKE. WD BAUER 1977 The role of lectins in plant-microorganism interactions. 1. Binding of soybean lectin to rhizobia. PlantPhysiol 60: 486-491

2. BIEBERDORE FW 1938 The cytology and histology of root nodules of someleguminosae. J Am Soc Agron 30: 375-389

3. DARr PJ 1977 Infection and development of leguminous nodules. In RWFHARDY, WS SILVER. eds. A Treatise on Dinitrogen Fixation. Section 111. WileyInterscience, New York, pp 367-412

4. GREEN PB 1965 Anion-exchange resin spheres as marking material for wet cellsurfaces. Exp Cell Res 40: 195-196

5. NEWCOMB W, D SIPPELL, RL PETERSON 1979 The early morphogenesis of Glicinemax and Pisum sativum root nodules. Can J Bot 57: 2603-2616

6. NUUMAN PS 1959 Some observations on root-hair infection by nodule bacteria.J Exp Bot 10: 250-263

7. PURCHASE HF, PS Nur.MAN 1957 Studies on the physiology of nodule formation.VI. The influence of bacterial numbers in the rhizosphere on nodule initiation.Ann Bot 21: 439-454

8. RAO RV, DL KEISrER 1978 Infection threads in root hairs of soybean (Glvcinemax) plants inoculated with Rhizobium japonicum. Protoplasma 97: 31 1 -316

9. VINCENr JM 1970 A Manual for the Practical Study of Root Nodule Bacteria.IBP Handbook No. 15. Blackwell Scientific Publications, Oxford

10. WEAVER RW, LR FREDERICK 1972 A new technique for most probable numbercounts of Rhizobia. Plant Soil 36: 219-222

Plant Physiol. Vol. 66, 1980 1031

www.plantphysiol.orgon May 15, 2018 - Published by Downloaded from Copyright © 1980 American Society of Plant Biologists. All rights reserved.